HIV and ID Observations

I’m currently on the inpatient consult service and just saw a guy who fits the typical profile of many hospitalized HIV patients in 2012:

• Low CD4 (in this case, 120)
• Irregular to non-existent outpatient care before admission (lots of no-shows, cancellations, etc)
• Has received several prescriptions for antiretroviral therapy but for a whole variety of reasons, hasn’t been taking it

The medical specifics are otherwise unimportant — he didn’t have an HIV-related reason for being in the hospital. Instead, I want to focus on a question from the resident caring for him as she prepared his discharge papers:

Hi Dr. Sax, question on Smith — you mentioned he should restart HAART [ugh, I didn’t say HAART], but not Bactrim for PCP prophylaxis. Should we add that to his discharge meds?

Now the textbook answer is clearly yes, anyone with a CD4 < 200 should receive PCP prophylaxis, and that’s what these fine guidelines would say.

But I deliberately didn’t include it, for two key reasons. First, what this man needs to do is take HIV therapy, and I wanted the regimen to be a simple as possible. Why clutter it with that giant Bactrim tablet?

Second, assuming we can actually get him on ART, do we have any evidence whatsoever that primary prophylaxis for PCP is still necessary? All the studies of PCP prophylaxis were done way before we had effective HIV therapy — in fact, this one (for you history buffs) was done in the mid-1980s, before we had any antiretrovirals at all.

I posed this question to OI Guidelines guru John Brooks, who answered the following:

A randomized trial to address the question (i.e., PCP incidence among persons starting ART at CD4 <200 with vs. without Bactrim) would be ideal, but I would bet the number of participants required to demonstrate no difference in risk would be enormous, especially  since (we hope!) folks would remain “at risk” for only a short period of time … As you probably know, cohort studies have tried to address the issue; I think only the Swiss Cohort has been able to successfully complete an analysis. We have tried with HOPS data, but incidence of the key prophylaxed OIs (PCP and MAC) was so low recently that we can’t get enough endpoints!

And that bolded statement right there is exactly my point. (The emphasis is mine, but John included his own exclamation point.) Effective HIV therapy drops OI incidence so sharply that prophylaxis is probably not necessary, and certainly is much less important than ART.

So if you get the question on the ID boards — should someone with a CD4 of 120 be on PCP prophylaxis? — the answer is yes.

In real life, however, I’m not so sure it’s still the right thing to do.