Angiogenesis Inhibitor IM862 Is Ineffective Against AIDS-Kaposi's Sarcoma in a Phase III Trial, but Demonstrates Sustained, Potent Effect of
http://www.100md.com
《临床肿瘤学》
the Memorial Sloan-Kettering Cancer Center, New York, NY
USC Keck School of Medicine, Norris Cancer Institute, Los Angeles, CA
Massachusetts General Hospital, Harvard Medical School
Beth Israel Deaconess Medical Center, Boston, MA
AMC Operations and Statistical Center, University of Birmingham, Birmingham, AL
University of Washington
Virginia Mason Clinic, Seattle, WA
University of Toronto, Toronto Hospital, Toronto, Ontario, Canada
ABSTRACT
PATIENTS AND METHODS: Two hundred two HIV-positive patients were enrolled, 104 on IM862 and 98 on placebo.
RESULTS: Baseline characteristics were comparable except current antiretroviral therapy: 88% versus 96% (IM862 v placebo group; P = .042). The median treatment durations were 19.5 versus 24 weeks (IM862 v placebo). No significant difference was detected in response rate (IM862, 23%; 95% CI, 15% to 32% v placebo, 21%; 95% CI, 14% to 31%; P = .46), time to response (8.5 weeks v 14 weeks; P = .024), or duration of response. However, IM862 was associated with both a shorter time to response (8.5 weeks v 14 weeks; P = .024) and shorter median time to progression (16 weeks, 95% CI, 13 to 27 weeks v 35 weeks, 95% CI, 26 to 114 weeks; P = .012).
CONCLUSION: Despite promising phase I and phase II studies, IM862 5 mg every other day was not superior to placebo and may accelerate time to progression. Highly active antiretroviral therapy alone was associated with a substantial rate of sustained tumor response and may have contributed to prior estimates of IM862 response. Therapeutic trials for AIDS-Kaposi's sarcoma must account for ongoing immune reconstitution in the setting of concurrent highly active antiretroviral therapy that may confound estimates of therapeutic activity.
INTRODUCTION
IM862 is a synthetic dipeptide (L-glutamine L-tryptophan) with a half-life of minutes and no known toxic metabolites. It was originally isolated in Russia from bovine thymus during a search for immune adjuvants and was studied in Russia as an adjuvant to chemotherapy and in parasitic and fungal disorders. No toxicities were reported at doses in excess of 2,000-fold, the dose reported in the current trial.3
Although IM862 exhibits no direct cytotoxic effects, in in vitro studies3 IM862 exhibited dose dependent inhibition of angiogenesis in the chorioallantoic membrane assay in a range of 1 to 1000 μg/mL with complete inhibition of -FGF and VEGF-induced angiogenesis (1 to 100 μg/disc). The mechanism of action was not related to inhibition of endothelial proliferation as demonstrated in human umbilical vein endothelial cell or in Boyden chamber assays.
Although the exact mechanism of action remains unclear, IM862 exhibited activity in in vivo preclinical models.3 It inhibited melanoma growth in fetal mice, prevented neovascularization in a dose-dependent manner in the Lewis Lung tumor model, and prevented growth of sarcoma and hepatoma implants.3 In other experiments, IM862 was effective in a xenograft Kaposi's sarcoma (KS) model (Parkash Gill, personal communication, 2003). At least some of the activity has been ascribed to decreased production of VEGF and enhancement of natural killer (NK) cell function.4
Based on these findings, IM862 was studied in AIDS-related KS, a highly vascular tumor associated with human herpesvirus-8. While cytotoxic agents and interferon are active in KS, side effects may outweigh the benefits in patients with less advanced disease. Consequently, non-toxic, easily administered agents would be welcomed if effective and well tolerated over a long period of time. In animals, antitumor activity of IM862 was similar, irrespective of the route of administration: intranasal, subcutaneous, intravenous, or intramuscular.3 Given the ease of administration and a bioavailability of 71% after intranasal administration, this route was adopted in human trials.5
In a phase I/II trial, intranasal IM862 was administered at a dose of either 5 mg every other day (QOD; n = 26) or 5 consecutive days followed by 5 days without treatment (n = 18). Forty-seven percent of patients had more than 50 mucocutaneous lesions, CD4+ T-lymphocyte counts were > 200 cells/μL in 55%, and concurrent protease inhibitors had been used for a median of 10 months (range, 0 to 24 months) before study entry in 89% of patients. Of 44 patients, 16 (36%) patients had an overall response (complete response [CR] + partial response [PR]), with CR in five patients, PR in 11 patients, and stable disease (SD) in 21 patients. The median time to response was 6 weeks (range, 3 to 26 weeks), with the median response duration equal to 33+ weeks (range, 12+ to 95+ weeks). No significant toxicities were noted.6 Given the high benefit-to-risk ratio in the phase I/II study, a placebo-controlled, double-blind, phase III trial in AIDS-related KS was initiated using the 5 mg QOD intranasal dosing over a 24-week period.
PATIENTS AND METHODS
The institutional review boards of each participating institution reviewed and approved the protocol and consent form and all patients gave written informed consent before enrollment.
Patient Evaluation
Patients underwent a complete physical examination, WHO performance status evaluation, CBC, serum chemistries analysis, and toxicity assessment at study entry and then monthly for 6 months. CD4+ cell count and HIV viral load were followed prospectively every 2 months by standard methods at each individual institution.
Patients were staged according to a modification of the AIDS Clinical Trials Group–based tumor, immune, systemic staging system.7 As many as 50 cutaneous lesions were evaluated for their character (raised or flat) and size. If there were more than 50 cutaneous lesions, a representative anatomic area was selected and the lesions were counted and characterized. In addition, five discrete, nodular, indicator lesions were selected, and the sum of the products of their largest perpendicular diameters was recorded. All adverse events were graded using National Cancer Institute (NCI) Common Toxicity Criteria.8
Treatment Plan
IM862 was self-administered intranasally at a dose of 5 mg in 0.7 mL of diluent QOD for 24 weeks or until disease progression (PD). The diluent vehicle served as the placebo. Concurrent use of antiretroviral agents and prophylaxis against opportunistic infections was required as per standard of care. Responders could continue treatment in a blinded fashion for an additional 6 months. At 6 months, those with SD were unblinded and offered IM862 5 mg QOD (placebo group) or IM862 at either 5 mg QOD or 10 mg tid (already on IM862). The higher dose was added in a May 2000 amendment. At the time of PD, patients were also unblinded and offered IM862 5 mg QOD (placebo group) or IM862 10 mg tid (already on IM862). Patients whose treatment continued past 24 weeks were considered to have completed the blinded phase. Missed doses and treatment discontinuations are described through 24 weeks. Adverse events are reported up to 30 days past the end of 24 weeks.
Efficacy Criteria
Mucocutaneous response to therapy was evaluated using the AIDS Clinical Trials Group criteria.7 Complete response (CR) was defined as resolution of all disease for 4 weeks or more without evidence of new disease. In patients with 50 lesions, a CR required all 50 lesions to have resolved. In patients with residual pigmentation, all lesions had to be completely flat with pathologic remission documented in a representative lesion. PR was defined as a reduction in the product of the bi-dimensional measurements of the target lesions by 50% or complete flattening of ≥ 50% of the lesions, without evidence of new disease, lasting for 4 weeks. PD was defined as the development of new lesions, increase in the bi-dimensional measurements of the target lesions by 25% or more, development of new or worsening of existing tumor-associated edema, or development of visceral disease. SD was defined as not fulfilling CR, PR, or PD criteria.
Investigators prospectively recorded the nature, number, and size of the lesions. Each of the three members of the protocol team (A.N., P.G., and D.S.) made a clinical response assignment by reviewing individual data reported in an anonymous format as culled from the case report forms. The protocol team was blinded to the treatment assignments. The three authors discussed discrepancies and reached a consensus. In no cases was there an unresolved response assignment when reviewing cases for response.
Statistical Methods
The sample size was based on detecting a 20% increase in response rate from 15% on placebo to 35% on IM862 at the two-sided 0.05 significance level with power of 0.90.9 Two hundred patients, evenly divided between the placebo and IM862, were required for the study.
Randomization was done centrally at Cytran (Kirkland, WA). The randomization was stratified by CD4 count, HIV viral load, and prior systemic chemotherapy. Randomization was not stratified by site, as site of registration was not anticipated to affect the results.
The binomial proportion and its 95% CI were used to estimate the response rates. Response duration and time to disease progression were estimated using the Kaplan-Meier method,10 using Greenwood's formula11 for the standard error of the cumulative proportion progression-free.12 The 95% CIs for the median time to progression were estimated using the method of Brookmeyer and Crowley.13 For categoric data, the two treatment groups were compared using Fisher's exact test and {chi}2 tests.14 For continuous variables, the Wilcoxon rank sum test was used to compare the two treatment groups.15 The log-rank test was used to compare the two groups with respect to duration of response and time to PD.12
An interim analysis was performed after 50% of the patients had completed therapy and could be evaluated for response, to determine if IM862 demonstrated a significantly different response rate from placebo. Stopping boundaries were derived based on the Lan-Demets procedure16 with the O'Brien-Fleming stopping rule.17 To retain an overall significance level of 0.05, the statistical significance levels for the interim and final analyses were 0.00155 and 0.049, respectively. The interim analysis was reviewed by the NCI-designated Data Safety Monitoring Committee for the AIDS Malignancy Consortium.
An intent-to-treat analysis was used in this trial to assess response. The time to first response was calculated from the date of protocol initiation to the onset of PR or CR. Duration of response was calculated from the onset of PR or CR to PD in responding patients. Time to PD was calculated from beginning of therapy to first evidence of PD or loss to follow-up.
All adverse events reported by the patient or observed by the investigator were collected from the case report forms. An adverse event was defined as any adverse change from the patient's baseline condition, whether it was considered related to treatment or not. Each event was graded according to the NCI Common Toxicity Criteria grading system.18 Unexpected grade 3 and 4 adverse experiences were reported by either telephone, e-mail, or fax within 24 hours of occurrence. In addition, the investigator immediately reported all fatal, life-threatening, or serious expected or unexpected adverse experiences by telephone to the Cytran clinical monitor or regulatory director. Information was reported, as required, to the Center for Drug Evaluation and Research within the US Food and Drug Administration. Fatal, life-threatening, and serious adverse experiences were also reported to the individual institutional review boards by the principal investigator.
RESULTS
The two treatment groups (104 patients on IM862 and 98 on placebo) were comparable with respect to demographic and baseline characteristics except for edema (Table 1). Edema was present at baseline in 34 (32.7%) of 104 IM862 patients and 20 (20.4%) of 98 placebo patients (P = .057). All five patients in the intravenous drug user Centers for Disease Control risk group were in the placebo group. The median duration of KS from diagnosis to study entry was not recorded. The median entry CD4+ lymphocyte count was 295 and 274 cells/μL in the IM862 and placebo groups, respectively, with 39 (38%) on IM862 and 36 (37%) on placebo having CD4+ lymphocyte counts of < 200 cells/μL. Median HIV viral load did not differ between the two groups, nor did the percentage of patients with HIV viral load less than 500 or 5,000 copies/mL. A prior AIDS-defining OI was reported in 111 patients (58%), of whom 42 reported multiple prior AIDS-defining OIs. The two groups were balanced for prior AIDS-defining OIs: 54 IM862 patients (52%) and 57 placebo patients (48%; P = .398). The most common OIs included Pneumocystis carinii pneumonia in 19% patients and chronic herpes simplex ulcers, bronchitis, pneumonitis, or esophagitis in 16%. The two treatment groups were balanced for patients with > 50 lesions: 46 IM862 patients (44%) and 40 placebo patients (41%; P = .670).
Local therapy had been used by about half of the patients for their KS (Table 1). Prior investigational therapy was reported by 22% versus 26% in IM862 and placebo groups, respectively (P = .62). Noninvestigational therapy was also evenly distributed between the two arms (76% v 74%, IM862 v placebo; P = .86). Prior chemotherapy was also balanced, with the most commonly reported agents as liposomal doxorubicin or daunorubicin and paclitaxel.
Almost all patients received antiretroviral therapy (Table 1). The duration of the current antiretroviral therapy was required to be greater than 8 weeks. Additional information on the duration of therapy was not recorded, as its impact was not anticipated at the time of study design. However, while prior antiretroviral therapy was 87% in both treatment groups, current antiretroviral therapy was less common in the IM862 (88%) versus placebo group (96%; P = .042). Current protease inhibitors were reported for 63 IM862 (61%) and 68 placebo patients (69%; P = .238).
Tumor Response
There were no CRs. Both treatment arms were comparable (Table 2) with respect to PR rate (IM862, 23%; 95% CI, 15% to 32%) versus placebo (21%; 95% CI, 14% to 31%; P = .46). The median response duration was not reached in this study and the response durations did not differ between groups (29+ v 12+ weeks; P = .48). The median time to response was shorter with IM862 (8.5 v 14.0 weeks; P = .024), but the time to PD was statistically significantly shorter on IM862 then placebo (P = .012; hazard ratio = 1.66; 95% CI, 1.10 to 2.50), with a median of 16 weeks on IM862 (95% CI, 13 to 27 weeks) and 35 weeks (95% CI, 26 to 114 weeks) on placebo (Fig 1).13 The cumulative proportion progression free at 24 weeks was 42% on IM862 (95% CI, 33% to 52%) and 61% on placebo (95% CI, 51% to 71%). For the subset of patients who were currently on antiretroviral therapy, IM862 remained associated with a shorter time to PD (P = .031; hazard ratio = 1.60; 95% CI, 1.04 to 2.45).
Duration of Therapy and Adverse Events
Patients were to receive 24 weeks of therapy unless KS progression or an adverse event required discontinuation. The median durations of treatment were 19.5 and 24 weeks on IM862 and placebo, respectively (P = .38). Forty-nine IM862 (47%) and 53 placebo (54%) patients completed the 6-month blinded phase of the trial (Table 3). Grade 3 or 4 toxicities were reported in 16% of IM862 and 20% of placebo patients (P = .47; data not shown). Seven IM862 patients (6.7%) experienced infection-related adverse events (six grade 3 and one grade 4). Eleven infection-related events were reported in eight placebo patients (8.2%; nine grade 3 and two grade 4). One placebo patient had three grade 3 events and one patient had two grade 3 events. There was no statistical difference between the two arms with respect to the proportion of patients who reported infection-related adverse events (P = .79).
Only one IM862 patient and four placebo patients terminated study drug therapy due to adverse events. Six patients, three on each treatment arm, discontinued therapy for other reasons. Of the three IM862 patients, two were noncompliant and one was incarcerated without access to study medication. Of the three placebo patients, one was ineligible and never received study drug, one was noncompliant, and one withdrew from the study. One death on the placebo arm was attributed to pneumonia.
Missed doses during the 24-week blinded period were reported by 26 IM862 patients (25%) and 27 placebo patients (28%). Five IM862 patients and four placebo patients missed doses due to adverse events.
Effect on IM862 on CD4+ Cell Count and HIV Viral Load
Parameters for total CD4, CD3, and CD8 cells did not vary with time on study or between treatments (data not shown). HIV viral load data was required by protocol at baseline and at each 4-week follow-up. Data was only available for a limited number of patients at any time point. At any given time point, the percent with viral load < 500 copies/μL did not vary between the two treatment groups. Moreover, when individuals were evaluated for changes in viral loads, there was no correlation between KS response and HIV viral load measurements (data not shown).
DISCUSSION
Time to progression in the current trial was actually statistically significantly shorter on the IM862 arm as compared to placebo, suggesting that IM862 may accelerate time to KS progression. However, time to progression was censored for 104 patients (51.5%) as the study was designed to have a 24-week study duration, without a requirement for additional follow-up. It is notable, however, that with IM862, the shorter time to progression was most evident in those with viral loads < 5,000 copies/mL. The 24-week cumulative proportions progression free, and their 95% CI for each viral load/treatment group combination, is as follows: IM862 with viral load less than 5,000: 0.483 (95% CI, 0.359 to 0.606); placebo with viral load less than 5,000: 0.731 (96% CI, 0.617 to 0.844); IM862 with viral load greater or equal to 5,000: 0.314 (95% CI, 0.154 to 0.473); and placebo with viral load greater or equal to 5,000: 0.335 (95% CI, 0.154 to 0.515). It is plausible that highly active antiretroviral therapy (HAART) is most effective in slowing KS progression in patients with relatively low HIV viral loads, but IM862 might suppress the anti-KS activity of HAART through unknown mechanisms.
Several biologic end points considered successful for clinical trials were satisfied in preclinical testing: bioavailability, dose response relationships, and lack of toxicity. In vitro studies3 demonstrated dose-dependent inhibition of angiogenesis in the chorioallantoic membrane assay with complete inhibition of -FGF and VEGF-induced angiogenesis, as did in vivo preclinical models of other cancers including melanoma, lung, sarcoma, and hepatoma. Murine KS xenograft studies have also shown activity (Parkash Gill, personal communication, 2003). In addition, the mechanism of action of IM862 is now more clearly understood, with at least some of the activity ascribed to decreased production of VEGF19 and enhancement of NK cell function.4 However, it has been recently reported that NK cells are not infected by KS herpesvirus and that NK cell number does not correlate with the development of KS. Nonetheless, this does not abrogate the possibility that NK cell function is influential in KS regression.20 One potential explanation of IM862 inefficacy in the current trial may be dose related. Indeed, in vivo and in vitro experiments performed well in the execution of this protocol, suggesting that higher doses of IM862 in the range of 50 to 110 mg/kg may be associated with a greater antitumor effect.4 Alternatively, it is possible that the in vitro and in vivo KS models do not predict for human in vivo activity. Several other factors may have also played a role, including possible nonadherence to the dosing schedule in patients already taking multiple drugs, or forgetting to take the medication in a 5-day-on and 5-day-off schedule since patients cannot tie their medication intake to a specific activity of daily living. Nasal absorption may have been erratic due to inadequate placement of the drug in the nostril, inadequate amount being delivered or impedance of absorption due to congestion, irritation after repeated dosing, or usage of illicit drugs such as cocaine. Finally, it is possible that weight-based dosing may have been more appropriate and effective. Of the IM862 patients, the response rate was 12 (42.9%) of 28 of those weighing ≤ 65 kg and 12 (15.8%) of 76 of those > 65 kg (P = .007). Among placebo patients, the response rate was three (15.0%) of 20 in patients weighing ≤ 65 kg, and 18 (23.1%) of 78 of patients > 65 kg (P = .551).
The incidence of KS has dramatically decreased in the HAART era.21-24 This directly relates to the role of HIV in the pathogenesis of AIDS-KS.1,25,26 Consequently, treatment of HIV infection could lead to AIDS-KS regression through immune reconstitution and reduction in HIV cofactors which contribute to KS pathogenesis. As such, any imbalance in the two study arms with respect to HAART therapy could lead to a bias. Notably, prior antiretroviral therapy was 87% in both treatment groups, and almost all patients were receiving antiretroviral therapy with protease inhibitor usage being evenly distributed (Table 1). Current antiretroviral therapy was less common in the IM862 (88%) versus placebo group (96%; P = .042), possibly biased towards the IM862 group. However, when the analysis was restricted to patients currently on HAART, the treatment difference in time to progression persisted. Finally, 51% of patients in both groups had viral loads < 500 copies/mL at baseline, suggesting that those with optimal viral load were equally distributed. It is notable, however, that the proportion of IM862 patients with edema (32.7%) versus placebo (20.4%) had a trend towards statistical significance (P = .057). It is highly probably that patients with more advanced KS will not respond to HAART alone.27
Although small studies have documented KS response to HAART,28-30 our trial is one of the largest studies to date suggesting a significant ongoing response rate in AIDS-KS to HAART alone. The current study suggests that the PR rate to HAART alone is approximately 22% in a population such as ours. Patients had ongoing responses at 6 months in both arms (29+ v 12+ weeks, IM862 v placebo; 0.478; Table 2). CRs were not seen, but this may reflect the heavily pretreated patient population with 75% having received prior cytotoxic therapy and 24% prior investigational agents. While we hypothesize that these responses were a result of HAART, it would have been impossible to have a no HAART arm. We are also unable to correlate response and HIV viral load in the current study. Although this may be, in part, due to an incomplete data set for CD4 and HIV viral load, 8 weeks of stable antiretroviral therapy was an entry requirement, making it likely that most patients would have achieved a maximal suppression of HIV viral load before study entry. In contrast, immunologic reconstitution continues for many months after maximal HIV suppression. It is generally only after 8 weeks of HAART that thymic generation of new T cells becomes substantial, leading to increased numbers of naive T cells and expansion of the T-cell receptor repertoire. Neither parameter was assessed in the course of our study, so it is difficult to definitively state that an ongoing change in immunity accompanied prolonged HAART. However, the known immunologic responsiveness of KS and prior studies on prolonged immunologic regeneration on HAART strongly argue for a sustained HAART-related impact on AIDS-KS. It is highly unlikely that this spontaneous regression is simply a reflection of the natural course of the disease. In addition, our response rate is similar to 18% placebo response rate reported in a 12-week, randomized, double-blind 268 patient study of topical alitretinoin (9-cis-retinoic acid) 0.1%. No information is available on HIV viral loads due to the timeframe of the study, but 75% were receiving HAART therapy and one third had a change in their antiviral therapy during the study period.
Several factors might be implicated in the improvement of KS after HAART alone. Protease inhibitors may have a direct action on HHV-8 replication,31 although this is in dispute.32 Second, HAART mediated HIV replication suppression reduces the production of HIV-Tat protein implicated in the pathogenesis of KS.33,34 Perhaps, most significantly, AIDS-KS is related to a persistent herpesvirus infection. Similar to other human diseases associated with persistent herpesviruses, it is likely that immune control of infected cells is critical to control of the disease. Immune restoration during HAART may be reasonably assumed to indirectly affect the neoplasm through altered targeting of HHV-8 expressing cells known to compose AIDS-KS lesions.35
At 6 months, patients on our trial with SD were unblinded and offered IM862 5 mg QOD (placebo group) or IM862 at either 5 mg QOD or 10 mg tid (already on IM862). This may appear not to have been such a beneficial action in view of the trial results. Unblinded availability of IM862 after the 24-week study period was incorporated in to the trial design under the assumption that the study drug would prove to be active and that placebo patients with SD would be given an opportunity to receive the study drug to affect a response. Similarly, patients on study drug with SD would be given an opportunity to either prolong the SD phase (continued 5 mg QOD) or possibly achieve a response at a higher dose (10-mg tid dose). The risk of toxicity in these unblinded patients beginning, continuing, or escalating study drug was anticipated to be low based on previous studies; continued monitoring was mandated by protocol. Finally, physician investigators and HIV community representatives alike felt that the option of receiving study drug after unblinding would assist with study accrual in a placebo-controlled trial. Perhaps in future placebo-controlled studies with unblinded cross-over, it should be emphasized in the consent form that subjects may be receiving an inactive compound if they choose to receive the study drug after unblinding.
In summary, IM862 at the dose and schedule used did not add a benefit beyond that seen with HAART alone. In this study, HAART alone was associated with a prolonged, potent response in AIDS-KS. These data do not preclude that other antiangiogenesis strategies, or even this compound at other dose levels or schedules, might be active in KS. However, this study points to the necessity of considering the impact of HIV control on AIDS-KS when designing therapeutic trials, and particular attention should be given to the duration of the concurrent HAART therapy. Trials could specify a longer time on the current HAART regimen or a stratification for various durations of therapy. Further, these data argue for initiating or optimizing HAART therapy in patients with AIDS-KS and uncontrolled HIV viral loads before initiating other therapy.
Appendix
AMC sites: Beth Israel Deaconess Medical Center: B. Dezube and J. Proper; Boston Medical Center: T. Cooley and R. Haivanis; Case Western Reserve University: S. Remick and K. Coviello; The Johns Hopkins University: R. Ambinder and L. Post; Massachusetts General Hospital: D. Scadden and K. Shea; Memorial Sloan-Kettering Cancer Center: S. Krown, A. Noy, and A. Martelli; Northwestern University: J. Von Roenn and L. Volini; New York University: A. Friedman-Kien and T. Spoto; Ohio State University: M. Caligiuri and P. Snider; Pennsylvania Oncology Hematology Associates: D. Henry and H. Cohen; San Francisco General Hospital: L. Kaplan and L. Marsco; University of California Los Angeles: S. Miles; University of California San Diego Cancer Center: M. Saville and K. Smith; University of Medicine and Dentistry of New Jersey: T. Cheung and T. Saunders; University of Miami: M. Fischl and L. Thompson; University of Southern California Norris Cancer Hospital: A. Tulpule and B. Espina; Vanderbilt University Medical Center: P. Browning and K. Willenbert; Virginia Mason Medical Center: D. Aboulafia and C. Weaver; and Washington University: L. Ratner.
Non-AMC sites: AIDS Healthcare Foundation: C. Farthing; Alfred Hospital: M. Chipman and J. Roney; British Columbia Cancer Agency: B. Melosky and S. Monkman; Centre Hospitalier Universitaire St-Pierre, Medecine Interne/Maladies: N. Clumeck and E. O'Doherty; Institute of Tropical Medicine: R. Colebunders; Markus Conant,: M. Conant and C. Eden; Montreal General Hospital: C. Tsoukas and M. D'Astous; Prince of Wales Hospital: D. Goldstein and S. Ryan; St. Vincent's Hospital: S. Milliken and G. Dolan; Sunnybrook & Women's College Health Science Center: A. Rachlis and M. Bast; Taylor Square Private Clinic: N. Bodsworth and R. Richardson; Ottawa Hospital: S. Kravcik and N. Foss; Toronto General Hospital: S. Walmsley and A. Foster; Tower ID Medical Associates: P. Ruane.
Authors' Disclosures of Potential Conflicts of Interest
NOTES
Supported by the Ontario HIV Treatment Network Career Scientist Award (S.W.), grants from the National Cancer Institute (U01CA083118, U01CA070079, U01CA070081, U01CA070068, U01CA070075, U01CA070047, U01CA070054, U01CA070072, U01CA070080, U01CA070081, U01CA083035, U01CA071375, U01CA083038, U01CA070062, U01CA070019), and Cytran, Inc.
Presented in part at the International Conference on Malignancies in AIDS and Other Immune Deficiencies, April 22-24, 2002. Bethesda, MD.
Terms in blue are defined in the glossary, found at the end of this issue and online at www.jco.org.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Scadden DT: AIDS-related malignancies. Annu Rev Med 54:285-303, 2003
2. Folkman J: Role of angiogenesis in tumor growth and metastasis. Semin Oncol 29:15-18, 2002
3. Investigator Brochure: IM862 (small peptide). Ed 6. Kirland, WA, Cytran Ltd., 1998
4. Smith DL, Cai J, Zhu S, et al: Natural killer cell cytolytic activity is necessary for in vivo antitumor activity of the dipeptide L-glutamyl-L-tryptophan. Int J Cancer 106:528-533, 2003
5. Tulpule A, Espina B, Cabriales S: IM-862 nasal solution is an active antiangiogenic agent in the treatment of AIDS-related Kaposi's sarcoma. Proc Am Soc Clin Oncol:18:535a, 1999 (abstr 2065)
6. Tulpule A, Scadden DT, Espina BM, et al: Results of a randomized study of IM862 nasal solution in the treatment of AIDS-related Kaposi's sarcoma. J Clin Oncol 18:716-723, 2000
7. Krown SE, Testa MA, Huang J: AIDS-related Kaposi's sarcoma: Prospective validation of the AIDS Clinical Trials Group staging classification. AIDS Clinical Trials Group Oncology Committee. J Clin Oncol 15:3085-3092, 1997
8. Green S, Weiss GR: Southwest Oncology Group standard response criteria, endpoint definitions and toxicity criteria. Invest New Drugs 10:239-253, 1992
9. Fleiss J, Tytun A, Ury S: A simple approximation for calculating sample sizes for comparing independent proportions. Biometrics:343-346, 1980
10. Kaplan E, Meier P: Nonparametric estimator from incomplete observations. J Am Stat Assoc 53:457-481, 1958
11. Clark AGV: Survival distributions: Reliability Applications in the Biomedical Sciences. New York, John Wiley & Sons, 1975
12. Lee E: Statistical methods for survival data analysis. Belmont, California, Lifetime Learning Publications, 1980
13. Brookmeyer R: A confidence interval for the median survival time. Biometrics 38:29-41, 1982
14. Fleiss J: Statistical methods for rates and proportions, 2nd edition. New York, John Wiley & Sons, 1981
15. Armitage P: Statistical methods in medical research. New York, John Wiley & Sons, 1974
16. Lan KKG, DeMets DL: Discrete sequential boundaries for clinical trials. Biometrika 70:659-663, 1983
17. O'Brien PC, Fleming TR: A multiple testing procedure for clinical trials. Biometrics 35:549-556, 1979
18. Cancer Therapy Evaluation Program: Common Toxicity Criteria, version 2.0. Bethesda, MD, National Cancer Institute, 1998
19. Ferrario A, von Tiehl KF, Rucker N, et al: Antiangiogenic treatment enhances photodynamic therapy responsiveness in a mouse mammary carcinoma. Cancer Res 60:4066-4069, 2000
20. Stebbing J, Gazzard B, Flore O, et al: Natural killer cells are not infected by Kaposi's sarcoma-associated herpesvirus in vivo, and natural killer cell counts do not correlate with the risk of developing Kaposi's sarcoma. AIDS 17:1998-2000, 2003
21. International Collaboration on HIV and Cancer: Highly active antiretroviral therapy and incidence of cancer in human immunodeficiency virus-infected adults. J Natl Cancer Inst 92:1823-1830, 2000
22. Ives NJ, Gazzard BG, Easterbrook PJ: The changing pattern of AIDS-defining illnesses with the introduction of highly active antiretroviral therapy (HAART)in a London clinic. J Infect 42:134-139, 2001
23. Grulich A, Li Y, McDonald A, et al: Decreasing rates of Kaposi's sarcoma and non-Hodgkin's lymphoma in the era of potent combination anti-retroviral therapy. AIDS 15:629-633, 2001
24. Portsmouth S, Stebbing J, Gill J, et al: A comparison of regimens based on non-nucleoside reverse transcriptase inhibitors or protease inhibitors in preventing Kaposi's sarcoma. AIDS 17:F17-F22, 2003
25. Noy A: Update in Kaposi sarcoma. Curr Opin Oncol 15:379-381, 2003
26. Krown SE: Therapy of AIDS-associated Kaposi's sarcoma: Targeting pathogenetic mechanisms. Hematol Oncol Clin North Am 17:763-783, 2003
27. Krown SE: Highly active antiretroviral therapy in AIDS-associated Kaposi's sarcoma: Implications for the design of therapeutic trials in patients with advanced, symptomatic Kaposi's sarcoma. J Clin Oncol 22:399-402, 2004
28. Paparizos VA, Kyriakis KP, Papastamopoulos V, et al: Response of AIDS-associated Kaposi sarcoma to highly active antiretroviral therapy alone. J Acquir Immune Defic Syndr 30:257-258, 2002
29. Cattelan AM, Calabro ML, Gasperini P, et al: Acquired immunodeficiency syndrome-related Kaposi's sarcoma regression after highly active antiretroviral therapy: Biologic correlates of clinical outcome. J Natl Cancer Inst Monogr:44-49, 2001
30. Gill J, Bourboulia D, Wilkinson J, et al: Prospective study of the effects of antiretroviral therapy on Kaposi sarcoma–associated herpesvirus infection in patients with and without Kaposi sarcoma. J Acquir Immune Defic Syndr 31:384-390, 2002
31. Varthakavi V, Browning PJ, Spearman P: Human immunodeficiency virus replication in a primary effusion lymphoma cell line stimulates lytic-phase replication of Kaposi's sarcoma-associated herpesvirus. J Virol 73:10329-10338, 1999
32. De Milito A, Catucci M, Venturi G, et al: Antiretroviral therapy with protease inhibitors in human immunodeficiency virus type 1- and human herpesvirus 8-coinfected patients. J Med Virol 57:140-144, 1999
33. Deregibus MC, Cantaluppi V, Doublier S, et al: HIV-1-Tat protein activates phosphatidylinositol 3-kinase/ AKT-dependent survival pathways in Kaposi's sarcoma cells. J Biol Chem 277:25195-25202, 2002
34. Barillari G, Ensoli B: Angiogenic effects of extracellular human immunodeficiency virus type 1 Tat protein and its role in the pathogenesis of AIDS-associated Kaposi's sarcoma. Clin Microbiol Rev 15:310-326, 2002
35. Lederman MM, Valdez H: Immune restoration with antiretroviral therapies: Implications for clinical management. JAMA 284:223-228, 2000(Ariela Noy, David T. Scad)
USC Keck School of Medicine, Norris Cancer Institute, Los Angeles, CA
Massachusetts General Hospital, Harvard Medical School
Beth Israel Deaconess Medical Center, Boston, MA
AMC Operations and Statistical Center, University of Birmingham, Birmingham, AL
University of Washington
Virginia Mason Clinic, Seattle, WA
University of Toronto, Toronto Hospital, Toronto, Ontario, Canada
ABSTRACT
PATIENTS AND METHODS: Two hundred two HIV-positive patients were enrolled, 104 on IM862 and 98 on placebo.
RESULTS: Baseline characteristics were comparable except current antiretroviral therapy: 88% versus 96% (IM862 v placebo group; P = .042). The median treatment durations were 19.5 versus 24 weeks (IM862 v placebo). No significant difference was detected in response rate (IM862, 23%; 95% CI, 15% to 32% v placebo, 21%; 95% CI, 14% to 31%; P = .46), time to response (8.5 weeks v 14 weeks; P = .024), or duration of response. However, IM862 was associated with both a shorter time to response (8.5 weeks v 14 weeks; P = .024) and shorter median time to progression (16 weeks, 95% CI, 13 to 27 weeks v 35 weeks, 95% CI, 26 to 114 weeks; P = .012).
CONCLUSION: Despite promising phase I and phase II studies, IM862 5 mg every other day was not superior to placebo and may accelerate time to progression. Highly active antiretroviral therapy alone was associated with a substantial rate of sustained tumor response and may have contributed to prior estimates of IM862 response. Therapeutic trials for AIDS-Kaposi's sarcoma must account for ongoing immune reconstitution in the setting of concurrent highly active antiretroviral therapy that may confound estimates of therapeutic activity.
INTRODUCTION
IM862 is a synthetic dipeptide (L-glutamine L-tryptophan) with a half-life of minutes and no known toxic metabolites. It was originally isolated in Russia from bovine thymus during a search for immune adjuvants and was studied in Russia as an adjuvant to chemotherapy and in parasitic and fungal disorders. No toxicities were reported at doses in excess of 2,000-fold, the dose reported in the current trial.3
Although IM862 exhibits no direct cytotoxic effects, in in vitro studies3 IM862 exhibited dose dependent inhibition of angiogenesis in the chorioallantoic membrane assay in a range of 1 to 1000 μg/mL with complete inhibition of -FGF and VEGF-induced angiogenesis (1 to 100 μg/disc). The mechanism of action was not related to inhibition of endothelial proliferation as demonstrated in human umbilical vein endothelial cell or in Boyden chamber assays.
Although the exact mechanism of action remains unclear, IM862 exhibited activity in in vivo preclinical models.3 It inhibited melanoma growth in fetal mice, prevented neovascularization in a dose-dependent manner in the Lewis Lung tumor model, and prevented growth of sarcoma and hepatoma implants.3 In other experiments, IM862 was effective in a xenograft Kaposi's sarcoma (KS) model (Parkash Gill, personal communication, 2003). At least some of the activity has been ascribed to decreased production of VEGF and enhancement of natural killer (NK) cell function.4
Based on these findings, IM862 was studied in AIDS-related KS, a highly vascular tumor associated with human herpesvirus-8. While cytotoxic agents and interferon are active in KS, side effects may outweigh the benefits in patients with less advanced disease. Consequently, non-toxic, easily administered agents would be welcomed if effective and well tolerated over a long period of time. In animals, antitumor activity of IM862 was similar, irrespective of the route of administration: intranasal, subcutaneous, intravenous, or intramuscular.3 Given the ease of administration and a bioavailability of 71% after intranasal administration, this route was adopted in human trials.5
In a phase I/II trial, intranasal IM862 was administered at a dose of either 5 mg every other day (QOD; n = 26) or 5 consecutive days followed by 5 days without treatment (n = 18). Forty-seven percent of patients had more than 50 mucocutaneous lesions, CD4+ T-lymphocyte counts were > 200 cells/μL in 55%, and concurrent protease inhibitors had been used for a median of 10 months (range, 0 to 24 months) before study entry in 89% of patients. Of 44 patients, 16 (36%) patients had an overall response (complete response [CR] + partial response [PR]), with CR in five patients, PR in 11 patients, and stable disease (SD) in 21 patients. The median time to response was 6 weeks (range, 3 to 26 weeks), with the median response duration equal to 33+ weeks (range, 12+ to 95+ weeks). No significant toxicities were noted.6 Given the high benefit-to-risk ratio in the phase I/II study, a placebo-controlled, double-blind, phase III trial in AIDS-related KS was initiated using the 5 mg QOD intranasal dosing over a 24-week period.
PATIENTS AND METHODS
The institutional review boards of each participating institution reviewed and approved the protocol and consent form and all patients gave written informed consent before enrollment.
Patient Evaluation
Patients underwent a complete physical examination, WHO performance status evaluation, CBC, serum chemistries analysis, and toxicity assessment at study entry and then monthly for 6 months. CD4+ cell count and HIV viral load were followed prospectively every 2 months by standard methods at each individual institution.
Patients were staged according to a modification of the AIDS Clinical Trials Group–based tumor, immune, systemic staging system.7 As many as 50 cutaneous lesions were evaluated for their character (raised or flat) and size. If there were more than 50 cutaneous lesions, a representative anatomic area was selected and the lesions were counted and characterized. In addition, five discrete, nodular, indicator lesions were selected, and the sum of the products of their largest perpendicular diameters was recorded. All adverse events were graded using National Cancer Institute (NCI) Common Toxicity Criteria.8
Treatment Plan
IM862 was self-administered intranasally at a dose of 5 mg in 0.7 mL of diluent QOD for 24 weeks or until disease progression (PD). The diluent vehicle served as the placebo. Concurrent use of antiretroviral agents and prophylaxis against opportunistic infections was required as per standard of care. Responders could continue treatment in a blinded fashion for an additional 6 months. At 6 months, those with SD were unblinded and offered IM862 5 mg QOD (placebo group) or IM862 at either 5 mg QOD or 10 mg tid (already on IM862). The higher dose was added in a May 2000 amendment. At the time of PD, patients were also unblinded and offered IM862 5 mg QOD (placebo group) or IM862 10 mg tid (already on IM862). Patients whose treatment continued past 24 weeks were considered to have completed the blinded phase. Missed doses and treatment discontinuations are described through 24 weeks. Adverse events are reported up to 30 days past the end of 24 weeks.
Efficacy Criteria
Mucocutaneous response to therapy was evaluated using the AIDS Clinical Trials Group criteria.7 Complete response (CR) was defined as resolution of all disease for 4 weeks or more without evidence of new disease. In patients with 50 lesions, a CR required all 50 lesions to have resolved. In patients with residual pigmentation, all lesions had to be completely flat with pathologic remission documented in a representative lesion. PR was defined as a reduction in the product of the bi-dimensional measurements of the target lesions by 50% or complete flattening of ≥ 50% of the lesions, without evidence of new disease, lasting for 4 weeks. PD was defined as the development of new lesions, increase in the bi-dimensional measurements of the target lesions by 25% or more, development of new or worsening of existing tumor-associated edema, or development of visceral disease. SD was defined as not fulfilling CR, PR, or PD criteria.
Investigators prospectively recorded the nature, number, and size of the lesions. Each of the three members of the protocol team (A.N., P.G., and D.S.) made a clinical response assignment by reviewing individual data reported in an anonymous format as culled from the case report forms. The protocol team was blinded to the treatment assignments. The three authors discussed discrepancies and reached a consensus. In no cases was there an unresolved response assignment when reviewing cases for response.
Statistical Methods
The sample size was based on detecting a 20% increase in response rate from 15% on placebo to 35% on IM862 at the two-sided 0.05 significance level with power of 0.90.9 Two hundred patients, evenly divided between the placebo and IM862, were required for the study.
Randomization was done centrally at Cytran (Kirkland, WA). The randomization was stratified by CD4 count, HIV viral load, and prior systemic chemotherapy. Randomization was not stratified by site, as site of registration was not anticipated to affect the results.
The binomial proportion and its 95% CI were used to estimate the response rates. Response duration and time to disease progression were estimated using the Kaplan-Meier method,10 using Greenwood's formula11 for the standard error of the cumulative proportion progression-free.12 The 95% CIs for the median time to progression were estimated using the method of Brookmeyer and Crowley.13 For categoric data, the two treatment groups were compared using Fisher's exact test and {chi}2 tests.14 For continuous variables, the Wilcoxon rank sum test was used to compare the two treatment groups.15 The log-rank test was used to compare the two groups with respect to duration of response and time to PD.12
An interim analysis was performed after 50% of the patients had completed therapy and could be evaluated for response, to determine if IM862 demonstrated a significantly different response rate from placebo. Stopping boundaries were derived based on the Lan-Demets procedure16 with the O'Brien-Fleming stopping rule.17 To retain an overall significance level of 0.05, the statistical significance levels for the interim and final analyses were 0.00155 and 0.049, respectively. The interim analysis was reviewed by the NCI-designated Data Safety Monitoring Committee for the AIDS Malignancy Consortium.
An intent-to-treat analysis was used in this trial to assess response. The time to first response was calculated from the date of protocol initiation to the onset of PR or CR. Duration of response was calculated from the onset of PR or CR to PD in responding patients. Time to PD was calculated from beginning of therapy to first evidence of PD or loss to follow-up.
All adverse events reported by the patient or observed by the investigator were collected from the case report forms. An adverse event was defined as any adverse change from the patient's baseline condition, whether it was considered related to treatment or not. Each event was graded according to the NCI Common Toxicity Criteria grading system.18 Unexpected grade 3 and 4 adverse experiences were reported by either telephone, e-mail, or fax within 24 hours of occurrence. In addition, the investigator immediately reported all fatal, life-threatening, or serious expected or unexpected adverse experiences by telephone to the Cytran clinical monitor or regulatory director. Information was reported, as required, to the Center for Drug Evaluation and Research within the US Food and Drug Administration. Fatal, life-threatening, and serious adverse experiences were also reported to the individual institutional review boards by the principal investigator.
RESULTS
The two treatment groups (104 patients on IM862 and 98 on placebo) were comparable with respect to demographic and baseline characteristics except for edema (Table 1). Edema was present at baseline in 34 (32.7%) of 104 IM862 patients and 20 (20.4%) of 98 placebo patients (P = .057). All five patients in the intravenous drug user Centers for Disease Control risk group were in the placebo group. The median duration of KS from diagnosis to study entry was not recorded. The median entry CD4+ lymphocyte count was 295 and 274 cells/μL in the IM862 and placebo groups, respectively, with 39 (38%) on IM862 and 36 (37%) on placebo having CD4+ lymphocyte counts of < 200 cells/μL. Median HIV viral load did not differ between the two groups, nor did the percentage of patients with HIV viral load less than 500 or 5,000 copies/mL. A prior AIDS-defining OI was reported in 111 patients (58%), of whom 42 reported multiple prior AIDS-defining OIs. The two groups were balanced for prior AIDS-defining OIs: 54 IM862 patients (52%) and 57 placebo patients (48%; P = .398). The most common OIs included Pneumocystis carinii pneumonia in 19% patients and chronic herpes simplex ulcers, bronchitis, pneumonitis, or esophagitis in 16%. The two treatment groups were balanced for patients with > 50 lesions: 46 IM862 patients (44%) and 40 placebo patients (41%; P = .670).
Local therapy had been used by about half of the patients for their KS (Table 1). Prior investigational therapy was reported by 22% versus 26% in IM862 and placebo groups, respectively (P = .62). Noninvestigational therapy was also evenly distributed between the two arms (76% v 74%, IM862 v placebo; P = .86). Prior chemotherapy was also balanced, with the most commonly reported agents as liposomal doxorubicin or daunorubicin and paclitaxel.
Almost all patients received antiretroviral therapy (Table 1). The duration of the current antiretroviral therapy was required to be greater than 8 weeks. Additional information on the duration of therapy was not recorded, as its impact was not anticipated at the time of study design. However, while prior antiretroviral therapy was 87% in both treatment groups, current antiretroviral therapy was less common in the IM862 (88%) versus placebo group (96%; P = .042). Current protease inhibitors were reported for 63 IM862 (61%) and 68 placebo patients (69%; P = .238).
Tumor Response
There were no CRs. Both treatment arms were comparable (Table 2) with respect to PR rate (IM862, 23%; 95% CI, 15% to 32%) versus placebo (21%; 95% CI, 14% to 31%; P = .46). The median response duration was not reached in this study and the response durations did not differ between groups (29+ v 12+ weeks; P = .48). The median time to response was shorter with IM862 (8.5 v 14.0 weeks; P = .024), but the time to PD was statistically significantly shorter on IM862 then placebo (P = .012; hazard ratio = 1.66; 95% CI, 1.10 to 2.50), with a median of 16 weeks on IM862 (95% CI, 13 to 27 weeks) and 35 weeks (95% CI, 26 to 114 weeks) on placebo (Fig 1).13 The cumulative proportion progression free at 24 weeks was 42% on IM862 (95% CI, 33% to 52%) and 61% on placebo (95% CI, 51% to 71%). For the subset of patients who were currently on antiretroviral therapy, IM862 remained associated with a shorter time to PD (P = .031; hazard ratio = 1.60; 95% CI, 1.04 to 2.45).
Duration of Therapy and Adverse Events
Patients were to receive 24 weeks of therapy unless KS progression or an adverse event required discontinuation. The median durations of treatment were 19.5 and 24 weeks on IM862 and placebo, respectively (P = .38). Forty-nine IM862 (47%) and 53 placebo (54%) patients completed the 6-month blinded phase of the trial (Table 3). Grade 3 or 4 toxicities were reported in 16% of IM862 and 20% of placebo patients (P = .47; data not shown). Seven IM862 patients (6.7%) experienced infection-related adverse events (six grade 3 and one grade 4). Eleven infection-related events were reported in eight placebo patients (8.2%; nine grade 3 and two grade 4). One placebo patient had three grade 3 events and one patient had two grade 3 events. There was no statistical difference between the two arms with respect to the proportion of patients who reported infection-related adverse events (P = .79).
Only one IM862 patient and four placebo patients terminated study drug therapy due to adverse events. Six patients, three on each treatment arm, discontinued therapy for other reasons. Of the three IM862 patients, two were noncompliant and one was incarcerated without access to study medication. Of the three placebo patients, one was ineligible and never received study drug, one was noncompliant, and one withdrew from the study. One death on the placebo arm was attributed to pneumonia.
Missed doses during the 24-week blinded period were reported by 26 IM862 patients (25%) and 27 placebo patients (28%). Five IM862 patients and four placebo patients missed doses due to adverse events.
Effect on IM862 on CD4+ Cell Count and HIV Viral Load
Parameters for total CD4, CD3, and CD8 cells did not vary with time on study or between treatments (data not shown). HIV viral load data was required by protocol at baseline and at each 4-week follow-up. Data was only available for a limited number of patients at any time point. At any given time point, the percent with viral load < 500 copies/μL did not vary between the two treatment groups. Moreover, when individuals were evaluated for changes in viral loads, there was no correlation between KS response and HIV viral load measurements (data not shown).
DISCUSSION
Time to progression in the current trial was actually statistically significantly shorter on the IM862 arm as compared to placebo, suggesting that IM862 may accelerate time to KS progression. However, time to progression was censored for 104 patients (51.5%) as the study was designed to have a 24-week study duration, without a requirement for additional follow-up. It is notable, however, that with IM862, the shorter time to progression was most evident in those with viral loads < 5,000 copies/mL. The 24-week cumulative proportions progression free, and their 95% CI for each viral load/treatment group combination, is as follows: IM862 with viral load less than 5,000: 0.483 (95% CI, 0.359 to 0.606); placebo with viral load less than 5,000: 0.731 (96% CI, 0.617 to 0.844); IM862 with viral load greater or equal to 5,000: 0.314 (95% CI, 0.154 to 0.473); and placebo with viral load greater or equal to 5,000: 0.335 (95% CI, 0.154 to 0.515). It is plausible that highly active antiretroviral therapy (HAART) is most effective in slowing KS progression in patients with relatively low HIV viral loads, but IM862 might suppress the anti-KS activity of HAART through unknown mechanisms.
Several biologic end points considered successful for clinical trials were satisfied in preclinical testing: bioavailability, dose response relationships, and lack of toxicity. In vitro studies3 demonstrated dose-dependent inhibition of angiogenesis in the chorioallantoic membrane assay with complete inhibition of -FGF and VEGF-induced angiogenesis, as did in vivo preclinical models of other cancers including melanoma, lung, sarcoma, and hepatoma. Murine KS xenograft studies have also shown activity (Parkash Gill, personal communication, 2003). In addition, the mechanism of action of IM862 is now more clearly understood, with at least some of the activity ascribed to decreased production of VEGF19 and enhancement of NK cell function.4 However, it has been recently reported that NK cells are not infected by KS herpesvirus and that NK cell number does not correlate with the development of KS. Nonetheless, this does not abrogate the possibility that NK cell function is influential in KS regression.20 One potential explanation of IM862 inefficacy in the current trial may be dose related. Indeed, in vivo and in vitro experiments performed well in the execution of this protocol, suggesting that higher doses of IM862 in the range of 50 to 110 mg/kg may be associated with a greater antitumor effect.4 Alternatively, it is possible that the in vitro and in vivo KS models do not predict for human in vivo activity. Several other factors may have also played a role, including possible nonadherence to the dosing schedule in patients already taking multiple drugs, or forgetting to take the medication in a 5-day-on and 5-day-off schedule since patients cannot tie their medication intake to a specific activity of daily living. Nasal absorption may have been erratic due to inadequate placement of the drug in the nostril, inadequate amount being delivered or impedance of absorption due to congestion, irritation after repeated dosing, or usage of illicit drugs such as cocaine. Finally, it is possible that weight-based dosing may have been more appropriate and effective. Of the IM862 patients, the response rate was 12 (42.9%) of 28 of those weighing ≤ 65 kg and 12 (15.8%) of 76 of those > 65 kg (P = .007). Among placebo patients, the response rate was three (15.0%) of 20 in patients weighing ≤ 65 kg, and 18 (23.1%) of 78 of patients > 65 kg (P = .551).
The incidence of KS has dramatically decreased in the HAART era.21-24 This directly relates to the role of HIV in the pathogenesis of AIDS-KS.1,25,26 Consequently, treatment of HIV infection could lead to AIDS-KS regression through immune reconstitution and reduction in HIV cofactors which contribute to KS pathogenesis. As such, any imbalance in the two study arms with respect to HAART therapy could lead to a bias. Notably, prior antiretroviral therapy was 87% in both treatment groups, and almost all patients were receiving antiretroviral therapy with protease inhibitor usage being evenly distributed (Table 1). Current antiretroviral therapy was less common in the IM862 (88%) versus placebo group (96%; P = .042), possibly biased towards the IM862 group. However, when the analysis was restricted to patients currently on HAART, the treatment difference in time to progression persisted. Finally, 51% of patients in both groups had viral loads < 500 copies/mL at baseline, suggesting that those with optimal viral load were equally distributed. It is notable, however, that the proportion of IM862 patients with edema (32.7%) versus placebo (20.4%) had a trend towards statistical significance (P = .057). It is highly probably that patients with more advanced KS will not respond to HAART alone.27
Although small studies have documented KS response to HAART,28-30 our trial is one of the largest studies to date suggesting a significant ongoing response rate in AIDS-KS to HAART alone. The current study suggests that the PR rate to HAART alone is approximately 22% in a population such as ours. Patients had ongoing responses at 6 months in both arms (29+ v 12+ weeks, IM862 v placebo; 0.478; Table 2). CRs were not seen, but this may reflect the heavily pretreated patient population with 75% having received prior cytotoxic therapy and 24% prior investigational agents. While we hypothesize that these responses were a result of HAART, it would have been impossible to have a no HAART arm. We are also unable to correlate response and HIV viral load in the current study. Although this may be, in part, due to an incomplete data set for CD4 and HIV viral load, 8 weeks of stable antiretroviral therapy was an entry requirement, making it likely that most patients would have achieved a maximal suppression of HIV viral load before study entry. In contrast, immunologic reconstitution continues for many months after maximal HIV suppression. It is generally only after 8 weeks of HAART that thymic generation of new T cells becomes substantial, leading to increased numbers of naive T cells and expansion of the T-cell receptor repertoire. Neither parameter was assessed in the course of our study, so it is difficult to definitively state that an ongoing change in immunity accompanied prolonged HAART. However, the known immunologic responsiveness of KS and prior studies on prolonged immunologic regeneration on HAART strongly argue for a sustained HAART-related impact on AIDS-KS. It is highly unlikely that this spontaneous regression is simply a reflection of the natural course of the disease. In addition, our response rate is similar to 18% placebo response rate reported in a 12-week, randomized, double-blind 268 patient study of topical alitretinoin (9-cis-retinoic acid) 0.1%. No information is available on HIV viral loads due to the timeframe of the study, but 75% were receiving HAART therapy and one third had a change in their antiviral therapy during the study period.
Several factors might be implicated in the improvement of KS after HAART alone. Protease inhibitors may have a direct action on HHV-8 replication,31 although this is in dispute.32 Second, HAART mediated HIV replication suppression reduces the production of HIV-Tat protein implicated in the pathogenesis of KS.33,34 Perhaps, most significantly, AIDS-KS is related to a persistent herpesvirus infection. Similar to other human diseases associated with persistent herpesviruses, it is likely that immune control of infected cells is critical to control of the disease. Immune restoration during HAART may be reasonably assumed to indirectly affect the neoplasm through altered targeting of HHV-8 expressing cells known to compose AIDS-KS lesions.35
At 6 months, patients on our trial with SD were unblinded and offered IM862 5 mg QOD (placebo group) or IM862 at either 5 mg QOD or 10 mg tid (already on IM862). This may appear not to have been such a beneficial action in view of the trial results. Unblinded availability of IM862 after the 24-week study period was incorporated in to the trial design under the assumption that the study drug would prove to be active and that placebo patients with SD would be given an opportunity to receive the study drug to affect a response. Similarly, patients on study drug with SD would be given an opportunity to either prolong the SD phase (continued 5 mg QOD) or possibly achieve a response at a higher dose (10-mg tid dose). The risk of toxicity in these unblinded patients beginning, continuing, or escalating study drug was anticipated to be low based on previous studies; continued monitoring was mandated by protocol. Finally, physician investigators and HIV community representatives alike felt that the option of receiving study drug after unblinding would assist with study accrual in a placebo-controlled trial. Perhaps in future placebo-controlled studies with unblinded cross-over, it should be emphasized in the consent form that subjects may be receiving an inactive compound if they choose to receive the study drug after unblinding.
In summary, IM862 at the dose and schedule used did not add a benefit beyond that seen with HAART alone. In this study, HAART alone was associated with a prolonged, potent response in AIDS-KS. These data do not preclude that other antiangiogenesis strategies, or even this compound at other dose levels or schedules, might be active in KS. However, this study points to the necessity of considering the impact of HIV control on AIDS-KS when designing therapeutic trials, and particular attention should be given to the duration of the concurrent HAART therapy. Trials could specify a longer time on the current HAART regimen or a stratification for various durations of therapy. Further, these data argue for initiating or optimizing HAART therapy in patients with AIDS-KS and uncontrolled HIV viral loads before initiating other therapy.
Appendix
AMC sites: Beth Israel Deaconess Medical Center: B. Dezube and J. Proper; Boston Medical Center: T. Cooley and R. Haivanis; Case Western Reserve University: S. Remick and K. Coviello; The Johns Hopkins University: R. Ambinder and L. Post; Massachusetts General Hospital: D. Scadden and K. Shea; Memorial Sloan-Kettering Cancer Center: S. Krown, A. Noy, and A. Martelli; Northwestern University: J. Von Roenn and L. Volini; New York University: A. Friedman-Kien and T. Spoto; Ohio State University: M. Caligiuri and P. Snider; Pennsylvania Oncology Hematology Associates: D. Henry and H. Cohen; San Francisco General Hospital: L. Kaplan and L. Marsco; University of California Los Angeles: S. Miles; University of California San Diego Cancer Center: M. Saville and K. Smith; University of Medicine and Dentistry of New Jersey: T. Cheung and T. Saunders; University of Miami: M. Fischl and L. Thompson; University of Southern California Norris Cancer Hospital: A. Tulpule and B. Espina; Vanderbilt University Medical Center: P. Browning and K. Willenbert; Virginia Mason Medical Center: D. Aboulafia and C. Weaver; and Washington University: L. Ratner.
Non-AMC sites: AIDS Healthcare Foundation: C. Farthing; Alfred Hospital: M. Chipman and J. Roney; British Columbia Cancer Agency: B. Melosky and S. Monkman; Centre Hospitalier Universitaire St-Pierre, Medecine Interne/Maladies: N. Clumeck and E. O'Doherty; Institute of Tropical Medicine: R. Colebunders; Markus Conant,: M. Conant and C. Eden; Montreal General Hospital: C. Tsoukas and M. D'Astous; Prince of Wales Hospital: D. Goldstein and S. Ryan; St. Vincent's Hospital: S. Milliken and G. Dolan; Sunnybrook & Women's College Health Science Center: A. Rachlis and M. Bast; Taylor Square Private Clinic: N. Bodsworth and R. Richardson; Ottawa Hospital: S. Kravcik and N. Foss; Toronto General Hospital: S. Walmsley and A. Foster; Tower ID Medical Associates: P. Ruane.
Authors' Disclosures of Potential Conflicts of Interest
NOTES
Supported by the Ontario HIV Treatment Network Career Scientist Award (S.W.), grants from the National Cancer Institute (U01CA083118, U01CA070079, U01CA070081, U01CA070068, U01CA070075, U01CA070047, U01CA070054, U01CA070072, U01CA070080, U01CA070081, U01CA083035, U01CA071375, U01CA083038, U01CA070062, U01CA070019), and Cytran, Inc.
Presented in part at the International Conference on Malignancies in AIDS and Other Immune Deficiencies, April 22-24, 2002. Bethesda, MD.
Terms in blue are defined in the glossary, found at the end of this issue and online at www.jco.org.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Scadden DT: AIDS-related malignancies. Annu Rev Med 54:285-303, 2003
2. Folkman J: Role of angiogenesis in tumor growth and metastasis. Semin Oncol 29:15-18, 2002
3. Investigator Brochure: IM862 (small peptide). Ed 6. Kirland, WA, Cytran Ltd., 1998
4. Smith DL, Cai J, Zhu S, et al: Natural killer cell cytolytic activity is necessary for in vivo antitumor activity of the dipeptide L-glutamyl-L-tryptophan. Int J Cancer 106:528-533, 2003
5. Tulpule A, Espina B, Cabriales S: IM-862 nasal solution is an active antiangiogenic agent in the treatment of AIDS-related Kaposi's sarcoma. Proc Am Soc Clin Oncol:18:535a, 1999 (abstr 2065)
6. Tulpule A, Scadden DT, Espina BM, et al: Results of a randomized study of IM862 nasal solution in the treatment of AIDS-related Kaposi's sarcoma. J Clin Oncol 18:716-723, 2000
7. Krown SE, Testa MA, Huang J: AIDS-related Kaposi's sarcoma: Prospective validation of the AIDS Clinical Trials Group staging classification. AIDS Clinical Trials Group Oncology Committee. J Clin Oncol 15:3085-3092, 1997
8. Green S, Weiss GR: Southwest Oncology Group standard response criteria, endpoint definitions and toxicity criteria. Invest New Drugs 10:239-253, 1992
9. Fleiss J, Tytun A, Ury S: A simple approximation for calculating sample sizes for comparing independent proportions. Biometrics:343-346, 1980
10. Kaplan E, Meier P: Nonparametric estimator from incomplete observations. J Am Stat Assoc 53:457-481, 1958
11. Clark AGV: Survival distributions: Reliability Applications in the Biomedical Sciences. New York, John Wiley & Sons, 1975
12. Lee E: Statistical methods for survival data analysis. Belmont, California, Lifetime Learning Publications, 1980
13. Brookmeyer R: A confidence interval for the median survival time. Biometrics 38:29-41, 1982
14. Fleiss J: Statistical methods for rates and proportions, 2nd edition. New York, John Wiley & Sons, 1981
15. Armitage P: Statistical methods in medical research. New York, John Wiley & Sons, 1974
16. Lan KKG, DeMets DL: Discrete sequential boundaries for clinical trials. Biometrika 70:659-663, 1983
17. O'Brien PC, Fleming TR: A multiple testing procedure for clinical trials. Biometrics 35:549-556, 1979
18. Cancer Therapy Evaluation Program: Common Toxicity Criteria, version 2.0. Bethesda, MD, National Cancer Institute, 1998
19. Ferrario A, von Tiehl KF, Rucker N, et al: Antiangiogenic treatment enhances photodynamic therapy responsiveness in a mouse mammary carcinoma. Cancer Res 60:4066-4069, 2000
20. Stebbing J, Gazzard B, Flore O, et al: Natural killer cells are not infected by Kaposi's sarcoma-associated herpesvirus in vivo, and natural killer cell counts do not correlate with the risk of developing Kaposi's sarcoma. AIDS 17:1998-2000, 2003
21. International Collaboration on HIV and Cancer: Highly active antiretroviral therapy and incidence of cancer in human immunodeficiency virus-infected adults. J Natl Cancer Inst 92:1823-1830, 2000
22. Ives NJ, Gazzard BG, Easterbrook PJ: The changing pattern of AIDS-defining illnesses with the introduction of highly active antiretroviral therapy (HAART)in a London clinic. J Infect 42:134-139, 2001
23. Grulich A, Li Y, McDonald A, et al: Decreasing rates of Kaposi's sarcoma and non-Hodgkin's lymphoma in the era of potent combination anti-retroviral therapy. AIDS 15:629-633, 2001
24. Portsmouth S, Stebbing J, Gill J, et al: A comparison of regimens based on non-nucleoside reverse transcriptase inhibitors or protease inhibitors in preventing Kaposi's sarcoma. AIDS 17:F17-F22, 2003
25. Noy A: Update in Kaposi sarcoma. Curr Opin Oncol 15:379-381, 2003
26. Krown SE: Therapy of AIDS-associated Kaposi's sarcoma: Targeting pathogenetic mechanisms. Hematol Oncol Clin North Am 17:763-783, 2003
27. Krown SE: Highly active antiretroviral therapy in AIDS-associated Kaposi's sarcoma: Implications for the design of therapeutic trials in patients with advanced, symptomatic Kaposi's sarcoma. J Clin Oncol 22:399-402, 2004
28. Paparizos VA, Kyriakis KP, Papastamopoulos V, et al: Response of AIDS-associated Kaposi sarcoma to highly active antiretroviral therapy alone. J Acquir Immune Defic Syndr 30:257-258, 2002
29. Cattelan AM, Calabro ML, Gasperini P, et al: Acquired immunodeficiency syndrome-related Kaposi's sarcoma regression after highly active antiretroviral therapy: Biologic correlates of clinical outcome. J Natl Cancer Inst Monogr:44-49, 2001
30. Gill J, Bourboulia D, Wilkinson J, et al: Prospective study of the effects of antiretroviral therapy on Kaposi sarcoma–associated herpesvirus infection in patients with and without Kaposi sarcoma. J Acquir Immune Defic Syndr 31:384-390, 2002
31. Varthakavi V, Browning PJ, Spearman P: Human immunodeficiency virus replication in a primary effusion lymphoma cell line stimulates lytic-phase replication of Kaposi's sarcoma-associated herpesvirus. J Virol 73:10329-10338, 1999
32. De Milito A, Catucci M, Venturi G, et al: Antiretroviral therapy with protease inhibitors in human immunodeficiency virus type 1- and human herpesvirus 8-coinfected patients. J Med Virol 57:140-144, 1999
33. Deregibus MC, Cantaluppi V, Doublier S, et al: HIV-1-Tat protein activates phosphatidylinositol 3-kinase/ AKT-dependent survival pathways in Kaposi's sarcoma cells. J Biol Chem 277:25195-25202, 2002
34. Barillari G, Ensoli B: Angiogenic effects of extracellular human immunodeficiency virus type 1 Tat protein and its role in the pathogenesis of AIDS-associated Kaposi's sarcoma. Clin Microbiol Rev 15:310-326, 2002
35. Lederman MM, Valdez H: Immune restoration with antiretroviral therapies: Implications for clinical management. JAMA 284:223-228, 2000(Ariela Noy, David T. Scad)