2, 4, 6, 8, Who's Afraid to Phosphorylate?

Factors other than viral cross-resistance, such as pharmacological mechanisms, may also be important to take into consideration when planning a long-term strategy. The nucleoside analogs are actually prodrugs that must be converted into their active, triphosphate form by cellular enzymes in the human body. This is a three-step process.

Articles appearing in two journals last summer described problems with this process that limit AZT's clinical efficacy (A. Lavie et al. Nature Medicine. Aug. 1997; 3(8):922-4; A. Lavie et al. Nature Structural Biology. Aug. 1997; 4(8):601-4). After AZT enters the cell, it is efficiently converted (or phosphorylated) to AZT monophosphate (AZT-MP). A bottleneck occurs at the second phosphorylation step, causing a buildup of high concentrations of inactive AZT-MP, but little of the triphosphated form (AZT-TP). AZT-MP hinders the normal activity of the cellular enzymes and inhibits the conversion of AZT to its active form. This inability to achieve a sufficient intracellular concentration of AZT-TP leads to incomplete suppression of HIV, even in the absence of high-level drug resistance.

There is some controversy as to whether the rate of phosphorylation decreases after prolonged exposure to a particular drug over time and if prior use of one compound will affect the subsequent ability of the cell to phosphorylate new drugs. At a late-breaker presentation at the this winter's Retrovirus Conference, David Back, M.D., of the University of Liverpool and North Manchester Hospital in England, reported preliminary findings on 10 patients receiving treatment with AZT/ddI or AZT/3TC over the course of 12 months (late-breaker 10). No evidence of systemic decreases in intracellular phosphorylation of AZT over time or difference between treatment-naïve and -experienced patients was found. However, Dr. Back noted that AZT-TP levels were very variable. One AZT-experienced patient had no measurable AZT triphosphates at months two and three, with a corresponding increase in viral load. Another patient had been heavily pretreated with AZT, initially on monotherapy and then on combination therapy. He was still phosphorylating reasonably well after nearly five years of treatment. The ten patients Dr. Back studied were responding to AZT treatment. Results might have been different had he looked at AZT-treatment failures.

Jean-Pierre Sommadossi, Pharm.D., of the University of Alabama, recently completed in vitro research on six patients who failed AZT therapy after long-term treatment. In a small study called RETROPHAR, cells from AZT-naïve patients and cells from AZT-treatment failures were exposed to d4T and AZT to measure phosphorylation. There was a four- to five-fold decrease in both AZT and d4T phosphorylation in the treatment failures.

At the 1998 Retrovirus Conference, Dr. Sommadossi presented data from ACTG 881, a substudy of participants from ACTG 290 who were treated with a combination of AZT and d4T (abstract 3). ACTG 290 randomized patients with prior AZT experience to four treatment arms: ddI alone, AZT/ddI, d4T alone or AZT/d4T. There was little or no antiviral effect seen in those treated with AZT/d4T, and CD4 cell counts decreased (see table 2). The objective of ACTG 881 was to reveal any underlying pharmacological causes for this poor response.

The possibility of an unexpected drug reaction between d4T and AZT causing malabsorption was ruled out. There was no difference in d4T blood levels in six patients treated with either d4T alone or in combination with AZT. In contrast, the intracellular concentrations of the active triphosphate form of d4T (d4T-TP) were reduced in patients receiving the combination (although only four patients were studied). Patients treated with d4T/AZT had six to seven times less d4T-TP than patients treated with d4T monotherapy. Dr. Sommadossi concluded that these results might explain the reduced virologic response to the combination of AZT/d4T.

Dr. Sommadossi also examined the effect of d4T/3TC use in patients who had failed long-term AZT-based therapies. He presented results from ALTIPHAR, a pharmacological follow-up study of the ALTIS 1 and 2 trials (poster 362). The ALTIS studies observed that treatment-naïve patients experienced a ten-fold better response to d4T/3TC than treatment-experienced patients (prior use of AZT/ddI or AZT/ddC), which could not be explained by genotype analysis. Dr. Sommadossi was able to correlate intracellular phosphorylation with antiviral effect in most cases. He found that the cells from AZT-experienced patients (termed "non-responders" due to less than a one log viral load decrease on d4T/3TC therapy) had a lower level of phosphorylation than cells from AZT-naïve patients (termed "responders" due to a one log or greater viral load decrease on d4T/3TC therapy). Long-term prior use of AZT apparently reduced the subsequent ability of cells to phosphorylate d4T and 3TC. However, it is still unclear whether it was AZT specifically, or some other factor, that caused reduced intracellular phosphorylation/activation of d4T and 3TC. Dr. Sommadossi stated, "We have to go to a much larger scale to understand if indeed there was a direct effect of AZT on the phosphorylation of other nucleosides by elucidating the underlying molecular basis of this effect. We need to demonstrate the relevance of the individual data with in vivo results obtained from controlled clinical trials."

Multivariate analysis did not show any effect of AZT, ddI or ddC resistance mutations on HIV RNA response in the ALTIPHAR participants (V. Calvez et al. 1998 Retrovirus Conference poster 676). Every patient studied (both the responders and non-responders) had the 3TC (184) mutation after 24 weeks of treatment. The poor antiviral effect in the non-responders looked similar to that seen with 3TC monotherapy. Dr. Sommadossi pointed out that while d4T-TP levels were decreased in these patients, he could not conclude that their poor response was because d4T was having no effect at all. He felt that the emergence of the 184 mutation was inevitable, regardless of phosphorylation levels, as dual nucleoside combination therapy usually allows for viral replication that inevitably leads to the development of viral resistance.

Dr. Sommadossi cautioned that the data are not black and white. "All we can say is that in situations like ALTIPHAR and ACTG 881 this is what occurs. To find the key to drug failure we need to look at both the viral genotype and phenotype as well as drug exposure. Adherence will also play a major role in success or failure." Dr. Sommadossi will continue his research by compiling data on patients who have had prior d4T use. d4T is metabolized into its active form by the same enzymes as AZT but because it has a different molecular structure, it has more difficulty with the first phase of the phosphorylation process than the second. It is, as yet, unclear if patients who start on d4T will have a reduction in the subsequent phosphorylation of AZT or other nucleoside analogs.

In a presentation at the Retrovirus Conference in Chicago, Charles Flexner, M.D., of Johns Hopkins, warned the audience to "beware of pharmacologists bearing small data sets." While acknowledging that the decreased levels of d4T phosphorylation would be an attractive potential explanation for d4T treatment failure (since data linking patients with inadequate d4T responses to d4T resistance mutations are lacking), he said "I think it is premature to try and incorporate these findings into clinical practice until we understand them further." (lecture S47)

Calvin Cohen, M.D., of Community Research Initiative of New England is, nonetheless, starting to look to this early data when considering the best possible "Plan A" for a patient that will also maximize the success of a "Plan B," should one be needed. "The metabolism of d4T is altered by AZT exposure. The one key unknown is whether this is bi-directional (in other words, whether d4T affects subsequent AZT use as well as AZT affecting d4T). In theory you would try and use d4T first, assuming it's an equally effective drug. We have to make sure the drugs are equally potent and you're not going to increase risk of failure by substituting d4T for AZT. There were studies presented in Chicago that suggest they are fairly interchangeable." (See "Reevaluating Initial Therapy" in this issue)

Table 2: ACTG 290 and ACTG 298 Results

ACTG 290 (48 weeks)				ACTG 298 (48 weeks)
Baseline characteristics				Baseline characteristics
Median CD4: 401/mm3				Median CD4: 407/mm3
Median Viral Load: 6,300 copies/ml				Median Viral Load: 10,000 copies/ml
Number of Subjects: 144				Number of Subjects: 128
Group	Mean Change in CD4	Mean Log Change in Viral Load	%BLQ*	Group	Mean Change in CD4	Mean Log Change in Viral Load	%BLQ*
AZT/d4T	-45	-0.01	0%	AZT/d4T	+14	-0.33	8%
d4T	+20	+0.13	6%	d4T	+65	-0.47	31%
AZT/ddI	+5	-0.46	36%	AZT/3TC	+101	-0.91	38%
ddI	+70	-0.44	46%
*Percent of participants below the limit of quantification, i.e. 500 copies/ml, using the Chiron bDNA assay (equivalent to 1500 copies/ml by PCR assay) All participants in ACTG 290 were AZT-experienced with a median duration of 33.6 months of prior AZT use. In addition, 12% of the participants had some previous ddC treatment. ACTG 298 participants were treatment-naïve. At the time these companion studies were designed, nucleoside analogs were the only commercially available antiretroviral treatment. Many patients had been treated with AZT monotherapy and prior trials suggested that these patients would benefit from switching to a ddI-based regimen. d4T was also becoming available, but the activity of ddI and d4T had not been compared in AZT-experienced individuals. In addition, AZT/d4T was being used in clinical practice without any efficacy studies to back this combination. There was concern that AZT and d4T would actually prove to be antagonistic due to early lab tests. Data from 290 and 298 suggest that AZT/d4T does not have a more potent antiviral effect that d4T alone. The combination is associated with CD4 cell decline in AZT-experienced patients. ddI-containing regimens produced greater viral suppression than d4T in AZT-experienced patients. AZT/ddI was not associated with greater antiviral activity than ddI monotherapy in AZT-experienced patients. In treatment-naïve patients, d4T alone has better antiviral activity than the combination of AZT/d4T and AZT/3TC was more effective than both d4T alone and AZT/d4T. Diane Havlir, M.D., from University of California at San Diego, presented these data at the 5th Conference on Retroviruses and Opportunistic Infections (abstract 2).