Like many people I know, I started antiretroviral therapy shortly after the Vancouver conference in the autumn of 1996. With T-cell counts that had see-sawed between 400 and 500 for nearly a decade, I'd always sort of been on the cusp of start/don't start. It was the three-years-to-eradication talk, which began early in the spring of 1996 and then reached its frenzied crescendo by late summer, that tipped the balance for me in favor of starting. After I'd taken my first dose I remember wandering the streets of the East Village, zombie-like, imagining the end of AIDS. Even if the big talk didn't pan out, at least I would give my immune system a break. And so that was the deal I made with myself: jump on the "It's the virus, stupid" bandwagon, see where it leads, give it a couple of years, then reassess the situation.
Contrary to the neatly organized stories of viral "set points," my viral load measurements had always been all over the map. Eighty thousand one month, a hundred and fifty the next; one hundred thousand in the spring, one hundred and eighty in the fall. Half a year of monthly DNCB paintings had appeared to bring my viral load down to 50,000 just before I started taking the pills. (No, there wasn't a control group.)
At that time there were only three protease inhibitors available. Ritonavir's reputation terrified me, and saquinavir seemed like protease-lite. Given the alternatives, indinavir seemed the safest bet. And I liked the twice-a-day schedule of d4T and 3TC, so that's where I began. Yet I under-estimated how intrusive the fasting requirements and thrice daily indinavir dosing would be in my life. I was obsessively compliant but was driving my boyfriend crazy. I lived my life by the ticking hands of the clock.
My viral load fell below 400 copies within a month. A few months later, when I changed doctors in order to gain access to the ultra-sensitive 50 copy assay, my virus was still detectable -- and took nearly six months to go below 50. My T-cells never really budged; five or six hundred was the most I ever got.
A couple months into the combo my skin and lips began to harden, dry and crack -- to the point where it was as if the skin from my elbows or heels had overgrown to the rest of my body. Under the sheets it was like sharing a bed with a pint-sized pachyderm. "The skin is the body's largest organ -- and a window to the immune system," I remember my dermatologist friend telling me. If this is what Crixivan is doing to my skin, I mused, god only knows the effect it's having on my insides! My boyfriend urged me to jump ship. He'd been one of those ritonavir people who, incredibly, had suffered no ill effects at all, not even the first few weeks. "Only twice a day -- and with food," he lobbied me to join him, but I declined.
Nelfinavir came along late that winter -- and I jumped. What a relief it was to have no more fasting requirements. Even for an urban gay man living on the outer reaches of a Barbie doll six-pack culture, planning for a full stomach was much more exciting than contemplating its opposite. Apart from the early diarrhea and the chronic loose stools, nelfinavir was a veritable vacation compared to indinavir. Some months later I would talk with friends who had also made the switch. We shared stories of shitting our pants in the middle of Manhattan, no time to search out a toilet. (Imodium anyone?) Wasn't life on Viracept grand.
The year 1996 had not even come to a close when researchers began explaining why the Perelson/Ho predictions were a fantasy. Memory T-cells would pose a significant hurdle, as would the existence of viral reservoirs untouched by treatment. The Diamond team's assumption of 100% viral shut-down was also to prove naive. As experiments at Hopkins, the NIH and UC-San Diego proceeded, the three-year cure timeline had been quickly extended to 60. The bandwagon, it seemed, had encountered an impasse.
In the spring of 1998, faithful to my earlier bargain with myself, I bid my triple combo farewell and opted for an experimental foray with a less intensive "maintenance therapy" regimen of ddI and hydroxyurea. ("High genetic barrier to resistance," and all that.) Never mind that the handful of such trials (ACTG 343, Trilège, ADAM) were not panning out. Most of them had been poorly designed, almost as if they had been set up to show failure. None of them had included hydroxyurea as part of the maintenance regimen. Little by little my viral load became detectable again. It plateaued at around 1,000 copies and remained there for many months.
Over the course of the past three years I have gone on and off my ddI/hydroxyurea bitherapy for increasing periods of time. First just a week, then two; later for a month, then longer. Each time I would test my viral load obsessively and go back on my souped-up monotherapy when the virus threatened to break out. Strangely, it always seemed to rise at a snail's pace. Emboldened, I began to take longer treatment breaks: two months, three months, five. But I'd grown complacent: my RNA results came back at 180,000; CD4s hadn't budged. I pulled the pills back out for a couple of months.
I am currently logging the progress of my longest time off therapy since this experiment began some five years ago. My March 2001 viral load, nine months off therapy, was a mere 12,000 copies; my CD4s, at 660, were higher than ever -- even on the protease triple combo.
Of course, it's possible that with only a moderately high baseline viral load I hadn't needed a triple combo in the first place. (Remember though, back in 1996 a mere 20,000 copies was considered immediately life-threatening, thanks to the pip-squeak arriviste John Mellors and his poorly preserved plasma samples from the MACS.) Perhaps ddI/HU would have been sufficient from the start. My doctors and I conceded that possibility in a letter to The Lancet (3 October 1998, 352:1149). The dinitrochlorobenzene pre-treatment spree was omitted from the Lancet letter. Why risk the ridicule?
Now, of course, we learn there may be individual characteristics -- genetic makeup -- which pre-determine the course of one's HIV infection. Chemokine receptors mutations (e.g., CCR2, SDF-1) and HLA subtypes. I didn't even know what HLA was, and yet researchers now talk of "advantageous" and "disadvantageous" alleles. (See "Human Leukocyte Antigens (HLA) and HIV Disease Progression" in this issue.) With all the progress the field has made, it boggles the mind to think how much we still don't understand.
My physician friend in Berlin talks of how shamefully crude our methods are some fifteen years into HIV care. The fact that we are still relying on CD4 cell counts and plasma viral loads in 2001 seems absolutely barbaric to him. Will we, a year or two from now, be routinely measuring patients' viral diversity, the breadth of their CTL response, cytokine signaling and antigen presentation?
I performed my own maintenance therapy experiment because I was less afraid of developing resistance to ddI than to a protease inhibitor or a non-nuke. I also wanted to thumb my nose at an oligopolistic biomedical industrial complex that delights in the stranglehold it exercises over the AIDS research agenda.
Who knows what my counts will be when I test again in a few weeks. Maybe it will be time to haul out the Videx again. It's even possible that one day ddI's high but not nonexistent genetic barrier will be reached, and I'll have to re-think this entire strategy. For the time being, though, I will keep on experimenting, "N of 1" or not. Sometimes self-experimentation's all a guy's got.