Sunday, January 21, 2007

AIDS drugs cause AIDS I

Yesterday, I was puttering through the hospital when I came across a J, pathologist friend of mine. Whenever I see a pathologist, I always ask to see whatever slide they're working on, which usually results in an interesting story.

J told me about a case from his residency in 1994. A 20 year old man presented to the morgue with multiple Kaposi's Sarcomas in his GI tract. KS is a cancer of an unknown, probably blood vessel cell which is caused by the Kaposi's Sarcoma Herpes Virus (KSHV, HHV8). Prior to 1981, the stereotypical patient with KS was an 85 year old Italian male.

In '94 the stereotypical patient was this patient, a young gay man. Since the cancer was in his digestive tract instead of the skin, he didn't have a chance to get treatment.

Not that there was much treatment.

J kind of breaks off and says, "We don't really see cases like that anymore, not since '96."

Boards question:
A 37 year old male heroin user has noticed multiple 0.5- to 1.2-cm plaque-like, reddish-purple, skin lesions on his face, trunk, and extremeties. Some of the larger lesions appear to be nodular. These lesions have appeared over the past 6 months and have slowly enlarged. The most effective treatment for his condition is:

A. Vincristine, irinotecan, and cisplatin.
B. Valcyclovir, gancyclovir, and foscarnet.
C. Indapamide, metoprolol, and prazosin
D. Zidovudine, lamivudine, and efavirenz.
E. Ticarcillin, clavulanate, and gentimicin.


Correct answer: D. The common thread in all of this is the Human Immunodeficiency Virus, and the cocktail indicated is an ideal treatment. The release of protease inhibitors in 1996 meant that the number of patients with full blown AIDS and AIDS defining illnesses cratered, hence the dearth of KS biopsies.

This brought to mind a story on about South Africa's stand against providing AIDS drugs to AIDS patients. For instance at an international HIV/AIDS conference, the South African booth featured that infamous reverse transcriptase inhibitor, the African Potato. Anyway, part of the South African government's argument is that, as the title of my piece indicates, AIDS drugs cause AIDS.

How can we reconcile J's observations with this claim?

Prior to 1987, there were no AIDS drugs in general circulation. One can only conclude that gay men, injection drug users, and hemophilliacs were secretly stealing AZT and snorting it starting in the mid 1970s. Of course the world's known supply was sitting on some lab shelf in Research Triangle Park. The only reasonable explanation is therefore that the Columbian cartels must have stolen the structure and synthesized a bunch of the stuff, and used it to cut the blow.

Explaining the sudden drop of KS after 1996 is relatively trivial. As it turns out, AZT causes AIDS, but the 3TC and Efavirenz are actually antidotes, so as long as people take them as a cocktail, nothing too horrible happens.

Essential resources:
AIDS reappraisal - Wikipedia (it's almost painful how hard they worked to keep this article neutral)
A little piece of history. Check the date.
Hopefully I'll get to address other aspects of the 'reappraisal' in a later piece.

Among my many flaws:

I: I seem to be missing the gene that makes you want to go to a bar, have one drink, and say, "Hey, let's go to a different bar!

She: SRY?

Thursday, January 04, 2007

A problem possibly solved

Yesterday we started anemia, particularly the iron, B12, and folate deficiencies. It reminded me of a patient I saw last year who had a very confusing constellation of symptoms, but which I now realize were all anemia caused by a rash that had appeared on his chest 3 years earlier. The rash contained, presumably, rapidly proliferating skin cells and immune cells, which would suck of the folate and B12, both of which are involved in DNA synthesis. This was compounded by the patient's poor eating habit. Now if I could just figure out what the autoimmune condition was.

Something sort of similar is maybe in this paper.

Problem Solving Skills Dos

We're doing CPP&T this quarter, which is the part of medical school where they actually teach medicine. Each lecture takes the following format: "Here's a disease, here's how you get it, here's how you recognize it, here's how you treat it." Then we have lab twice a day, which is a series of cases where patients have the illness and we have to establish differentials and suggest labs and such.

We're also taking Physical Diagnosis, part of which is to run around the hospital, interviewing patients and doing as much of the physical we have learned, then trying to do diagnosis and present them to an attending physician.

It occurred to me today that medicine is very much like school problem solving - you are presented a patient/problem, you attempt to solve it in a stereotyped way based on practice problems of a similar appearance. You either get it right or wrong, but either way, you quickly move on to the next problem. If you do a good or bad job overall, it will definitely matter, but success on one problem, or even one type of problem is not necessary.

This is unlike other areas of work, which I can't think of right now, where either you can't move on, or all your problems are interrelated in ways that are simply impossible to entangle.

Anyway, that makes medicine comparatively comfortable, as opposed to, say, research, where you may be stuck on a problem for years, never knowing if you have the right answer.