Second Intl Workshop on Salvage Therapy for HIV Infection

Second Intl Workshop on Salvage Therapy for HIV Infection
May 19-21 Toronto, Canada

Jules Levin
NATAP
www.natap.org

I'm writing to you from beautiful downtown Bethesda, MD.I arrived here about 7pm after leaving the Salvage Therapy Meeting in Torontothis afternoon. Its good to be back in the good old USA after having to pass thru a number of immigration checkpoints. The Salvage Therapy Meetingwas mostly a repackaging of information that had been previously reportedbut was formatted to address salvage therapy issues. However, there werea few interesting and some new presentations. I will briefly mention someof these points and important new information I heard tonite at the openingdinner session at the Immune Reconstitution & Surrogate Markers in HIV/AIDSmeeting in Bethesda.

IL-2 Did Not Purge HIV From Reservoir

In Bethesda Anthony Fauci of the NIH talked about the patientsfor whom they could not find HIV in the memory CD4 reservoir after receivingIL-2+HAART. You should recall that for 6/14 patient they could not findHIV in this reservoir. For 3/14 they couldn't HIV after looking at a lotof cells with sensitive testing. They looked at the lymph tissue of 2 ofthese 3 and could not find HIV. Tonite Fauci announced that they discontinuedtherapy in 10 patients recently. You may recall I sent out an email abouta study from Rick Davey at the NIH. 10 patients discontinued HAART. Althoughthey were undetectable all 100 had their viral load rebound. Unfortunately2 of the 10 were the two mentioned above for whom they could not find HIVin lymph tissue or the reservoir. This means that HIV was not eradicatedfrom the reservoir. The hope was that IL-2 added to HAART might purge HIVfrom the reservoir. This new development shows they did not purge HIV fromthis reservoir although the amount of HIV was greatly reduced. I hope toreport additional information reported at this meeting to you.

Interferon-a2b (Intron-A) + HAART

17 asymptomatic patients received daily interferon-alfa(5 million units) by subcutaneous injection for 28 days before initiatingHAART. The treatment-naïve patients had CD4 >200 and viral load>5,000 copies/ml. Of 14 evaluable patients, one had a viral load reduction<0.5 log, (non-responder), 5 had intermediate (0.5-1.0 log) viral loaddecreases and 8 showed declines of 1 to 2.5 log. The average viral loaddeclines were 1.0, 1.2, 1.1 and 1.0 log at days 7, 14, 21 and 28 days, respectively.Maximum viral load suppression was associated with lower viral load at baseline.Study author reported all patients experienced transient "mild"flu-like symptoms. I think he underestimated the side effects of inerferonbut if its effective in lowering viral load it should be tolerable for 28days. The author concluded this pilot study suggests that IFN-a 5 MIU dailyexerts potent suppressive effects against HIV and should be explored assalvage therapy. Interferon is currently used as therapy for hepatitis Ceither alone or in combination with ribavarin.

DOTC (BCH-10652) Cross-Resistance

Promising in vitro data has previously been reported suggestingthat this new NRTI from BioChem Pharma would suppress HIV resistant to AZT/3TC,AZT, or 3TC. At this meeting additional data was reported. Clinical isolateswere obtained from patients previously treated for 32 weeks with 3TC andAZT. RF Rando reported isolates with genotypic resistance to 3TC or AZT/3TCexhibited an approximate 2-fold increase in resistance to dOTC.. They concludedthe presence of 3TC or AZT mutations did not affect the potency of dOTC.15 isolates with RT genotypic mutations of V184 or M184/V184 and >100fold 3TC resistance were used. And 8 isolates with V184, L41, and Y215 wereused. In a second experiment in mice researchers reported that dOTC wascapable of significantly inhibiting a M184V mutant clone when 3TC was ineffective.Virus with reduced susceptibility to dOTC (M184, K65R) were obtained invitro, and researchers reported it took longer for resistance to dOTC todevelop than for 3TC resistance to develop suggesting resistance to dOTCmay take longer to develop than resistance to 3TC. A clinical trial (phaseI/II) is underway. Data is expected at ICAAC (Sept '99).

Drug Holidays May Reverse Resistance?

At the Salvage Therapy meeting I think the most interestingdiscussion centered around a presentation made at the Retrovirus Conferenceby Veronica Miller which was reported in the NATAP Reports newsletter Aprilissue. She and her German colleagues gave patients a drug holiday of 2-3months followed by Mega HAART (>5 drugs) for salvage therapy. These werevery treatment experienced individuals who were failing therapy. After adrug holiday 26/39 had their virus return to wild-type. Drug resistancewas not detectable. In 13/39 they did not revert to wild-type. The revertersreceived a Mega HAART regimen of at least 5 drugs or more and on averagehad a maximum 2.80 log reduction. Baseline susceptibility was associatedwith a virological response (achieving and maintaining VL<500 c/ml).67% of responders had viral populations which were sensitive to at least4 of the drugs in their regimen; whereas, 70% of patients of failing regimens(never achieving <500 c/ml) had viral populations which were sensitiveto a maximum of 3 drugs in their regimen. During drug holiday's viral loadincreased by a median of 0.71 log and CD4s declined by a median of 89 cells.Treatment with drug regimens resulted in a median week 8 VL reduction of2.9 log for patients with shift to wild-type and 0.78 log in patients withouta shift to wild type. Researchers reported CD4s increased subsequent torestarting therapy but slowly. 19/24 patients with shift to WT reached <500c/ml within 24 weeks compared to 1/9 without shift to WT. Miller concludedcomplete treatment interruptions may lead to shifts in the viral population,with a change to WT as the dominant variant in some patients. These changesin susceptibility may be associated with better treatment responses. However,CD4 counts need to be closely monitored during drug holidays. The Germangroup also said that if a person did not respond to the drug holiday andMega HAART the first time they would give them a second drug holiday andtry another Mega HAART regimen. They said this worked for some. Howard Grossman,a doctor in private practice in NYC said he devised a bid Mega HAART regimenand found that patients were able to tolerate and be adherent to these regimens.

It was previously generally accepted that once a personhad resistance to a drug the resistance would re-emerge after restartingthe drug after stopping it. In fact, John Mellors expressed skepticism aboutthis newly reported development. Although creating controversy, this discussionwas probably the highlight of the meeting. I personally spoke with the Germanresearchers and they have an ongoing program utilizing the drug holidayapproach, and they have confidence that the approach works for some individuals.Bare in mind this approach is used for people who are failing therapy whohave resistance to drugs they are using. I think additional studies willbe developed to address this question.

CD4 Predicts Viral Rebound

Veronica Miller and the same German group reported in aseparate abstract that the risk of viral load rebound (after reaching undetectable)was associated with CD4 count at baseline and with the change in CD4 countdue to HAART. Achieved CD4 count was the most important for predicting viralrebound. Individuals with CD4 <20 were at the greatest risk for rebound.This suggests a case for using IL-2 to raise CD4s if HAART reduces viralload to undetectable but CD4s don't increase enough. Miller concluded lowerbaseline CD4 counts are associated with an increased risk of viral reboundbut this risk is significantly reduced in patients with low baseline CD4swho experience substantial rises in CD4 counts during due to HAART. In Bethesdatonite Tony Fauci showed an analysis suggesting that IL-2+HAART have a betterclinical outcome than HAART alone.

Alternating Therapy Regimens

V Soriano reported data from a small study of monthly cyclingof salvage therapy regimens. The goal was to see if alternating administrationof two regimens including 3-4 drugs (one combination per month) in heavilypretreated individuals with virological failure might be useful becausetime for accumulation of mutations in escape mutants is limited. The studyenrolled seventeen individuals with <300 CD4s, viral loads >10,000and previously exposed to NRTIs, NNRTIs, and Pis for at least 2 years. CombinationA was administered at 1, 3, 5 and 7 months; combination B was administeredat 2, 4, and 6 months. The most common regimens were d4T/ddI/nelfinavir/efavirenz(A) and AZT/3TC/ritonavir/indinavir (B). Six patients did not complete 24weeks and withdrew due to toxicity or intolerance. Six of the remaining11 reduced viral load <500 copies/ml by month 6. Four of the remaining5 reduced viral load >1 log but did not reach <500. The authors concludedthat limiting the time for the accumulation of drug-resistant mutationsin a single genome seems to be the mechanism explaining the success of alternatingtreatment regimens. And that cycling therapy is a useful salvage strategyin individuals for whom toxicity and quality of life are a problem withMega HAART.

Virological Effect of Mega HAART

Julio Montaner reported data from using Mega HAART in Vancouver,Canada. In this observational study, patients who experienced failure withseveral regimens were offered multi-drug rescue therapy (MDRT) includingup to 9 drugs: up to 4 NRTIs, 2 Pis, 2 NNRTIs and hydroxyurea. Regimenswere tailored to prior history of antiretroviral (ART) use, tolerabilityand lab profile. 82 individuals started MDRT. Median prior ART exposurewas 7 drugs for 45 months. Median baseline HIV-RNA was 50,000 copies/ml.Median baseline CD4 count was 185. Median follow-up was 12 months. On anintent-to-treat analysis 56%, 46%, 33%, 23% and 16% of patients achievedviral load <400 copies/ml in 1, 2, 3, 4, and 5 consecutive determinationsof viral load. Out of 44 patients with a VL<400 copies/ml at least once,45% remained suppressed after a median follow-up of 30 weeks; 39% and 16%rebounded >400 before and after 16 weeks from the first undetectableVL. Phenotypic resistance data was available for 60 patients at baseline,Sensitivity to 3 or more drugs in the MDRT regimen (p=0.088) and a low baselineVL (p=0.005) were associated with a favorable response. Montaner concludedMDRT achieved a substantial anitiviral effect in up to 40% of patients.Adverse drug effects are frequent. Relying exclusively on historical, clinicaland lab evidence may not be sufficient to rule out a possible antiviralresponse. Several researchers have suggested that although an individualhas extensive resistance in all classes of drugs and would not be expectedto respond to a salvage regimen these individuals can respond to Mega HAARTor MDRT. Resistance testing may help select a more effective and efficientregimen.

Montaner reported results from a second observational studyof 67 similarly experienced individuals starting MDRT. On an intent-to-treatanalysis 63%, 50%, and 34% of patients achieved VL<400 in 1, 2, and 3consecutive determinations. On an on-treatment analysis 70% (n=59), 55%(n=51) and 41% (n=32) of patients achieved 1, 2, and 3 consecutive VL<400copies/ml, respectively. Montaner concluded MDRT achieved a substantialantiviral effect in up to 60% of patients in this second cohort. Adversedrug effects, however, remained a frequent problem in this group. But asI mentioned above Howard Grossman reported that he could improve patientability to remain on regimen by designing bid regimens that may be moretolerable. Schlomo Staszewski from the German group said that after a patientreduces VL to undetectable with Mega HAART he might reduce the regimen toa protease sparing regimen to improve tolerability. For some this reducedregimens holds VL below detection.

One abstract reported phenotypic resistance tests resultspredicted better resistance than genotypic testing results, but a secondstudy reported their data suggested that genotypic assays are adequate topredict phenotypic resistance with regards to 3TC, AZT, NVP, and Pis. Whetherthis is adequate for abacavir, ddI, or ddC remains to be established. Therewas a strong correlation between phenotypic and genotypic resistance for3TC, AZT, IDV, NFV, RTV, SQV and NVP. Moderate correlation was detectedbetween abacavir and d4T. There was not a significant relationship betweenthe 2 resistance measures for ddC or ddI.

Delavirdine Containing Salvage Regimens

Trip Gulick reported preliminary data from ACTG 359. 277individuals with a median of 14 months indinavir use, baseline VL of 31,746c/ml and CD4s of 229 were randomized to receive one of six regimens-

(A) SQV 400mg, RTV 400 mg and DLV 600mg all twice daily.

(B) SQV/RTV/adefovir

(D) SQV/NLF/DLV

(E) SQV/NLF/adefovir

(F) SQV/NLF/DLV/adefovir

At 16 weeks patients in the DLV arms (C and F) had themost individuals below 500 Copies/ml. The percent of patients with VL<500overall was 30% (77/254) with 33% in arm A, 20% in Arm B, 31% in Arm C,47% in Arm D, 16% in Arm E and 38% in Arm F. Unexpected interactions betweenDLV and adefovir were found in this study. Adefovir reduced DLV blood levelsby 50% that in turn reduced SQV blood levels when combined with DLV andadefovir because delavirdine substantially increases SQV blood levels. Thiscould explain why arms C and F did not do as well.

Howard Grossman reported results from using Mega HAARTin an open label, non-randomized study of 35 highly therapy experiencedgroup treated for up to 15 months with 6 to 8 drug regimens. The regimensincluded 2 or 3 NRTIs, 1 NNRTI, 1 nucleotide RTI (adefovir), 2 Pis and hydroxyurea.21 patients had genotyping sometime prior to this study. Patients had 73months prior treatment experience with a mean of 9 drugs (5-11). No patientswere lost to follow-up. One patient discontinued therapy due to toxicityand 2 due to failure. Of the remaining 34, 23 (68%) achieved VL<500 on2 or more determinations. 23 (68%) had at least 2 VLs <20 c/ml (or <40)and 17 (50%) had >2 readings undetectable on ultrasensitive assays. 13patients (38%) have had VL<20 or 40 c/ml for more than 6 months, thelongest being undetectable by ultrasensitive for 10 months. Patients reportedgood tolerability and adherence. As mentioned above Grossman designed bidregimens saying that this helped with tolerability and adherence. Grossmanreported recycling of previously used medications to which patients developedresistance, given with new drugs, may overwhelm small sub-populations ofHIV and prevent the occurrence of resistance to the new medications.

PR Harrigan reported that the presence of the 215 mutation,associated with AZT resistance, was associated with an attenuated responseto two different treatments: hydroxyurea+ddI; d4T+3TC. Then S Becker reportedresults from an analysis of VL and CD4 response among patients on AZT containingregimens with or without prior d4T experience. And among patients receivinga d4T regimen with or without prior AZT therapy. This study called the CHORUScohort is an observational study funded by Glaxo Wellcome. The authors showeddata that prior d4T therapy was associated with impaired VL response tosubsequent AZT therapy when compared to no prior d4T therapy to a greaterextent than AZT to prior d4T therapy. Researchers at the meeting raisedquestions about some disparities between the treatment arms that may haveeffected the results. They questioned the conclusions reached based on thesedisparities.

Week 36 Update of Salvage Therapy Study IncorporatingHydroxyurea+ddI and Indinavir+Ritonavir

This is my fourth update on Mike Youle's ongoing salvagetherapy study in which he used indinavir+ritonavir as part of a multi-drugregimen. He reported data starting at the Virgin Islands meeting in December'98, the Retrovirus meeting in Feb '99, the RIGHT meeting in April '99 andreported week 36 data for the first time at the SalvageTherapy Workshop.Youle's data reports a high level of success for the patients in this studyin reducing viral load to undetectable. It appears as though a key componentto this success is a Therapy Support Package (described below) providinga good deal of personal support in the patient's efforts. Sixty-three patientson PI containing regimens for >6 months with a viral load >5,000 copies/mlwere treated with a 5-7 drug regimen consisting of a backbone of efavirenz,hydroxyurea, ddI, and ritonavir+indinavir. Median baseline CD4 was 120,and viral load 63,000 copies/ml, but 31 individuals had a viral load >750,000copies/ml. The study participants were "heavily" experienced withantiretrovirals and had been exposed to a median of 5 drugs, and 85% (n=54)were given a median drug holiday of 8 weeks before starting their new regimen.Only 14% (n=9) had prior NNRTI experience.

Over 70% of patients had prior experience with each ofthe following NRTIs: AZT, d4T, ddI, and 3TC. Study participants were alsoexperienced with PIs: indinavir (n=43), nelfinavir (n=21), ritonavir (n=30+),saquinavir (n=32). Median duration was 8-11 months for each of the drugs.

The regimens were tailored to what people could tolerateconsidering prior experiences. All patients initiated treatment with efavirenzalthough 8 switched to nevirapine after the study started. 90% initiatedtreatment with ddI and about 82% with hydroxyurea. The incidence of otherdrugs used in study regimens were d4T (35%), indinavir (71%), ritonavir(66.7%), nelfinavir (15.9%), saquinavir (14.3%). All patients received eitherritonavir+indinavir or nelfinavir+saquinavir. The choice was based on whatmight be more tolerable for the individual.

Study patients received counseling and support in an effortto help them succeed with their therapy. Through the Therapy Support Packagepatients could contact study personnel any hour of the day or night. Theimportance of this therapy was discussed with the patients and Youle thinksthe drug holiday gave patients an opportunity to reflect upon how crucialto their health this therapy was.

The median follow-up in this study is 27.5 weeks. 13 patientsstopped all drugs at some point and 3 restarted within 1 month of stopping.During 31 patient years of follow-up of 63 subjects 36 drugs were stoppedin 25 subjects. People didn't always stop everything. Reasons for stopping-10just for not wanting to continue, 2 neutropenia, 2 lipoemias, 10 CNS sideeffects (8 switched from efavirenz to nevirapine), 1 diarrhea, 1 nausea,2 intercurrent illness, 3 had viral load rises and stopped therapy (allwere compliant). HU toxicity mild (4 stopped and 4 reduced dose).

Viral Load Reductions at Weeks 28 and 36

Youle said the data analysis is an intent-to-treat analysiswith statistically significant viral load reductions. The median viral loadreduction at week 28 was 3.08 log (n=20), with 75% achieving better than1 log drop and 85% <400 copies/ml.. At week 16 (n=42) viral load reductionwas 2.37 log, with 64% <400 copies/ml. At week 36, 62% had <50 c/ml(39/63), and 21% are >50 but <400 (13/63); 3 stopped-2 poor compliance.Two persons stopped therapy after <50. One of these two stopped with<50 and restarted therapy a month later but had developed NNRTI mutationsand never re-suppressed. He had acquired 181, 190 (NNRTI mutations) and74 mutations. This raises the issue mentioned before that the efavirenzlong half-life can be of concern if stopping all drugs abruptly. If a personstops their regimen efavirenz may still be present in the blood for a fewdays but full suppression on viral replication is no longer present. 6 patientsof 63 did not suppress viral load at all-- 3 were Africans who were verycompliant and 3 were not compliant. 5/63 >1 log >400 (1 lost to follow-up-2poor compliance). It's suggested that the 3 Africans didn't respond dueto a different virus.

Median CD4 rise was only 120 so far.

Factors associated with reducing VL<400 c/ml: 70% lesslikely to suppress <400 if not white; for every log higher at baselineyou were 40% less likely to suppress <400 c/ml; 15% more likely to suppressfor every month of treatment break or drug holiday.

Only the prior number of antiretroviral drugs was associatedwith rebound in 16 subjects after being below 400 c/ml, but Youle cautionedthere were too few patients too rely much on this conclusion. Although only30 patients were included in this analysis, the amount or dose of ritonavirwas associated with the amount of lipid elevations. Subjects were on varyingdosing combinations of RTV/IDV: 100/800, 200/800, 400/800. The higher thedose of RTV the higher lipids (cholesterol and tryglicerides) were.

Finally, the subject of combining NVP with efavirenz wasinformally discussed. Studies exploring this combination is expected inboth treatment naïve individuals as well as for salvage therapy. Preliminarydata suggests they should not be antagonistic, may complement each other,and should not effect blood levels of each drug. However, using this combinationis still experimental as here is no clinical data yet.