1
00:00:09,110 --> 00:00:16,580
Hello and welcome to episode six of Gut Instinct. Research Updates Bringing you the latest research in Gastroenterology and Hepatology.

2
00:00:17,310 --> 00:00:21,200
I'm Tamsin Cargill. Lecturer in gastroenterology at Oxford.

3
00:00:21,200 --> 00:00:24,290
Interested in hepatology. Follow hepatitis vaccine.

4
00:00:25,100 --> 00:00:28,130
And I'm Michael Fitzpatrick, known as Fitz to pretty much everyone.

5
00:00:28,370 --> 00:00:34,130
And as well as Thomson's podcast sidekick, clinical lecturer in Gastroenterology at Oxford.

6
00:00:34,620 --> 00:00:39,570
My research interests in coeliac disease, clinical nutrition.

7
00:00:40,140 --> 00:00:49,410
So we started this podcast to bring you some of the most interesting GI related papers that have come out recently, and we're already on episode six.

8
00:00:50,250 --> 00:00:57,780
Each episode we'll talk you through normally a couple of primary research papers in a bit of detail, one clinical and one translational.

9
00:00:57,780 --> 00:01:02,700
Although today we've got two cracking clinical endoscopy papers for you.

10
00:01:03,000 --> 00:01:08,280
So it'll give we'll give you our take on the research and what we think of it.

11
00:01:09,600 --> 00:01:12,270
Clearly there are loads of great papers coming out every month.

12
00:01:12,390 --> 00:01:19,260
So in addition, we'll give you a slightly more Rapidfire rundown of what else is out there in the gastro world.

13
00:01:19,260 --> 00:01:26,010
In our five in 15 section, we're aiming to give you some context and critical appraisal of the papers we talk about.

14
00:01:26,550 --> 00:01:35,610
And and while we both love gastro, I'm more interested in liver disease, whereas Fitz is more interested in IBD small bowel and nutrition.

15
00:01:36,030 --> 00:01:39,390
So hopefully this podcast will have something for everyone.

16
00:01:41,250 --> 00:01:45,360
Now, as we say, each episode, no podcast should constitute medical advice.

17
00:01:45,360 --> 00:01:52,480
If you are a patient, go and consult your medical practitioner for advice about your medical management and for doctors, please.

18
00:01:52,710 --> 00:01:57,480
I'm sure you would not base your medical management solely on what some people told you on the Internet,

19
00:01:57,480 --> 00:02:03,390
and you will verify this with your own reading and research to let us know what you think of the podcast.

20
00:02:03,900 --> 00:02:12,720
We're on Twitter. You can rise. Write us a lovely review on Apple Podcasts, Spotify or whatever streaming platform you use,

21
00:02:12,870 --> 00:02:17,460
and we'd really love to hear from you with any suggestions or thoughts.

22
00:02:23,170 --> 00:02:26,980
Now. Tamsin, do you want me to start today?

23
00:02:28,090 --> 00:02:31,120
Yeah. Go for it. Okay, cool. So I'm.

24
00:02:32,780 --> 00:02:37,790
I read too many papers. It's not often a problem. But I realised I had.

25
00:02:37,880 --> 00:02:46,850
I had. I had five that I wanted to talk about today. And there were two too I think really interesting endoscopy papers.

26
00:02:46,850 --> 00:02:53,510
I don't think we spent enough time talking about endoscopy, so they're nice clinical trial papers in endoscopy.

27
00:02:53,810 --> 00:03:01,280
And interestingly, I was reading one of them this afternoon while an endoscopy and the consultant I was scoping with,

28
00:03:01,490 --> 00:03:04,940
he's one of the authors on the paper. You go multitasking.

29
00:03:05,390 --> 00:03:10,790
I interviewed them about the paper while while chatting over a light kaleidoscope.

30
00:03:10,970 --> 00:03:17,510
And this is about colonoscopy. And this paper was in gastroenterology just a few days ago,

31
00:03:17,840 --> 00:03:25,520
and it is entitled The Impact of Artificial Intelligence on Ms. Rate in Colorectal Neoplasia.

32
00:03:26,690 --> 00:03:38,329
And this is from a group of authors from Florida, from Italy, a few other centres, and also from Oxford, where we are based and so on.

33
00:03:38,330 --> 00:03:47,660
The author is James East, Rebecca Palmer and and Beth Burt Lieberman, amongst others from the Oxford Endoscopy Group,

34
00:03:48,380 --> 00:04:00,200
said the background to this is that we know that there is a significant Ms. rate in colorectal rectal neoplasia.

35
00:04:00,200 --> 00:04:06,380
So say polyps after an index surveillance colonoscopy.

36
00:04:06,590 --> 00:04:15,590
And it's one of the key drivers for you know, that the real problem in colorectal cancer screening, which is post colonoscopy, colorectal cancer.

37
00:04:15,950 --> 00:04:22,670
So that's a that's a cancer occurring after a screening colonoscopy that wasn't picked up.

38
00:04:23,000 --> 00:04:27,590
And the the incidence for that is around 1% at ten years.

39
00:04:27,890 --> 00:04:38,420
So quite significant. And studies have suggested that around half of that is due to the fact effectively we've missed smaller polyps at that index

40
00:04:38,420 --> 00:04:45,260
colonoscopy and that that's progressed along that adenoma carcinoma pathway and developed into a kind of rectal cancer.

41
00:04:46,730 --> 00:04:52,219
So in studies that have Yeah. Have looked at the missed rate of co-direct neoplasia.

42
00:04:52,220 --> 00:04:54,830
So if you for instance do back to back colonoscopy,

43
00:04:55,040 --> 00:05:03,320
the Ms. rate is surprisingly high in the range of 25% missed polyps at these these screenings since

44
00:05:03,590 --> 00:05:09,229
effectively they can either be due to the fact that the mucosa is been inadequately visualised,

45
00:05:09,230 --> 00:05:14,780
we've not been able to see around a certain fold or particular areas like in the caecum or the hepatic

46
00:05:14,780 --> 00:05:20,810
or splenic lecture that which are harder to visualise are just not seen or we've visualised it.

47
00:05:20,960 --> 00:05:28,910
But the endoscopy has not recognised that there's pathology there and there are a variety of factors that can drive that.

48
00:05:28,910 --> 00:05:34,250
So inexperienced operators, lack of training, distraction, tiredness,

49
00:05:34,250 --> 00:05:40,520
all of these reasons why a polyp may flash up on screen, but the colonoscopy doesn't see this.

50
00:05:41,420 --> 00:05:49,249
So it would be great if there was a way we could augment our detection of polyps during colonoscopy.

51
00:05:49,250 --> 00:05:58,100
And certainly there's been a lot of research over that over the last decade or so using artificial intelligence platform.

52
00:05:58,110 --> 00:06:07,670
So sort of deep learning platforms to analyse the the data from the colonoscopy and try and highlight areas that may contain a pull out.

53
00:06:08,420 --> 00:06:16,819
And studies of these technologies have been pretty successful in that they are in the range of sort

54
00:06:16,820 --> 00:06:25,160
of 95% sensitive and 88% specific for finding colorectal neoplasia finding polyps on colonoscopy.

55
00:06:25,910 --> 00:06:32,120
But what we're really interested in is how they affect either adenoma detection rates.

56
00:06:32,120 --> 00:06:39,650
So the percentage of surveillance, colonoscopy is where and that's oedema is detected, but also adenoma.

57
00:06:39,980 --> 00:06:43,820
Ms. Right. So how many astronomers are missed?

58
00:06:44,450 --> 00:06:46,400
And now there are quite a lot.

59
00:06:46,430 --> 00:06:56,480
There are a number of trials using AI technologies and in particular AI technology called GI Genius, which is made by Medtronic.

60
00:06:57,350 --> 00:07:01,790
And they've looked at adenoma detection rates. So that's the number of surveillance.

61
00:07:01,790 --> 00:07:03,470
Colonoscopy is where you find it out today.

62
00:07:03,860 --> 00:07:15,080
And and studies have shown that the RET that there's a relative increase in your added no mat detection rate of about 44%.

63
00:07:15,230 --> 00:07:17,240
So it increases your adenoma detection rate,

64
00:07:17,240 --> 00:07:31,960
which is great but and post colonoscopy colorectal cancers will be also driven by your missed your your adenoma Ms. rates and they're.

65
00:07:32,250 --> 00:07:37,500
Really different things because your adenoma detection rate is per patients.

66
00:07:37,950 --> 00:07:41,190
So like we've either found a polyp in this patient or we happens.

67
00:07:41,880 --> 00:07:45,210
But your adenoma, Miss Rate, is per polyp.

68
00:07:45,600 --> 00:07:49,260
So let's say the patient has three polyps, but you only found two.

69
00:07:49,500 --> 00:07:55,500
You've missed one polyp. So you might you might have detected one polyp in that patient, but actually missed the other two.

70
00:07:56,520 --> 00:08:07,139
So they are both important metrics when we consider trying to prevent post polyp back to me and colorectal cancers and this is

71
00:08:07,140 --> 00:08:20,690
what this trial was trying to trying to test was whether using this I can increase or sorry can decrease the adenoma miss rate.

72
00:08:21,270 --> 00:08:27,840
Okay. So the study design, this was performed in three countries and in eight centres.

73
00:08:28,200 --> 00:08:31,710
And in each centre there were at least three endoscopies.

74
00:08:32,010 --> 00:08:34,079
So I really like this sort of trial design.

75
00:08:34,080 --> 00:08:45,200
So we're controlling for we're controlling for, for sort of Europe, UK and United States in terms of location, but also different endoscopies.

76
00:08:45,480 --> 00:08:51,330
And actually they controlled to some extent of further kind of, I guess, the quality and experience of these endoscopy.

77
00:08:51,330 --> 00:08:52,500
So they were all pretty experienced.

78
00:08:52,500 --> 00:09:00,960
They'd all done at least 2000 colonoscopy and they all had adenoma detection rates in the range of between 20 to 40%.

79
00:09:01,290 --> 00:09:05,030
So yeah, pretty good. They're, they're pretty good endoscopies.

80
00:09:05,190 --> 00:09:14,610
But it's not like some trial just, you know, just with with to, you know, to incredible endoscopies or indeed just with novel.

81
00:09:14,610 --> 00:09:20,669
And so I think this is kind of representative of, you know, kind of high quality, high quality endoscopy.

82
00:09:20,670 --> 00:09:30,000
But but in the kind of more real world setting, all of these the patients were in the trial were patients over 45 years old

83
00:09:30,240 --> 00:09:36,720
who were having a screening or surveillance procedure for colorectal cancer.

84
00:09:36,840 --> 00:09:42,330
So these had previous polyps or they met whatever the criteria was for screening in that country.

85
00:09:42,570 --> 00:09:46,740
And they excluded people with polyposis syndromes.

86
00:09:46,740 --> 00:09:52,980
So they were sort of standard, sort of and, you know, adenoma risks as opposed to specific polyposis syndromes.

87
00:09:53,790 --> 00:10:00,780
And they randomised the patients to undergo one of two arms of a back to back design.

88
00:10:01,020 --> 00:10:07,470
So these lucky patients got one colonoscopy turns into two to back to back.

89
00:10:08,050 --> 00:10:11,750
Of course, you know, you just you to pull out. That's the end of the procedure, sir.

90
00:10:12,420 --> 00:10:17,310
And then and now we're going to do it again. We're going to do it all over again or roll over and let's start.

91
00:10:18,120 --> 00:10:22,330
So let's all go. Wow.

92
00:10:22,350 --> 00:10:25,860
I'm I'm surprised anyone consented to that, but that's great.

93
00:10:26,760 --> 00:10:29,850
Yeah, well, you'd be delighted to know. 230 patients,

94
00:10:30,210 --> 00:10:36,810
200 230 subjects are included in the analysis and they were split in either have standard

95
00:10:36,810 --> 00:10:43,080
white light endoscopy followed by a II augmented white light endoscopy or vice versa.

96
00:10:43,380 --> 00:10:52,160
I kind of effectively they were they were kind of measuring is how many polyps were found on the second endoscopy and the the

97
00:10:52,230 --> 00:11:01,410
operator for both endoscopy was was the same the same I presume on the first endoscopy they didn't remove any polyps they saw.

98
00:11:02,220 --> 00:11:06,240
No, I think they locked them off if they if they were okay to be to be removed.

99
00:11:06,450 --> 00:11:09,540
So it was about okay anymore. Okay, fine.

100
00:11:10,320 --> 00:11:13,889
And they they clearly randomise patients.

101
00:11:13,890 --> 00:11:25,799
They the randomisation was very sensibly stratified by by age, by indication for endoscopy, by study site in country and so on.

102
00:11:25,800 --> 00:11:31,560
So it was all all very it's all very sensible in terms of design and.

103
00:11:32,760 --> 00:11:39,410
I am not an expert, Tamsin, but I will basically just read this paragraph from the paper.

104
00:11:39,990 --> 00:11:48,389
I'm not either. So guy genius by Medtronic has has been evaluated in a number of different trials before.

105
00:11:48,390 --> 00:12:01,530
But it's a technology on a convolutional neural network that was trained on a series of 2684 histologically confirmed polyps in a previous

106
00:12:01,530 --> 00:12:09,930
high quality randomised trial and the idea of the GI genius is that it can recognise these polyps and also classify what it thinks they are.

107
00:12:10,440 --> 00:12:16,500
So whether it thinks they're adenomas or hyperplastic or serrated or things like that, so basically whether you should remove them or not.

108
00:12:17,430 --> 00:12:24,390
And the output is quite nice because some of the early AI tools had a whole separate screen.

109
00:12:24,780 --> 00:12:28,140
So you were trying to look at two screens at once. So that's that's not great.

110
00:12:28,320 --> 00:12:37,320
But all this does is effectively it puts a little, little red box around the it puts a little box around the what it thinks is the lesion.

111
00:12:37,560 --> 00:12:41,460
And so it just attracts your eye and then you can evaluate it as you as you usually would.

112
00:12:41,670 --> 00:12:45,720
And it's currently approved for use in both the US and Europe.

113
00:12:46,740 --> 00:12:54,930
Okay, um, that sounds really very good. I think the, I mean, the only limitations of the colonoscopy procedure was that they,

114
00:12:55,170 --> 00:13:01,530
the endoscopes were not allowed to use MBI or display narrowband imaging or display for the detection of the polyps.

115
00:13:01,530 --> 00:13:09,629
They could use NPI to characterise a lesion further if they wanted to once they spotted it, but not not for anything thing.

116
00:13:09,630 --> 00:13:14,490
So this was not using any of those other and those other tools.

117
00:13:15,360 --> 00:13:26,400
Okay. So they screen 253 patients, randomise 249 and 230 were included in the main analysis.

118
00:13:27,570 --> 00:13:32,880
And, you know, there were a few people who didn't complete their that that colonoscopy,

119
00:13:33,360 --> 00:13:37,919
a few patients who dropped out for technical reasons or withdrawal or adverse events.

120
00:13:37,920 --> 00:13:45,990
But the vast majority were included in the analysis, and they did a sensible power calculation.

121
00:13:46,860 --> 00:13:51,780
So the results. So the key one was the adenoma miss rate,

122
00:13:51,780 --> 00:13:59,729
which was calculated by how many polyps were found on the second colonoscopy in the back to back

123
00:13:59,730 --> 00:14:05,520
colonoscopy as a percentage of the total number of polyps found in both procedures together.

124
00:14:06,240 --> 00:14:12,660
So the total number of polyps that's I guess as a as a sort of a ground truth for how many polyps we think there actually were.

125
00:14:13,170 --> 00:14:22,770
Um, and the adenoma miss rate was significantly lower in the air compared to the standard white light endoscopy arm.

126
00:14:22,770 --> 00:14:32,639
So it was 16.6% in the eye versus 34.6% in the white light.

127
00:14:32,640 --> 00:14:37,240
Endoscopy. Um, so that sounds like quite a big absolute difference.

128
00:14:37,240 --> 00:14:44,490
So that's, that's a reduction. Yeah, about about half the rate of of misstated.

129
00:14:44,490 --> 00:14:53,850
I was using the AI technology. Um, so that's the headline used that this technology works effectively and in a slightly more real world setting.

130
00:14:54,270 --> 00:15:00,320
Um. What's interesting is when you start to look at the kind of the size and what kind of lesions they were.

131
00:15:00,330 --> 00:15:07,080
So the way this was driven. In the main by small polyps.

132
00:15:07,590 --> 00:15:13,680
So the detection of polyps less than five millimetres in in diameter.

133
00:15:14,220 --> 00:15:16,170
So that was significantly difference.

134
00:15:16,650 --> 00:15:26,400
But actually polyps between six and ten millimetres or polyps greater than ten millimetres were not significantly difference between the two arms.

135
00:15:26,610 --> 00:15:31,469
Now, the trial wasn't powered to detect those subgroup differences,

136
00:15:31,470 --> 00:15:39,180
but it does seem that the major effect seems to be with smaller polyps, and these were mainly little sessile lesions.

137
00:15:39,780 --> 00:15:44,850
It seemed to augment the detection in all parts of the colon, the right colon and and the left colon.

138
00:15:46,710 --> 00:15:49,710
Which is which is encouraging.

139
00:15:51,990 --> 00:16:01,740
And that then affects how those patients have managed, because for a number of those patients, they had a more more polyps detected.

140
00:16:01,890 --> 00:16:05,700
And that then puts them into a shorter surveillance interval.

141
00:16:06,210 --> 00:16:13,800
Yes. So actually, it's not just that when lopping of those polyps and preventing them progressing potentially to colorectal cancer,

142
00:16:13,950 --> 00:16:17,250
but actually it changes their surveillance interval as well.

143
00:16:18,080 --> 00:16:22,760
And of course, that would have a different that would have a different impact depending on which country

144
00:16:22,760 --> 00:16:26,540
you're in with the different surveillance guidelines in different areas of the world.

145
00:16:26,570 --> 00:16:27,230
Yeah, definitely.

146
00:16:27,440 --> 00:16:39,620
And potentially in the UK with less and less frequent surveillance compared to, for example, the US, that could potentially have a bigger impact.

147
00:16:39,740 --> 00:16:44,930
Yeah, it could put someone from not having follow up endoscopes to having them.

148
00:16:45,080 --> 00:16:48,320
So actually, you know, for those patients could make a big difference.

149
00:16:48,830 --> 00:16:52,400
I think the other thing that's helpful is there's a nice forest plot on Figure two.

150
00:16:52,550 --> 00:16:59,330
And although a lot of the these differences in larger polyps and so on were not significance,

151
00:16:59,780 --> 00:17:06,050
all of them were trending in some some towards A.I. being superior.

152
00:17:07,400 --> 00:17:15,889
And I suppose if the if the study isn't powered to detect those subgroup analysis, then, you know, you'd expect that.

153
00:17:15,890 --> 00:17:21,020
But maybe they don't find a significant difference when in the real world there might well be.

154
00:17:21,110 --> 00:17:24,200
So overall, I think this is a really interesting technology.

155
00:17:24,200 --> 00:17:33,140
And and I guess when the costs come down, when it gets just integrated within systems rather than being an expensive add on,

156
00:17:33,470 --> 00:17:42,020
I think I can't imagine anybody doing surveillance endoscopy and not wanting to just turn on the extra button and and having this and it sounds great.

157
00:17:42,350 --> 00:17:45,770
Has it ever been compared directly to other methods like in the I?

158
00:17:45,830 --> 00:17:52,800
They haven't mentioned it in this paper. I'm not sure I know the literature well enough to say for certain, but it doesn't, it doesn't look like it's,

159
00:17:53,120 --> 00:17:59,210
um, I guess, I guess most people don't use MBI as a kind of routine to all the way round.

160
00:17:59,360 --> 00:18:04,610
No, no. Whereas this would just literally sort of highlight when you're there,

161
00:18:04,610 --> 00:18:10,579
it's like you don't have to do anything differently really as the yeah, the operator definitely.

162
00:18:10,580 --> 00:18:18,710
And, and in terms of display, I think everybody would love to find evidence that display isn't needed because it's,

163
00:18:19,010 --> 00:18:21,710
you know, it's it's a pain in the proverbial. Yes.

164
00:18:21,980 --> 00:18:30,230
So it would be great to be able to, you know, have have something that was as good as display and didn't involve all the fuss.

165
00:18:30,590 --> 00:18:36,230
So that's really that's really encouraging, I guess my.

166
00:18:37,480 --> 00:18:39,190
My caveat is that.

167
00:18:40,690 --> 00:18:49,390
The polyps that it seem to detect most often, the the ones that really drive this difference in miss rate are these really kind of diminutive polyps,

168
00:18:49,660 --> 00:18:55,090
which are the least likely to progress to invasive colorectal cancer over a short period of time.

169
00:18:55,600 --> 00:18:59,860
And certainly in countries that have a bowel cancer screening programme or a colonoscopy programme.

170
00:19:00,310 --> 00:19:06,240
And I guess there is a question of how in, in, in real world data,

171
00:19:06,610 --> 00:19:14,380
what's finding a few more of these tiny polyps would actually would actually do to missed colorectal cancers.

172
00:19:14,940 --> 00:19:22,929
And so I think I think if I were, you know, having to pay a huge amount of money to implement this technology,

173
00:19:22,930 --> 00:19:26,740
I can imagine maybe being a bit sceptical.

174
00:19:27,760 --> 00:19:34,090
Having said that, I think it's always a patient being scoped or if I was the endoscope doing surveillance,

175
00:19:34,360 --> 00:19:38,800
I can't, you know, I can't really see any great downsides what they had once clipped in.

176
00:19:39,250 --> 00:19:48,430
And I think it's very cool and I think it really shows one kind of where the future of aspects of endoscopy are going.

177
00:19:48,490 --> 00:19:57,670
And I think these sort of focussed, focussed uses of sort of A.I. technology to help us with certain sets and tasks.

178
00:19:58,900 --> 00:20:05,950
But also, I think just a really, really nice study which is being done in a kind of a rigorous manner to evaluate these technologies,

179
00:20:05,950 --> 00:20:10,900
where I think sometimes the hype can maybe exceed the reality.

180
00:20:10,930 --> 00:20:14,290
Yeah, no, I think it's great paper. Thank you for presenting that one.

181
00:20:14,380 --> 00:20:24,950
It's a good one. Shall I do a sciencey paper?

182
00:20:25,760 --> 00:20:31,320
Yeah, I do sciencey one and then I'll do an endoscopy sandwich and industry sandwich with that second second.

183
00:20:31,610 --> 00:20:38,149
So my sciencey paper is entitled Next Generation Sequencing of file cell free

184
00:20:38,150 --> 00:20:42,410
DNA for the early detection of patients with malignant biliary strictures.

185
00:20:43,040 --> 00:20:47,000
And this was published last month in gut.

186
00:20:47,750 --> 00:20:57,120
And the lead author, I hope I pronounce this correctly, is Archer Deira and it's Al, and the group is from Spain.

187
00:20:57,680 --> 00:21:03,950
So the background for this paper is that, as we all know,

188
00:21:04,220 --> 00:21:09,170
the diagnostic accuracy of the current methods for diagnosing a malignant

189
00:21:09,170 --> 00:21:15,560
biliary stricture as opposed to a benign biliary stricture is relatively poor.

190
00:21:16,310 --> 00:21:19,790
So sort of looking at studies,

191
00:21:19,790 --> 00:21:37,430
overall sensitivity using cytology with biliary Russians plus SCA 99 plus maybe fish analysis on a pathology specimen ranges from between 14 to 60%.

192
00:21:38,150 --> 00:21:46,730
So not fabulous. So that means that we see a stricture that we think might be latent.

193
00:21:47,210 --> 00:21:54,110
And the the diagnosis is either a it's a benign stricture when actually it's malignant

194
00:21:54,110 --> 00:21:56,930
or it's unclear whether it's benign or malignant when it's actually malignant.

195
00:21:58,670 --> 00:22:06,770
And this sort of problems with that is that that feeds into some of the poor prognosis of patients

196
00:22:06,770 --> 00:22:12,230
with malignant biliary strictures because it contributes to late diagnosis of malignancy,

197
00:22:12,620 --> 00:22:16,760
and therefore the treatment options for them are more limited.

198
00:22:18,940 --> 00:22:28,960
And so we know that there are molecular changes in both pancreatic ductal adenocarcinoma and cholangiocarcinoma,

199
00:22:29,080 --> 00:22:32,770
which can both lead to malignant biliary strictures.

200
00:22:33,730 --> 00:22:41,460
And and they've been identified using either whole genome sequencing or targeted platforms,

201
00:22:41,470 --> 00:22:46,810
looking at particular cancer associated genes in tumour DNA.

202
00:22:47,830 --> 00:22:57,220
And whether you've actually taken either a proper and histological sample of tumour or a biopsy and sequenced the tumour itself.

203
00:22:57,610 --> 00:23:06,790
And there are some recurrent ulcerations in oncogene and tumour suppressor genes and they can be associated with the anatomical site of the tumour.

204
00:23:07,030 --> 00:23:16,480
So for example, intrahepatic mangos are associated with a mutations in something called IDR,

205
00:23:16,690 --> 00:23:26,919
one in 13% of patients and FGF or two in 20% of mutations and extrahepatic palanga and

206
00:23:26,920 --> 00:23:33,850
gallbladder cancer are associated with KRAS or EOB two mutations in up to 20% of patients.

207
00:23:35,530 --> 00:23:41,270
There are also differences between tumours with a different background aetiology.

208
00:23:41,290 --> 00:23:50,650
For example, patients that have PSC related biliary cancer or liver fluke related BEN-ARI cancer are associated with particular mutations.

209
00:23:51,640 --> 00:23:56,740
And but and the importance of this overall is that some of these mutations are

210
00:23:56,740 --> 00:24:02,080
now molecular targets for oncology drugs that are now approved for treatment.

211
00:24:02,710 --> 00:24:11,860
So for example, with that idea, one mutation in intrahepatic cholangiocarcinoma is there's a drug called either cediranib,

212
00:24:12,580 --> 00:24:23,020
which is a small molecule inhibitor of IDR one, and that's been shown to improve progression free survival in addition to the usual treatment

213
00:24:23,380 --> 00:24:31,150
for non surgically respectable and Billary cancers that have that mutation in them.

214
00:24:31,540 --> 00:24:36,730
But in order to have that drug, you need to have had a biopsy that shows you have that mutation.

215
00:24:37,810 --> 00:24:48,280
So previous pilot studies have shown that you can identify cancer, meet these cancer mutations using next generation sequencing technologies,

216
00:24:48,670 --> 00:24:57,970
and that this can improve the diagnostic accuracy of diagnosing a malignant as opposed to a benign biliary stricture.

217
00:24:58,600 --> 00:25:07,870
And these previous sequencing methods, have they been looking at Bille or have they been looking at circulating various different things?

218
00:25:07,900 --> 00:25:13,990
So there's been about four different studies.

219
00:25:14,170 --> 00:25:27,820
So two of them have looked at circulating tumour DNA in the blood, and these were published 2015, one of them in 2019 another.

220
00:25:28,630 --> 00:25:40,420
And basically they take blood from patients with biliary strictures and they both use it used different but targeted panels genetic

221
00:25:40,750 --> 00:25:51,400
panels that use next generation sequencing to identify between 15 and 54 different cancer mutations or common cancer mutations.

222
00:25:52,300 --> 00:25:59,740
And often these studies paired that with analysis of mutations detected in tumour biopsies as well as the blood.

223
00:26:00,580 --> 00:26:08,470
And they've basically shown that the sequencing of cell free DNA in the blood is feasible, is accurate, and it's sensitive.

224
00:26:08,650 --> 00:26:16,240
So the concordance between mutations found in the blood and in the tumour itself was pretty good in those studies.

225
00:26:16,240 --> 00:26:27,399
They haven't actually tested or investigated their diagnostic accuracy in the clinical sense of diagnosing a malignant versus a binary stricture.

226
00:26:27,400 --> 00:26:38,770
These were all proven malignant strictures. And now a more recent paper that was published in Gut two years ago has sequenced use next generation

227
00:26:38,770 --> 00:26:50,409
sequencing of biliary bushings or biopsies in the diagnostic algorithm alongside usual imaging pathology.

228
00:26:50,410 --> 00:27:02,920
And see a 99. And this study basically showed that using a 28 gene targeted next generation sequencing panel in a lot of patients actually.

229
00:27:03,880 --> 00:27:14,890
So over 250 patients this improved malignant stricture detection when they used the sequencing versus looking at the biopsy or cytology alone.

230
00:27:15,990 --> 00:27:28,510
And that increased the sensitivity. To about 73% and versus pathology alone being about 48%, both with high specificity of nearly 100%.

231
00:27:29,740 --> 00:27:38,430
And the other thing it can do in a small majority of patients was identify and some of them for targeted therapy.

232
00:27:38,440 --> 00:27:44,910
Like I was saying before, there are certain new drugs that are approved for the treatment.

233
00:27:44,920 --> 00:27:49,720
If they can, the patient can be shown to have that mutation in the tumour.

234
00:27:50,260 --> 00:27:59,290
So these methods also allow that allow tailored treatment for specific patients looking for cell free tumour.

235
00:27:59,290 --> 00:28:06,340
DNA and bile has also been done but a long time ago and not using next generation sequencing methods.

236
00:28:07,180 --> 00:28:13,839
So we know that there is cell free tumour DNA in bile in patients with biliary strictures, malignant killer restrictions.

237
00:28:13,840 --> 00:28:20,170
And in the early 2000s, there were several papers that looked at cell free DNA,

238
00:28:20,710 --> 00:28:30,250
and they looked at two two common mutations or two common genes that are mutated in

239
00:28:30,550 --> 00:28:36,310
biliary cancer K and TP53 And because they didn't use next generation sequencing,

240
00:28:37,560 --> 00:28:44,590
you know, it took a long time and they only focussed on a few codons only and that all of

241
00:28:44,590 --> 00:28:49,470
those lacked sensitivity so didn't improve the diagnostic algorithm at all.

242
00:28:49,480 --> 00:28:54,220
And that was all kind of left and left a bit at that at that time.

243
00:28:54,640 --> 00:29:00,730
And have any of these previous methods progressed towards any kind of diagnostic test,

244
00:29:00,730 --> 00:29:06,280
or have they really stayed as a sort of a translational as a translational study?

245
00:29:06,970 --> 00:29:15,220
So none of them are used in any diagnostic algorithm that I know of in clinical practise.

246
00:29:15,640 --> 00:29:18,340
And that's really what this study set to address.

247
00:29:18,910 --> 00:29:29,110
So the panels that were available commercially and sort of the the targeted the targeted panels, a variety of different ones,

248
00:29:29,440 --> 00:29:42,370
and they're made by science, technology companies, you know, essentially that one of us could buy and use in a lab for research purposes only.

249
00:29:43,720 --> 00:29:49,990
So what the study set out to do was to evaluate two commercially available assays of

250
00:29:50,530 --> 00:29:55,470
of two different targeted next generation sequencing platforms to identify cancer.

251
00:29:55,510 --> 00:30:03,280
So genetic mutations in cell free DNA from bile samples in patients with biliary strictures undergoing LCP.

252
00:30:04,450 --> 00:30:11,440
So it's a pilot study to see whether this is feasible and also see whether

253
00:30:11,680 --> 00:30:17,470
clinically this improves diagnostic accuracy in the usual diagnostic algorithm.

254
00:30:18,610 --> 00:30:23,229
We know from that 2020 study that using these kind of things can improve the diagnostic

255
00:30:23,230 --> 00:30:27,370
accuracy when you've got cytology and brushing and you use this technology on them.

256
00:30:27,370 --> 00:30:38,020
But can this improve things even more if you just look at the cell free DNA in bile samples and which could potentially open it up to more patients?

257
00:30:38,380 --> 00:30:45,460
Because, of course, often you have an insufficient cytology or pathological sample for analysis.

258
00:30:46,600 --> 00:30:55,840
I thought it might be good just very briefly to talk about what next generation sequencing is and what it means as a little scientific aside.

259
00:30:57,160 --> 00:31:04,810
Do you think that would be helpful? Interesting. I think a couple of minutes on it is a is a good one because I'll be honest,

260
00:31:04,820 --> 00:31:15,520
it was it was a couple of years into my Ph.D. really only when I was about to run some NGS experiments that I actually bothered to learn how it works.

261
00:31:16,060 --> 00:31:21,459
Same. And actually the history. History is quite interesting, I think.

262
00:31:21,460 --> 00:31:33,910
So basically if you go back through the history of how we have learnt as a scientific community to sequence genes and then whole genomes,

263
00:31:35,410 --> 00:31:44,170
there was a big kind of leap forward as such in the seventies with the development of Sanger sequencing,

264
00:31:44,530 --> 00:31:47,200
which uses something called the chain termination method.

265
00:31:48,270 --> 00:31:59,170
And and basically what this does is it takes a primer which targets a specific start code, one of interest within a genome.

266
00:31:59,920 --> 00:32:03,190
So you have to know what you're looking for.

267
00:32:03,580 --> 00:32:16,630
You have to know where you want to start. That primer, you add in some DNA polymerase, you add in some new nucleic acids, and you also add in some.

268
00:32:17,470 --> 00:32:20,200
Actually some special nucleic acids, I guess.

269
00:32:20,470 --> 00:32:28,870
And so the terminal nucleic acids with a fluorescent detached and initially it was actually radial labels and then it uses PCR to amplify.

270
00:32:29,350 --> 00:32:34,780
So you start with start with a DNA sample, you add all this stuff,

271
00:32:35,440 --> 00:32:42,360
the primer and then and the DNA polymerase then leads to the addition of these nucleic acids.

272
00:32:42,700 --> 00:32:49,660
And then finally you'll get this terminal nucleic acid, which you can detect either by radio labour or later with a fluorescent dye.

273
00:32:50,630 --> 00:32:59,440
And because that labelled one is a terminal nucleic acid from this one primer,

274
00:32:59,440 --> 00:33:10,210
you will get gene reads of different lengths and or DNA strands of different lengths, and you can then work out if you order them all by length.

275
00:33:10,930 --> 00:33:18,720
And what's each nucleic acid at that terminal point is because the first dye or the radial

276
00:33:18,730 --> 00:33:25,000
label is different for the four different nucleic acids and nucleic acid possibilities.

277
00:33:25,300 --> 00:33:29,620
And then you can build up what you think the genetic code of that sequence is.

278
00:33:30,630 --> 00:33:35,520
The basic method is still used now in next generation sequencing.

279
00:33:36,570 --> 00:33:48,510
The main difference with next generation sequencing is you can do it much quicker or much more sample in a massively parallel way.

280
00:33:49,200 --> 00:33:57,540
So rather than you just targeting one specific start coding and just looking at, you know,

281
00:33:57,630 --> 00:34:03,870
one gene that you're interested in, you can sequence an entire genome pretty quickly.

282
00:34:05,400 --> 00:34:09,870
So in order to do that, the setup is a little bit different.

283
00:34:11,160 --> 00:34:14,300
But but essentially it's the same.

284
00:34:14,310 --> 00:34:16,470
It's a very, very similar process.

285
00:34:16,950 --> 00:34:26,010
But you're not restricted to a particular you can be restricted to a particular gene of interest, but you can just sequence everything.

286
00:34:26,280 --> 00:34:29,940
So that's a quick overview of next generation sequencing.

287
00:34:31,200 --> 00:34:40,260
And you know, and as as you said, like the scale and power of these technologies has just, you know, absolutely, absolutely.

288
00:34:40,680 --> 00:34:46,380
You know, lapse over the last 25, 30 years is transformed so much.

289
00:34:46,710 --> 00:34:48,870
And it has so many different applications.

290
00:34:49,380 --> 00:34:58,260
And and in the context that we're talking about now with these sort of gene panels, the known cancer mutations.

291
00:34:58,620 --> 00:35:07,439
This wouldn't have been possible really without next generation sequencing to bring to a clinical content context or to have basically

292
00:35:07,440 --> 00:35:16,290
a kit you can buy from a bioscience company that you can use on on patient samples in a lab because it would have taken too long.

293
00:35:16,500 --> 00:35:25,260
And so these panels, I'm sure you probably want to go into this, but all the panels looking for just detection of a particular gene in the vial,

294
00:35:25,260 --> 00:35:30,690
or are they detecting mutations in those genes within the body?

295
00:35:31,230 --> 00:35:34,710
So I think they are detecting specific mutations.

296
00:35:35,100 --> 00:35:43,140
Okay. They looked at two different commercially available ones, both produced by some official of the same company.

297
00:35:44,190 --> 00:35:50,880
Both were not designed specifically to look for biliary malignancy.

298
00:35:51,660 --> 00:36:00,540
They are just panels of genes. And so one of them is a 52 gene panel and one of them's 161 gene panel.

299
00:36:01,260 --> 00:36:06,180
One of the assays was called the Oncogene Pan Cancer Cell Free Assay.

300
00:36:06,570 --> 00:36:12,390
They called it in this paper, the bile nuit assay. This looks at 52 genes.

301
00:36:13,530 --> 00:36:21,600
And the panel was developed to detect sort of cell free DNA fragments and in blood.

302
00:36:21,900 --> 00:36:27,210
And the cell free DNA fragments that are found in blood in general do tend to be small.

303
00:36:27,540 --> 00:36:34,170
So it's a bit more sensitive. And they also used something called the oncoming comprehensive assay,

304
00:36:35,280 --> 00:36:45,059
which looks for more genes or gene mutations 161 But this is has been developed to detect larger

305
00:36:45,060 --> 00:36:51,900
DNA fragments from histological specimen specimens that are sort of paraffin embedded or,

306
00:36:52,980 --> 00:36:56,190
you know, in slides or frozen sections.

307
00:36:57,330 --> 00:36:59,190
So the initial part of this paper,

308
00:37:00,390 --> 00:37:08,129
they took bile and plasma ran these two different commercially available panels on them and found that the first one,

309
00:37:08,130 --> 00:37:15,930
so the bone meat assay detected a higher number of mutations than the other one in the bile and it was more sensitive.

310
00:37:15,930 --> 00:37:21,390
So that's the one they decided to test more extensively in their patient sample.

311
00:37:21,780 --> 00:37:32,549
What they then set out to do was really a pilot study of diagnostic accuracy of this ball metre assay and the the cohort

312
00:37:32,550 --> 00:37:40,620
they tested it in was prospective and in 68 patients with below restrictions that clinicians were suspicious for,

313
00:37:41,040 --> 00:37:45,330
you know, they could be malignant, they could be benign in a single centre in Spain.

314
00:37:45,750 --> 00:37:51,690
That was the inclusion criteria then they had to be undergoing in the LCP and they had to have an undiagnosed biliary stricture.

315
00:37:52,140 --> 00:37:56,520
There weren't any specific exclusion criteria, there wasn't a detailed consort diagram,

316
00:37:56,520 --> 00:38:00,240
and it wasn't how it but I think this was just because it was a pilot study

317
00:38:00,420 --> 00:38:05,340
trying to understand whether these methods were feasible in in this context.

318
00:38:06,450 --> 00:38:13,350
So the the reference standard they used was the usual care diagnostic pathway, which as we know,

319
00:38:13,680 --> 00:38:21,719
includes imaging LCP, if the patient is appropriate for that plus if possible, cytology.

320
00:38:21,720 --> 00:38:25,410
So from biliary brushing or histology from biopsies,

321
00:38:25,410 --> 00:38:32,370
if again that's technically possible and actually the reference standard really was using that

322
00:38:32,370 --> 00:38:38,070
diagnostic pathway and what their final diagnosis was at a year after their initial LCP.

323
00:38:38,610 --> 00:38:43,110
And we'll talk about why a year not immediately later.

324
00:38:44,400 --> 00:38:48,390
And the index tests they compared that to was this ball meet assay.

325
00:38:48,600 --> 00:38:59,249
So the cell free DNA mutation detection in file and and the usual diagnostic pathway.

326
00:38:59,250 --> 00:39:09,180
But the initial diagnosis so that diagnosis after their initial LCP imaging and cytology, the primary outcome wasn't specified,

327
00:39:09,180 --> 00:39:16,200
but they reported the sensitivity and specificity of each diagnostic test calculated for the initial fine diagnosis.

328
00:39:16,200 --> 00:39:19,950
So I presume that was what the main outcome was.

329
00:39:20,340 --> 00:39:29,790
Looking at the 68 patients with the biliary strictures all underwent an LCP and their initial diagnosis after the LCP.

330
00:39:29,960 --> 00:39:33,320
With cytology or histology.

331
00:39:33,410 --> 00:39:38,900
Histology 26 are benign, nine with an indeterminate and 33 were malignant.

332
00:39:39,680 --> 00:39:43,190
And of those 33 that were diagnosed as being malignant.

333
00:39:44,220 --> 00:39:47,630
And their final diagnosis at year was they were all malignant.

334
00:39:48,140 --> 00:39:53,420
So if you're labelled as having a malignant stricture and that initial diagnosis with the current diagnostic pathway,

335
00:39:53,600 --> 00:40:04,640
we know it's probably pretty much cancer. But the final diagnosis for the 26 benign strictures at a year.

336
00:40:05,570 --> 00:40:09,050
Actually 14 of them turned out to be malignant.

337
00:40:09,650 --> 00:40:19,640
And of the nine indeterminate strictures. Eight of them turned out to be malignant, just using a two year just using the normal diagnostic pathways.

338
00:40:20,120 --> 00:40:25,930
And what happened between the initial and the final diagnosis in these patients?

339
00:40:25,940 --> 00:40:37,730
A Completely ignoring the bomb me to say this is just using normal clinical practise is patients had more imaging, more A.L.S. pains,

340
00:40:38,120 --> 00:40:48,170
maybe more more histology, etc., etc. and clinically they progressed clinically or ideologically they progressed or died or whatever.

341
00:40:48,380 --> 00:41:05,750
And the diagnosis was made clear. So the crux of this is actually does the bottle meat assay improve the diagnostic accuracy of the initial diagnosis?

342
00:41:05,930 --> 00:41:16,700
So for the index test being usual care initial diagnosis that it turns out a year there were 13 benign strictures and 55 malignant strictures,

343
00:41:16,700 --> 00:41:20,630
of which 24 worse linggo and nine were pancreatic.

344
00:41:20,840 --> 00:41:27,350
And the sensitivity of that usual care initial diagnosis was 60% not great.

345
00:41:27,680 --> 00:41:30,380
The specificity, as we said before, was 100%.

346
00:41:32,330 --> 00:41:42,709
When they looked at the follow meat assay and they classified a positive result as any positive detected mutation on that value.

347
00:41:42,710 --> 00:41:51,260
To say the initial point of diagnosis, of the benign of the strictures that turned out to be actually benign at a year.

348
00:41:53,730 --> 00:42:01,740
Nine of them didn't have any mutations in the bomb to say, but four of them were false positives, so they had mutations.

349
00:42:02,100 --> 00:42:05,820
So they would have been classified as malignant by that allowed me to say.

350
00:42:06,210 --> 00:42:11,460
But in fact, at that year they did not have malignant stricture.

351
00:42:12,540 --> 00:42:15,780
But for the detection of malignant, malignant strictures,

352
00:42:16,380 --> 00:42:27,210
the bell meat assay identified 53 of 55 truly malignant strictures, and there were two false negatives.

353
00:42:28,740 --> 00:42:37,950
Now that working out the numbers, that means that the sensitivity using the ball meat assay increases to 96.4%,

354
00:42:38,880 --> 00:42:43,410
but the specificity does drop to about 70%.

355
00:42:44,370 --> 00:42:46,350
But this is not in combination.

356
00:42:47,100 --> 00:42:57,330
And the other thing to say for those four false positives, one of them actually on longer follow up turned out to have pancreatic cancer,

357
00:42:57,990 --> 00:43:05,010
which is interesting because it had essentially probably picked up a mutation that was the very,

358
00:43:05,010 --> 00:43:08,399
very early of a cancer that probably wasn't already there.

359
00:43:08,400 --> 00:43:12,030
But of course, that can't be present. Now, that's really interesting.

360
00:43:12,810 --> 00:43:18,600
So this assay most frequently identified KRAS and TP53 mutations.

361
00:43:19,200 --> 00:43:27,120
And it's these these mutations are not molecular therapeutic targets currently.

362
00:43:28,020 --> 00:43:37,140
And there are also mutations that occur in lots of different types of cancers, not just malignant biliary cancers.

363
00:43:38,250 --> 00:43:46,170
And it did identify Ida H one and FGFR two in a well in two in one patient.

364
00:43:47,640 --> 00:43:51,150
And these are molecular targets and this was a small sample.

365
00:43:51,690 --> 00:43:58,740
But for these particular patients that would that would change their management because it meant they were eligible for different drugs.

366
00:44:00,500 --> 00:44:06,500
And those targets they've detected in the bio, did they have match tissue and did they show that?

367
00:44:07,560 --> 00:44:10,950
The mutations in the vial correspond to the mutations in the tumour.

368
00:44:12,480 --> 00:44:20,760
Yeah. So this was in a supplementary figure and they did have tissue samples for all of the bile samples.

369
00:44:21,300 --> 00:44:27,000
In total, 68 mutations were identified and 50% of them were nearly 50%.

370
00:44:27,000 --> 00:44:33,420
And they were common to both the tissue and the bile. But 35% of them were found in the bottle and not the tissue.

371
00:44:33,720 --> 00:44:40,380
So it's not the concordance is not as high as, for example,

372
00:44:40,710 --> 00:44:56,130
it was in previous studies of these type of assays looking at cell free DNA in the blood versus a tumour sample, which is interesting, but I guess.

373
00:44:57,550 --> 00:45:10,210
The point of the paper is not to look at that so much because it does overall probably improve the sensitivity of diagnosis earlier,

374
00:45:10,300 --> 00:45:14,800
which might increase diagnostic accuracy in these patients.

375
00:45:15,160 --> 00:45:21,130
Just in terms of diagnostic accuracy, I mean, clearly, they've evaluated it compared to standard,

376
00:45:21,610 --> 00:45:28,030
which is it's great for a pilot study, but in reality, we're not going to ignore other things like,

377
00:45:28,030 --> 00:45:37,720
you know, radiological features or or the brushing those or other other aspects that push you towards,

378
00:45:38,230 --> 00:45:40,540
you know, labelling to me as a benign or malignant thing.

379
00:45:40,780 --> 00:45:48,250
Do they do any kind of exploratory analyses about how how it might be integrated into the is the care pathway?

380
00:45:49,060 --> 00:45:53,620
So there is a figure where they suggest how it could be integrated.

381
00:45:53,620 --> 00:46:01,749
But what they don't do specifically is an integrated analysis, you know, in adding it to usual care.

382
00:46:01,750 --> 00:46:05,409
But I mean, that's how they envision it being used, obviously,

383
00:46:05,410 --> 00:46:12,670
because you wouldn't want to drop that specificity that's already good in there from the usual diagnostic pathway.

384
00:46:13,000 --> 00:46:20,350
You just want to improve the sensitivity so it makes sense that you add it rather than use it instead of a plus.

385
00:46:20,350 --> 00:46:24,580
We'd still do the imaging, plus we'd still do the you have to do the ACP for this,

386
00:46:24,940 --> 00:46:28,930
but you'd still do the brushing or the spyglass biopsy or whatever it was.

387
00:46:29,440 --> 00:46:30,840
And I guess, I mean, what,

388
00:46:31,090 --> 00:46:40,390
what seems to be really attractive about this is whether it can deal with that half of patients where we either get a benign in inverted commas label,

389
00:46:41,020 --> 00:46:48,550
half of which will go on and develop malignancy at a year, or that indeterminate label, which is almost certainly cancer.

390
00:46:49,120 --> 00:46:54,040
But we probably don't have enough to necessarily go straight to surgery.

391
00:46:54,340 --> 00:47:02,760
So like, if you can if you can augment the diagnosis for those and either have like a,

392
00:47:02,890 --> 00:47:07,840
you know, a more rapid follow up test or repeat or whatever it is that's actually,

393
00:47:08,440 --> 00:47:15,100
you know, that's massive for those patients because that period of uncertainty and often that delay for surgery is,

394
00:47:15,340 --> 00:47:19,740
you know, is really clinically important. Yeah. Yeah, absolutely.

395
00:47:19,770 --> 00:47:23,570
So I think I think it's really promising and a really useful pilot.

396
00:47:23,580 --> 00:47:33,180
But obviously, as you said, it requires validation and together with the other tests,

397
00:47:34,980 --> 00:47:41,870
but also in prospective studies with larger patient numbers, the needs, the population they look.

398
00:47:41,880 --> 00:47:47,460
I mean, this wasn't a controlled study and they hadn't powered it for a primary outcome.

399
00:47:47,880 --> 00:47:53,610
And they assessed sensitivity and specificity, but they didn't look at, you know,

400
00:47:55,530 --> 00:48:03,180
the positive or negative predictive values or number, the numbers needed to treat and numbers needed to test in this case.

401
00:48:03,960 --> 00:48:08,430
And and in terms of the generalisability of the population they looked at.

402
00:48:08,610 --> 00:48:12,990
So it was a single centre in Spain. Most of these patients had a distal stricture.

403
00:48:13,230 --> 00:48:19,110
Okay. And, and this actually matters basically because the distal strictures,

404
00:48:20,400 --> 00:48:31,059
especially if this have a broadly different mutational profile to intrahepatic ones and also the treatment options are often slightly different.

405
00:48:31,060 --> 00:48:36,510
Like it's easier to operate on a on a distal stricture than is on an intra partic stricture.

406
00:48:36,870 --> 00:48:41,699
Plus, they mentioned that there wasn't a high proportion of patients in the cohort.

407
00:48:41,700 --> 00:48:46,379
For example, I couldn't tell from the table that they provided.

408
00:48:46,380 --> 00:48:51,240
They didn't say who had PSC and who didn't, but that also changes the mutational profile.

409
00:48:51,240 --> 00:48:57,660
So there's there's a few different things in that sense that could change the accuracy of this.

410
00:48:58,350 --> 00:49:06,120
And then the there were obviously some practical things to address, like is the technology scalable?

411
00:49:06,150 --> 00:49:12,930
Like ultimately at the moment these tests are validated in a lab only, not in a clinical setting.

412
00:49:14,760 --> 00:49:21,720
And we would want it to be performed in a hospital lab in a sort of automated way.

413
00:49:24,030 --> 00:49:33,480
Would this actually impact on treatment? Because, yes, it is been shown in a minority of patients that it picks up certain mutations.

414
00:49:34,050 --> 00:49:45,780
But um, you know, these cancers because it requires a C LCP and because this LCP is being prompted by clinical stuff plus imaging,

415
00:49:45,780 --> 00:49:48,749
you know, they're not, these patients aren't just rocking up for surveillance.

416
00:49:48,750 --> 00:49:56,250
The OCP will actually diagnose it that much earlier enough to make a difference.

417
00:49:56,250 --> 00:49:59,280
That's the key. Yeah. Will it diagnose it early enough?

418
00:49:59,880 --> 00:50:03,090
Early enough that they could have surgery. But that's the key.

419
00:50:03,090 --> 00:50:10,260
And I and I don't know the answer to that. And you do have to get you do still have to have the LCP.

420
00:50:10,680 --> 00:50:16,290
And some patients cannot have any LCP for various reasons who probably wouldn't be

421
00:50:16,440 --> 00:50:21,929
suitable for the surgery anyway if they're not suitable for any LCP for whatever reason.

422
00:50:21,930 --> 00:50:27,149
But nevertheless, it doesn't help those that particular group of patients.

423
00:50:27,150 --> 00:50:33,570
But certainly it's, it's promising. Yeah, it's, it's, it's very much so and interesting.

424
00:50:33,570 --> 00:50:38,760
So, yeah. Has anyone just tried sampling like small bowel contents?

425
00:50:39,240 --> 00:50:41,549
Because, you know, my, my anatomy is not great,

426
00:50:41,550 --> 00:50:47,910
but even I am aware that the bile goes into the small bowel and it is a relatively straightforward plumbing.

427
00:50:48,910 --> 00:50:56,040
But that's you know, you can you could do an OCD, you know, and, you know,

428
00:50:56,070 --> 00:51:04,469
weighing down a tube and aspirate aspirate some small bowel contents as long as your distal to the distal to the,

429
00:51:04,470 --> 00:51:08,490
um, puller and there'll be a bit of bit of bile around.

430
00:51:09,300 --> 00:51:17,250
Yeah, yeah, that's possible I assume. But it's a total assumption that you can't detect things in stool.

431
00:51:17,280 --> 00:51:26,910
Otherwise people would have looked just because presumably the DNA is degraded by the time you've choked through the rest of the GI tract.

432
00:51:27,720 --> 00:51:33,090
I guess the other thing about doing both of these techniques is that.

433
00:51:34,110 --> 00:51:45,280
And. Well, it's the same with plasma, isn't it? You will pick up all sorts of mutations that might be due to a different tumour somewhere else.

434
00:51:45,390 --> 00:51:54,570
Any kind, colon, stomach, whatever. You need some A.I. to pick it up on your colonoscopy first before you do the stool test.

435
00:51:56,700 --> 00:52:06,929
Cool. That's it's really interesting. And it's sometimes sometimes you see these sequencing papers and and you go, that is absolutely lovely.

436
00:52:06,930 --> 00:52:12,630
And that super interesting. That is of no use to man or beast for for a while.

437
00:52:12,900 --> 00:52:21,210
But actually but I like this because it's it's tried it's trying to make it relevant to clinical practise.

438
00:52:21,510 --> 00:52:31,830
And and I think there's enough evidence from this to go on to potentially do bigger trials or develop of the assay itself.

439
00:52:32,010 --> 00:52:37,530
Yeah. There's maybe a bit more targeted for biliary tree malignancy.

440
00:52:37,830 --> 00:52:45,940
Yeah. So that someone. So we've done colonoscopy.

441
00:52:46,180 --> 00:52:49,870
We've now done LCP, albeit for stretches.

442
00:52:50,500 --> 00:52:53,800
So I've got I've got two more endoscopy papers. Okay, cool.

443
00:52:54,030 --> 00:52:59,890
Okay. So I've got. We're now getting on to uppers now offers for GI bleeding.

444
00:53:00,160 --> 00:53:05,650
So this was a paper in guts that I think is really interesting and I think it's quite cool technology.

445
00:53:05,860 --> 00:53:09,910
I have to say I'm slightly sceptical about this study, I think.

446
00:53:11,260 --> 00:53:18,520
And if any of the authors are amongst our, you know, our small coterie of listlessness, I'd love to I'd love to hear from you.

447
00:53:18,930 --> 00:53:24,490
And this is a trial that was in got just a few weeks ago.

448
00:53:25,210 --> 00:53:29,680
This first author is Benjamin Meyer. This is a German group,

449
00:53:29,980 --> 00:53:38,800
and the title of the paper is Over the scope clips versus standard treatment for high risk patients with acute non-venomous heel upper GI bleeding.

450
00:53:39,010 --> 00:53:42,700
A randomised controlled trial. The Sting two trial.

451
00:53:42,880 --> 00:53:50,070
It's good name and it's good name. Non-Racial upper GI bleeding is really common.

452
00:53:50,110 --> 00:53:55,690
So it's around 100 cases per 100,000 population in Europe, in the US,

453
00:53:55,960 --> 00:54:03,340
and it's still associated with a pretty high mortality somewhere in the region of 7 to 10%, depending on what studies you look at.

454
00:54:04,180 --> 00:54:07,030
Most of that is actually not driven by the bleeding itself,

455
00:54:07,030 --> 00:54:13,720
but the fact that the people who bleed are generally old, frail, co-morbid and on anticoagulants.

456
00:54:14,050 --> 00:54:24,940
And now the the major treatment for for vast upper GI bleeding is endoscopic therapies,

457
00:54:24,940 --> 00:54:36,729
a diagnostic endoscopies diagnostic endoscopy and then endoscopic therapy for haemostasia and standard therapy for that is based

458
00:54:36,730 --> 00:54:45,910
on based on previous trials pretty successful with rates of haemostasia considerably greater than 90% in previous trials.

459
00:54:46,150 --> 00:54:48,730
And usually, as you know, we use dual therapy,

460
00:54:48,880 --> 00:55:00,790
so we inject in diluted adrenaline and then we either we the use mechanical haemostasia we clip it or we use coagulation,

461
00:55:00,790 --> 00:55:04,809
thermal therapy, we burn it, say, inject it, clip it.

462
00:55:04,810 --> 00:55:09,100
And burnet butts and.

463
00:55:10,250 --> 00:55:19,729
We know that if you have insufficient haemostasia or ongoing bleeding, then a repeat endoscopy is often unlikely to be successful.

464
00:55:19,730 --> 00:55:26,570
And then you need to have various types of salvage therapy so you can either have an angiographic therapy,

465
00:55:26,570 --> 00:55:32,360
so interventional radiology treatments or in some cases surgical treatments.

466
00:55:32,690 --> 00:55:38,630
And I think the key thing is, is that those those salvage therapies are associated with a greater mortality,

467
00:55:38,780 --> 00:55:46,910
whether that's because of the procedure itself, because of the cohort of patients it is, or because the ongoing bleeding is is not entirely clear.

468
00:55:47,090 --> 00:55:50,810
But the mortality is, you know, is higher sort of 10 to 30%.

469
00:55:51,890 --> 00:55:58,730
Now, there was a sort of a new clip on the block, which is the over the scope clip.

470
00:55:58,790 --> 00:56:04,190
And I have to say, I've only actually ever seen one deployed once, so I've never seen them.

471
00:56:04,430 --> 00:56:14,210
They kind of look like a bear trap, one of those kind of comedy bear traps that you see in a cartoon with two big kind of jaws with claws on them,

472
00:56:14,480 --> 00:56:17,540
that snapshots, that sort of that kind of sprung load.

473
00:56:17,550 --> 00:56:26,510
And they snapshots. And what you do is you put put effectively a cap which has the over scope clip loaded in the open position on the top.

474
00:56:26,690 --> 00:56:29,690
A little bit like a speaker band.

475
00:56:29,960 --> 00:56:32,330
Yeah, exactly. Yeah. And very much like the band. And there's a little,

476
00:56:32,510 --> 00:56:39,680
little thread or something that goes up the working channel of the scope and effectively you find whatever

477
00:56:39,680 --> 00:56:48,110
you want to what want to bear a bear trap and suck it up into your cap and then deploy this massive clamp.

478
00:56:48,230 --> 00:56:55,700
The thing that that grabs grabs hold of it. So over over the scope clips were developed by a company called Invesco, which are a German company,

479
00:56:56,600 --> 00:57:03,920
and they were originally developed for the closure of G.I fistulas and GI perforation.

480
00:57:03,930 --> 00:57:10,280
So the idea is you made a whacking great hole in the bowel and then effectively you just close it trying to reduce the,

481
00:57:10,770 --> 00:57:15,350
the, the, the need for, for surgery for those kind of situations.

482
00:57:15,890 --> 00:57:27,140
But they have been increasingly used in case series in clinical trials for haemostasia in certain situations because their,

483
00:57:27,640 --> 00:57:35,510
their physical kind of reach, their bite is, is bigger than most standard clips.

484
00:57:35,720 --> 00:57:44,660
And also the sheer kind of strength of them and the fact that they've got these ruddy great teeth that really kind of dig in means that,

485
00:57:44,810 --> 00:57:47,600
you know, in a kind of more sort of fibrotic setting,

486
00:57:47,810 --> 00:57:58,520
you can kind of oppose the edges of an ulcer and gets that kind of that mechanical pressure on the vessel to allow it to coagulate and stop bleeding.

487
00:57:59,870 --> 00:58:06,200
So there have been some retrospective studies that have shown that they are in case controlled series

488
00:58:07,760 --> 00:58:15,170
highly effective for the management of severe arpeggio bleeding in certain patient populations.

489
00:58:16,670 --> 00:58:25,280
However, it's unclear whether of scope these over the scope clips are a useful treatment as first

490
00:58:25,280 --> 00:58:31,160
line for high risk non partial API bleeds and that's what this trial is trying to do.

491
00:58:32,100 --> 00:58:38,100
Okay. So. And is this still used with dual therapy?

492
00:58:38,110 --> 00:58:42,100
So if you use about it. Oh, not necessarily.

493
00:58:42,790 --> 00:58:47,580
Okay. But I think I think most of these patients.

494
00:58:47,590 --> 00:58:50,650
Yeah. So they injected adrenaline for most of these patients.

495
00:58:51,280 --> 00:58:58,030
Okay. And that's all. And so this trial, the sting two trial, that was a sting one.

496
00:58:58,870 --> 00:59:03,640
I have some. I forgot what Sting one was. Presumably the same thing, but in a different cohort.

497
00:59:04,180 --> 00:59:09,159
So Sting two was in 13 academic referral centres in Germany.

498
00:59:09,160 --> 00:59:13,360
So this is all in Germany. And each centre there was only one or two endoscopies.

499
00:59:13,370 --> 00:59:15,390
So this is a little bit different from the previous trial.

500
00:59:15,400 --> 00:59:20,350
So we've got effectively a whole clinical trial and you can probably get them around a small table,

501
00:59:20,860 --> 00:59:28,540
all of the documents who were doing these procedures and they were all highly experienced in the management of the GI bleeding.

502
00:59:28,540 --> 00:59:29,560
And there's an interesting line.

503
00:59:29,830 --> 00:59:37,810
They are all highly experienced in the application of over the scope clips greater than 20 over the scope clips per year.

504
00:59:38,170 --> 00:59:41,580
So that means they use them normally, they are using them normally.

505
00:59:41,590 --> 00:59:43,840
And I would say that they are converts force.

506
00:59:43,840 --> 00:59:50,980
And it is interesting to look at the competing interests statement because although this was not funded by the company,

507
00:59:51,700 --> 00:59:58,929
several authors are either consultants for of ASCO have received fees for INVESCO anything like that

508
00:59:58,930 --> 01:00:06,400
to say that these these people are converts to the technology which comes into relevance later.

509
01:00:06,790 --> 01:00:12,040
So these are all participants who were hospitalised for acute non-volatile upper GI bleeding.

510
01:00:12,340 --> 01:00:13,510
They had the endoscopy,

511
01:00:13,510 --> 01:00:22,270
they were consented for the trial before they had the endoscopy and then they were included in the trial and basically during the endoscopy.

512
01:00:22,390 --> 01:00:27,760
So what they did is they did the endoscopy and then they did the complete rock score.

513
01:00:27,940 --> 01:00:31,240
So including the endoscopic elements of the recalled score.

514
01:00:31,540 --> 01:00:35,860
And if they had a recall score, a complete recall score of seven or greater,

515
01:00:36,280 --> 01:00:46,240
then they met the criteria and then the little envelope was opened and they were randomised to one of two arms standard therapy or a vesco.

516
01:00:47,290 --> 01:00:53,169
Okay. Yeah. So that was the plan. I mean, you wouldn't do that in real life, would you?

517
01:00:53,170 --> 01:00:59,990
But. Okay, well, you know, a little envelope halfway through, and then I might do that.

518
01:01:00,220 --> 01:01:03,730
Just got some letters. I just need to see what's in my correspondence.

519
01:01:03,970 --> 01:01:07,450
Please help escape that. Yeah. No, no, no.

520
01:01:07,450 --> 01:01:11,230
Calculate the ruffle. But no, you know, but it's.

521
01:01:11,500 --> 01:01:16,450
It's a trial note and they need to do that. But effectively, they, you know, they're assessing the forest score.

522
01:01:16,480 --> 01:01:22,380
These people have got a forest one or one or two way.

523
01:01:22,390 --> 01:01:33,370
So effectively they've either got her actively bleeding, actively bleeding ulcer, or they have got a visible vessel in the presence of clot.

524
01:01:34,060 --> 01:01:37,900
Okay. So these are sort of high, high risk bleeds that we need to intervene in.

525
01:01:38,110 --> 01:01:43,480
And they've got a circle of seven or more. There are a couple of things that are worth noting.

526
01:01:43,510 --> 01:01:48,190
They were not allowed to use a couple of other potential treatments.

527
01:01:48,400 --> 01:01:51,549
So in particular, they were not allowed to to use chemo.

528
01:01:51,550 --> 01:01:55,390
Pray he might spray topical therapies that we've discussed.

529
01:01:55,600 --> 01:01:59,469
You can't call it that anymore. We can't call it because it was a great trial, wasn't it?

530
01:01:59,470 --> 01:02:03,490
Because it was great. Yeah. But again, it comes up into, into relevance here.

531
01:02:03,730 --> 01:02:08,680
Um, so they weren't allowed to use topical therapies like chemo spray.

532
01:02:09,100 --> 01:02:20,020
Um, and they said that for ethical reasons people were allowed to crossover between the arms if they, if haemostasia was not achieved.

533
01:02:20,440 --> 01:02:24,769
And I'm not sure I desperately agree with that. But that's interesting.

534
01:02:24,770 --> 01:02:33,729
That's what the trial says. You know, I presume they mean crossover to having a basically if the rest hasn't worked.

535
01:02:33,730 --> 01:02:37,840
Yeah, yeah, yeah. You're seeing where this is going.

536
01:02:37,840 --> 01:02:42,040
Tamsin Cool. So there were some, there were some,

537
01:02:42,520 --> 01:02:48,280
some power calculations in which they've ended up with a made up made up number of people that they're going to put in that trial.

538
01:02:48,580 --> 01:02:51,850
So assessed 246 patients.

539
01:02:52,240 --> 01:02:56,830
Um, and 100 of those were randomised.

540
01:02:56,830 --> 01:03:03,580
So most of them didn't meet the criteria for it because either they couldn't,

541
01:03:03,580 --> 01:03:11,260
they couldn't get consent or when they did the procedure they didn't have severe enough bleeding and didn't meet the criteria.

542
01:03:11,380 --> 01:03:15,970
So that's fine. So they randomise them roughly once,

543
01:03:15,970 --> 01:03:28,210
once a standard group 52 and over the scope clip 48 and then they and all of the ones who are randomised are included in the analysis which is good.

544
01:03:29,560 --> 01:03:34,720
The primary end point of this study was clinical success.

545
01:03:35,080 --> 01:03:42,520
So clinical haemostasia from the as decided by the managing endoscopic.

546
01:03:43,000 --> 01:03:44,350
So endoscopy opens,

547
01:03:44,350 --> 01:03:52,990
the envelope goes either standard therapy or back in and then they apply the therapy and then they decide whether that was clinical success or not.

548
01:03:53,980 --> 01:04:12,490
So and and the outcome was that in the standard therapy arm, 38 of the patients had clinical success, which is only 73%, only 73%, which is low.

549
01:04:12,820 --> 01:04:18,220
That is a low number compared to previous studies of standard dual therapy.

550
01:04:18,730 --> 01:04:28,630
And in the over the scope, it was 44 patients with a clinical success rate of 91.7% in 92%.

551
01:04:30,190 --> 01:04:34,840
So it sounds like quite a big difference, 73% compared to 92% big difference in that.

552
01:04:34,870 --> 01:04:44,500
But first, the 73% is quite low. And second, actually, that only corresponds to six patients, six patients difference between the two arms.

553
01:04:44,920 --> 01:04:51,030
And in those six patients, the endoscopies, having applied a couple of clips of that,

554
01:04:51,040 --> 01:05:01,000
that standard therapy has decided that the therapy is not working and has switched to the Skype, which in all six of them.

555
01:05:01,970 --> 01:05:05,920
I provided him instant haemostasia. So.

556
01:05:09,060 --> 01:05:17,580
I'm trying to put this in a very diplomatic way. There is an element of subjectivity by the endoscope test to decide that they've done enough standard

557
01:05:17,580 --> 01:05:24,270
therapy and that there is now not haemostasia and then they've switched over to the other other therapy.

558
01:05:25,710 --> 01:05:31,230
That is the key. Significant results in this in terms of re bleeding rates.

559
01:05:32,810 --> 01:05:45,620
There was a there was a nonsignificant trend for increased recurrent bleeding within the first seven days in the standard of therapy group.

560
01:05:45,620 --> 01:05:51,919
So eight patients replied compared to four patients, so 15% compared to eight patients 8%.

561
01:05:51,920 --> 01:06:01,100
But that was not significant. However, there were two delayed bleeds in the over the scope clip on.

562
01:06:02,290 --> 01:06:08,320
And it gets slightly convoluted about re bleeding and different interventions.

563
01:06:08,590 --> 01:06:19,720
But I think that the only person who actually ended up going for surgery for ongoing bleeding was in the over the Skype clip creep in terms of like,

564
01:06:20,110 --> 01:06:25,210
you know, big headline, you know, the things that people really care about like dying.

565
01:06:25,600 --> 01:06:30,370
And there was no difference in overall 30 day mortality between the two arms.

566
01:06:30,370 --> 01:06:38,170
It was around 7%. So that's to me, half saying they picked the wrong.

567
01:06:39,870 --> 01:06:44,020
Have they picked the wrong primary outcome? Um.

568
01:06:44,870 --> 01:06:48,200
I don't know. It is a plexi. I'm thinking.

569
01:06:48,260 --> 01:06:52,370
I'm just thinking out loud. Yeah. No, I think you're.

570
01:06:52,370 --> 01:07:02,510
I think you're spot on. So I agree with you that I think that this is a proxy and point that makes a certain amount of sense.

571
01:07:03,050 --> 01:07:07,010
But one isn't actually what patients care about.

572
01:07:07,010 --> 01:07:10,069
They care about whether they keep on bleeding over the following days.

573
01:07:10,070 --> 01:07:14,930
And then that leads to them either requiring something else or them having some kind of bad outcome.

574
01:07:15,380 --> 01:07:20,570
But also it is an outcome that is a subjective measure, very subjective.

575
01:07:20,570 --> 01:07:30,290
And it's not like that was a sort of essential reader or somebody who who decided and made an independent person who knew decided that.

576
01:07:30,470 --> 01:07:35,240
So there's a bit of judgement there and these people are proponents of this treatment.

577
01:07:35,240 --> 01:07:39,650
Otherwise they wouldn't be doing this study and they wouldn't be doing to over the scope, perhaps a year each.

578
01:07:39,980 --> 01:07:48,200
So these are people who think that it works well and therefore are probably likely to lean towards it in certain circumstances.

579
01:07:48,980 --> 01:07:59,150
I think there are a couple of other things here. Um, I am concerned that the clinical success rate in the standard arm was only 73% because first,

580
01:07:59,150 --> 01:08:03,950
that doesn't tally with other trials of chil therapy, which suggest it's well over 90%.

581
01:08:04,220 --> 01:08:10,400
What's interesting is that, um, they pretty much only used clips.

582
01:08:10,880 --> 01:08:17,870
So this pretty much, I think there was one patient who ended up getting a little bit of Bernie with the, with the heater probe.

583
01:08:19,350 --> 01:08:24,860
Pretty much everybody had had clips rather than adrenaline.

584
01:08:25,170 --> 01:08:28,950
And that includes what's a single therapy? Not dual therapy.

585
01:08:28,980 --> 01:08:32,100
No. Sorry. No. So they had adrenaline plus clips.

586
01:08:32,460 --> 01:08:35,820
Yeah, but but very few people had thermal therapy.

587
01:08:36,240 --> 01:08:44,970
And that interested me because there were 16 or 17 ulcers that were over 20 millimetres wide.

588
01:08:45,980 --> 01:08:58,010
And my experience is that big old 20 or 30 millimetre ulcers are nigh on impossible to close with standard standard clips.

589
01:08:58,220 --> 01:09:03,770
You either need to get those giant ultra clips or you need to or you need to burn.

590
01:09:04,170 --> 01:09:11,180
So I wonder about those those those cases.

591
01:09:11,540 --> 01:09:14,870
I think the other one is is that they weren't allowed any haemostatic powders.

592
01:09:15,950 --> 01:09:23,210
And actually, as a if we're thinking about somebody who may have some ongoing issues after you've applied partial standard therapy,

593
01:09:23,510 --> 01:09:32,990
actually, you might well convert to a Hema spray and and add that on and get Haemostasia.

594
01:09:33,830 --> 01:09:39,530
Yeah. So and I guess that would be part of standard of care at the moment,

595
01:09:40,160 --> 01:09:47,720
although that's not sort of guideline based, but it would be an alternative that people would use.

596
01:09:48,020 --> 01:10:00,460
So are there any adverse ALK adverse outcomes or reasons not to use a backflip in in those cases rather than he was very apart from lack of evidence.

597
01:10:00,500 --> 01:10:05,630
So not in this trial. So I think I think it's fair to say that these are actually pretty safe things.

598
01:10:05,840 --> 01:10:13,250
However, I have seen some case reports and I've also seen some presentations where people have talked about complications from them,

599
01:10:14,330 --> 01:10:21,799
most obviously perforation. So you've got like a thinned duodenal ulcer and you put on this whacking great thing

600
01:10:21,800 --> 01:10:27,770
over your your fibrotic ulcer and you just just go through and you've perforated them.

601
01:10:28,440 --> 01:10:32,810
And so there are potential complications about it.

602
01:10:33,980 --> 01:10:41,530
I think it's a really interesting trial. And I think they've I think they've tried to answer a question about whether this should be standard of care,

603
01:10:41,540 --> 01:10:45,230
whether it is an alternate, a suitable alternative is standard of care.

604
01:10:46,280 --> 01:10:49,820
And I think I think you could argue that it is an alternative to standard of care.

605
01:10:49,910 --> 01:10:52,070
Like in this trial, it seems to work.

606
01:10:53,030 --> 01:11:04,249
I am not convinced, however, with the statement that it is superior to standard of care because of the caveat, all the caveats raised and from people.

607
01:11:04,250 --> 01:11:08,450
I don't know many people who've used it, but from hearing presentations where people.

608
01:11:09,020 --> 01:11:18,080
And speaking to a couple of people who have used them mean there are certain situations where they could be really useful for bleeding.

609
01:11:18,260 --> 01:11:24,170
So particularly if you've got a large ulcer that you can't just physically get any kind of clip around,

610
01:11:24,890 --> 01:11:27,379
and particularly if it's got a bigger vessel that you're not,

611
01:11:27,380 --> 01:11:34,010
you're a bit worried about, you know, touching with the heater probe and actually getting a big old clip over.

612
01:11:34,010 --> 01:11:42,170
It sounds a lot of sense. And also, if you re scoping someone and standard therapy has failed, particularly for one of those bigger ulcers,

613
01:11:42,560 --> 01:11:48,500
then that's maybe a nice alternative to prevent people go to angiography and then,

614
01:11:48,530 --> 01:11:52,670
you know, maybe in very safe areas in terms of putting on a whacking great clip.

615
01:11:52,880 --> 01:11:55,520
So like a stomach, a stomach bleed,

616
01:11:55,730 --> 01:12:02,330
particularly if you can't accurately localise exactly where it's coming from like a OD due to for a lesion or something like that.

617
01:12:02,330 --> 01:12:05,510
Maybe just grabbing it with one of these massive things might work well.

618
01:12:06,140 --> 01:12:13,900
So I think this is an interesting trial and I don't think it's answered my question, which is where is Invesco?

619
01:12:13,910 --> 01:12:21,680
Where do these oversight clips fit into to the therapeutic sort of armamentarium?

620
01:12:22,460 --> 01:12:28,400
I think you could argue that they're an alternative, and but I think it's probably a bit more complex than that.

621
01:12:28,670 --> 01:12:36,590
And I'm not sure in this trial they'd be any better than standard therapy plus hemo spray if initial statement doesn't work.

622
01:12:36,980 --> 01:12:41,180
Yeah. Thank you. That was another really interesting trial. So thank you for bringing that one as well.

623
01:12:41,240 --> 01:12:44,270
Brilliant. Let's do one more endoscopy. One, because it's quite quick.

624
01:12:44,280 --> 01:12:48,230
Yeah. Okay. Focus on it. Okay. And now for something completely different.

625
01:12:49,550 --> 01:12:55,550
This is also in goats. Oh. So Billy goat focussed on three in a row.

626
01:12:56,750 --> 01:12:58,309
So this is a bit off piste.

627
01:12:58,310 --> 01:13:05,870
Estimating the environmental impact of disposable endoscopic equipment and endoscopic is good that we're thinking about this now.

628
01:13:06,140 --> 01:13:17,240
Really good. I do. And there's a there's a awesome movement in that in the UK and there's a gastroenterologist in at King's he's

629
01:13:17,570 --> 01:13:24,770
Hussain Bely who has been involved in setting up this sort of green endoscopy group as part of the BSG,

630
01:13:24,920 --> 01:13:31,370
which I think is brilliant, to start to think about how we can make positive environmental changes in health.

631
01:13:31,400 --> 01:13:34,010
Yeah, I think it's because health care is, according to this paper,

632
01:13:34,010 --> 01:13:41,030
at least responsible for 4.4% of total greenhouse gas emissions and 8% within the EU.

633
01:13:41,450 --> 01:13:44,660
The US said we all we have terrible.

634
01:13:45,080 --> 01:13:51,670
And call that a disposable stuff. Things all the plastic, the heat is a lot of rubbish that we do.

635
01:13:51,700 --> 01:13:56,450
Yeah, there's a lot of rubbish we generate and in the USA we also probably do far too many procedures.

636
01:13:56,450 --> 01:14:06,200
In the US, 18 million endoscopies performed each year and a lot of the kit we use is single use and disposable.

637
01:14:06,560 --> 01:14:12,410
Yeah. Now you may well remember that there were a of people who got really horrible

638
01:14:13,310 --> 01:14:18,800
cholangitis with resistant organisms from contaminated duodenoscopes a few years ago.

639
01:14:20,420 --> 01:14:22,610
I don't. Thanks anyway. Yes.

640
01:14:23,090 --> 01:14:31,340
So there was this run of like people who got horrible, like carbapenems producing, you know, enterococcus, horrible, horrible bugs.

641
01:14:31,610 --> 01:14:37,820
And they were being transmitted because of inadequate cleaning of the bridge of the Duodenoscope.

642
01:14:38,450 --> 01:14:43,940
And because of that and concerns about infection, transmission and things like CJD and things like that.

643
01:14:44,300 --> 01:14:53,300
And uh, companies have started manufacturing single use endoscopes and the FDA has recently approved old endoscopes for,

644
01:14:54,770 --> 01:14:58,060
for gastrostomy, colonoscopy, duodenoscope.

645
01:14:58,820 --> 01:15:04,670
I've seen adverts for sister scopes. This is a bit of a movement.

646
01:15:04,670 --> 01:15:10,220
And from the joke and from the story, from the manufacturers, it's a win win moneywise.

647
01:15:10,520 --> 01:15:20,990
So it's a good line of business to encourage so and so despite the absolute risks of of endoscope

648
01:15:21,590 --> 01:15:29,180
acquired infection being like tiny in the one per million this there is a bit of a move towards.

649
01:15:29,180 --> 01:15:32,750
I mean I do feel the NHS will never be able to afford that.

650
01:15:32,760 --> 01:15:37,219
So from all you say that the nice thing about it is you buy a whole lot of

651
01:15:37,220 --> 01:15:41,240
single use stuff and then what you do is you sack all your staff who are clean.

652
01:15:41,900 --> 01:15:49,910
Yeah, yeah. So, you know, it's for some departments there is a financial thing you can ask,

653
01:15:50,000 --> 01:15:56,899
argue about the the ethics about this and about the shortsightedness about sacking staff.

654
01:15:56,900 --> 01:16:00,230
But that's often how I think it's sold to departments that, hey,

655
01:16:00,410 --> 01:16:07,820
we've all single use and then you don't have to have any of your your processing deacons staff.

656
01:16:08,030 --> 01:16:12,170
You can just get rid of that unit because, hey, you're just chucking it away afterwards.

657
01:16:13,670 --> 01:16:22,400
So what this study did actually, this was just a really nice basically an audit and they published it in gut.

658
01:16:22,490 --> 01:16:24,650
So well done then. I love this.

659
01:16:24,890 --> 01:16:35,090
And basically they audited two centres, a high volume endoscopy centre and a low volume endoscopy centre for a week in 2020, just before the pandemic.

660
01:16:35,330 --> 01:16:46,479
And they then used their audit data and measuring all of the rubbish that was generated from those from the procedure.

661
01:16:46,480 --> 01:16:50,870
Then they study hospitals to work out how much trash does endoscopy make.

662
01:16:52,610 --> 01:17:00,440
And so there's a whole lot of methods about how they calculated the the amount of rubbish that's all generated.

663
01:17:00,740 --> 01:17:10,610
And so they've they've calculated this the base of 278 endoscopies performed in this five day period at these two hospitals.

664
01:17:11,240 --> 01:17:23,390
And those 278 and the scoping procedures filled 190 20 gallon waste bins and generated 619 kilos of rubbish.

665
01:17:23,480 --> 01:17:27,350
Of which two thirds of which was destined for the landfill.

666
01:17:27,800 --> 01:17:35,150
Another 17% of it was biohazard waste that was headed for incineration.

667
01:17:35,950 --> 01:17:41,040
Um. And it's not good, is it?

668
01:17:41,160 --> 01:17:45,600
The vast majority of it was not. Yeah, I was going to say nothing was recycled.

669
01:17:45,660 --> 01:17:52,950
Yeah, that sounds so. The average scope produced 2.1 kilos of waste and filled half a waste bin.

670
01:17:54,480 --> 01:17:58,380
And these are standard. These are standard scopes. No, no serving single use scopes.

671
01:17:59,850 --> 01:18:09,270
So basically we make a lot of rubbish. So they then calculated based on the 18 million endoscopic procedures performed in the US each year,

672
01:18:10,200 --> 01:18:21,780
that the US fills 11 million waste bins and generates approximately 38,000 metric tonnes of rubbish each year,

673
01:18:22,050 --> 01:18:34,710
which would cover 117 soccer fields in a metre of rubbish and is equivalent to the weight of around 25,000 cars.

674
01:18:36,240 --> 01:18:40,110
So that's always depressing we generate.

675
01:18:40,260 --> 01:18:45,210
So we already, we already generate an inordinate amount of rubbish from doing it.

676
01:18:45,540 --> 01:18:52,320
Yeah. They've then estimated what would happen if we switched to single use endoscopes.

677
01:18:52,950 --> 01:18:57,450
And if you did it for. All single use and escape.

678
01:18:57,450 --> 01:19:04,740
So Cullen escapes, gastro escapes and duodenoscopes you increase the amount of rubbish by 40%.

679
01:19:05,310 --> 01:19:09,600
If you just did it for kaleidoscopes and duodenoscopes you would increase by 25%.

680
01:19:09,870 --> 01:19:12,929
So we're going to make a whole load more rubbish.

681
01:19:12,930 --> 01:19:18,120
So if you calculate it, you will do this. So that's an extra kilo of rubbish for every endoscope.

682
01:19:18,360 --> 01:19:19,950
It's a big difference.

683
01:19:20,310 --> 01:19:35,760
And that's not including the costs of making it the costs of the raw materials, some of which are not not recyclable and and and all of that.

684
01:19:35,760 --> 01:19:42,630
And that's, you know, and that's calculated and and factored in the kind of decon and reprocessing things.

685
01:19:43,230 --> 01:19:53,210
So, I mean, I, I feel pretty strongly that, you know, we should push, push back against single use and discuss.

686
01:19:53,270 --> 01:19:56,590
Yeah, it sounds like that's not from an environmental engineer.

687
01:19:56,610 --> 01:20:03,330
Yeah. Except in situations where there is clearly an infection control risk.

688
01:20:04,260 --> 01:20:07,350
You know, this seems to be a a no brainer.

689
01:20:08,130 --> 01:20:11,610
And I think we need to push back against companies that are, I think,

690
01:20:11,610 --> 01:20:18,750
trying to exploit a a line of arguments about patient safety, which I think is not really justified.

691
01:20:18,780 --> 01:20:24,689
And we need to really consider on a wider level, not just in endoscopy, but in clinical practise in general,

692
01:20:24,690 --> 01:20:30,299
about moving towards using less waste, moving towards recycling things.

693
01:20:30,300 --> 01:20:35,970
And you know, yeah, because there's so much waste generated that we could do better.

694
01:20:37,830 --> 01:20:43,980
And the other thing that they mentioned right off the bat, which actually I think is much more important,

695
01:20:44,400 --> 01:20:50,760
is that I think we should start to factor in not just kind of clinical harm in

696
01:20:50,760 --> 01:20:55,139
terms of when we when we arrange telescopes and thinking about the patient,

697
01:20:55,140 --> 01:21:00,690
the risk benefit, environmental loss of the patient is the fact that we're spending a whole load of, you know,

698
01:21:00,690 --> 01:21:06,990
the Government's money in an NHS system and we're also using some of Earth's valuable resources for that procedure.

699
01:21:07,260 --> 01:21:13,230
And there is no doubt that we do a whole load of scopes that if we're honest with ourselves and we were honest with the patients,

700
01:21:13,410 --> 01:21:17,400
have pretty, pretty borderline indications. Yeah, absolutely.

701
01:21:17,520 --> 01:21:25,440
You know, so do less do less tests is the best way to make your tests more environmentally friendly.

702
01:21:25,620 --> 01:21:29,520
And the second one is, don't use it to use a single use endoscope.

703
01:21:30,090 --> 01:21:30,990
If at all possible.

704
01:21:31,350 --> 01:21:39,389
Yes, we do need to think about all of these things and just the fact that it's starting to come into our consciousness as clinicians now,

705
01:21:39,390 --> 01:21:47,420
I think is a really good thing. So I've got two more.

706
01:21:47,420 --> 01:21:51,940
But I know you've got a couple of paper papers that I am going to try and be really, really quick with.

707
01:21:52,070 --> 01:21:54,380
So I'll let you do your ones and then I'll wait through these two.

708
01:21:55,760 --> 01:22:05,270
A while ago I discussed a modelling paper about minimum unit pricing and how that might impact on various different health,

709
01:22:05,270 --> 01:22:11,180
but also broader social outcomes if it were applied in England.

710
01:22:12,080 --> 01:22:24,740
So this is sort of going back to this little bit and this paper or report anyway is from commissioned by the Public Health Scotland

711
01:22:25,010 --> 01:22:30,950
and it's entitled Evaluating the impact of minimum unit pricing in Scotland on people who are drinking at harmful levels.

712
01:22:31,460 --> 01:22:35,780
And this was published on the 7th of June this year.

713
01:22:36,800 --> 01:22:43,490
So it's not in a peer review paper and it's it's 210 pages long.

714
01:22:43,490 --> 01:22:54,140
I didn't read all of it, but I did read quite a bit of it actually, and it was really interesting and definitely changed my view.

715
01:22:55,090 --> 01:23:00,760
About minimum unit pricing and perhaps what the evidence says about it,

716
01:23:01,510 --> 01:23:09,090
because obviously as a sort of lefty hepatologist, I'm like, yeah, minimum unit pricing, great.

717
01:23:09,100 --> 01:23:16,060
We need to get people to drink less. But is it impacting on the people we want it to impact the most?

718
01:23:16,940 --> 01:23:27,580
So quick background, I think most people know this minimum unit price for a unit of alcohol in Scotland was introduced on the 1st of May 2018.

719
01:23:28,150 --> 01:23:32,320
So the minimum price for a unit of alcohol in Scotland is 50 p.

720
01:23:34,160 --> 01:23:38,660
And one unit of alcohol is ten mils or eight grams of pure ethanol.

721
01:23:39,470 --> 01:23:44,330
This policy does have a sunset clause, so that means that six years after it started, say,

722
01:23:44,330 --> 01:23:51,260
2024, this policy will expire unless the Scottish Parliament votes for it to continue.

723
01:23:51,770 --> 01:23:59,870
Hence why Public Health Scotland has commissioned quite a few different studies to evaluate its impact.

724
01:24:01,200 --> 01:24:10,729
And so this piece of work was specifically targeted at the population of people that

725
01:24:10,730 --> 01:24:15,830
drink at harmful levels and including people with alcohol dependence in Scotland.

726
01:24:15,830 --> 01:24:17,510
And this is a specific group of people.

727
01:24:17,840 --> 01:24:26,670
So the definition in the UK of somebody that drinks at a harmful level is a person that drinks more than 35 units a week.

728
01:24:26,690 --> 01:24:29,480
If you're a woman or more than 50 units a week. If you're a man.

729
01:24:30,850 --> 01:24:38,290
And within that group, one in five of them will be classified as being alcohol dependent.

730
01:24:38,290 --> 01:24:44,800
So they have a physical and or and or psychological dependence on alcohol.

731
01:24:45,880 --> 01:24:53,530
And these this particular group, are most at risk of poor health outcomes from alcohol consumption,

732
01:24:53,530 --> 01:24:59,670
high alcohol consumption, and also some of the other social and.

733
01:25:00,580 --> 01:25:10,250
Negative outcomes that go hand in hand with it. So the intervention was obviously the introduction of the price and and I'll talk about the,

734
01:25:10,250 --> 01:25:16,870
the different outcomes they looked at that some of the outcomes they looked at when there was no control

735
01:25:16,870 --> 01:25:23,859
group just because of the nature of the methods used and where there was quantitative work done.

736
01:25:23,860 --> 01:25:30,460
They did use control populations in northern England as a as a control.

737
01:25:31,430 --> 01:25:34,280
So there were four work packages. So these are the outcomes they looked at.

738
01:25:35,210 --> 01:25:41,810
One was to look at the impact of the minimum unit pricing on people who use

739
01:25:41,810 --> 01:25:49,160
alcohol treatment services specifically with harmful and with healthy drinking.

740
01:25:49,760 --> 01:25:58,639
So they had to be in touch with those services work package to looked at the impact of it in the community and

741
01:25:58,640 --> 01:26:05,930
that included on speaking to families and carers of people that lived with those that drinker harmful levels.

742
01:26:06,830 --> 01:26:14,170
And both of these work packages used same qualitative better methods.

743
01:26:14,180 --> 01:26:20,180
They were package one used qualitative interviews with service users that use the alcohol treatment services,

744
01:26:20,180 --> 01:26:26,059
but also some quantitative surveys and work package to use interviews and focus

745
01:26:26,060 --> 01:26:30,680
groups with people with alcohol dependence and their families and carers.

746
01:26:31,550 --> 01:26:34,340
Then what? Package three, which I'll focus on a bit more,

747
01:26:34,550 --> 01:26:43,640
it's a probably more interesting one is looked at the impact of the minimum price on harmful drinking in the general population.

748
01:26:43,640 --> 01:26:52,670
So that's looking at just the prevalence of harmful drinking change after minimum pricing is introduced.

749
01:26:52,940 --> 01:27:02,900
And does the the consumption patterns of alcohol change often in pricing in this particular group work package for wasn't completed,

750
01:27:03,170 --> 01:27:10,040
but it aimed to look at the impact on within primary care.

751
01:27:10,420 --> 01:27:14,629
I aim to analyse the data set of linked electronic GP hospital death records,

752
01:27:14,630 --> 01:27:23,690
but as with previous IT projects in the NHS, it sounds like that linkage didn't work out.

753
01:27:23,690 --> 01:27:35,570
And so that hasn't it hasn't been completed yet. And so the headline findings really are from I'll start with the work package three.

754
01:27:37,300 --> 01:27:39,980
So. The consumption patterns.

755
01:27:40,520 --> 01:27:52,250
So after minimum unit pricing was introduced, there is evidence that people with alcohol dependence are paying more for alcohol.

756
01:27:52,970 --> 01:28:00,560
So the cost for them has gone up from about 49 P to about 59 per unit and that's in line with other

757
01:28:00,560 --> 01:28:05,420
studies that have looked at the population levels and not just people with alcohol dependence,

758
01:28:05,690 --> 01:28:10,670
but the population level. People are paying more for their alcohol as you would imagine.

759
01:28:12,650 --> 01:28:21,080
However, the prevalence of drinking at harmful levels did not change significantly after the minimum unit pricing was introduced.

760
01:28:21,710 --> 01:28:25,320
And. That was a key,

761
01:28:26,130 --> 01:28:33,090
a key kind of a wish of the Scottish Government by introducing this policy is

762
01:28:33,090 --> 01:28:38,489
that actually by introducing a minimum unit pricing that it should impact and

763
01:28:38,490 --> 01:28:43,350
reduce the number of people drinking harmful levels because they're the people

764
01:28:43,650 --> 01:28:50,129
that are often paying and buying cheap alcohol with the is high percentage.

765
01:28:50,130 --> 01:28:53,370
So it should impact them more than people that drink at lower levels.

766
01:28:54,870 --> 01:29:06,179
Now the dataset they used to evaluate this was a a market survey and this market

767
01:29:06,180 --> 01:29:11,460
survey used individual level data on alcohol consumption over a ten year period,

768
01:29:11,490 --> 01:29:21,030
and they used an interrupted time series design. And the control in this case was the same data that from individuals from northern England.

769
01:29:21,840 --> 01:29:25,829
It's a commercial market research survey. It's continuously collected.

770
01:29:25,830 --> 01:29:31,049
It's cross-sectional it's adults that a resident in Great Britain and the survey

771
01:29:31,050 --> 01:29:36,420
about 30,000 individuals a year and they have a weekly quota sample based on age,

772
01:29:36,420 --> 01:29:46,020
sex, social status and where they live in the UK. And the nature of the data they collect is that the participants first answer

773
01:29:46,020 --> 01:29:50,429
questions on the typical alcohol related behaviours and alcohol related attitudes,

774
01:29:50,430 --> 01:29:53,370
and then they complete a one week retrospective drinking diary.

775
01:29:54,240 --> 01:30:05,280
And so you get an idea of how often they're drinking, what they're drinking, how much they're drinking, and this has been collected for a long time.

776
01:30:06,840 --> 01:30:21,240
And so they used diary data that from the 1st of May 2018, anything after that was post and the placement and pricing data and before that.

777
01:30:22,210 --> 01:30:27,370
Was pre and the number of individuals was this was massive.

778
01:30:27,370 --> 01:30:32,500
So 38,000 plus individuals in Scotland.

779
01:30:32,770 --> 01:30:38,680
So that's an average of 267 people a month filling in this survey over the time.

780
01:30:38,950 --> 01:30:46,640
And in northern England, 71,000 people average 494 a month and.

781
01:30:48,050 --> 01:30:53,270
And and so they looked at using these diaries.

782
01:30:53,540 --> 01:31:04,010
They looked at how much people were consuming to define whether they were drinking a harmful level or not, i.e. over 35 or over 50 units per week.

783
01:31:04,520 --> 01:31:05,780
And that's how they worked out.

784
01:31:05,930 --> 01:31:14,460
The prevalence of drinking harmful levels in the last week based on the survey didn't change significantly after the introduction of minimum pricing.

785
01:31:15,470 --> 01:31:22,630
They did look at some other groups that don't fall into that harmful level.

786
01:31:22,640 --> 01:31:33,890
So there's another definition of people that drink hazardous levels, and these are people that drink about 14 units a week, but below 35 or 50.

787
01:31:35,060 --> 01:31:46,970
And the proportion of people drinking at that level did fall significantly between 2017 and 2018, superimposed by a 3.5%.

788
01:31:47,390 --> 01:31:55,580
And the people drinking at moderate levels. Didn't didn't change significantly, but increased a bit.

789
01:31:55,820 --> 01:32:01,160
So suggests that people drinking at hazardous levels fell and they just started drinking a bit less,

790
01:32:01,160 --> 01:32:04,670
maybe as an impact, maybe impacted by the minimum price.

791
01:32:05,570 --> 01:32:11,540
There's some evidence of a switch from high percentage lager in or cider to spirits,

792
01:32:11,720 --> 01:32:21,680
but it wasn't significant and there was no significant impact on the variation in prevalence or consumption patterns between urban and rural areas.

793
01:32:24,320 --> 01:32:30,380
Now, this is different. These findings are very different to what I.

794
01:32:31,390 --> 01:32:36,250
Thought they would find and also what some other studies have found.

795
01:32:37,720 --> 01:32:41,530
So I looked into this a bit more to understand perhaps why.

796
01:32:43,200 --> 01:32:48,329
So the previous studies that have been published looking at the minimum unit price of

797
01:32:48,330 --> 01:32:57,480
alcohol in Scotland have used similar time series analysis data to what these authors used.

798
01:32:57,870 --> 01:33:01,530
But they've used a different type of.

799
01:33:04,730 --> 01:33:11,720
Dataset. So they've looked at household level data on household purchasing rather than individual level data.

800
01:33:12,080 --> 01:33:15,260
Okay. And that's really the difference.

801
01:33:16,220 --> 01:33:23,780
So there's been several different studies since the introduction of minimum unit pricing

802
01:33:24,050 --> 01:33:29,390
that have all looked at household level data and found that alcohol purchasing.

803
01:33:30,360 --> 01:33:35,610
Has decreased at the population level in Scotland since the introduction of minimum

804
01:33:35,610 --> 01:33:43,640
unit pricing and the households that are in the top of sort of fifth so to

805
01:33:43,790 --> 01:33:48,830
the top 20% of households that buy the most alcohol have been most affected have

806
01:33:49,020 --> 01:33:53,850
have reduced their alcohol spend or the amount of alcohol they buy the most.

807
01:33:54,980 --> 01:34:00,680
But it doesn't tell you within that household how much someone's drinking.

808
01:34:00,710 --> 01:34:09,150
Doesn't say how many people are drinking at hazardous level, a moderate level or a harmful level who are alcohol dependence.

809
01:34:09,710 --> 01:34:15,290
And in that household, you know, is it one person in that household that drinks like that?

810
01:34:15,320 --> 01:34:18,810
Is it all of the adults in that house that drink like that?

811
01:34:18,830 --> 01:34:26,870
So this data set is different, quite different and quite it's important because it's looking at slightly different data.

812
01:34:28,490 --> 01:34:37,050
And and I think all of it needs to be looked at together in context with the other work packages.

813
01:34:37,470 --> 01:34:41,730
The outcomes that they looked at, I thought were really interesting and important as well.

814
01:34:42,540 --> 01:34:46,350
So and they were more mixed methods studies qualitative.

815
01:34:46,710 --> 01:34:57,870
And looking at how individuals for example how in a minute pricing had impacted on the daily lives of people who drink at a harmful level.

816
01:34:57,890 --> 01:35:07,140
So they found there was evidence of increased financial strain in in many and the coping mechanisms that people used to account for these

817
01:35:07,320 --> 01:35:16,559
the increased prices were that they used the for intensified say they bought less food they spent less money on gas and electricity.

818
01:35:16,560 --> 01:35:20,420
They borrowed money from friends and family and pawnbrokers and things like that.

819
01:35:20,430 --> 01:35:21,780
They relied on food banks.

820
01:35:22,690 --> 01:35:30,610
And but there were there were obviously unrelated factors that were occurring during this period that may have also impacted that.

821
01:35:30,620 --> 01:35:34,569
So, I mean, right now we've got lots of inflation and things in the economy,

822
01:35:34,570 --> 01:35:38,170
but there was a reduction in universal credit and that might have also contributed.

823
01:35:40,700 --> 01:35:45,350
There was no substantial evidence of the people within this group.

824
01:35:47,960 --> 01:35:54,590
Substituted alcohol for illicit drugs or that there was increased stealing or

825
01:35:54,590 --> 01:35:59,030
violence or alcohol related crimes and no increase in acute alcohol withdrawal.

826
01:35:59,690 --> 01:36:10,459
But there was some evidence in from from the carers and the family members that some of the some of the people with

827
01:36:10,460 --> 01:36:22,580
alcohol dependence or harmful use actually switched to spirits and they were more fearful of violence towards terrorism.

828
01:36:22,580 --> 01:36:28,010
Family members. But there wasn't any. Quantitative evidence of that.

829
01:36:29,090 --> 01:36:35,140
And then, of course, the health side of things wasn't really analysed. I'm in this particular study.

830
01:36:36,280 --> 01:36:43,629
So I thought this was really interesting because it looked at a specific population of people

831
01:36:43,630 --> 01:36:48,970
that we're really interested in what their outcomes are to minimum unit pricing in Scotland,

832
01:36:49,060 --> 01:37:01,060
because actually they're where the money is, the higher risk of negative impacts of alcohol dependence or harmful drinking.

833
01:37:02,020 --> 01:37:07,329
And and there are obviously some limitations and there are several explanations

834
01:37:07,330 --> 01:37:12,729
as to why their findings might be a little bit different to previous studies.

835
01:37:12,730 --> 01:37:28,240
So the methodological limitations obviously include non-random sampling and survey data, which is notoriously and often not not very accurate.

836
01:37:28,960 --> 01:37:34,780
And some of the work packages looked at particular populations within that group of alcohol dependence.

837
01:37:34,780 --> 01:37:38,319
So they had to be in touch with, you know, alcohol services, for example.

838
01:37:38,320 --> 01:37:49,570
So they're missing lots of people and. And it also required it also has heavily relied on data from people drinking at

839
01:37:49,570 --> 01:37:54,350
harmful levels that only participated in that online market research panel.

840
01:37:54,370 --> 01:37:59,170
So I bet there's a whole load of people that do not do that and therefore weren't captured in this.

841
01:38:00,570 --> 01:38:05,260
And so the way they interpreted the sort of main, main findings.

842
01:38:05,500 --> 01:38:12,159
Well, the first thing is they might have the findings might have underestimated the effect

843
01:38:12,160 --> 01:38:16,240
of minimum unit pricing on the consumption of alcohol in this particular population.

844
01:38:18,280 --> 01:38:27,580
It differs from other published data, as I've already talked about, the household data and modelling studies is the other thing to talk about.

845
01:38:27,850 --> 01:38:31,299
They consistently and I know they're modelling not real world studies,

846
01:38:31,300 --> 01:38:38,260
but they consistently model that there will be a positive impact of minimum unit pricing on alcohol consumption,

847
01:38:38,710 --> 01:38:45,820
particularly in these groups that drink the most harmful levels.

848
01:38:48,670 --> 01:38:54,219
But equally, another argument is that people with alcohol dependence might be less responsive

849
01:38:54,220 --> 01:38:57,280
and probably are less responsive to price changes than lighter drinkers,

850
01:38:58,990 --> 01:39:01,629
particularly if they're dependent not just harmful drinkers,

851
01:39:01,630 --> 01:39:10,330
because they've got a physiological, psychological, physical addiction essentially to alcohol.

852
01:39:10,370 --> 01:39:19,420
So they will put alcohol consumption above eating and buying fuel and all of those things.

853
01:39:20,260 --> 01:39:23,950
And it might be that the minimum unit price is too low, therefore,

854
01:39:24,190 --> 01:39:31,270
to have a have the impact we want it to have on the people that are drinking at that high level.

855
01:39:33,430 --> 01:39:36,160
And they also commented on, you know,

856
01:39:36,580 --> 01:39:47,320
these management strategies that people seem to be using to offset that increased price in alcohol might not be very sustainable,

857
01:39:47,780 --> 01:39:54,309
and especially in the current economic climate, you know, how long can you know?

858
01:39:54,310 --> 01:39:58,450
They're all they're all short term solutions and that might not.

859
01:39:58,840 --> 01:40:08,530
And actually, longer term, it might be that people do change their behaviour and it also might be that it might encourage people who

860
01:40:08,530 --> 01:40:18,290
are younger that would have become hazardous or sort of harmful drinkers to not drink to that level.

861
01:40:18,310 --> 01:40:24,310
So it hasn't looked at that particular group. And also the other important thing,

862
01:40:24,310 --> 01:40:28,660
and I guess the most interesting thing for us is it hasn't it's only this

863
01:40:28,660 --> 01:40:33,430
particular report hasn't really evaluated the health effects and a few of the.

864
01:40:34,680 --> 01:40:40,170
Reports have touched on it, but in terms of admissions, so.

865
01:40:40,530 --> 01:40:46,350
Yeah, exactly. Tendencies without. Yeah. Related problems or injuries or alcoholic hepatitis.

866
01:40:46,860 --> 01:40:53,459
Exactly. Right. I mean, the the Public Health Scotland plan to undertake health outcome analysis.

867
01:40:53,460 --> 01:41:04,140
So that should come and there will be a final over overarching report next June about all the different things that they've reported.

868
01:41:04,810 --> 01:41:15,390
And the other thing that will muddy the water obviously here is COVID, because that did change things that at times in I go through a whole episode.

869
01:41:15,570 --> 01:41:26,070
I know. I'm sorry. But so so I just I just thought it was really interesting mean to look at because I kind of thought oh yeah.

870
01:41:26,070 --> 01:41:30,210
Minimum unit pricing, right? Yeah. I mean, that is very new.

871
01:41:30,430 --> 01:41:38,640
Sold as a bit of a panacea because it's I think the the modelling data that it was a really it's a bit of a silver silver bullet.

872
01:41:39,000 --> 01:41:49,380
Yes. You know, and it's it's, you know, the Royal College of Physicians, you know, health inequalities and policy,

873
01:41:49,680 --> 01:41:56,069
you know, include support for minimum unit pricing and as as do as do many others.

874
01:41:56,070 --> 01:42:04,170
And. It will be interesting to see the kind of the real world effects hopefully trying to control,

875
01:42:04,470 --> 01:42:10,020
try and treat, trying to control for COVID and other other parallel effects.

876
01:42:10,050 --> 01:42:17,010
But yeah, actually, this is a slightly, slightly depressing outcome actually, though I was rather hoping that this is going to happen.

877
01:42:17,030 --> 01:42:27,370
Yeah, I know it wasn't. It wasn't what I was expecting, but yeah, really, really interesting.

878
01:42:27,370 --> 01:42:33,120
And I'll definitely look out for the further reports that are going to be published in the next year or so.

879
01:42:33,150 --> 01:42:46,080
Progress. Yeah. Yeah. So I've got a short, mildly depressing one, and then I've got a positive one to end on.

880
01:42:46,470 --> 01:42:49,670
Right. Okay. Well, it's good. As long as we end on a positive.

881
01:42:49,680 --> 01:42:52,799
That's fine. Yeah. Say something.

882
01:42:52,800 --> 01:43:02,570
Negative one. Is a commentary a commentary in gastroenterology by Lauren and Danielle Rabinovich.

883
01:43:02,630 --> 01:43:09,400
They are from Harvard. Beth Israel Deaconess Medical Centre and the Boston Children's Hospital.

884
01:43:09,760 --> 01:43:18,970
This is a paper about the disparities in inclusion of female, pregnant, lactating and older individuals in IBD clinical trials.

885
01:43:20,590 --> 01:43:25,840
And this is a nice and nice little read and I like the way they did the study.

886
01:43:26,290 --> 01:43:30,189
And so inclusion is really important.

887
01:43:30,190 --> 01:43:34,230
It is a priority for the National Institutes of Health Health,

888
01:43:34,240 --> 01:43:42,879
the niche in the US who have specifically said that they want health, you know, medical,

889
01:43:42,880 --> 01:43:48,700
medical trials and health research to include groups that have historically been excluded from that,

890
01:43:48,700 --> 01:43:54,070
including women, children, older adults and members of racial and ethnic minority groups.

891
01:43:55,510 --> 01:43:59,979
And it's really important in the context of clinical trials,

892
01:43:59,980 --> 01:44:09,630
because we want to know that the results of those clinical trials are generalisable to the populations that we serve and Crohn's.

893
01:44:09,750 --> 01:44:17,469
Quincy's is an interesting one because it's not it's a heterogeneous population,

894
01:44:17,470 --> 01:44:23,560
which includes some of those some of those harder to reach populations.

895
01:44:23,560 --> 01:44:30,970
So there is a female preponderance in IBD overall and more so in crimes than you see.

896
01:44:32,380 --> 01:44:43,690
But you know, overall, overall, 55% of people with IBD are female and IBD affects a whole range of ages.

897
01:44:43,690 --> 01:44:48,340
So this this this commentary isn't about paediatric space.

898
01:44:48,550 --> 01:44:55,780
Clearly, there's a considerable number of kids with IBD who are often excluded from from trials that focus just on adults.

899
01:44:56,140 --> 01:45:05,230
But we also have that bimodal age distribution, and we are increasingly diagnosing people later in life with de novo IBD,

900
01:45:05,410 --> 01:45:10,809
as well as having patients who've had IBD for a long time and are now entering their latter years.

901
01:45:10,810 --> 01:45:16,810
But 10 to 15% of new diagnoses of IBD are in the over sixties.

902
01:45:17,420 --> 01:45:25,030
So this study was looking at the effects of the how representative clinical trials were,

903
01:45:25,890 --> 01:45:34,810
and the other group that importantly are often excluded are pregnant women and women who are breastfeeding.

904
01:45:35,610 --> 01:45:40,450
And we know that this this, this group is quite large, quite important.

905
01:45:40,690 --> 01:45:46,060
And we have lots of our IBD patients who are either pregnant or are interested in becoming pregnant.

906
01:45:46,540 --> 01:45:55,209
Indeed. And so this study, they looked just they looked at clinical trials,

907
01:45:55,210 --> 01:46:02,500
dot gov and all IBD trials registered in a 20 year period up to June the 30th, 2020.

908
01:46:03,980 --> 01:46:07,040
And they found 146 trials that met their criteria.

909
01:46:07,250 --> 01:46:14,030
And then they went through the demographic data and compared it to the US population data.

910
01:46:14,300 --> 01:46:20,420
So basically, do U.S. IBD trials match the US population?

911
01:46:20,720 --> 01:46:24,910
And spoiler alert, they don't say.

912
01:46:25,550 --> 01:46:31,280
So 146 trials, 50,000 participants, paediatric and adults.

913
01:46:31,490 --> 01:46:41,480
Most of these are industry funded because most of these are therapeutic drug trials, although there were few other like fam t trials or device trials.

914
01:46:42,350 --> 01:46:47,509
80% industry funded and 75% are phase two or phase three trials.

915
01:46:47,510 --> 01:46:56,750
So these are the key registration trials upon which, you know, the key data upon which we then decide how we use these, these, these drugs.

916
01:46:57,020 --> 01:47:00,410
Yeah. So then we use them in all sorts of different populations.

917
01:47:00,710 --> 01:47:04,340
Yeah. But these are, you know, these are, these are the key results.

918
01:47:05,450 --> 01:47:14,060
So women and total trial participants and this is include all the paediatric trials as well.

919
01:47:14,870 --> 01:47:18,500
51.6 were female, so quite similar.

920
01:47:19,760 --> 01:47:33,770
But when we look at adult studies, female patients comprise 54.6% of Crohn's disease trials, but only 44% of U.S. trials.

921
01:47:33,770 --> 01:47:43,940
They slightly underestimate representing women compared to kind of, you know, the known IBD frequency in the U.S.

922
01:47:45,590 --> 01:47:51,530
Pregnant or lactating women were excluded from over 60% of the trials, you know, off the bat.

923
01:47:52,130 --> 01:48:03,770
And but what's also of relevance there is that even when they were allowed to be included, they often don't they weren't.

924
01:48:04,250 --> 01:48:08,389
Um, well, they either weren't. But also it's not possible to tell.

925
01:48:08,390 --> 01:48:11,870
So there's no subgroup analysis for those patients.

926
01:48:12,170 --> 01:48:20,840
So it's often, even if they don't seem to be a formal exclusion criteria, you then don't have the data that you kind of pull out to say,

927
01:48:20,840 --> 01:48:29,540
ah, there were, you know, three breastfeeding women in this group and, and, you know, with that this was the outcomes.

928
01:48:30,380 --> 01:48:47,900
Um, so when we start thinking we've got 10 to 15% of IBD patients diagnosed with IBD over 60 buts of the 105 trials in adults,

929
01:48:48,320 --> 01:48:59,900
the mean age is relatively low, which is, I guess what you would expect, say, in Crohn's disease, the median age was 37 and you see it was, it was 42.

930
01:49:00,380 --> 01:49:07,220
Um, only 39% of trials included patients over 65 years old.

931
01:49:07,640 --> 01:49:10,430
So all they did was just exclude them off the bars.

932
01:49:10,760 --> 01:49:19,640
But even where they were included, they often made up a small minority of of the the patients in the trial.

933
01:49:19,850 --> 01:49:28,310
And so overall, if we look at all Crohn's disease trials in this cohort, less than 1% of patients were over 65.

934
01:49:28,520 --> 01:49:31,700
And in the U.S. trials, less than 4%.

935
01:49:32,090 --> 01:49:39,200
And this is so important, actually, because, you know, we we don't have the data for these patients, these older patients.

936
01:49:39,200 --> 01:49:44,509
And yet a lot of our clinical decision making is, especially for biologics and things like that,

937
01:49:44,510 --> 01:49:50,840
is based on, oh, you know, they're over a certain age or have lots of co-morbidities.

938
01:49:51,920 --> 01:49:54,980
Let's not give them let's give them V do. Yeah.

939
01:49:55,010 --> 01:49:58,489
So, so we infer risk. Yeah.

940
01:49:58,490 --> 01:50:04,309
But it's not actually I but it's actually not. Yeah, it's not from, it's not from a great deal of clinical trial data is like, oh,

941
01:50:04,310 --> 01:50:10,160
they're at higher risk of, for instance, cancer or heart failure related to anti-TNF, let's say.

942
01:50:10,430 --> 01:50:16,940
But actually a lot of these trials are excluding these patients off the bat because of those risks.

943
01:50:17,090 --> 01:50:22,879
So self-fulfilling prophecy. So you don't include them because you haven't got the data about them.

944
01:50:22,880 --> 01:50:28,370
But the reason you have the data about them is because you didn't include them. So, yes, yeah, this is a problem.

945
01:50:28,910 --> 01:50:32,660
And then we come on to race and ethnicity.

946
01:50:33,050 --> 01:50:36,740
And in these trials this is a US population.

947
01:50:36,740 --> 01:50:51,020
So big difference in the UK. Um, US trials overall we've got nearly somewhere between 85 and 90% of trial participants are white, um, uh,

948
01:50:51,110 --> 01:51:03,050
depending on how you stratify it into Crohn's or you see, but only 76% of the US population is, is white the populations that are underrepresented.

949
01:51:03,480 --> 01:51:16,470
In particular are those of black and African American nationality, which make up more like 13% of the U.S. and they make up about 3%.

950
01:51:17,580 --> 01:51:26,250
In these trials, Hispanic communities are also lower, and I don't even think they were able to delineate populations like,

951
01:51:26,250 --> 01:51:32,250
you know, South Asian descent or First Peoples and, you know, Pacific Islanders.

952
01:51:32,250 --> 01:51:35,550
And and so this is poor. This is so.

953
01:51:35,580 --> 01:51:40,020
Yeah. So isn't it. And not surprising. Sadly, it's not.

954
01:51:40,020 --> 01:51:44,009
It's it's not it's cool. And, you know, they haven't actually gone on to to look at this,

955
01:51:44,010 --> 01:51:49,260
but I don't think they've scratched the surface in terms of thinking then about socio economic status.

956
01:51:49,560 --> 01:51:55,230
And I have, you know, no doubt that there is disparities along along those lines there.

957
01:51:55,710 --> 01:52:01,200
Yeah. So it's it's a little bit of a depressing kind of outcome.

958
01:52:02,550 --> 01:52:07,980
You know, I think if we were grading grading IBD clinical trials in the US,

959
01:52:07,980 --> 01:52:12,690
I think we would say, you know, in terms of inclusion, four out of ten could do better.

960
01:52:13,620 --> 01:52:21,390
We really need to make an effort to include these groups and if there are reasons not to include them,

961
01:52:21,600 --> 01:52:27,600
I think, you know, that needs to be looked at quite carefully and the default should not be,

962
01:52:27,600 --> 01:52:34,980
oh, let's just exclude all old people and let's exclude women because they might get pregnant and mock up our stance like those.

963
01:52:35,310 --> 01:52:38,430
Those are not good reasons to exclude people from trials.

964
01:52:38,730 --> 01:52:45,990
Yeah. And it perpetuates. And I suspect this is true of not just the IBD field, but.

965
01:52:47,400 --> 01:52:51,690
Much the wider medical field and it reflects wider society.

966
01:52:52,680 --> 01:52:57,540
Indeed, indeed. I suppose gastroenterology is a microcosm of society.

967
01:53:03,440 --> 01:53:08,150
Right. Happy news. Only take a few minutes. And also a positive paper.

968
01:53:08,720 --> 01:53:12,380
Positive, positive paper. Yeah, exactly. Yeah. New England Journal of Medicine.

969
01:53:13,580 --> 01:53:21,070
This paper is called to zap her tide once weekly for the treatment of obesity.

970
01:53:21,350 --> 01:53:27,020
So it has appetite is a drugs that I find incredibly difficult to say.

971
01:53:27,980 --> 01:53:32,120
But this is the main thing. Say obesity is bad.

972
01:53:32,360 --> 01:53:38,840
And unfortunately, most of the US and and Europe are overweight or obese.

973
01:53:39,560 --> 01:53:42,740
But we've had a couple of really exciting trials.

974
01:53:42,740 --> 01:53:52,250
We've spoken about one that came out a few months ago on this podcast before of Semaglutide, which is a GLP one receptor agonist.

975
01:53:52,460 --> 01:53:57,920
And that demonstrated some really nice efficacy for the full weight loss,

976
01:53:57,920 --> 01:54:04,100
I think in the region of sort of 12 to 14% body weight lost after a bit over a year on the drug,

977
01:54:04,340 --> 01:54:08,900
which was pretty amazing and is much better than previous drugs.

978
01:54:09,080 --> 01:54:17,420
It comes with some problems, particularly things like gall bladder related complications and and pancreatitis

979
01:54:17,690 --> 01:54:25,819
as well as a variety of GI side effects to cause appetite is it's quite cool.

980
01:54:25,820 --> 01:54:39,660
It's a peptide, it's a sort of it's engineered from the native GI GI sequence and it's an agonist both for GIP receptors but also GLP one receptor.

981
01:54:39,680 --> 01:54:46,550
So it's targeting both the GLP one pathway, but also the GP sort of gut hormone pathway.

982
01:54:47,060 --> 01:54:53,120
And it is licenced for the treatment of type two diabetes, I think.

983
01:54:54,080 --> 01:54:58,640
Well, it's certainly been used in trials for type two diabetes. I am not a diver ologist.

984
01:55:00,530 --> 01:55:07,430
Anyway, this is a phase three multi-center double blind randomised controlled trial in 119 sites.

985
01:55:07,490 --> 01:55:14,810
Nine countries in adults who have a BMI of 30 or more or 27 if they've got complications.

986
01:55:15,500 --> 01:55:19,220
So this is including people who are very overweight and obese,

987
01:55:19,430 --> 01:55:28,280
but also people who are sort of on the kind of borderline overweight or obese, but with some dyslipidemia or cardiovascular disease.

988
01:55:28,310 --> 01:55:33,139
Okay. This was not a trial, including people with diabetes.

989
01:55:33,140 --> 01:55:44,540
They were excluded and they were randomised 1 to 1 to 1 to 1 to either placebo or three different doses of this drug that I cannot say.

990
01:55:44,810 --> 01:55:51,830
And as an adjunct as an adjunct to the drug therapy, they were given lifestyle intervention.

991
01:55:51,830 --> 01:55:55,790
So they had regular lifestyle counselling sessions. They saw a dietician.

992
01:55:56,090 --> 01:56:04,160
They were helped with a healthy, balanced diet with a calorie deficit of 500 kilocalories a day.

993
01:56:04,310 --> 01:56:09,680
And they were also incentivised to exercise for 150 minutes of physical activity a week.

994
01:56:09,680 --> 01:56:15,440
So we did all of the right things and then we either gave them the drug or not for 72 weeks.

995
01:56:15,980 --> 01:56:23,629
Okay. So quite a long time as well. You say the Co-Primary endpoints were the percentage in body weight change from baseline

996
01:56:23,630 --> 01:56:28,280
to week 72 and also the proportion of people who had lost a certain increment of weight.

997
01:56:28,520 --> 01:56:34,340
And the key one that's become one of the key metrics in these trials is losing more than 5% of your body weight.

998
01:56:34,610 --> 01:56:36,170
Mm hmm. There's two reasons for that.

999
01:56:36,440 --> 01:56:45,290
One is that basically all trials of any kind of kind of conventional lifestyle therapy has have deltas of less than that.

1000
01:56:45,290 --> 01:56:48,860
People lose less than 5% over a year on average.

1001
01:56:49,130 --> 01:56:55,370
And two, that seems to be associated with tangible benefits in terms of blood pressure,

1002
01:56:55,370 --> 01:56:58,430
cardiovascular risk, fatty liver disease, blah, blah, blah, blah, blah.

1003
01:56:59,510 --> 01:57:02,750
So overall, this is a really well powered study.

1004
01:57:02,760 --> 01:57:10,430
They had 2539 participants, 86% adherence to the therapy.

1005
01:57:10,760 --> 01:57:19,010
And the demographics were much as they planned. These people have a meet mean BMI of 38.

1006
01:57:19,280 --> 01:57:26,840
That's sort of 105 kilos. And they have, you know, some 40% had prediabetes at baseline.

1007
01:57:28,640 --> 01:57:31,730
The results are stonking.

1008
01:57:32,420 --> 01:57:43,729
Absolutely stonking. So in the placebo arm, after a year, the percentage body weights, the mean percentage, body weight is -3%.

1009
01:57:43,730 --> 01:57:48,560
So people lost 3% of their body weight. Mm hmm. In the.

1010
01:57:49,570 --> 01:58:01,340
To his appetite arms, particularly in the ten and 15 mg arms of the two higher doses, they lost 19.5 and 20.9.

1011
01:58:01,610 --> 01:58:05,870
Wow. That's massive. That's just me.

1012
01:58:06,290 --> 01:58:14,890
That's the mean. So that's when we start looking at the people who've lost 5%, 10%, 15%, 20% within that time.

1013
01:58:15,640 --> 01:58:25,090
So basically, in the two top doses, the ten or 15 milligram doses, 90% of people are losing 5% body weight.

1014
01:58:26,110 --> 01:58:33,550
Well, that's pretty good, losing 10% body weight and two thirds losing 15%.

1015
01:58:33,580 --> 01:58:41,260
And, you know, as you would expect, half are losing 20% of their body weight by the end of the trial.

1016
01:58:41,440 --> 01:58:47,140
So what's the drawback? Well, honestly, this is this is pretty pretty amazing.

1017
01:58:47,440 --> 01:58:49,629
Yeah. The trial was really nicely designed.

1018
01:58:49,630 --> 01:58:58,750
And actually it's a good follow on from the previous one because 71% of the participants was white, which is exactly the proportion of people.

1019
01:58:58,840 --> 01:59:09,910
Yeah. It states they also have 7% black participants, a slightly less than the US average, but a bit better than the average.

1020
01:59:10,150 --> 01:59:21,460
And they've also quoted for people of Asian origin and also people who are American-Indian or Alaska Natives and also Hawaiian and Pacific Islander.

1021
01:59:21,470 --> 01:59:24,570
So that's stratified properly. That was good.

1022
01:59:25,720 --> 01:59:34,150
Or ethnic group, which is great. The downsides are actually not too bad.

1023
01:59:34,900 --> 01:59:37,940
So interesting. There really wasn't any kind of cholesterol.

1024
01:59:37,960 --> 01:59:49,900
Sorry, a cola. Stacy's gallbladder gallstone on the signal or a pancreatitis signal, which were the ones from things like Semaglutide and Liraglutide.

1025
01:59:50,410 --> 01:59:53,860
And is there a kind of. I don't know much about these drugs.

1026
01:59:53,860 --> 01:59:57,460
Is there a biological reason as to why that might be? I don't know.

1027
01:59:57,760 --> 02:00:04,430
It's a great question. I have no answer for you, Tamsin. The main thing is that loads of people get GI side effects.

1028
02:00:06,640 --> 02:00:14,740
They get GI side effects, say like, you know, in compared to the placebo arm, like 25 to 30% of people are getting nausea.

1029
02:00:15,460 --> 02:00:19,770
20% of people are getting diarrhoea. But did it stop them?

1030
02:00:19,780 --> 02:00:26,390
It didn't stop them taking so. So it looks like a lot of people have adverse events.

1031
02:00:26,410 --> 02:00:37,590
Most of them seem to be mild to moderate in terms of stopping their overall in the higher dose terms.

1032
02:00:38,150 --> 02:00:46,630
His appetite arms 6 to 7% are stopping because of adverse events within a year compared to 2.6% in the placebo arm.

1033
02:00:46,870 --> 02:00:52,090
So maybe four or 5% extra leaving the trial because of the side effects.

1034
02:00:53,950 --> 02:00:59,079
But they do have some quality of life data and some of that health related quality of life is improving.

1035
02:00:59,080 --> 02:01:08,020
So clearly this is balancing out. There's a bit of a signal for alopecia, which isn't great, but in terms of ones that are sort of,

1036
02:01:08,380 --> 02:01:15,160
you know, serious adverse events or death, there was no signal of harm in for for this treatment.

1037
02:01:15,850 --> 02:01:19,839
Well, overall, that's very, very positive indeed. Extremely positive.

1038
02:01:19,840 --> 02:01:22,840
And basically, we should put this in the water.

1039
02:01:24,130 --> 02:01:29,890
Yes. You know, honestly, this is absolutely brilliant. I think would be great to see more quality of life data.

1040
02:01:30,730 --> 02:01:40,300
They've got plexi outcomes for long term health benefits like blood pressure and you know, lipids and insulin levels and stuff like that.

1041
02:01:40,600 --> 02:01:41,590
But you know,

1042
02:01:41,680 --> 02:01:49,450
clearly we want to follow up with these cohorts to see why if they keep their weight off and to to they have less progression to diabetes,

1043
02:01:49,450 --> 02:01:54,249
heart attacks, strokes. But overall, I think this is really exciting.

1044
02:01:54,250 --> 02:01:57,790
And if you look back, you know, even two or three years ago,

1045
02:01:57,790 --> 02:02:02,860
you'd say there's a couple of really rubbish drugs and surgery as good treatments for obesity.

1046
02:02:03,160 --> 02:02:08,320
And now I think you can actually say that we've got multiple medical alternatives.

1047
02:02:08,590 --> 02:02:14,590
Yeah. Suitable for patients with obesity. That's actually really good for this.

1048
02:02:15,130 --> 02:02:18,760
Yes. Yes. Well, that's that's really great.

1049
02:02:18,760 --> 02:02:26,950
A great one to end on. Super. I think this is a world record for the longest podcast with FDA, which is very interesting types of reading really.

1050
02:02:27,250 --> 02:02:31,480
I'm going to say thanks for listening. We'll see you next time.