In this phase 2 study, the activity and tolerability of the combination of bevacizumab, an inhibitor of angiogenesis, and everolimus, an inhibitor of the mammalian target of rapamycin (mTOR), was evaluated in the treatment of patients with metastatic melanoma.
Patients with metastatic melanoma who had received up to 2 previous systemic regimens (chemotherapy and/or immunotherapy) were eligible. Previous treatment with angiogenesis or mTOR inhibitors was not allowed. All patients received bevacizumab at a dose of 15 mg/kg intravenously every 21 days and everolimus at a dose of 10 mg orally daily. Patients were re-evaluated every 6 weeks; those with an objective response or stable disease (according to Response Evaluation Criteria in Solid Tumors [RECIST]) continued therapy until tumor progression or unacceptable toxicity occurred.
Fifty-seven patients with metastatic melanoma received a median of 4 treatment cycles (range, 1-14+ cycles). Seven patients (12%) achieved major responses, whereas 33 patients (58%) were found to have stable disease at the time of first evaluation. The median progression-free and overall survivals were 4 months and 8.6 months, respectively. Approximately 43% of patients were alive after 12 months of follow-up. The treatment regimen was well tolerated by the majority of patients.
The incidence of melanoma has increased during the last decade, and currently accounts for approximately 60,000 new cases and 9000 deaths yearly in the United States.1 Patients who develop metastatic melanoma have a very poor prognosis, with a median survival of 6 to 8 months and a 5-year survival rate of <5%.
Metastatic melanoma is resistant to most cytotoxic agents. A few agents, such as dacarbazine, temozolomide, and paclitaxel, are used for palliative treatment, but these agents have response rates of <15% and have minimal impact on the natural history of the disease.2 As a result, the majority of current clinical research concerning systemic therapy for advanced melanoma is focused on novel targeted agents.
Overexpression of the vascular endothelial growth factor (VEGF) pathway is common in metastatic melanoma, and high serum levels of VEGF represent an adverse prognostic feature.3-6 Several angiogenesis inhibitors have demonstrated single-agent activity, either alone or in combination with cytotoxic agents, and clinical trials of many of these agents are continuing.7-14 The phosphatidyl inositol 3-kinase (PI3K)/AKT signal transduction pathway regulates many basic cellular properties, including apoptosis resistance, survival, and motility. This pathway is up-regulated, by a variety of mechanisms, in many cancers, including metastatic melanoma. Agents that inhibit mammalian target of rapamycin (mTOR), including temsirolimus and everolimus, are effective in inhibiting the PI3K/AKT pathway.15, 16 These agents also result in down-regulation of the VEGF receptor, and may be useful in combination with other antiangiogenesis agents. Although mTOR inhibitors have been to reported to have limited single-agent activity against melanoma,17, 18 the combination of temsirolimus with sorafenib demonstrated some activity in a preliminary report.19
In this phase 2 study, we evaluated the activity and tolerability of the combination of bevacizumab and everolimus in the treatment of patients with metastatic melanoma. We recently piloted this combination in the treatment of metastatic renal cell carcinoma, and found the regimen to be active and well tolerated in patients with that disease.20
MATERIALS AND METHODS
This multicenter, community-based phase 2 trial was initiated in February 2008. Selected sites in the Sarah Cannon Oncology Research Consortium participated in the trial (Table 1). Before patients were enrolled, the trial was approved by the institutional review boards of all participating sites.
Table 1. Sarah Cannon Oncology Research Consortium Participating Sites
Tennessee Oncology, PLLC
Florida Cancer Specialists
Fort Myers, FL
Chattanooga Oncology and Hematology Associates
Virginia Cancer Institute
Center for Cancer and Blood Disorders
Grand Rapids Oncology Program
Grand Rapids, MI
Florida Hospital Cancer Institute
Nebraska Methodist Cancer Center
Consultants in Medical Oncology and Hematology
Drexel Hill, PA
Northeast Georgia Medical Center
Watson Clinic for Cancer Research
Oncology Hematology Associates of Southwest Indiana
South Texas Oncology and Hematology
San Antonio, TX
Gulfcoast Oncology Associates
St. Petersburg, FL
St. Louis Cancer Care
To be eligible for this trial, patients were required to have histologically confirmed, unresectable metastatic melanoma. In addition to those with cutaneous melanomas, patients with ocular and mucosal melanomas were eligible. Patients were allowed to have received up to 2 previous chemotherapy or immunotherapy regimens. However, previous treatment with antiangiogenesis agents or mTOR inhibitors was not allowed. Additional inclusion criteria included measurable disease (using Response Evaluation Criteria in Solid Tumors [RECIST])21; an Eastern Cooperative Oncology Group (ECOG) performance status of 0, 1, or 2; adequate bone marrow function (an absolute neutrophil count ≥1500/μL and a platelet count ≥100,000/μL); serum creatinine ≤2.0 mg/dL; serum bilirubin ≤1.5 mg/dL; and transaminases ≤2.5 × the upper limit of normal (or ≤5 × the upper limit of normal in patients with liver metastases). All patients were required to provide written informed consent before study entry.
Patients were excluded if they had central nervous system involvement by melanoma. In addition, concurrent treatment with drugs known to be inhibitors or inducers of cytochrome P450 3A (CYP3A) enzymes was not allowed. Additional exclusion criteria included treatment with investigational agents within 4 weeks of study entry, female patients who were pregnant or lactating, acute myocardial infarction within 6 months of study entry, clinically significant cardiovascular disease (uncontrolled hypertension, unstable angina, class II or greater congestive heart failure, or serious cardiac arrhythmia), human immunodeficiency virus infection, and any impairment of gastrointestinal function that would alter the absorption or swallowing of oral medications. Specific exclusion criteria related to bevacizumab included a history of hemoptysis or hematemesis, major surgical procedures within 28 days or minor surgical procedures within 7 days of the initiation of treatment, patients with any type of gastrostomy tube, a urine protein:creatinine ratio ≥1.0, and a history of a bleeding diathesis or coagulopathy.
Before the initiation of therapy, all patients underwent a complete medical history, physical examination, complete blood counts, chemistry profile, prothrombin time, partial thromboplastin time, lipid panel, and urinalysis. If patients were found to have >1+ protein on a urinalysis dipstick test, either a urine protein:creatinine ratio or a 24-hour urine collection for protein was performed. Radiologic evaluation included computed tomography (CT) scan of the chest and abdomen, and either CT or magnetic resonance imaging scan of the head. Tumor measurements were performed in all patients before the beginning of therapy.
All patients received treatment with bevacizumab at a dose of 15 mg/kg, administered by intravenous infusion on Day 1 of each 21-day course. In addition, all patients received everolimus at a dose of 10 mg orally on a daily basis. Bevacizumab was infused according to standard guidelines. Everolimus was taken with a whole glass of water by the patient in a fasting state or after no more than a light, fat-free meal. No routine premedications were given with bevacizumab or everolimus.
Before each dose of bevacizumab, patients had complete blood counts and urinalysis performed; a chemistry profile was performed once every 6 weeks. Patients were seen and examined by their physicians on Day 1 of each cycle (every 3 weeks). After 6 weeks of treatment, patients were evaluated for response. At that time, repeat CT scans of the chest and abdomen were performed, as well as any additional scans necessary to include all measurable lesions. Patients with an objective response to treatment or stable disease continued to receive bevacizumab and everolimus. Patients with tumor progression were removed from the study.
For patients continuing treatment, re-evaluations were performed at 6-week intervals. Treatment was continued until disease progression occurred or the patient developed intolerable toxicity.
On the basis of previous clinical experience with bevacizumab, the commonly expected toxicities were proteinuria and hypertension. Before each dose of bevacizumab, patients were tested for proteinuria using dipstick analysis. If the dipstick indicated ≥2+ proteinuria, bevacizumab was withheld and a 24-hour urine collection for protein was obtained. Bevacizumab was continued as long as the 24-hour urine protein remained at <3500 mg/24 hours. If proteinuria was ≥3500 mg/24 hours, bevacizumab was withheld for 2 weeks, and 24-hour urine protein was remeasured. Bevacizumab was reinitiated at the full dose when proteinuria decreased to <3500 mg/24 hours. Patients who developed grade 3 hypertension while receiving bevacizumab were treated with standard antihypertensive therapies. If grade 4 hypertension developed, bevacizumab was discontinued. Bevacizumab was withheld for patients who had grade 3 or 4 bleeding, and was discontinued if any life-threatening hemorrhagic events occurred. With bleeding events of ≤grade 3, bevacizumab could be reinitiated at the discretion of the treating physician after the bleeding episode had resolved. Bevacizumab was also discontinued for any grade 4 toxicities believed to be related to the drug.
A single dose reduction of everolimus to 5 mg daily was allowed. For patients with grade 3 or 4 hematologic toxicity, everolimus therapy was interrupted until the toxicity recovered to ≤grade 2, and then reintroduced at a daily dose of 5 mg. For patients with grade 3 nonhematologic toxicity, everolimus was interrupted until the toxicity improved to ≤grade 1 and then was reinitiated with a dose reduction to 5 mg daily. Patients who developed grade 4 nonhematologic toxicity had everolimus discontinued. Nonhematologic toxicity that failed to improve after a treatment interruption of >2 weeks resulted in the discontinuation of everolimus.
If the discontinuation of bevacizumab or everolimus occurred because of toxicity, patients were allowed to continue the other drug if they were believed to be receiving clinical benefit from treatment.
Definition of Response
All patients were re-evaluated for response after the completion of 6 weeks of treatment, and response categories were assigned using RECIST criteria.21 All patients with major responses had confirmation of response with repeat scans performed at 6-week intervals. Patients with stable disease (ie, those who did not meet RECIST criteria for either partial response or disease progression) at 6 weeks were re-evaluated at 6-week intervals as treatment continued. The final response category assigned to these patients represented the best response obtained during the treatment course.
The primary objective of this phase 2 clinical trial was to assess the activity of the combination of bevacizumab and everolimus in the treatment of patients with metastatic melanoma. The primary endpoint was progression-free survival (PFS). Secondary endpoints included overall survival (OS), objective response rate, and toxicity evaluation.
On the basis of the low levels of activity observed with multiple other regimens in the treatment of metastatic melanoma, as well as the single-agent activity noted with everolimus (a PFS rate of 30% at 16 weeks), a PFS rate of at least 50% after 16 weeks was considered to be sufficient to continue the development of this regimen. A preliminary evaluation of activity was planned after treatment of the first 19 patients. At that time, if fewer than 7 patients remained free of disease progression at first re-evaluation, the trial was to be stopped. If >7 patients remained free of disease progression at the time of the first re-evaluation (after 6 weeks of treatment), the trial was scheduled to continue to a total of 55 patients. If ≥22 patients were free of disease progression at 12 weeks, the study would be considered to have demonstrated a positive result for this population of patients based on a 1-sided optimal 2-stage test of proportion at the 5% significance level and 80% power.
PFS was measured from the date of first study treatment to the date of disease progression or death. OS was measured from the date of first study treatment until the date of death. Survival curves were calculated using the method of Kaplan and Meier.22 All patients who received at least 1 dose of treatment were included in the toxicity analysis. Toxicity was graded according to the National Cancer Institute Common Toxicity Criteria (version 3.0).
Between February 2008 and October 2008, 57 patients were enrolled in this clinical trial. Patient characteristics are detailed in Table 2. The median age of the patients was 70 years (range, 36-88 years); 68% of patients were male. The large majority of patients had good performance status (ECOG performance status of 0 or 1), although 29 patients (51%) had stage M1c disease and 26% of patients had liver metastases. Thirty-two patients (56%) had not received previous systemic therapy.
Fifty-three of the 57 patients (93%) received at least 2 cycles (6 weeks) of treatment and were evaluated for response. Four patients discontinued treatment during the first 6 weeks for the following reasons: toxicity (interstitial pneumonitis; 1 patient), intercurrent illness (fall with rib fractures; 1 patient), patient request (1 patient), and death (possibly treatment-related) because of myocardial infarction (1 patient).
The median number of treatment cycles received was 4 (range, 1-14+ cycles). At the time of last follow-up, 4 patients were continuing treatment after treatment durations of 9 to 13 months. A total of 53 patients had discontinued treatment, 35 because of tumor progression; additional reasons for discontinuing treatment included treatment-related toxicity (5 patients), intercurrent illness (8 patients), and patient/physician decision (5 patients).
Seven patients (12%) achieved major responses to treatment (a complete response in 1 patient and partial responses in 6 patients). Five of the 7 patients with major responses had metastases in visceral locations (M1b disease in 1 patient and M1c disease in 4 patients). An additional 33 patients were classified as having stable disease as per RECIST criteria. Twenty-three of these patients had objective shrinkage of measurable lesions (between 5-30%), but did not meet the criteria for a partial response. Overall, 30 patients (53%) had objective improvements in tumor measurements, as shown in Figure 1.
The estimated PFS for all patients is shown in Figure 2. After a median follow-up of 13 months, the median PFS was 4 months (95% confidence interval, 2.8-5.3 months), with a 1-year PFS rate of 7%. The median OS and OS rate for the entire group were 8.6 months and 43%, respectively (Fig. 3).
The objective response rates were similar in previously untreated patients compared with those who had received 1 or 2 previous regimens (16% vs 8%, respectively). For responding patients, the median duration of response was 7 months (range, 4-10+ months); 3 patients remained free of disease progression after 6, 9, and 10 months of treatment, respectively.
Treatment-related toxicities are listed in Table 3. The combination of bevacizumab and everolimus was well tolerated by the majority of patients, although grade 1/2 toxicities were relatively common. The most common grade 3/4 toxicities included fatigue (12%), hypertension (11%), and stomatitis (11%). Serious hematologic toxicity was uncommon, with only 4 patients developing grade 3 thrombocytopenia. Five patients were required to discontinue treatment because of toxicity (interstitial pneumonitis in 2 patients, proteinuria in 1 patient, epistaxis in 1 patient, and abdominal pain in 1 patient). An additional 24 patients required dose reductions of everolimus at some time during therapy, but were able to continue treatment with a modified dose. There was 1 death reported during this study (from myocardial infarction) that was considered to be possibly related to treatment.
Toxicity was graded according to the National Cancer Institute Common Toxicity Criteria (version 3.0).
Metastatic melanoma remains 1 of the few advanced malignancies for which systemic therapy has failed to alter the disease course. Traditional cytotoxic agents have been ineffective, and current efforts are focused almost exclusively on novel, targeted agents. New classes of agents currently of interest in the treatment of advanced melanoma include the angiogenesis inhibitors and agents affecting the PI3K/AKT pathway, including mTOR inhibitors. Several inhibitors of angiogenesis have demonstrated single-agent activity against melanoma, including bevacizumab, sorafenib, PTK787, sunitinib, axitinib, and aflibercept.7-10, 14 Limited experience with mTOR inhibitors has failed to demonstrate substantial single-agent activity17, 18; however, the down-regulation of the VEGF receptor induced by these agents provides a rationale for testing an mTOR inhibitor in combination with an antiangiogenesis agent. Our previous experience with the combination of bevacizumab and everolimus in the treatment of patients with advanced renal cell carcinoma had demonstrated this combination to be active and well tolerated,20 and therefore this combination was chosen for evaluation in patients with metastatic melanoma.
In the group of 57 patients with metastatic melanoma included in this clinical trial, the combination of bevacizumab and everolimus demonstrated activity. Major responses occurred in 13% of patients, whereas 53% had objective improvement in their tumor measurements noted during therapy (Fig. 1). The median PFS of 4 months met the targeted endpoint for activity in this trial. However, the majority of the responses in this trial were brief, and the estimated PFS rate at 1 year was only 7%. As in our previous experience, delivery of both drugs at full doses was possible with this regimen, which was well tolerated by most patients. Treatment-related toxicity was consistent with the known toxicity profiles of each of these agents when used separately.
It is worth mentioning that the clinical characteristics of the patients in the current trial differed in several ways from those reported in other recent trials. First, patients in our trial had a median age of 70 years, which to our knowledge is approximately 10 years older than the median ages reported in other trials. Second, the proportion of patients with M1c disease (51%) was lower than in recently reported trials, in which this proportion was usually reported to be >70%.13, 23-26 The reasons for these differences in patient characteristics are unknown, but may reflect the community-based patient population. Finally, nearly half of the patients in the current study had received at least 1 previous regimen for the treatment of metastatic melanoma. The impact of these differences on the observed activity of the combination of bevacizumab and everolimus is also speculative.
Although the activity observed in this clinical trial was only modest, further exploration of inhibitors of these pathways is most likely indicated in the treatment of patients with malignant melanoma. Several other trials have evaluated angiogenesis inhibitors in combination with chemotherapy, and have reported mixed results. In a randomized phase 2 trial, patients with metastatic melanoma received first-line treatment with the combination of paclitaxel and carboplatin, with or without bevacizumab.27 Although this trial did not meet its survival endpoint, there was a trend toward improved PFS (5.6 months vs 4.2 months) and OS (12.3 months vs 9.2 months) in the group treated with bevacizumab. However, a similar trial adding sorafenib instead of bevacizumab to the combination of paclitaxel and carboplatin indicated no benefit in the sorafenib group.23 Given the demonstrated modest activity of the angiogenesis inhibitors in the treatment of advanced melanoma, it is possible that the activity observed in this trial was produced by bevacizumab alone.
The PI3K/AKT/mTOR pathway can be activated by several mechanisms including: 1) activation of PI3K via signaling through receptor tyrosine kinases or Ras; 2) overexpression of AKT; or 3) loss or inactivation of the suppressor gene phosphatase and tensin homolog (PTEN).28-30 Activation of AKT leads to multiple downstream effects, including aberrant signaling through mTOR and phosphorylation of multiple other substrates contributing to a malignant phenotype (including murine double minute protein [MDM2], nuclear factor κβ, bcl 2-associated death promoter, glycogen synthase kinase 3β [GSK3β], human telomerase reverse transcriptase, and p27).28 To our knowledge, the best way to inhibit this pathway is not currently clear, and it is possible that inhibitors of PI3K or AKT will prove more effective than mTOR inhibitors.
Interactions between the PI3K/AKT pathway and the mitogen-activated protein kinase (MAPK) pathway may provide another treatment strategy. The MAPK pathway is activated in 60% to 80% of melanomas, because of activating mutations in either BRAF or NRAS.31 Inhibition of the MAPK pathway with either BRAF or mitogen-activated protein kinase (MEK) inhibitors has recently shown great promise in the treatment of patients with metastatic melanoma.32, 33 However, activation of the MAPK pathway frequently results in the concurrent activation of the PI3K/AKT pathway.34 In addition, inhibition of the mTOR pathway can induce activation of the MAPK pathway via a PI3K-dependent feedback loop.35 The interconnection of these pathways suggests that the combined inhibition of the PI3K/AKT and MAPK pathways may greatly increase treatment efficacy in patients with melanoma, and several such combinations currently are being actively investigated.
After years of frustration, it appears finally that treatment advances are on the horizon for patients with metastatic melanoma. The availability of new agents with novel targets, coupled with the increased understanding of the unique biology of this cancer, provide realistic hope for significant advances in the near future.
CONFLICT OF INTEREST DISCLOSURES
Supported in part by grants from Genentech and Novartis. Dr. Hainsworth has received research funding from Genentech. Dr. Spigel has served as an uncompensated advisor for, and has received research funding from, Genentech and Novartis. Dr. Burris has received honoraria from Genentech and Novartis.