Gastroenterology Controversies: Long Term Safety of Proton Pump Inhibitors

Gastroenterology Controversies: Long Term Safety of Proton Pump Inhibitors



so well welcome everybody thank you for coming today to GI Grand Rounds on controversies on the long term safety of proton pump inhibitors Thank You dr. Rubin for inviting me and for had for helping me to choose this topic as it is timely uncontroversial thank you two of my panel of experts there's also a screen up here for you to look at but if you guys are uncomfortable feel free to turn around and look at the big screen I don't want to watch this stream it's a little bit smaller out there so first on our panel our very own Mark Chapman an expert in gastric physiology we also have dr. Daniel Freiburg who came to join us from Columbia University who's an expert in PPI use and wrote the aga expert review and best practices advisory that was published this year and dr. M Manuela AoE from the Department of Epidemiology she is the director of translational epidemiology and she's here to help us interpret some of these other studies that we've been discussing in the past year so that outline of this presentation I'll first place I'll first start talking about the efficacy of proton pump inhibitors versus h2 receptor antagonists I will interview the indications for PBIS we will critically evaluate the literature on PPS and the risk of CKD then we'll discuss the strength of this evidence with the help of our expert panel will review the relative risks of dementia fractures and gastric cancer and then I will go over the age a best practice of advice so the key points that you'll see me highlight multiple times throughout this presentation are that we want to reassure and educate our patients if PPS are actually indicated we want to reduce the dose to the lowest possible dose in all patients and if there's no indication we want to discontinue PBS and I think wherever we land on the literature however you however you interpret the literature and these three points will maintain throughout so again I'll start with the first case here everyone already had their phones app so again in case you're just logging on for the first time in the to box where you usually type a cell phone number you type 2 2 3 3 3 and then in the text message instead of saying hey when fgi conference you type in Charles tonight at six oh three and that will log you into the poll so the first case is a 62 year old obese white man with symptomatic GERD and c5 m2 Barrett's esophagus he comes to you and asks if he should continue his once daily PPI he reports that his reflux recurs if he misses a single dose of his PPI he also has a printed email that says PPI is caused kidney disease which is from one from his golf buddies and he really wants to talk to you about this is very good sir so we have a poll everywhere don't don't don't vote yet I will activate the poll and make it fullscreen and you can again read the read the question and then just text message a B or C for your answer choice and we'll see how the room votes good it's working we got one I'll hopefully get about ten of these all right at 10 well great so good exactly what I wanted is a split between a and B so this question was written to be somewhat controversial hopefully you felt a little dissonance when you were trying to answer a or b I think it really it really depends on where when you land on on your your own understanding or your own interpretation of the literature so I picked this case because Shivani Gupta just presented the grounds two weeks ago on Eric's esophagus and one of the points she presented was that there's a value in PPI there's a shown benefit to chemo prevention so decreasing prevention from low grade dysplasia to high grade dysplasia and and adenocarcinoma in patients who have Barrett's esophagus and so yes you should continue the PPI and really and is just kind of your your own interpretation so that's great exactly exactly was hoping for again the key points this one doesn't line up but the key point for this one whether this one hit was reassure and educate your patient if a PPI is educated so you can reassure this guy with Barrett's that he does need it so why not use h2 receptor antagonists I want to briefly review the mechanism and then some of the original literature that some of the fellows we haven't we haven't looked at so much so there are three major pathways of parietal cell stimulation when thinking about acid secretion the first is the vagus nerve stimulating the parietal cell directly from the neural pathway the second is the hormonal pathway when gastrin secretes is secreted into the into the blood vessels and then works in a hormonal pathway to impact riedel cells inscri acid and the third is local enteric chromaffin like cells that act in a paracrine manner to stimulate parietal cells and this is really the one we know that impacts each to that h2 receptor antagonists work on they block the history I mean acting on the parietal cells whereas a tap proton pump inhibitors are inhibited downstream by PBS so you bypass that upstream hormonal neuro and petheram pathway which is why theoretically a pee pee eyes are more effective than h2 blockers at decreasing acid secretion increasing pH and thus treatment of gerd and esophagitis so this is some literature a lot of us probably take for granted at least the fellows that that she has worked better than h2-receptor antagonist this is a meta-analysis from 1997 from gastroenterology where they looked at patients both with heartburn and patients with esophagitis who are treated with PP is verse h2 receptor antagonists and you can see that 80% of patients who are treated with a PPI where were free of symptoms and 3-4 weeks at 6 to 8 weeks compared to h2 receptor antagonists who only had about a 50 percent at most improvement symptoms and as well as in esophagitis patients were treated with PPA I have almost 100% healing at 12 weeks compared to HT receptor antagonists where they only had healing at about a 60% of 12 weeks and there's other literature that supports that long term maintenance healing of esophagitis is maintained and from these authors they suggested that a tenet that an n NT of 3/8 of a number needed treated three patients with PPI / hg factors we needed to continue healing after Ed's and their appropriate long-term indications so this is from a group in Italian group of gastrin just general practitioners and pharmacists who put together a lot of appropriate long-term PPI indications led by authors Cara ping nada last year in 2016 they similar to what we mentioned Barrett's esophagus high-grade erosive esophagitis AoE and then some of the other indications that we usually talk about like high risk bleeding patients who end set with history of upper GI bleeding or antiplatelet so the history about the GI bleeding they also commented on inappropriate long-term uses I added this one here but this is no indication so this is patients that come to you and they have no idea why they're taking a PPI someone gave it to them or it's on the hospital discharge they continued it so not only are there in a pro clear clear in inappropriate indications but no indication I think it's a very should I left this for dr. Chapman any comments on the mechanisms of PPI or h2 receptor antagonists that we should discuss probably lasts so they're definitely fact sometimes I wonder why were using each reason we're using more now you're and I didn't I didn't put a graph in here on Barrett's I think for Vaughn I include some of the center is our presentation a couple weeks ago where I think their own the only trials looking at each two receptor antagonists and Barrett's were added on to PPS there's no clear I don't think there's any clear data that it shows up to Robert alone can decrease progression we just don't have progression appearances in that comment and it may still be a negative studies so there's no all right and this is what our patients are coming to us they're concerned obviously we're seeing the literature but they're seeing the headlines talking about bone breaks and dementia and so I wanted to use PPI and CKD and risk of CKD is kind of a framework for how to think about these papers and how we evaluate them especially the epidemiology representation to help us interpret the first paper that I wanted to go over was from 2016 this gets a lot of press and a lot of discussion from our patients PPI is and the risk of CKD so this was out of the Department of Epidemiology at the johns hopkins university by led by dr. Lazarus and what they did was they took two separate populations this was all data that was reviewed retrospectively but there was a prospective cohort where that where the actual data was collected prospectively this was from the a rec study the atherosclerosis risk in communities then about 10,000 10,000 patients that were followed in 1996 to 2001 they also replicated their findings in a large geisinger Health cohort of 250,000 patients this was all data that was collected retrospectively the prospective cohort from this a rig study had the benefit that they were actually looking directly at pill bottles so patients the first visit and at future visits they actually made they actually were seeing what the patients were taking or hearing them read out of the list and in terms of assessing CKD at baseline they had lab values that proved that all their patients had normal couldn't function at the start and then they used incident CKD was determined based on ICD codes so this might seem a little a little bit odd you say how can you really assess whether a patient has CKD based an ICD code there are previous studies that have validated using this that show that there's a high specificity to ICD codes at hospital discharge first CKD upwards of ninety five percent but the sensitivity for picking up some KD based on the NiCd code is quite low I'm all over 30 to 35 percent this was timely so at the baseline of the study only about three and a half percent of patients were on a peep yeah but in the early 2000s PPI used drastically increased men over the course of the study it also drastically increased and this this study showed that their base orbit didn't show but their baseline populations have had differences so some of the main differences and I think this is what concerns a lot of people about these retrospective observational studies is that the PPI users were more frequently white more frequently insured they were fatter they were more likely to have hypertension they were more likely to be on antihypertensives or diuretics and the EGFR at baseline was different among these groups so they performed multiple different different analyses they did multivariate analyses using sex BMI comorbidities medications used and they adjusted they adjusted for all those variants and what they found was that an unadjusted and an adjusted analyses there is an increased hazard ratio of a risk associated with patients who took PPI they also did propensity score matching to attempt to alleviate some of these co-founders and they found that that this association held they also used the negative control by looking at baseline h2-receptor antagonist use course no h2 receptor antagonist use and this did not show us this is really significant increase in CMC KD with patients who are just dying h2 subtitles so what I was trying to show here is that there they basically use multi very analysis that use propensity scoring models they use multiple different techniques to create the highest level of evidence possible here and then in the larger 250,000 replication cohort this was all collected retrospectively so they were looking at prescriptions from prescribers the benefit to this cohort was they actually assessed CKD using I used to the trend in lab values over time I again there were major baseline differences in this larger 250,000 patient cohort which we'll talk about shortly and again with the multiple forms of analysis that they performed they still saw a significant association with CKD there there were other end said yeah I didn't I didn't include it from the baseline and the negative control again did not show and increases association with CKD when you went with h2 receptor antagonists they also looked at twice daily dosing there's once daily dosing and so on increased association with CKD and twice daily dosing and overall incidence CKD was really rare so in the prospective cohort calculated it was point three three percent risk of difference over one year and in the replication cohort of 0.17% absolute risk difference over one year so these are really really small numbers and blue glow incidents per year per patient and again the HT receptor and tigers were not associated with it since the strengths were this was a large population it was replicated in an even larger cohort it was long follow-up and they use multiple sensitivity analyses the limitations are we can't prove causality here so we can't say that there unmeasured co-founder that they're definitely not other unmeasured co-founders they could be impacting the results here I talked about the low sensitive that ICD codes and we can't completely capture over-the-counter prescriptions I just want to briefly touch on this next Peter before we turn to the panel this was from the Journal of American Society of Nephrology from 2016 is a VA study also from the Department of Epidemiology where they were looking not just an incidence CKD but progression of CKD the ESRD also large numbers 20,000 h2 receptor antagonists 170,000 PP eyes again they had baseline differences in age GFR and comorbidities but they did control for GI indication to yourself as we as we would expect more of the patients who have GERD GI bleeding ulcers or Barrett's were on PPS and they were able to adjust for this and again they they they did multiple different analyses but they had different definitions so instead just incident see kitty they define this as to GFR readings less than 60 mm mmm creatinine drop in GFR of over 30% yes Rd so with multiple different definitions of CKD and looking at CKD progression they still show that there's an Associated risk of PPI users versus h2 receptor antagonists and they they trended this over time so they looked at patients who runs who are on PPI is at 31 in 90 days up to the Year over two years and what they saw is that every definition here of how they just had it probably defined at CKD progression over time the risk increased over time although that changed at about the two-year point where there's a drop-off so maybe PPI is our protective after two years or maybe there's a survival benefit that we can talk about yes boots so again right and then again large-scale data multiple definitions here but the same limit limitations possible confounder that we don't know about the associations were significant but very infrequent and this is a VA population so they recommended judicious use of PPI which I also recommended of the Dominions for this discussion limiting the exposure to the minimum dose necessary which I also recommend and they talked about close monitoring of renal function which I'm not sure is necessary and we'll talk about that shortly sure so there's there are many studies in the past that have looked at why they why zucchini might be resolved at PPI use there's there's been studies on hypomagnesemia which is a very rare event but can also be linked to some TV there's been just there's been studies on acute interstitial nephritis and possible idiosyncratic reaction applications again very rare so we don't have a clearly defined I didn't read about that sure so we have dr. Taylor from Department of Epidemiology and I was wondering if she could help us piece apart kind of the strengths and limitations of this study that's the comment of the day but I think a strong card especially first studied its really wealthy but you still see that there are mutations I think the first is the mutation studies that they keep asking the patient with severe sleep in 12 so they have it shut up what's going on because a lot of course studies ask the question once at the beginning and then they would put the follow-up so then they assume what they see at the beginning is what happens over 20 years of time they study actually kept asking what's going on and difficulty is diagnosis – not that strong as a very good diagnosis because they're measuring kidney function so they can catch somebody what's going down exposure because they have to keep the prescription within 90 days so to say that they support to are commenting change in practice the other issue that all the studies that try to have shown so clearly that people who take these drugs are different from the rest because the other issue so even if you can't just for compound drugs and even one of mice clinical trial which is what everybody thinks as the best you're going to create a selection of patients because compared to gross if you need to select them so not necessarily you will have the representativeness of the patients who take the drug so it's difficult it's a difficult topic to solve it with illogically but I think these data considering also that is a dead away the number of cases of of chronic disease which there is all tournaments so it's very even if that means in the fact that 20% is from a rare disease it's not going to try so so that tells me that my concern my concern with reading just the literature is that it doesn't seem like checking periodic kidney function testing doesn't cause either you kind of fall off this cliff or not like just noting that the humanik reading went up a little bit doesn't seem to work I was very surprised at that article the Lazarus article was published with editorial encouraging periodic check in renal function oh they didn't test that periodic checking and renal function produces different results why nobody's moving we can figure out why within sad nobody's trying to minimize it those barrels well barracks and such matter is we can treat and then dr. friebert the AG a clinical practice updates community use the great criteria and rated the quality of these observational studies is generally very low or some of them some low is the was even the best rating that that the committee gave gave these studies but we're still we still seem to be making small amounts of clinical changes in terms of reducing the dose considering cutting half a lot black of indications it's questions and it's not the because the we could get that so my last question for the panel was will where we see higher quality data in the future in order to link cause and he breeds the commented on this whipped up without her lane study with is that that's the best we can expect in the future there's no way to ever do so have a vacation some reason so I want to go back to the pole so the second taste is 77 year old one she has CKD stage 3 symptomatic Gert with a history of le grade C esophagitis in 2014 she's been taking on members all 40 milligrams twice daily this was prescribed at the time their ambassador three years ago and she never decreased it she misses her afternoon doses but never has symptoms and she again asked if she should continue for current vit does but she she hasn't seen the literature so she just asking this one specific question don't go yet hold on one second okay go ahead a B or C all right and we view the results here good so we had already discussed the advance date esophagitis le grade C or D is good notification an appropriate indication for long term PPI use but again this hits the point that we want to reduce to the lowest possible this so in terms of tools out there this is actually from dr. Friedberg who had the this expert review and best practice advice I won't go too much into it designer coming back in January don't wanna steal too much of your discussion what the one table that I really did want to highlight at least a few of his looking at the relative risks and the absolute sxs risk of these different adverse effects and the point the point that I took away from this hopefully trying to make also is that the absolute excess risk is extremely low so I went through this and I just highlighted this and bigger font there was an absolute excess risk of dementia order of 0.07% to 1.5% per patient for a year the risk of bone fracture 0.1 to 0.5% per patient per year versus see death 0% to 0.1% per patient per year so these are really small numbers here and then this is one that just just came out this past September that some people may have already seen and got which is long term proton pump inhibitor and the risk of gastric cancer development so one of the inherent difficulties in studying gastric cancers that explores the biggest confounder so this group tried to try to control for that by taking patients who had known h pylori that was treated successfully and then followed them over about nine years to see who how many developed gastric cancer and the overall risk overall gastric cancer incidence was quite a lot of 0.24% but again they found the hazard ratio of two point four four and patients who are on PPI which I calculated myself as 0.04 percent per patient per year of having gastric cancer after successful eradication of a triple or it see this this is the last case world over this is a 35 year old woman from Hong Kong that was a study out of Hong Kong with a history of age Gloria gastritis who was successfully treated with quadruple therapy her symptoms have resolved yes she could continue for PPI and go ahead and answer this all right good and we'll show this one was obvious discontinue PPI does beginning at the end of the presentation discontinue PPI information Laurie present the eradication and this hit this hit the third main point that I was trying to make that we should just continue ppi win no indication and also we just discussed so again this is a tool that we that we can all use dr. friebert wrote this AG a clinical practice update when talking to patients it's a good guide and here it's a little bit too small to read but there's 10 best practice advice generations that don't want to go over it explicitly because with that I leave that for you in the future but the main points are that it aligns aligns with some of the indications that I talked about front of my time you're good with the acid complications Barrett's esophagus high risk for GI bleeding and also that for other tools we go to a gastro battle occurred whether there's a constantly updated for patient education and provider education which it seems like each new study is critically evaluated so that we can have talking points for our patients I get very much panel at anything else that let's go thank you

Leave a Reply

Your email address will not be published. Required fields are marked *