AIDS therapy with two, three or four agent combinations, applied in short sequences, differing from each other by drug rotation. I. First of two parts: a phase I trial equivalent, concerning five virostatics: AZT, ddI, ddC, acriflavine and an ellipticine analogue.

We have individually treated ten AIDS patients whose CD4 numbers were inferior to 200/mm3, with the five following HIV1 virostatics: a) azido-deoxythymidine (AZT), dideoxyinosine (ddI) and dideoxycytidine (ddC), which affect the same viral target, retrotranscriptase, b) acriflavine (ACF) and methyl-hydroxy-ellipticine (MHE) which we have discovered to be strong virostatics in vivo, in mice, against Friend's virus, and in man, against AZT resistant HIV1. We have shown that their combinations with AZT, hitting three viral targets, reduces in mice, the blood Friend's virus load below detectable level. Due to the short doubling time of HIV1, AIDS therapy must be continuous, and to allow the best tolerance, the five virostatic combinations were applied in short, three-week sequences, each differing as much as possible from the former and from the following one, due to drug rotation [1]. Among the ten patients, a) three received the two-drug combinations for 15 to 30 months, followed by the three-drug combinations, b) three received the three-drug combinations from the beginning, c) four received the four-drug combinations also from the beginning, two having less than 10 CD4/mm3 at initiation of treatment, and two having more than 100. The tolerance was remarkable: the only side-effect being macrocytosis. The application of the two-drug combination sequences maintained stable CD4 levels in two subjects whose viral load (the evaluation of which had became available) was, at the end of this period, of 4,486 and 39,238 RNA copies. The third subject who had received, an intensive UV irradiation for a psoriasis, presented an irreversible decrease in his CD4 count and a high viral load (1,352,495 RNA copies/mL) at the end of the two-drug period. Fifteen to 25 months after the shift to the three-drug combinations, the viral load decreased, from 39,328 to 13,291 in one of the non-UV irradiated subjects, and from 1,352,495 to 314,387 in the irradiated one. No subject had an increase in CD4 number. In the three patients having initially received the three-drug combinations, a very strong decrease of viral load was registered after periods of observation varying from 77 to 40 months, while the CD4 counts increased moderately in two subjects, and noticeably in the third (from 126 to 266). Out of the four subjects initially treated with four-drug combinations, the two with less than 10 CD4/mm3 had a moderate decrease in viral load in about three months, and the CD4 increased from 9 to 34/mm3 in one. But the two subjects, because of opportunistic infections and psychological reasons, abandoned their treatments. In the two subjects who had more than 100 CD4/mm2 at initiation of the four-drug combination treatment, the viral load decreased to undetectable levels after four months: but their CD4 counts, after some oscillations, had very moderately increased at the end of the observation period (respectively, from 200 to 222, and from 129 to 134). In practice, these results suggest the interest of conducting phase II or III studies of AIDS treatment protocols, starting with the four-drug combination model, and attempting to maintain the effect with the three-drug combination one. As for theoretical considerations, one must underline the contrast between the remarkable reduction of the viral load and the usually moderate increase of the CD4 counts. The study but not the trial has been interrupted, due to the unavailability of three antiproteases, saquinavir, ritonavir and indinavir, which are now introduced in the same type of combinations, one by one, in replacement of one of the studied agents as shown in figure 1. The effect of increasing the total number of virostatics from five to eight will be published in the second part of this article series.