Phase 1/2 trial of BMS-275291 in patients with human immunodeficiency virus-related Kaposi sarcoma†‡
A multicenter trial of the AIDS Malignancy Consortium
See editorial and companion article on pages 962–5 and 1147–52, this issue.
The following investigators participated in the study: Memorial Sloan-Kettering Cancer Center, New York, NY (S. Krown); Montefiore Hospital, Bronx, NY (J. Sparano); Northwestern University School of Medicine, Chicago, IL (J. Von Roenn); Ohio State University, Columbus, OH (M. Shah); Pennsylvania Hospital, Philadelphia, PA (D. Henry); University of California, San Francisco, CA (L. Kaplan); and University of Miami, Miami, FL (M. Fischl)
Matrix metalloproteinases (MMPs) are overexpressed in Kaposi sarcoma (KS). The safety and efficacy of a novel, orally bioavailable MMP inhibitor, BMS-275291, was evaluated in patients with human immunodeficiency virus-associated KS and the correlation between changes in the percentage of apoptotic cells in tumor biopsies and response was explored.
Cohorts of 6 patients were to be treated with BMS-275291. The initial cohort received 1200 mg once a day; subsequent doses were to be escalated to 600 mg twice daily and 1200 mg twice daily, or decreased to 600 mg/day. Tumor biopsies for apoptosis assays were collected pretreatment and on Day 29. Prospectively defined dose level adjustments were to be based on dose-limiting toxicity (DLT), tolerability, changes in the percentage of apoptotic cells, and treatment response.
Sixteen patients were enrolled; 15 received the study drug and could be evaluated. The median duration of treatment was 20 weeks (range, 3–54 weeks). A dose of 1200 mg once a day was well tolerated but induced only 1 response. A DLT occurred in 3 patients treated with 600 mg twice daily, and included grade 3 fatigue, grade 3 allergic reaction, and grade 3 arthralgias; 2 responses were noted at this dose level (toxicity was graded according to the National Cancer Institute Common Toxicity Criteria [version 2.0]). Based on predetermined endpoints, the trial was closed after accrual of 15 treated patients. Assessment of biologic response for dose escalation/de-escalation decisions utilizing the apoptosis assay was not feasible.
BMS-275291 given at a dose of 600 mg twice daily induced unacceptable toxicity. The better-tolerated schedule of 1200 mg once a day demonstrated inadequate efficacy in patients with human immunodeficiency virus-associated KS. The apoptosis assay was not helpful in predicting response. Cancer 2008. © 2008 American Cancer Society.
Kaposi sarcoma (KS) is a multifocal angioproliferative malignancy with a variable clinical presentation. Although its incidence has declined in developed countries since the introduction of highly active antiretroviral therapy (HAART), the standardized incidence ratio for KS among people with the acquired immunodeficiency syndrome (AIDS) remained >3600-fold higher in the post-HAART period (1996–2002) than in the general population in the U.S,1 and recent data suggest that incidence rates may be increasing.2 Although active treatments for KS are available,3 the growing understanding of KS pathogenesis provides a rationale for new therapeutic approaches.4, 5
KS lesions are characterized by aberrant angiogenesis, inflammation, and proliferation of endothelial-derived spindle cells.6, 7 The KS-associated herpesvirus (KSHV) is required for KS development. KSHV encodes novel proteins as well as viral homologs of human proteins known to be involved in signaling and cancer. These induce signaling events that lead to induction of cytokines, such as vascular endothelial growth factor (VEGF), and induce constitutive expression of matrix metalloproteinases (MMPs).7–12 In a murine model, the tissue inhibitor of MMP-2 was shown to inhibit the development of KS-like lesions, suggesting that inhibition of MMP activity might have therapeutic value for patients with AIDS-associated KS.13
MMPs are a group of more than 20 structurally related zinc-dependent endopeptidases that degrade proteins of the extracellular matrix and basement membrane. They mediate normal tissue remodeling and are essential for embryogenesis, wound healing, and angiogenesis; MMP-2 and MMP-9 are of particular interest because preclinical models suggest that they are essential components of the metastatic process.14, 15
The potential of MMP inhibitors (MMPIs) as therapeutic agents for cancer has been investigated for more than 20 years. Early-generation MMPIs demonstrated modest antitumor activity, but their usefulness was limited because of poor oral bioavailability and dose-limiting polyarthritis.16–18 MMPI-induced arthralgias and myalgias are believed to be caused by inhibition of sheddases, which are MMPs that regulate shedding of cell surface mediators of inflammatory cytokines such as tumor necrosis factor (TNF)-α, the TNF-α receptor, and the interleukin (IL)-6 receptor.
BMS-275291 is an orally bioavailable, rationally designed, nonpeptidomimetic MMPI. Inhibition is selective for MMPs-1, -2, -7, -9, -13, and -14, while sparing sheddases.19 BMS-275291 induced a dose-dependent reduction in angiogenesis and tumor metastasis in animal models20, 21 and induced limited toxicity in a phase 1 study in patients with nonhuman immunodeficiency virus (HIV)-related cancers who received doses of up to 2400 mg/day.22
The major objectives of the current study were to evaluate the safety and toxicity of BMS-275291 in patients with AIDS-associated KS, to explore the relation between tumor response and the change in the percentage of apoptotic cells in KS biopsies, and to determine whether this biologic endpoint could be used as a criterion for dose escalation or de-escalation.
MATERIALS AND METHODS
Patients were required to have biopsy-proven KS with at least 5 measurable cutaneous lesions; serologic documentation of HIV infection; age ≥16 years; life expectancy ≥3 months; a Karnofsky performance status ≥60; and adequate hepatic, renal, and hematopoietic function. Women of childbearing potential were required to have a negative serum β-human chorionic gonadotropin within 72 hours of study entry and to practice adequate birth control.
HAART was permitted but not required; patients receiving HAART were required to be on a stable regimen for at least 4 weeks without evidence of KS regression. Antineoplastic therapies within 3 weeks of study entry were not permitted.
Subjects with symptomatic visceral KS requiring chemotherapy, concurrent infection, another neoplasm, treatment with corticosteroids or investigational drugs, or previous local therapy to KS indicator lesions were excluded. Because MMPs influence bone development, nursing women were excluded. All patients provided written informed consent in accordance with the human experimentation guidelines of the U.S. Department of Health and Human Services and the Human Investigations Committee at each participating site.
BMS-275291 was supplied as 600-mg tablets by the Cancer Therapy Evaluation Program of the National Cancer Institute (NCI) through a collaborative agreement with Bristol Myers-Squibb (Wallingford, CT). Cohorts of 6 subjects were to be sequentially assigned to 1 of 3 dose levels: level I, a single dose of 1200 mg/day; level II, 600 mg twice daily; and level III, 1200 mg twice daily. Doses were to be reduced for a dose-limiting toxicity (DLT), defined as a ≥grade 3 nonhematologic toxicity, grade 4 thrombocytopenia, or recurrent grade 4 granulocytopenia or hemoglobin <8 g/dL while receiving growth factor support. For subjects enrolled on level I, the dose could be de-escalated to level 0 (600 mg/day).
The decision to escalate the dose to the next level, or to expand the number of patients in a given cohort, was to be based on the number of DLTs, the number of objective tumor responses, and the number of patients whose tumor biopsies demonstrated a biologic response, as measured by the percent change in apoptotic cells. The occurrence of DLT in more than 1 of 6 subjects within the first 28 days of treatment halted further dose escalation. If fewer than 2 of 6 subjects demonstrated objective tumor responses or fewer than 5 of 6 subjects demonstrated a biologic response, the dose level was considered ineffective. If there was evidence for efficacy at level I, additional patients would be entered at level 0.
Treatment was discontinued for progressive disease or unacceptable toxicity. With this study design, the probability of terminating the study due to toxicity was .89 if the underlying rate of a DLT was .50.
Schedule of Events
Baseline assessments included a history and physical examination, a complete blood count with differential, serum chemistries, prothrombin time, and partial thromboplastin time. Blood tests were repeated every 2 weeks for 3 months, and then every 4 weeks for 6 months. CD4 counts and HIV-1 plasma RNA levels were measured before treatment, on Day 29, and every 3 months thereafter. A chest X-ray was performed at baseline and 30 days after the last treatment. Additional tests to evaluate visceral KS were performed as required.
Punch biopsies (4-mm) of KS lesions were obtained from nonindicator lesions at baseline and on Day 29. Formalin-fixed, paraffin-embedded blocks were sectioned for morphologic examination and immunohistochemistry for KSHV LANA (ORF73) as reported previously.23 A terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay was used for in situ detection of apoptosis at the single cell level24 using the DermaTACS kit from Trevigen (Gaithersburg, MD). The total number of apoptotic cells was counted in the spindle cells in the dermis (excluding keratinocytes) in a single 5-μM thick section of the 4-mm biopsy. Apoptotic keratinocytes served as an internal control to ensure nucleic acid integrity. Cases in which no apoptotic keratinocytes were identified in spite of 3 independent TUNEL assays were considered noninformative.
Descriptive statistics were used in this study. The Wilcoxon signed rank test was used to evaluate changes from baseline in CD4 count. Toxicity was assessed using the NCI Common Toxicity Criteria (version 2.0). Tumor assessments were performed at baseline, on Days 15 and 29 of treatment, and every 28 days thereafter. Tumor assessments were based on measurements of 5 cutaneous marker lesions and on the overall number and characteristics of cutaneous lesions. Response to treatment was assessed according to previously described criteria.25
Sixteen patients from 8 participating AIDS Malignancy Consortium (AMC) sites were enrolled in the study between June 26, 2001, and July 25, 2002. One patient never received study treatment and was excluded from all analyses. Baseline characteristics are shown in Table 1. All participants were male, with a median age of 40.9 years and a median CD4 count of 263 cells/μL. Six patients (40%) were receiving HAART, 5 (33.3%) had ≥50 cutaneous lesions, 3 (20%) had tumor-associated edema, and none had visceral disease. All but 1 patient had previously received treatment for KS (Table 2).
Table 1. Baseline Characteristics
| White, non-Hispanic||10|
| Black, non-Hispanic||1|
| Asian/Pacific Islander||1|
|Median age (range), y||40.9 (32–60)|
|Median CD4+ T-cell count (range), cells/μL||263 (15–707)|
|No. (%) receiving HAART||6 (40)|
|KS lesions (no.)|
| ≥50 skin lesions||5|
| Visceral disease||0|
| Tumor-associated edema||3|
| Oral cavity lesions||0|
Table 2. Prior Therapy for Kaposi Sarcoma
|Any prior therapy||14 (93)|
|Local therapy||3 (20)|
| Radiotherapy||2 (13.3)|
| Cryotherapy||1 (6.7)|
| Intralesional biologic therapy||1 (6.7)|
| Topical therapy||1 (6.7)|
|Systemic therapy||11 (73)|
| Interferon-α||5 (33)|
| Doxorubicin||1 (6.7)|
| Liposomal doxorubicin||9 (60)|
| Paclitaxel||1 (6.7)|
| Etoposide||1 (6.7)|
| Liposomal daunorubicin||1 (6.7)|
| Other||1 (6.7)|
|Investigational therapy||7 (47)|
| IM862||5 (33)|
| 9-cis-retinoic acid||1 (6.7)|
| Other||1 (6.7)|
The first 6 subjects were enrolled on dose level I (1200 mg once a day). Because no DLTs developed during the first 28 days of treatment, the next 7 patients were enrolled on dose level II (600 mg twice daily). Because DLT was observed in 3 patients at this dose level, the next 2 patients were enrolled on dose level I. Based on the prospectively defined parameters, the trial was stopped due to the number of DLTs at dose level II and the documentation of only 1 response at dose level I.
Subjects treated at level I were found to have no grade 3 toxicity during the first 4 weeks of treatment. Three subjects who received divided-dose treatment (level II) experienced DLTs: grade 3 fatigue, grade 3 allergic reaction (including fevers), and grade 3 arthralgias in 1 patient each. No grade 4 toxicities were reported. Neither HAART treatment nor baseline CD4 counts was found to be correlated with the development of DLTs.
One study subject died during treatment with BMS-275291. The death was accidental and unrelated to HIV infection, KS, or study treatment.
Few laboratory abnormalities were observed. Grade 3 toxicities at level I included thrombocytopenia and elevated transaminase levels in 1 patient each. Grade 3 toxicities at level II included granulocytopenia, neutropenia, and anemia in 1 patient each. No grade 4 laboratory abnormalities were reported.
Effects on CD4 + T-cell Counts
Eleven patients had paired baseline and Day 29 CD4 T-cell counts. The median change was 26 (range, −262 to 230) (P = .814). Insufficient numbers of patients had paired baseline and on-study measurements of HIV RNA levels with which to reach meaningful conclusions regarding the effects on viral load.
Response to Therapy
Of the 16 study subjects, 1 (Patient 12) never received study drug and 1 patient received <4 weeks of treatment, leaving 14 subjects whose KS response could be evaluated (Table 3). Three of 14 patients (21%) achieved a partial response, 1 of whom was treated with single-dose 1200 mg/day and 2 of whom were treated with 600 mg twice daily. The median duration of response was 9 months (range, 6–12 months). Eleven of 14 patients (73%) had stable disease, with median treatment duration of 3 months (range, 2–12 months). The median treatment duration for all patients was 5 months (range, <1–12 months). Six patients remained on therapy at the time of study closure, after a median treatment duration of 7 months (range, 4–12 months).
Table 3. Response Summary
|1||1||Partial response||12||Study discontinued|
|Time to PR: 12 mo|
|PR duration: 1+ mo|
|2||1||Stable disease||3||Disease progression|
|3||1||Stable disease||10||Study discontinued|
|4||1||Stable disease||6||Patient withdrawal|
|5||1||Stable disease||10||Study discontinued|
|6||1||Stable disease||2||Disease progression|
|7||2||NA||<1||Grade 3 fatigue|
|8||2||Partial response||8||Disease progression|
|Time to PR: 1 mo||Grade 3 arthralgia|
|PR duration: 8 mo|
|9||2||Stable disease||3||Grade 3 allergic reaction|
|10||2||Stable disease||8||Study discontinued|
|11||2||Partial response||6||Study discontinued|
|Time to PR: 1 mo|
|PR duration: 7+ mo|
|13||2||Stable disease||3||Disease progression|
|14||2||Stable disease||4||Study discontinued|
|15||1||Stable disease||3||Death (train accident)|
|16||1||Stable disease||2||Disease progression|
Effects on Apoptosis
Of 28 biopsies evaluated using the TUNEL assay, 17 were informative. The average number of apoptotic cells in pretreatment biopsies was 7, and in Day 29 biopsies it was 5.5. Five patients had paired informative biopsies: 1 patient (Patient 6, whose best response was stable disease) demonstrated a decrease from 18 to 4 apoptotic cells; 2 patients (Patients 4 and 11, with stable disease and a partial response, respectively) demonstrated an increase in the number of apoptotic cells from 5 to 15 and from 2 to 6, respectively; and 2 patients (Patients 1 and 2, with partial response and stable disease, respectively) were found to have no change, and had 1 to 3 apoptotic cells in both Day 0 and 29 biopsies. Thus, there was no correlation noted between the absolute number or change in the number of apoptotic cells and response to therapy.
MMPs have been implicated in the progression of malignant disease, and their overexpression has been correlated with tumor progression and disease recurrence in several cancer types.26–28 Although MMP inhibitors reduced tumor size and metastasis in animal models,20, 21 they have rarely induced regression of advanced cancers in humans.29–32 However, because of their potential effects on tumor vasculature and metastasis, MMPIs are being evaluated as maintenance therapy and for treatment of relatively indolent malignancies.
In the current study, 15 patients were treated with BMS-275291 at a dose of 1200 mg/day, administered either as a single dose (level I) or a divided dose of 600 mg twice daily (level II). Only 1 objective response was noted at level I, a dosage associated with acceptable toxicity, whereas 2 objective responses were documented among patients enrolled on dose level II, which proved unacceptably toxic. Therefore, the study was closed based on predetermined criteria. The reason for the increased toxicity observed when the same total daily dose was given in divided doses is unknown; the significance of this finding is difficult to interpret given the small number of patients treated, and may have occurred by chance.
Hepatic and hematologic toxicities were the major laboratory toxicities observed, and may have been related to the underlying HIV infection or its treatment. In other studies, the toxicity profile for this agent has been inconsistent. In a phase 1 trial, no DLT was observed in 40 healthy volunteers at doses ranging from 150 to 1200 mg/day for 14 days.33 In a phase 1 trial in 44 patients with advanced cancers, DLTs were observed at a dose of 600 mg/day (grade 3 transaminase elevation) and 1200 mg/day (grade 3 rash and grade 4 shortness of breath), but were not observed in patients treated at doses of 900 mg/day, 1800 mg/day, or 2400 mg/day.22
Severe adverse clinical events in this trial included fatigue, an allergic reaction, and arthralgias. Dose-limiting polyarthritis, a well-recognized toxicity of early-generation peptidomimetic MMPIs, was not observed in this study, nor was it observed in a study of patients with hormone-refractory prostate cancer who received BMS-275291 doses of 1200 mg once or twice daily.32 Although the lack of polyarthritis was presumed to be a consequence of BMS-275291's nonpeptidomimetic structure and its increased specificity for selected MMPs, transient grade 1 and 2 myalgia/arthralgias were the most frequently reported adverse events among 44 patients with advanced cancers treated in a phase 1 trial22 and grade ≥2 arthralgia was reported in 32% of women with early-stage breast cancer treated with BMS-275291, compared with only 16.7% of subjects who received placebo.34
Although the objective response rate of HIV-associated KS to BMS-275291 in this trial was low (3 of 14 patients; 21%), 11 patients were found to have stable disease that had persisted for 4 months, 8 months, 10 months, and 10 months, respectively, in 4 patients who remained on treatment at the time of study closure. Long-term stable disease in patients with early KS might provide clinical benefit, but seeking evidence of such an effect would require extensive resources, particularly in a disease that often has a long natural history. Nonetheless, BMS-275251 was apparently less active than another MMPI, COL-3, which induced objective responses, including some complete responses, in 41% of KS patients treated at a dose of 50 mg/day.35
Unfortunately, apoptotic activity in KS biopsies could not be reliably quantified in a subset of biopsies, most likely due to nucleic acid degradation or tissue overfixation, and in those cases in which it could be reliably quantified, no predictive value was found. Thus, we could not validate apoptosis as an alternative to standard phase 1 endpoints as a criterion for dose modification. Established biologic endpoints to test the efficacy of antiangiogenic agents are lacking, but their identification would allow more rapid and accurate clinical evaluation of biologic agents. Alternative strategies aimed at gaining a better understanding of KS biology and identifying reliable surrogate endpoints for response are being evaluated in ongoing AMC trials.