- There Is More Risk Than Value in Treatment Interruptions (Controversy 194, 195)
Moderator: Gregg Gonsalves, Gay Men's Health Crisis (GMHC), New York, U.S.
Pro Speaker: Diane V. Havlir, University of California, San Francisco, U.S.
Con Speaker: Bernard Hirschel, Geneva University Hospital, Switzerland
Full transcript of debate
Strategic or structured treatment interruption (STI) is a term that is firmly entrenched in the lexicon of HIV treatment. However, there are many different kinds of STIs and it is undertaken for many different reasons. This point was actually brought out by a member of the audience in a comment after this debate was completed. To fully discuss treatment interruptions requires separating the general categories and reasons for STI.
1. STI on a fixed schedule to purposefully allow cycles of HIV replication and stimulate the immune response against HIV. There have been reports of the potential usefulness of this tactic with acute HIV infection; studies on chronic HIV infection for the most part have shown negative results.
2. STI on a fixed schedule (long or short) to reduce patient exposure to HIV medications. The hope is that, by using less medication, side effects will be reduced, medication costs will be reduced, and patient quality of life will be improved and treatment fatigue lessened. The hope is also that the immunologic and clinical benefit of antiretroviral therapy (ART) will occur with this approach. The risks, of course, are the development of drug resistance, compromised future treatment options and reduction in clinical benefit overall. Long cycle STI (see Dybul, 2003 CROI) has been shown to result in the development of resistance and enthusiasm for this strategy has waned. Short-cycle STI -- week on, week off -- has had some positive results (see Dybul, PNAS, IAS 2003 poster 597) and some negative results (see Hirschel LB-4).
3. STI on a variable schedule. Several studies are underway that use a CD4-driven treatment approach. The idea is that ART is given when CD4 cells drop below a certain predetermined level and then stopped when CD4 counts rise above a predetermined level. This is the study design for the SMART study, the international CPCRA protocol that compares CD4-driven treatment algorithm to traditional therapy with continuous administration of ART. In this STI treatment paradigm, there will likely be relatively long periods (months to years) on and off treatment.
4. STI in salvage setting to produce reversion to wild type virus prior to restarting treatment. The idea behind this is that in patients who have multi-drug-resistant virus, and thus have limited or no treatment options, STI may lead to partial or complete reversion to wild-type virus. If this happens, the efficacy of salvage therapy may be enhanced. Several studies have been done, with mixed results, to test this hypothesis; however, most evidence suggest lack of benefit (see abstract 119 this meeting).
5. STI for management of possible HIV-related side effects. In this setting, a temporary STI is planned to determine if a patient's symptoms are related to ART or not. Some medication-related side effects may have an insidious onset and are nonspecific in nature. In this scenario, medications are stopped. If symptoms resolve, this suggests that one of the antiretroviral medications was responsible and an alternative regimen is prescribed. If the symptoms remain unchanged, the same ART is typically resumed.
Pro Speaker: Diane V. Havlir, University of California, San Francisco, U.S.
Dr. Havlir structured her discussion for the pro position with a countdown of the 10 overall reasons not to stop ART.
10. Treatment Works
This topic was begun with a photo of a patient who had overwhelming cutaneous and mucus membrane involvement with Kaposi's sarcoma. In fact, because of ART, these patients are rarely seen anymore. Treatment has resulted in major declines in morbidity and rates of opportunistic infections and malignancies have dropped. Overall mortality rates have declined in all settings in which individuals have access to ART. Dr. Havlir also made the point that there are benefits to ART that go beyond these well-established improvements in morbidity and mortality. Pre-ART, we used to see "the dwindles," which are minor AIDS-related complications that compromise quality of life. There is also improvement in skin and hair integrity as well as better dental health with ART. Also, when HIV goes untreated, in addition to the inevitable CD4 count decline, there is decreased responsiveness to vaccines and immunologic activity is compromised. ART also reduces perinatal transmission of HIV, and many contemporary ART regimens also treat hepatitis B infection (i.e., 3TC [lamivudine, Epivir], tenofovir [TDF, Viread], FTC [emtricitabine, Emtriva]). However, following treatment interruption, there have been case reports of flares of hepatitis B infection (some fatal).
9. Drug Resistance in STI
Antiretroviral medications have differing half lives and this may effectively lead to exposure to dual or monotherapy. This may lead to the emergence of pre-existing mutants or the development of resistance as a result of viral replication during STI cycles. For some antiretroviral medications a single point mutation may lead to high level resistance that eliminates an entire antiviral class (K103 for NNRTI). Dr. Havlir then referenced several STI studies illustrating this phenomenon: One study with nevirapine (NVP, Viramune) for acute HIV infection showed 6/7 patients developed resistance to this agent; Marc Dybul's eight week on, four week off (presented at 2003 CROI) showed 3/8 developed resistance to efavirenz (EFV, Sustiva, Stocrin).
8. STI May Produce an Acute Retroviral Syndrome
Many studies have shown that varying proportions of patients develop a flu-like illness within several weeks after STI. These HIV flare reactions may be quite severe in some cases. Dr. Havlir presented two case studies to illustrate this point. One patient, who had originally been quite ill with toxoplasmosis and a low CD4 count, interrupted treatment. Acute retroviral syndrome developed and CD4 count rapidly dropped from 1267 to 247. Another patient with a CD4 count nadir of 2 and CMV retinitis had an excellent initial response to ART: the patient's CD4 count rose to 850 at year four after onset of illness. This patient interrupted therapy, rapidly developed dementia and depression, and has never completely recovered.
7. Reticence to Resume Therapy
After an STI, some patients decide not to resume therapy even though, based on CD4 counts and viral load, they really should be on therapy. This issue has been reported in clinical trials of STI and probably is more of an issue in clinical practice.
6. Transmission Could Be Increased After STI
HIV-RNA levels are the strongest predictor of vertical (Quinn, NEJM, 2000) and horizontal (Wawer, 2003 CROI) transmission of HIV. In view of these data, STI has public health implications as well.
5. Reservoirs Will Be Reseeded
Cerebrospinal fluid viral load increases in concert with plasma HIV-RNA levels. HIV is present in semen one month after treatment interruption. Some data indicate that time to viral load rebound may decrease with STI cycles (possibly related to reseeding of reservoirs).
4. Autoimmunization Theory Is Dead for Chronic Infection
Dr. Havlir referenced a study by Dr. Hirschal (CROI 2001) that showed STI did not work for autoimmunization. There does not appear to be any correlation between control of HIV viral load, or CD4 response, and cycles of STI.
3. STI Studies in Chronic Infection Are Concerning
Many STI studies are not randomized and have short follow up. Dr. Havlir mentioned that at this meeting there were 15 STI studies presented: three were randomized, follow up was two weeks to one year. CD4 count nadir does seem to predict risk: with the lower CD4 counts there is a greater risk of problems with STI. Only a few of the studies presented looked at the resistance issue -- this should be a standard assessment. One theme is that STI may reverse metabolic complications, but Dr. Havlir suggested that we are not seeing metabolic miracles in STI studies. The cost savings are often mentioned, but what about other costs such as a possible increased risk of transmission?
2. STI in Salvage Setting Does Not Work
Data from this meeting was referenced -- CPCRA study 064 looked at STI for patients with a viral load more than 5,000 and a multi-drug-resistant virus. This study demonstrated lower CD4 counts, together with a higher rate of clinical events in the STI arm (22) vs. the no STI arm (12). These clinical events were either devleopment of an opportunistic infection or death. (Eight patients in each group died, five had opportunistic infections in the no STI arm versus 14 patients who had opportunistic infections in the STI group.)
1. HIV Does Not Need a Holiday
Dr. Havlir completed her discussion by reviewing the main points of the pro argument. Overall she did an immensely impressive job of summarizing the many arguments against STI as well as presenting a formidable case for the pro position.
Con Speaker: Dr. Bernard Hirschel, Geneva University Hospital, Switzerland
There Is Not More Risk Than Value in Treatment Interruptions
The task of defending the con position was carried out by Dr. Bernard Hirschel, the Chief of HIV/AIDS service at Geneva University Hospital in Switzerland. Dr. Hirschel also presented oral abstract LB4 at this meeting demonstrating the failure of week-on, week-off STI.
Dr. Hirschel began his defense with three case studies: A probably lucky patient, a probably unlucky patient, and an extremely unlucky patient.
The probably lucky patient was recently diagnosed with HIV and was thought to have been recently infected, but records indicated that patient had actually tested HIV positive in 1990 and had a prolonged period with stable clinical course off of therapy.
The probably unlucky patient was diagnosed with HIV in 1995 and treated with a total of eight years of ART which included both PIs and NNRTIs. This patient had a cholesterol of over 300 and had a myocardial infarction in 2003 with a CD4 count of more than 500 cells. Dr. Hirschel referenced the D:A:D study showing a 27 percent increase in the risk of myocardial infarction per year of ART.
The very unlucky patient was a 28-year-old female with a CD4 of 520 and a viral load of 80,000. In April of 1997, she was treated with d4T, ddI and ritonavir. Over a several day period she took about 10 tablets of an ergotamine-containing migraine medication. As a result of the interaction with ritonavir (RTV, Norvir), she developed peripheral gangrene and required partial amputation of both feet.
The point of these case studies: ART carries risks, some quite substantial; some patients with HIV infection remain well for many years without treatment.
Dr. Hirschel then said that he planned to summarize all of the clinical trials that have demonstrated clinical benefit of treatment in HIV-infected patients with more than 300 CD4 cells. A slide went up and it was blank. His point: There have been no outcome studies showing improved survival in patients with greater than 300 CD4 cells. He also mentioned that no studies have shown benefits from ART in acute infection. Although there is potential benefit to patients treated during acute infection based on improved immunologic function, possibly lower set point, prevention of widespread dissemination of virus, etc. there have been no double-blind, placebo-controlled trial showing long-term benefit.
He briefly discussed the SSITT trial, a STI trial that used a viral load of more than 5,000 copies/mL and a CD4 count of less than 300 as criteria for restarting ART. Only 50 percent of the patients met these criteria and had to restart medications at 56 weeks of follow up. There was a fairly rapid decline in CD4 counts of about 200 cells in the first four to 12 weeks of this study after treatment discontinuation and then a decline of about 50-150 cells per year afterward. However, a large variation was noted between individual patients.
The Swiss HIV cohort study was discussed with regard to STI. In this cohort patients are characterized as either currently on treatment, never treated, or treatment interrupted. In January of 1999, 7.6 percent of patients were on treatment interruption. In January of 2003, 14.8 percent were in this category. So there appears to be a trend for more treatment interruption in real world treatment of HIV infection.
Dr. Hirschel used some comparisons to demonstrate the realities of life-long treatment. A 39-year-old male would be expected to live for a median of 44 more years based on actuarial analysis, which is a long time to be on HIV treatment. So we are willing to sentence patients to a 44-year sentence of antiviral therapy, yet, as Dr. Hirschel pointed out, the average person convicted of murder in Switzerland only gets a sentence of 11 years. His point: perhaps our sentencing guidelines are too strict for HIV infection. Dr. Hirschel used this analogy to challenge the paradigm that we must continuously treat HIV infection for a person's lifetime.
To further illustrate the issue of long-term ART toxicity, he displayed a slide showing two photomicrographs of mitochondria. One was severely distorted and an example of mitochondrial toxicity from a patient on ART; the other was normal. He then asked the question, which mitochondria would you rather have?
Dr. Hirschel then discussed the topic of the potential economic benefits of STI with a simplified economic case study.
|Total Population||7 million||12 million|
|HIV Infected||12,000||1.5 million|
|Gross National Product (GNP)||US$245 billion||US$6.6 billion|
|ART Cost/Year||US$144 million||US$450 million|
|Percent GNP for ART||0.06||6.6|
He indicated that typically, a country spends about 10 percent of its GNP for total health expenditures; 6.6 percent for ART alone is not within the realm of possibility for any country. If ART costs could be reduced through an intermittent treatment program, this would allow more patients to be treated in a real world scenario.
Dr. Hirschel completed his discussion with the point that drug resistance now is not what it used to be. In the STI studies he was involved with, the most common mutation that developed was M184V, and most patients re-suppressed on their original regimen despite the presence of this mutation.
Overall, Dr. Hirschel carried out an extremely interesting discussion for the con side of this argument. The global perspective and economic analysis was particularly persuasive.
In summary, both discussants led a thoughtful discussion that explored the many reasons that STI strategies are so attractive to some and so much of a concern to others. Certainly, one of the key points is that all STIs are not the same. Future discussion of this topic should clearly define the sub-category of STI that is being examined.
For previous research on this subject, click here.