Case Discussion: A 28 year old man with cough and fever, two weeks after starting Antiretroviral Therapy.
Sorry to Rafa and the other bored-by-medicine folk, but the comments on this case are really too good to leave hidden in the comments section of that last post. For those who missed them, Doctora Lola (Pulmonary Critical Care Specialist and ICU Director at one of NYC´s largest medical centers) said:
The differential is broad, although several elements of the history stand out to me as important; first he had 3 negative smears for AFB upon his initial diagnosis, second he had presumably been taking is PCP prophylaxis, third, he does not sound acutely ill on presentation to the MSF clinic, and lastly his symptoms began a couple weeks after the initiation of antiretroviral therapy. His CXR is certainly abnormal, with lower lobe predominant bilateral interstitial and micronodular infiltrates (one could use the ubiquitous radiologists term "reticulonodular infiltrates"). While the image quality is not perfect, he doesn't appear to have significant lymphadenopathy on the film. This radiographic pattern in an HIV patient is non-specific and could be many things, both infectious and non-infectious. Given his clinical course, infectious causes seem most likely, particularly with a possible element of immune reconstitution with him recently starting HIV meds. The infections I think are most likely are 1. PCP (despite prophylaxis), 2. TB, 3. Community acquired pneumonia 4. MAI (CD4 is a little high). Non-infectious etilogies to consider (although much less likely) are lymphocytic interstitial pneumonia, hypersensitivity pneumonitis and drug reaction.
Doctora Magda (Infectious Disease Consultant and HIV Researcher at another NYC mega-medical center) said:
I agree with Lola and think that IRIS is a real possibility, although his initial CD4 cell count was fairly high at 160. The risk of immune reconstitution increases with lower CD4 cell counts at the time of ART initiation. Of course, there are never any absolutes. I was curous whether you rechecked his eosinophils at his second presentation - I would also add strongyloides (hyperinfestation) to his ddx
Amazingly, the Puerto Barrios Community doc had arrived at the same differential diagnosis as the New York academic specialists! [Perhaps not all that amazing afterall, as I learned almost everything I know from these two bad-asses.)
I thought that infectious etiology was most likely, given the patient´s low grade fever and cough, along with the CXR results. The reticulonodular pattern on CXR was most suggestive of PCP, and the patient´s CD4 count of 160 would be consistent with this diagnosis. However, the pateint insisted that he had been taking his Trimethroprim, and I believed him. This would make PCP much less likely. Even though the patient had been AFB negative x 3 prior to starting his antiretrovirals, we all know that AFB smear is only about 50% sensitive in HIV positive patients with TB. I thought the likelihood of immune-reconstitution syndrome (IRS) with recognition of a previously undiagnosed TB infection was a real possibility. Of course Dr. Magda´s comment that IRIS is most commonly seen in patients with very low CD4 counts crossed my mind, making this diagnosis a bit less likely in our patient with CD4 of 160. I did not think about strongoloides hyperinfection. It is a great thought though, because we see a lot of strongoloides here (stool smears for parasites are one of the few strengths of our lab here in PB), and immunosuppresed patients are at high risk for hyperinfection.
So what is Immune Reconstitution (IRIS) anyway? I am no expert. My understanding is basically that as AIDS patients replenish their CD4 counts and their immune systems recover, the T-cells start fighting infectious that they previously did not have the energy to notice. A 2004 review in Clinical Infectious Diseases describes IRIS as:
an inflammatory syndrome occurring days to months after the start of ART, with outcomes ranging from minimal morbidity to fatal progression. This syndrome can be elicited by infectious and noninfectious antigens. Microbiologically, the possible pathogenic pathways involve recognition of antigens associated with ongoing infection or recognition of persisting antigens associated with past (nonreplicating) infection.
About 25-50% of IRIS cases are thought to be due to mycobacterial disease that was subclinical prior to starting antiretroviral therapy.
So what happened with the patient? I started him on empiric treatment for PCP as an outpatient and told him to drop off as many sputum samples as possible at the Centro De Salud for AFB analysis. He came back the next week with three samples +++ for AFB. I stopped the PCP treatment and told him to go back to the Centro de Salud STAT to get TB treatment. He is now on four drugs for tuberculosis and continues on Anti-retrovirals.
Thanks Lola and Magda! I miss you.