Putting the Risk in Perspective

I would like to go through the process of evaluating this potential risk, to put it in perspective, and discuss how you might potentially respond to this warning for your patients taking PPIs. The way I evaluate this is to look at the scientific hypothesis -- what is the biologic plausibility, what is the evidence base, what does the literature say, and what do I see in my patients? So, it is a 4-step process for evaluating the risk/benefit profile of any pharmacologic or intervention strategy. Let's walk through the process for PPIs and CDAD and see what we can garner from this.

The biologic plausibility. The scientific hypothesis is that the gastric acidity protects the gastrointestinal tract from ingested bacteria. A gastric pH less than 4 is fairly profoundly bacteriocidal, so most things do not survive that level of gastric acidity under normal conditions. The scientific hypothesis is that if you impair the gastric acid by means of an acid-suppressive agent, that you may diminish the protective effect of the gastric acidity and thereby raise risk for gastrointestinal infection. The idea that PPIs and even H₂-receptor antagonists may do this has been reported and evaluated in the literature.^[2]

The scientific hypothesis is that gastric pH diminution may pose increased risk for infection by loss of the bacteriocidal barrier of the gastric acidity. C difficile has 2 forms. One is a vegetative form, and this is acid-sensitive, so, under a normal acidic pH, the vegetative form is fairly quickly killed by the normal gastric acidity. The other is a spore form that is not killed by the normal gastric acidity, and can potentially transit [through] the gastrointestinal tract and pose risks. These spores can turn into a vegetative form once they get into the small intestine and colon. So the biologic plausibility is that the diminishment of acidity may affect the ability to kill the vegetative form, but not the spore form, which is acid-resistant.

To add one more factor to this, we do know that PPIs have an effect on white blood cell function. In vitro studies suggest that there may be alterations of chemotaxis, phagocytosis, and expression of adhesion molecules. This in vitro evidence suggests that PPIs (acid inhibition) may alter white cell function. The biologic plausibility, therefore, makes some sense.

Weighing the evidence. The scientific evidence is very controversial. I found 27 articles in the world literature to date that have looked at the association of C difficile and PPI therapy; 17 of these have suggested harm/potential risk. The risk ratios range from 1.1 to nearly 5 with respect to the incremental risk for C difficile in patients taking PPIs. None of these are prospective trials. Ten studies showed no associated harm. Four of the 17 studies that showed harm suggested an association between recurrent C difficile infection and PPI use, and 3 studies showed an increased risk with higher than standard dose PPIs. So the scientific evidence is fairly mixed, but, clearly, there is more evidence of an association than of a lack of association. Recognizing that these are all retrospective studies, it is very difficult thereby to risk-stratify. These patients are perhaps sicker and they have multiple exposures. All but 5 of these studies were hospital-based, so these were not outpatients. Extrapolation to the normal population of outpatients that you and I see in our clinics is hard to do on the basis of the available evidence.

Assess your practice. The evidence is fairly mixed, so I return to the fourth tenet, which is, what do I see in my practice? When I see a patient with C difficile, do the alarms go off that this patient is more likely to be on a PPI? I don't necessarily see in my practice that the patients who are positive for C difficile are more likely to be on a PPI. I would ask you to do the same analysis in your practice to see if that plays out.

Talking With Patients About PPIs

How do we put this in perspective? The news is out there, so it potentially needs to be discussed with patients. If you see a patient with diarrhea who is on a PPI, if they have risk factors, you should screen them for C difficile. It is always a healthy point to reanalyze, and consider: does this patient need to be on a PPI?. Many of these patients are on therapies that they don't need. However, in patients who are taking the PPI for nonsteroidal antiinflammatory drug (NSAID) prophylaxis, severe gastroesophageal reflux disease, or recurrent upper gastrointestinal bleeding, clearly the risk ratios are in favor of continuing that PPI. Nonetheless, reevaluate, potentially stop, or reduce the dose if you can, or change to alternative therapies if appropriate. It is always a healthy discussion to have with patients, and we talk about multiple risks associated with PPIs: bone fractures, hypomagnesemia, the clopidogrel interaction. It gives us the opportunity to discuss whether they really need the medication.

The International Society of Travel Medicine warns that PPIs may increase risk for multiple enteric infections and traveler's diarrhea.^[3] So, recognize that PPIs have received a lot of press with respect to enteric infections, the most recent of which is C difficile. The evidence base to support this is, I think, very tenuous at best, and very difficult to analyze using retrospective studies. If we wanted to study this in a prospective way, it would be helpful to also nest a cohort of achlorhydria patients (those with vagotomies or gastric surgeries, following which they make no acid, patients with pernicious anemia, or atrophic gastritis). Those patients would be the ones who would be potentially at greatest risk, with no acid production. At the best case, PPIs may prolong gastric pH reduction to a pH higher than 4 for up to 18 hours. We never make these patients achlorhydric or anything close to it.

Beware of the new evidence, and that this is something that has now been reported by the FDA in the United States and by Health Canada as well. Put this in perspective, look at the evidence base, and, hopefully, this will help you as you begin to discuss these new alerts with your patients. I will leave this to your best clinical judgment and I look forward to chatting with you again soon. I am Dr. David Johnson.

# posted by roodbont @ 3:29 PM