When to change treatment
Patients change therapy for many reasons. These include development new HIV-related illnesses, or ones they have had in the past; a rise in viral load or fall in CD4 count; drug side-effects; problems with adherence or just for the perception that a new therapy may be better.
Some cliSnics test viral load after just a month or two on treatment in order to check whether the expected fall in viral load is starting to happen. Seeing a good viral load response at this point can help many people with adherence, because it shows clearly that therapy is working and that it is worth sticking with. A poor viral load response at this point may be a sign of poor adherence, or low drug levels and the reasons for this need to be looked into, and therapy improved. These early results may not however reflect the medium-term response to treatment. The Frankfurt Cohort, for example, found that only 70% of individuals had undetectable viral load at week eight, rising to 80% at week 16 [1].
Another reason for a change in therapy is to switch to an alternative formulation or dosing regimen. A number of studies have provided encouraging evidence that people can maintain undetectable viral load when changing regimens. Nevertheless, a proportion of people with undetectable viral load do experience viral rebound after changing therapy. Viral load at the time of switch may be an influential factor. For example switching a PI for an NNRTI whilst viral load is suppressed below 500 copies rather than below 50 copies/mL appears a risky strategy and several studies have reported a high proportion of virological failures in these circumstances. A large, randomized study looked at people who had viral loads below 50 found 84% of the group that substituted efavirenz maintained viral suppression compared to 73% of the group who stayed on PI therapy [2] where as in a study of 64 Germans 95% remained below 50 copies [3].
When subjects in the ATHENA study cohort switched from PI's to either a nevirapine or a second PI the former was more likely to result in sustained viral suppression on the new regimen [4]. In 446 subjects of whom 125 switched to nevirapine there was a 5 fold lower risk in the nevirapine group primarily due to lower discontinuation. Certainly one of the negative outcome factors when switching appeared to be a prior episode of switching for drug toxicity [5]. Continuing the Spanish penchant for randomized PI to RTI switch studies the NEFA study examined 460 subjects stable on a PI who then randomly were switched to nevirapine (n = 155), efavirenz (n = 156) or abacavir (n = 149) [6,7]. All measure of virologic stability showed no difference across arms at 18 months with rates of suppression of 75% in the nevirapine arm, 69% in the efavirenz arm and 65% in those on abacavir, however fewer abacavir treated individuals, 9% versus 16% and 17%, stopped randomized treatment. Some success was seen in reducing lipids and improving insulin resistance but this was not markedly different across the three arms, however HDL levels rose in both NNRTI arms and fell in the abacavir arm.
In a Swiss study which switched 84/163 subjects on PI's to the triple NRTI regimen of abacavir/zidovudine/lamivudine adverse events occurred in 20% of the continuation group versus 7% in the switch group (p = 0.02). There was trend however to more virologic failure in the simplified therapy arm 15% versus 6% (p = 0.8) [8]. The same result was seen in the Trizal study where patients were either randomized to stay on a PI arm (n = 103) or were switched to Trizivir (zidovudine/lamivudine/abacavir as a fixed dose combination) (n = 106) [9]. Although patients found it significantly easier to take the latter (p < 0.0001) and had better lipid profiles three of six virologic failures in this group had prior dual NRTI therapy. There was no difference in failure between the two arms, 17% in the HAART arm and 20% on Trizivir over 24 weeks. A further study in the same vein (CNA30017) looked at continued PI exposure (n = 106) versus swapping the PI component of a HAART regimen to abacavir (n = 105) [10]. At week 48 twenty-six versus thirteen had a viral RNA rise (p = 0.022) but all the failures were rescued with novel PI based regimens.
Changing due to toxicity or tolerability
Toxicity is the most common reason for stopping the first HAART regimen prescribed; an Italian study found that 21% of patients stopped their first regimen within a year due to side-effects, compared to 5% who stopped due to viral rebound and 7% who stopped due to difficulties with adherence. Women were almost twice as likely to stop their first regimen due to intolerance when compared to men [11].
The development of long term toxicity such as lipodystrophy also has been a major reason for change in therapy as two recent reviews examine [12,13] and attempts have been made by prevent this by switching off PI's to reduce lipaemia [8,14]. The PIILR study evaluated 81 PI patients with lipodystrophy that in a 3:2 randomization switched their PI to abacavir, nevirapine, adefovir and hydroxyurea or continued as they were. Whilst lipaemia declined so did limb fat in these predominantly lipoatrophic patients, no virologic failures occurred.
A switch to nevirapine in a Spanish study of 106 subjects on PI's for > 6 months gave a significant benefit in lipid levels without losses of virologic control [15] as did a similar French study of 74 subjects [16]. Nevirapine gave a superior blood lipid level outcome compared to that when switching from a PI to a second PI option in a further randomized study of 77 subjects from the same research group [17]. This switching strategy should not be attempted in individuals with prior sub suppressive nucleoside analogue regimens since NNRTI's mutations may emerge along with dormant NRTI ones. A team from Bordeaux showed this to be so in 34 subjects switched from PI's to nevirapine. In the group with prior NRTI therapy 41% rebounded versus none in the naïve arm (p = 0.003) [18]. Estrada and co-workers seemed to show the futility of switches from PI's to efavirenz where in 41 subjects followed for one year had an increase in sub-cutaneous fat loss and no other significant beneficial alterations in lipodystrophic features [19]. These findings were confirmed in 38 subjects in a study from Toulouse [20].
The largest study to examine abacavir substitution for nucleoside analogs in patients with HIV lipodystrophy has been performed by the predominantly Australian MITOX group [21]. They studied 111 individuals (109 males) who were stable on zidovudine (n = 26) or stavudine (n = 85) containing regimens who were then randomized to continue these thymidine analogs or to switch to abacavir whilst maintaining their other agents. The primary endpoint was limb fat mass, measured by dual-energy x-ray absorptiometry and computed tomography and the substitution group showed significant benefits in both these measures (Fig. 1). Two other studies of switching stavudine to abacavir have demonstrated similar findings [22,23] although the study with 1-year follow-up found only a 7% increase in peripheral fat. However, another small trial reported a 10% further peripheral fat loss on DEXA scan at 24 weeks after a switch from stavudine to abacavir so results are not consistent [24].
Other toxicity exists for boosted PI's and two studies reported on the effect of switching to a low ritonavir dose (100mg bid) boosting regimen. In a small study American investigators observed outcomes in subjects switching from saquinavir/ritonavir 400/400mg bid to 1000/100mg bid [25]. There were marked decreases in lipaemia and median peak saquinavir level increased by 100% while the median trough level did change. The MaxCmin 1 trial also evaluated switches of patients already on PI's to boosted PI's, indinavir/ritonavir or saquinavir/ritonavir [26]. It showed that tolerability in the boosted PI's were reasonably well tolerated and produced comparable efficacy although more changes of drug due to adverse events occurred in the indinavir arm.
Early change of agent or regimen may result in better response and in a small study [N = 16] fasting lipids and insulin returned to normal after switch from PI to abacavir [27].
Changing due to viral failure
When to switch from a failing combination will depend on various individual factors such as previous treatment history and available treatment options, as well as the definition of ‘treatment failure’.
The most common definition of ‘treatment failure’ is viral rebound above 50. While the viral load remains low (e.g. below 10 000) there is little risk of health deterioration. However, the problem of continuing on a combination with detectable viral load is the risk of drug resistance, which may ultimately undermine future response to this and other combinations [1].
A study of 984 people at various European clinics who switched from a first PI to a second PI found that people who switched when their viral load was still low were much more likely to achieve undetectable viral load on their new protease inhibitor [28]. This study also found that people who switched at higher CD4 counts did better on their second PI, and that people who could also add one or more new nucleoside analogues had better rates of undetectable viral load.
Viral load blips
Once it has been suppressed below 50 copies/ml, viral load may sometimes rise above 50 copies without indicating that treatment is failing. There had been concern that a viral blip above 50 copies or 200 copies would be more likely to signal the emergence of drug-resistant virus in patients with extensive drug resistance, and the subsequent failure of the regimen. However, an analysis of three studies including ACTG 398 and 359, salvage studies there was no relationship between single blips above 50 copies and subsequent treatment failure (defined as two consecutive measurements above 50 copies) [29,30]. Even when the threshold was raized to 500 copies/ml, viral load blips still failed to predict rebound above 1000 copies/ml after 38 weeks of follow up. This has implications for definitions of success or failure in future clinical trials. In particular, defining virologic failure as the occurrence of one or more viral load measurements above 50 copies/ml, after initial suppression to below 50 copies/ml is probably too stringent.
An Italian group, which looked at 123 patients followed over approximately two and half years, found that the mean level of viral load reached during a blip above 50 copies was 165 copies/ml, and half of patients experienced blips no higher than 110 copies/ml [31]. The average interval of testing in this cohort was every 31 days, and 77% of the cohort was found to have at least one blip during the follow up period when tested at this high frequency. On average, one in ten viral load measurements found a blip, but a relatively small number of patients within the group accounted for the majority of the blips detected. Blips did not become more frequent as the length of time on treatment increased, suggesting that the random pattern of viral load blips is more likely to reflect occasional changes in drug concentration. These may be driven by other factors such as dietary changes or concomitant medications, or intercurrent infections, rather than declining adherence or efficacy of treatment.
Variability in the accuracy of viral load assays may also contribute to the appearance of viral blips. Samples from 249 patients with first viral load blips after suppression below 50 copies/ml were re-tested and 59% were found to have re-suppressed to undetectable (below 50 copies) [32]. The remainder still had HIV RNA levels above 50 copies/ml, and a confirmed blip was predictive of subsequent viral rebound above 400 copies/ml.
These studies emphasize the importance of continuing to monitor viral load rather than switching immediately after it increases above 50 copies/ml. However, the drugs being taken at the time will also be a factor in this decision. One study using an ultrasensitive resistance test found that mutations associated with resistance to 3TC could already be detected at a median viral load of 76 copies/ml in six individuals experiencing a viral load blip on their first regimen [33]. The authors ruled out the possibility that these individuals had been infected with drug-resistant virus due to the rapid viral load reductions that had occurred when they started treatment. In two other patients exposed to monotherapy or dual therapy prior to commencing HAART, other mutations emerged rapidly. In three patients apparently wild type viruses were detected: these individuals were also receiving 3TC and were fully adherent, and the authors suggest that these patients may have been experiencing bursts of replication from recently activated but long-infected cells.
GIGHAART, a randomized study of the effect of treatment interruption on the response to a new HAART regimen found that people who took an eight week break prior to a new regimen were significantly more likely to experience a one log drop in viral load after 12 weeks. In this study, the average CD4 count was already very low, around 27 cells. Use of HIV treatments was also very advanced — over 80% of participants had high level resistance to AZT, at least three AZT/d4T resistance mutations, at least one NNRTI mutation, and at least two protease mutation [34]. Sixty-eight people with extensive treatment experience were enrolled. 79% of patients had three or more AZT and d4T-associated mutations, and 10% had nucleoside analogue multidrug resistance mutations (e.g. Q151M or 69SS). 60% had two or more non-nucleoside reverse transcriptase inhibitor mutations. 42% had more than three major protease mutations and 82% had two or more protease mutations. Seventy per cent of participants receiving six drugs and 31% receiving eight drugs. Intent-to-treat analysis at week 12 showed 59% of the interrupters versus 26% of those who switched directly to a new combination had at least a one log drop in viral load. 35% versus 15% had a viral load below 400 at week 12 (non-significant). Of those on treatment at week 12, the average reduction in viral load was — 1.9 log in interrupters and — 0.4 log in non-interrupters. The outcome at 12 weeks appeared to be significantly linked to adequate drug trough levels and reversion of mutations in the interruption arm, 2.36 versus 0.44 log drop in viral load (0.018) [35]. An innovative Italian study looked at constantly switching regimens in patients with multiclass resistance and viral load > 10 000 and using genotyping to control the dominant population. The CD4 cell count in this group of 81 subjects has increased significantly by 24 months (p < 0.005) and median viral load has fallen [36].
Conclusions
The reality of antiretroviral therapy in 2002 is that changing and switching for a variety of reasons is commoner than might otherwise be thought. It is clearly not possible to accurately predict which agents a patient will best tolerate to which they will best respond, although with some drugs, such as abacavir, this may be possible [37]. The likelihood is that most therapeutic combinations of three or more agents will be effective at achieving an initial suppression of HIV RNA to below the current level of detection and that substitutions will continue to be required for linked or perceived toxicity. Further randomized studies with adequate sample size to examine the best switch strategies will help future generations of physicians and patients to best make those decisions. The availability of new agents with different adverse event profiles will also widen the therapeutic options available.
References
1. Cozzi-Lepri A, Miller V, Phillips AN, et al. The virological response to highly active antiretroviral therapy over the first 24 weeks of therapy according to the pre-therapy viral load and the weeks 4–8 viral load. AIDS 2001, 15:47–54.
2. Becker S, Rachils A, Gill J, et al. Successful substitution of protease inhibitors with efavirenz in patients with undetectable viral loads – a prospective, randomized, multicenter, open-label study (DMP-049). Eighth Conference on Retroviruses and Opportunistic Infections. Chicago, 2001 [abstract 20].
3. Knechten H, Sturner KH, Hohn C, Braun P. Switch to efavirenz in a protease inhibitor containing regimen. HIV Clin Trials 2001, 2:200–204.
4. Dieleman JP, Sturkenboom MC, Wit FW, et al. Low risk of treatment failure after substitution of nevirapine for protease inhibitors among human immunodeficiency virus-infected patients with virus suppression. J Infect Dis 2002, 185:1261–1268.
5. Dieleman JP, Jambroes M, Gyssens IC, et al. Determinants of recurrent toxicity-driven switches of highly active antiretroviral therapy: The ATHENA cohort. AIDS 2002, 16:737–745.
6. Martinez E, Podzamczer D, Riber E, et al. Switching protease inhibitors to nevirapine (NEV), efavirenz (EFV) or abacavir (ABA): a randomized, multicentre, open label simplification trial. Ninth Conference on Retroviruses and Opportunistic Infections. Seattle, 2002 [abstract LB15].
7. Martinez E, Podzamczer D, Riber E, et al. Switching protease inhibitors to nevirapine (NEV), efavirenz (EFV) or abacavir (ABA): a randomized, multicentre, open label simplification trial. XIV World AIDS Conference. Barcelona, 2002 [abstract WeOrB1262].
8. Opravil M, Hirschel B, Lazzarin A, et al. A randomized trial of simplified maintenance therapy with abacavir, lamivudine and zidovudine in human immunodeficiency virus infection. J Infect Dis 2002, 185:1251–1260.
9. Katalama C, Clumeck N, Fenske S, et al. Use of trizivir (abacavir, lamivudine, zidovudine) to simplify therapy in HAART experienced patients with longterm suppression of HIV-RNA. TRIZAL study (AZL 30002): 24-week interim analysis. Eighth Conference on Retroviruses and Opportunistic Infections. Chicago, 2001 [abstract 316].
10. Clumeck N, Geobel F, Rozenbaum W, et al. Simplification with abacavir-based triple nucleoside therapy versus continued protease inhibitor-based highly active antiretroviral therapy in HIV-1 infected patients with undetectable plasma HIV-1 RNA (on behalf of the CNA30017 study team). AIDS 2001, 15:1517–1526.
11. d'arminio Monforte A, Lepri AC, Rezza G, et al. Insights into the reasons for discontinuation of the first highly active antiretroviral therapy (HAART) regimen in a cohort of antiretroviral naïve patients. AIDS 2000, 14:499–507.
12. Mallon PW. Switch therapy studies: a review. J HIV Ther 2001, 2:45–48.
13. Murphy RL, Smith WJ. Switch studies: a review. HIV Med 2002, 3:146–155.
14. Wensing AM, Reedijk M, Richter C, et al. Replacing ritonavir with nelfinavir or nelfinavir/saquinavir as part of highly active antiretroviral therapy leads to an improvement of triglyceride levels. AIDS 2001, 15:2191–2193.
15. Ruiz L, Ribera E, Bonjochl A, et al. Virological and immunological benefit of a salvage therapy that includes Kaletra plus Fortovase preceded or not by antiretroviral therapy interruption in advanced HIV-infected patients (6 month follow-up). Ninth Conference on Retroviruses and Opportunistic Infections. Seattle, 2002 [abstract 421].
16. Buisson M, Grappin M, Piroth L, et al. Simplified maintenance therapy with NNRTI (nevirapine) in patients with long-term suppression of HIV-1 RNA. First results of a cohort study. 40th, ICAAC. Toronto, 2001 [abstract 1541].
17. Negredo E, Cruz L, Paredes R, et al. Virological, immunological and clinical impact of switching from protease inhibitors to nevirapine or to efavirenz in patients with human immunodeficiency virus infection and long-lasting viral suppression. Clin Infect Dis 2002, 34:504–510.
18. Masqualier B, Neau D, Chene G, et al. Mechanism of virologic failure after substitution of a protease inhibitor by nevirapine in patients with suppressed plasma HIV RNA. J Acquir Immun Defic Syndr 2001, 28:309–312.
19. Estrada V, De Villar NG, Larrad MT, et al. Long-term metabolic consequences of switching from protease inhibitors to efavirenz in therapy for human immunodeficiency virus-infected patients with lipoatrophy. Clin Infect Dis 2002, 35:69–76.
20. Bonnet EPP, Yovanovitch JD, Mularczyk MA, et al. Switching from protease inhibitors to efavirenz does not improve clinical lipodystrophy and hyperlipidaemia in HIV infected patients. 40th, ICAAC. Toronto, 2001 [abstract 1538].
21. Carr A, Workman C, Smith D, et al. Abacavir substitution for nucleoside analogs in patients with HIV lipodystrophy: a randomized trial. JAMA 2002, 288:207–215.
22. John M, James I, McKinnon E, et al. A randomized, controlled, open-label study of revision of antiretroviral regimens containing stavudine (d4T) and/or a protease inhibitor (PI) to zidovudine(ZDV)/lamivudine(3TC)/abacavir(ABC) to prevent or reverse lipoatrophy: 48 week data. Ninth Conference on Retroviruses and Opportunistic Infections. Seattle, 2002 [abstract 700-T].
23. McComsy G, Lonergan T, Fisher R, et al. Improvements in lipoatrophy (LA) are observed after 24 weeks when stavudine (d4T) is replaced by either abacavir (ABC) or zidovudine (ZDV). Ninth Conference on Retroviruses and Opportunistic Infections. Seattle, 2002 [abstract 701-T].
24. Moyle G, Baldwin C, Gazzard B, et al. A randomized open label study of substitution of PI/NNRTI/d4T or both in persons with virologic suppression and significant metabolic changes and/or fat redistribution. Third International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV. Athens, 2001 [abstract 97].
25. O'Brien WA, Rojo D, Acosta E, et al. Switch of saquinavir 400 mg/ritonavir 400 mg to saquinavir 1000 mg/ritonavir 100 mg during bid four drug antiretroviral therapy in patients with viral load less than 200 copies/mL. XIV World AIDS Conference. Barcelona, 2002 [abstract WeOrB1263].
26. Cahn P, Dragsted UB, Obel N, et al. Safety and efficacy evaluation of a phase IV randomized, open-label, multicentre trial of indinavir/ritonavir (800/100 mg bid) vs. saquinavir/ritonavir (1000/100 mg bid) in adult HIV-1 infection: the MaxCmin 1 trial. 14th World AIDS Conference. Barcelona, 2002 [abstract WeOrB1265].
27. Walli R, Huster K, Bogner JR, et al. Switching from PI to ABC improves insulin sensitivity and fasting lipids: 12 month followup. Eighth Conference on Retroviruses and Opportunistic Infections. Chicago, 2001 [abstract 672].
28. Mocroft A, Phillips AN, Miller V, et al. The use of and response to second-line protease inhibitor regimens: results from the EuroSIDA study. AIDS 2001, 15:201–209.
29. Havlir D, Bassett R, Levitan D, et al. Prevalence and predictive value of intermittent viraemia with combination HIV therapy. JAMA 2001, 286:171–179.
30. Havlir D, et al. Are episodes of transient viraemia (‘blips’ in HIV RNA) predictive of virologic failure in heavily treatment-experienced patients? Ninth Conference on Retroviruses and Opportunistic Infections. Seattle, 2002 [abstract 93].
31. Di Mascio M, Author please provide up to 3 author names, et al. Viral blip dynamics during HAART. Ninth Conference on Retroviruses and Opportunistic Infections. Seattle, 2002 [abstract 94].
32. Mazen Y, Author please provide up to 3 author names, et al. Evidence of low level viral replication (< 50 copies/ml) predicts eventual virological failure. Eighth Annual Conference of the British HIV Association. York, 2002 [abstract 011].
33. Cohen-Stuart W, Wensing AM, Kovacs C, et al. Transient relapses (‘blips’) of plasma HIV RNA levels during HAART are associated with drug resistance. J Acquir Immune Defic Syndr 2001, 28:105–113.
34. Katlama C, Dominguez, S, Duvivier, C, et al. Benefits of treatment interruptionas (TI) in patients with multiple therapt failures, CD4 cells < 200/mm3 and HIV RNA > 50000 cp/ml (GIGHAART ANRS 097). XIV World Conference on AIDS. Barcelona, 2002 [abstract WePeB5887].
35. Peytain G, Legrand M, Duvivier C, et al. Relationship between trough plasma concentrations of HIV antiretroviral drugs and virological response and mutations reversions in Gigahaart trial (anrs 097). XIV World AIDS Conference. Barcelona, 2002 [abstract Tu PeB4568].
36. Maggiolo F, Callegaro A, Ripamonti D, et al. The use of strategic selective pressure for salvage therapy. XIV World AIDS Conference. Barcelona, 2002 [abstract Tu PeB4589].
37. Mallal S, Nolan D, Witt C, et al. Association between presence of HLA-B* 5701, HLA-DR7 and HLA-DQ3 and hypersensitivity to HIV-1 reverse transcriptase inhibitor abacavir. Lancet 2002, 359:727–732.
© 2002 Lippincott Williams & Wilkins, Inc.