you might not like the pills, but the alternative seems worse
Written for NATAP by David Margolis, MD, University of North Carolina
This session at CROI highlighted the results of an array of studies that will probably require extensive analysis over months to come. Treatment interruptions (TIs) were tested or observed in a variety of patient populations. In many clinical settings, one has no choice but to perform a TI. But in reporting of TI studies from CROI and other meetings in the past, the point has been made by this author and others that the term "STI" is premature, as a clinically proven "strategy" for TIs that is superior to continuous therapy (CT) has yet to be defined. The overall opinions that this author came away with from the array of studies were:
1) Current guidelines are borne out. You are better off on continuous therapy, if therapy was begun in the context of current guidelines.
2) A TI can be done for toxicity or intolerance, but the optimum technique to allow a TI without limiting future treatment options requires further study. TIs carry risk and should only be done to reduce clear and present harm caused by therapy. Given the wide array of therapies, one might hope that if antiviral treatment is indicated, a tolerable therapy can be found.
3) The idea of cycling on and off HAART driven by a CD4 count threshold is clearly inferior if that threshold is 250 cells/ul. And studies with other thresholds have not been carried out far enough to be certain that the cost is less than the benefit.
4) There may be an increased rate of morbid events, particularly cardiovascular ones, in patients who are off HAART. If this proves to be true, we must find the cause. One possibility cause is simple inattention to routine healthcare maintenance, but the idea that secondary effects of chronic viremia and immune activation lead to specifically increased risk is biologically plausible.
Note from Jules Levin: another possible reason suggested for an increase in adverse events after HAART interruption - HIV increases inflammation & HAART reduces inflammation. After stopping HAART inflammation may return.
And now to the reports.......
Stopping HAART if you never needed it:
Dan Skiest from Baystate Medical Center, MA presented the results of ACTG 5170. By way of disclosure, the author was co-chair of this study. Patients who had a CD4 count of 350 cells/ul or more prior to ART initiation and within 45 days of study entry, HIV RNA < 55,000 copies/mL within 45 days of study entry, ART (≥ 2 drugs) for ≥ 6 months, and who desired to stop ART, stopped ART at entry and were followed for 96 weeks. (Note from Jules Levin. Baseline values; 71% <50 c/ml; 14% <350 CD4, 59% 351-500 CD4, 27% >500 CD4; pre-ART viral load, median: 17% <400 - 5,000 c/ml; 36% 5,001 - 55,000 c/ml; 19% >55,000 c/ml; 28% not available). Given that patients wished to stop their ART, this study sought to define the safety of this maneuver, with the hope of better defining predictors of future stability off HAART, or groups in which TI should be avoided.
ART was restarted at the discretion of the patient and health care provider. The primary endpoint was time to CDC Category B or C event or death or CD4 count ≦ 250 cells/ mm3. Secondary endpoints included clinical, virologic, and immunologic predictors of progression.
To be eligible subjects had to be asymptomatic and have no prior history of CDC category B or C events.
167 subjects were enrolled in this study. The median entry CD4+ cell count was 833 cells/mm3 and 137 subjects had HIV-1 RNA ≦ 400 copies/mL. The median nadir CD4+ cell count was 436 cells/mm3. Twenty-three subjects reported a CD4 nadir < 350 cells/mm3, while the remainder were > 350 cells/mm3. The median length of ART interruption was 96 weeks. CD4 cell decline following ART cessation occurred in two phases: 162 cells/mm3 in the first 8 weeks and 2.0 cells/mm3 per week from week 8 through 96.
By week 96, 17 subjects had confirmed CD4 cell counts ≦250 cells/mm3 and 46 subjects had reinitiated ART. Twenty-nine subjects with HIV-1 RNA ≦400 copies/mL at study entry chose to reinitiate ART and stayed on ART for longer than 16 weeks. Twenty-four of these 29 subjects achieved HIV-1 RNA ≦400 copies/mL by week 16 of ART reinitiation.
Following ART interruption, only 1 subject had acute retroviral rebound syndrome, and 4 subjects developed thrombocytopenia (platelet count < 55,000). Two subjects developed CDC category B events and 2 subjects developed CDC category C events, but none clearly HIV-related (Note from Jules. Week 48-96: 3-recurrent bacterial pneumonia, 1 sensory neuropathy, 1 varicella zoster, 2 additional deaths). These events all occurred after subjects had been off antiretroviral therapy for greater than 60 weeks. (note from Jules; to week 48 12 patients had <250 CD4, no CDC B or C events, 3 deaths).
Five subjects died during the study, but none of the deaths were related to HIV/AIDS infection. Four of these were possibly related to MI or arrhythmia in patients between 41 and 53 years of age; 3 had CAD histories. While these events may be unrelated to TI, they are concerning.
Univariate analyses showed that shorter time to primary endpoint (confirmed CD4 cell counts ≦250 cells/mm3, ART reinitiation, AIDS-related events or Death) at weeks 48 and 96 was predicted by lower CD4 nadir, and higher pre-ART HIV-1 RNA entry. In multivariate analyses lower CD4 nadir and pre-ART HIV-1 RNA (note from Jules: >50 c/ml) was predictive. The best predictor of CD4 decline or the development of clinical events following ART cessation was lower CD4 nadir.
Note from Jules. Grade 3-4 signs adjusted for follow-up (event rate/patient yrs): off ART, 0.25; ART restarted, 0.56, RR 2.23. Number of clinical diagnosis (after ART interruption, After ART reinitiation): herpes zoster 9/1; retroviral syndrome 1/0; thrombocytopenia 4/0; bacterial infection 7/3; mucosal HSV 2/1; sensory neuropathy 1/0; hypertension 5/1; cadidiasis/volvivag 2/2. There was a rapid CD4 loss after stopping ART, then gradual loss. Few CDC B or C clinical events (4 of 167). Predictors of 96 week outcome: nadir CD4 >400, entry VL <50 copies/ml.
- that it is safe to stop ART once in patients whose CD4 counts were above 350 cells/mm3 when ART was stopped, and in whom CD4 was above 350 around the time ART was started.
- While only a few HIV related clinical events occurred in study subjects who stopped ART, a CD4 nadir (lowest ever CD4 count) of below 400 cells/mm3 was associated with an increased risk of significant CD4 decline off ART and an increased likelihood that ART would be reinitiated within 2 years.
- Studies are ongoing to identify other factors (eg, genetic) associated with more rapid CD4 decline once ART is stopped, or failure of ART if it is restarted.
- Detailed studies of the evolution of NNRTI resistance after interruption are underway. This is still of potential concern, and the best way to stop a regimen, especially one that contains an NNRTI, is not ideally defined.
- As other studies below, especially SMART, the possibility that morbid events not classically associated with HIV infection occur at an increased rate in patients off ART is of potential concern, and worthy of further study.
CD4-guided TIs: "Staccato" played mostly in Thailand:
Jintanat Ananworanich presented the findings of this next TI study, in which 548 patients with CD4 > 350 cells/ul, and HIV RNA <50 copies/ml were randomized to receive either continuous therapy (CT, n = 154) or a TI (n = 299). Initially, this study also contained an arm (44 patients) who took WOWO therapy (week on-week off); which did not impress anyone. This arm of the study was "prematurely" (or perhaps mercifully) terminated due to frequent virologic failures.
284 patients remained for evaluation in the "staccato" TI arm, stopping therapy when CD4s rose above 350 cells/ul, and restarting when CD4 cells fell below 350. 146 patients were studied while on continuous ART (CT). Median baseline CD4 counts at entry were 470 cells/ul (STI) and 507 cells/ul (CT). The study groups were well-matched: median age 35 years, 40% male, 80% heterosexual/15% MSM, mean pre-ART CD4 cell count of 253-267 cells/ul, median duratio prior ARV 13-15 months.. Pre-ART HIV RNA was about 4.7 logs in both groups. However, 80% of the cohort was treated with ritonavir-boosted saquinavir, and 80% of the cohort was from Thailand. This may be of relevance as drug levels in the (often smaller) Thai population might have been higher and affected the outcome of the study.
No AIDS-defining events occurred in either group. Two deaths, from stroke (CT), and colon CA (TI) occurred. The CD4 slopes of the two groups looked very similar to ACTG 5170: the CT group, as expected, had a gradual increase of median CD4 over 108 weeks from ca. 500 cells/ul to 600 cells/ul. The TI group had a rapid drop of median CD4 from ca. 500 to 400 cells/ul, with relatively stable but slightly declining CD4 counts over 108 weeks.
Of significance 60.5% of TI subjects maintained CD4 > 350 cells/ul as compared with 96% in CT. At the end of the study TI patients restarted ART, and CD4 counts rose such that 86% of the TI group obtained CD4 > 350 after 12-24 weeks of ART (Jules: vs 96.9% in CT arm; proportion of patients with VL <50 c/ml after 12-24 weeks of HAART re-treatment: 91% STI, 92 CT). At the end of study, similar proportions of subjects had achieved HIV RNA < 50 copies (ca. 90% in both arms), and although there were 9 virological failures in the TI arm as compared with 6 in the CT arm, this was statistically no different.
Not surprisingly, less neuropathy or diarrhea was seen in the TI arm (1.9% vs. 4.6% and 16 vs 24%, respectively), and acute viral rebound syndrome (5.7%), low platelet counts (2.5%), and oral/vaginal candidiasis (3.5% vs 0 in CT arm) was exclusively seen in the TI arm.
Cholesterol levels were slightly lower in the TI arm, but the changes clinically insignificant. More patients in the CT arm self-reported lipodystrophy. There were low & similar resistance rates reported.
- The major benefit of this TI seemed to be a 62% savings in time on ART during the study. No differences were reported in drug resistance or treatment failure. However the trend towards a difference might have become more marked had follow-up been longer than 12-24 weeks. Also unclear, was the possible role of pharmacokinetics in the population studied (about 50% Thai women) with boosted saquinavir in this trial. One might wonder if the results would be similar in all populations.
- Although no deaths attributable to STI were seen, one seems to be trading one set of adverse events for another, at the cost of lower CD4 counts.
- 40% of the TI group spent time with CD4 counts below 350 cells/ul. It is not clear that the study was long enough or large enough to detect differences in event rates between this subset of the TI group and the CT group.
- It seems that both sides of the treat/don’t treat camps could find support in this study, but true validation of this management strategy would seem to require longer follow-up and examination of longer term outcomes.
The Italian version:
The ISS-PART study, performed in 68 Italian clinical sites, compared 24 months of CT to 3 month periods of ART with interruptions of 1, 1, 2, 3, and 3 months in between. Due to the long half-life of NNRTIs, efavirenz was stopped 6 days prior to other drugs, and nevirapine stopped 3 days prior. This was another cohort with early prior therapy, with a median age of 40, CD4 counts between 710-770 cells/ul at TI, and a prior nadir CD4 count of more than 400 cells/ul. HIV RNA was <400. The study was powered to prove non-inferiority for an endpoint of CD4 >500 cells/ul after 24 months of study.
273 participants with an average of 2 years of first-line therapy without prior failure enrolled, 70% who were on NNRTI-based HAART, the remainder on unboosted PI-based HAART. 137entered the CT arm, and 136 the intermittent arm. Preliminary results were first reported in 2003, and now after dropouts over the years 112 patients were reported from the CT arm, and 56 from the intermittent arm.
Over 24 months, the CT arm (with high CD4s already) gained 6 cells, while the intermittent arm lost 26 cells. VL < 400 at 24 months was the same in both arms (91-92%). Male sex and higher CD4 nadir were multivariate predictor of success. 14 SAEs were reported in both arms, and 27 Grade 3-4 AEs in the CT arm, vs 12 in intermittent arm. (note from Jules. % with CD4 >500 at 24 months: 86% in CT arm, 69% in IT arm).
However 38 of 136 pts on intermittent therapy developed genotypic drug resistance mutations, a 30% cumulative risk at 24 months and a 2.6-fold increased risk of virological failure when they resumed therapy! 30% of these were M184V, and 8% were K103N. Most patients receiving NNRTIs acquired mutations. The emergence of mutations was predicted by the use of an unboosted PI regimen (50% of those on unboosted PIs developed mutations and 20% on NNRTIs), and by the presence of archived mutations at baseline in this cohort thought to be on first-line therapy at entry. Overall, even in this group with CD4> 700, intermittent therapy could not be shown to be non-inferior by the rather lax CD4 endpoint. Clearly resistance was increased by TI.
- No strong evidence that TI is superior
- Resistance developed in many patients
Note from Jules: authors summarized we were unable to demonstrate the non-inferiority of STI in terms of proportion of subjects with >500 CD4 after 2 years
A Window from France seems unclear:
Bruno Marcho presented the Window study performed by the French ARNS. Over 2 years 391 patients cycled therapy (a fixed schedule): two months on, two months off (IT). Median CD4 count at entry was 745 cells/mm3 and median HIV RNA < 200 copies/ml. The study enrolled mostly men with a median of 5 years of ART experience. Roughly half used a PI and the other half used efavirenz. Patients with nadir CD4 cell count below 100 cells/ul were excluded, and mean nadir was 280 cells/ul & was required to be > 100 cells/ul, and median CD4 nadir was 280 cells/ul. Abacavir or nevirapine was not allowed, and those coinfected with hepatitis B virus could not take 3TC or tenofovir (due to fear of HBV rebound). When therapy was interrupted, those on efavirenz stopped it 7 days before the other drugs.
92% of subjects were followed to week 96; 27% of subjects in the CT arm, and 16 % in the IT arm left the study. The primary endpoint of the study was decline of CD4s below 300 cells/ul. By intent-to-treat, 7 subjects in the cycling IT arm (3.6%) and 3 subjects (1.5%) in CT fell below 300 cells. This difference was less than the study-defined threshold of "non-inferiority", and so the study could claim that the cycling strategy was shown to be non-inferior to CT. However, the non-inferiority threshold was at most a 7% difference between the two strategies. Therefore, as the number of events (CD4 declines below 300) were small, a extra handful of events in the IT arm would have made it inferior. In fact, by a per patient analysis, the difference between the two arms was 6.5%.
Not unexpectedly, there were no AIDS events, and only 1 death each arm. However, 9 in the IT and only 2 in the CT developed low platelet counts (grade 3 or 4), while 10 vs. 6 had thrush. Three in the IT arm had the acute retroviral rebound syndrome. (note from Jules Levin: herpes simplex: 4 in IT, 2 in CT; lymphadenopathy: 10 in IT, 3 in CT; hemooagic signs: 2 in IT, 0 in CT). 10 patients had reported 1 or more episodes of thrombocytopenia before enrollment).
The first 200 patients were enrolled in a virologic sub-study to examine the development of drug resistance. The proportions of patients failure (viral load ≥1000 copies/mL) after only six or more weeks in the study were similar in both groups (17% IT, 14% CT), but surprisingly high in a cohort of theoretically stable, drug-experienced patients. (note from Jules: in reviewing the study data I noticed at week 96, n=352, the % of patients with <400 c/ml was 81% in the IT group & 90% in the CT group). One wonders if the study selected patients hoping for the IT arm, driving up failure rates in the disappointed ones assigned to the CT arm. The presence of resistance mutations was little different between those who failed in the two arms. It would have been interesting to compare the frequency of archived drug-resistance mutations in the cells of patients in the two arms over time; perhaps this information is being developed.
- Again both sides of the treat/don’t treat camps could find support in this study.
- Again, longer follow-up would provide more reassurance that resistance will not develop with this IT strategy.
- Although "non-inferior," one must ask if twice as many failures in this clinical trial setting is acceptable, particularly in view of the toxicities encountered during the interruptions.
- If the main impetus for TI is cost savings, this strategy is a winner, but other costs are incurred.
Treatment interruption in a resource-poor setting:
Christine Daniel presented the Trivacan study for the French ARNS (ARNS 1269). Performed in West Africa, this non-inferiority study compared two TI plans to CT. 840 adult drug-naïve patients (77% women!, 35% with advanced disease by WHO staging) with CD4 counts between 150-350 cells/ul (inclusion criteria, & ARV-naïve) were given CT for 6 months. The vast majority were treated with Combivir (AZT/3TC) and Sustiva (efavirenz). (note from Jules; To be randomized in study patients had to be on continuous HAART >6 months, CD4>350, HIV RNA <300. Baseline characteristics: CD4 median 460 at randomization; CD4 nadir 273).
Then if CD4 >350/ul and VL < 300 copies/ml, 1 in 6 study volunteers was randomized to CT, half the volunteers were randomized to two months on ART and 4 months off, and the remaining third of the volunteers to CD4 guided therapy (off if CD4 >350, on if CD4 <250). Patients were followed for 96 weeks (2 years).
The primary endpoint was to be the proportion of patients with CD4>350/ul, or grade 3-4 toxicities, or death. However in October 2005, a data safety monitoring board stopped the CD4-guided arm as the incidence of serious illnesses was 2.6-times higher in the CD4-guided interruption arm compared with those on CT. At 19 months of follow-up the mortality was 0.6 per pt/yr in the CT arm, and the IT arm had twice the rate of death (1.2 per pt/yr). At 16 months the probability of serious morbidity in the CD4-guided group was 17.7% per 100 patient-years in the CD4-guided group vs. 6.7% in the CT group. These events were mostly a more frequent occurrence of bacteremia (invasive bacterial diseases) and oropharyngeal candiaisis, and TB. Most (85%) of the bacterial events were due to organisms with Bactrim resistance, as many of the patients continued Bactrim prohylaxis. (note from Jules-- invasive bacteria—overall 0.6 in CT vs 9.0 in CD4-guided incidence/100 person-years, IRR, incidence rate ratio, 15.9); with bacteriemia: 0.0 in CT vs 5.1 in CD4-guided). Also the CD4-guided group incurred more clinic visits and hospital days. A non-significant trend toward increased antiretroviral resistance (5% vs 11%) was seen in the CD4-guided arm. No data was presented on outcomes in the other arm of the study, testing the fixed TI plan of 2 months off ART and 4 months on ART; this arm of the study and the CT arm are still ongoing.
- This CD4-guided strategy was inferior in a resource-poor setting.
No Joking about SMART:
Cardiologists seem to pick catchy names for their studies, but the acronym for the Strategies for Management of Antiretroviral Therapy trial was a unique choice in the HIV field. So far, it seems like continuous therapy is the strategy that proved to be smart. This very large, multinational, and very important study originally intended to enroll 6,000 HIV-positive patients with T-cell counts above 350 cells/ul. Patients, either ART-experienced or naïve, were randomized to: 1) continuous treatment (CT) or 2) an episodic treatment group (TI), in which treatment would be delayed or discontinued until the CD4 cell count fell below 250, and then ART given until the count rose above 350, followed by another TI. The investigators’ motivation was the hypothesis that intermittent therapy would preserve CD4 cells and prevent disease, but minimize exposure to ART and thereby reduce drug-related toxicities. A huge amount of data was very rapidly analyzed to allow findings to be presented at CROI. More in-depth analysis is ongoing.
SMART enrolled patients in 33 countries at 318 sites: 57% in the USA, 26% in Europe, and 10% South America. The mean age of volunteers was 46 years, 27% were women, 30% were black. The study did not exclude patients who had prior AIDS, or failed therapies in the past and/or who had drug resistance. In fact, 25% of patients in the study had a CD4 nadir below 150 cells and 25% had an AIDS-defining diagnosis.
Expecting that ART would increase cardiac and other non-HIV-related disease risks, the investigators defined a unique compound primary endpoint for SMART: HIV clinical disease progression, death, or a defined group of specific severe adverse events (including strokes, heart attacks, kidney or liver damage). When recruitment was stopped by the safety monitoring board in January, SMART had enrolled 5472 of a planned 6000 patients. The patients had been followed up for a median of 14 months, during which there had been 164 recorded primary endpoints (HIV disease progression, death, or other serious medical complication). Remarkably, only 2% of the subjects were lost to follow-up.
Most of the volunteers were taking ART at enrollment (82%) and of those only 69% had viral loads below 400 copies/ml. The mean CD4 count at entry was 598 cells/ul, but the mean nadir CD4 count was 253 cells/ul. Overall, the patients in the TI arm spent 33% of the time on treatment, compared with 93% in the CT arm. Although the mean follow-up time was reported to be 14 months, it was also reported that the TI arm had a median of three TIs, with a median length of 18 months. A full reporting of the data is awaited to see how many patients interrupted, at what frequency and duration, and whether reinitiated ART allowed CD4 counts to rise above 350. It was reported that TI arm subjects spent only 3.1% of the time with CD4 counts below 200 cells/ul, 8.2% time below 250/ul, and 31.7% below 350/ul. CT arm patients spent < 9% of the time below 350 cells/ul. (note from Jules: 70% of patients had <400 c/ml at baseline; prior clinical AIDS: 24%; 4.6% ART naïve; 6 yrs of prior ART).
To the surprise of many, there were 117 events in the TI arm compared with 47 events in the CT arm, for rates of 3.7 vs 1.5 per 100 person-years (relative risk 2.5). The relative risk of events was highest in years 1 and 2, and then declined thereafter, but this trend was not statistically significant. Recall that mean followup was only 14 months, so longer outcomes are derived from observations of a minority of study subjects.
ALL the endpoints---the primary composite endpoint, death, serious events, non serious HIV disease progression, serious progression+death---favored the CT arm. Of 114 severe complications, 31 were cardiovascular/renal/liver events. CT was superior for both races and sexes. CT was favored in all entry CD4 strata, and all nadir CD4 strata, and all strata of baseline HIV RNA. (note from Jules: the relative risk was higher in the interruption arm for cardiovascular, liver or renal complications, for non-fatal cardiovascular events, for non-fatal hepatic events, and for non-fatal renal events).
There were 93 AIDS/death events in the TI arm vs. 44 in the CT arm (RR 2.15, p <0.0001), 47 vs. 29 deaths (RR 1.63, p =0.04), 59 episodes of MI/stroke/cardiac surgery/renal failure/cirrhosis in the TI arm vs. 37 in CT arm (RR 1.62, p=0.02). There were 54 AIDS-defining events (mostly oro/esophageal candidiasis) in TI arm vs. 17 in CT arm.
SMART study investigators examined whether events were linked to the "proximate" CD4 cell count, that is the CD4 cell count measured at last visit prior to an event. Dr. El-Sadr, who presented the study, stated that the proximate value was not predictive; events happened at high and low CD4 counts. However, as return visits could be as infrequent as every four months, the proximate-ness of these data will have to be carefully examined.
- The SMART study found that a cyclical strategy of stopping ART when the CD4 count rises above 350 cells/mm3 and starting it again when CD4 count dropped below 250 cells/mm3 was inferior to the usual practice of continuous ART.
- Of particular, novel, concern was the recognition that unpleasant events not typically thought of as related to HIV infection seem to occur more frequently in HIV-infected people off antiretroviral therapy. This finding is particularly powerful as it is directly in opposition to the hypotheses that the SMART study investigators sought to prove.
- Caution and attention to the general medical health of HIV-infected people is warranted, particularly if they have chronic viremia without antiretroviral therapy.
- As usual, more study is needed.