Abstract
Purpose
In limited institution Phase II studies, thalidomide and temozolomide has yielded response rates (RR) up to 32% for advanced melanoma, leading to the use of this combination as “standard” by some. We conducted a multi-center Phase II trial to better define the clinical efficacy of thalidomide and temozolomide and the immune modulatory effects of thalidomide, when combined with temozolomide, in patients with metastatic melanoma.
Patients and Methods
Patients must have had stage IV cutaneous melanoma, no active brain metastases, Zubrod PS 0–1, up to 1 prior systemic therapy excluding thalidomide, temozolomide or dacarbazine, adequate organ function, and given informed consent. The primary endpoint was 6-month progression-free survival (PFS). Secondary endpoints included survival (OS), RR, toxicities, and assessment of relationships between biomarkers and clinical outcomes. Patients received thalidomide (200mg/d escalated to 400mg/d for patients <70, or 100mg/d escalated to 250mg/d for patients ≥70) plus temozolomide (75mg/m2/d × 6 weeks then 2 weeks rest).
Results
Sixty-four patients were enrolled; 2 refused treatment. The 6-month PFS was 15% (95% CI, 6%–23%); 1-year OS was 35% (95% CI, 24%–47%); RR was 13% (95% CI, 5%–25%), all partial. One treatment-related death occurred from myocardial infarction; 3 other Grade 4 events occurred including pulmonary embolism, neutropenia and CNS ischemia. There was no significant correlation between biomarkers and PFS or OS.
Conclusion
This combination of thalidomide and temozolomide does not appear to have a clinical benefit that exceeds dacarbazine alone. We would not recommend it further for phase III trials or for standard community use.
Keywords: Phase II, Thalidomide and Temozolomide, Southwest Oncology Group
INTRODUCTION
Systemic treatment of metastatic malignant melanoma remains suboptimal. Response rates have traditionally been no better than 10% to 20%. This includes cytotoxic chemotherapy, whether in combination or with dacarbazine alone, and biological agents such as interferon alfa and interleukin-2.1,2 High-dose bolus interleukin-2 systemic immunotherapy can induce durable remissions in a very small percentage of patients, but is toxic and thus generally reserved for patients with excellent performance status and organ function.2 Combinations of immunotherapy with systemic cytotoxic chemotherapy, (biochemotherapy), initially showed promise as superior to chemotherapy alone, until randomized trials revealed equivalent survival.3 Better tolerated, less toxic, and more efficacious treatments in this disease are needed.
It is estimated that up to 70% of patients who die of metastatic melanoma have either known or subclinical brain metastases. Dacarbazine, other standard chemotherapy agents and immunotherapeutic agents do not cross the blood-brain barrier. Temozolomide, an agent approved for the treatment of refractory anaplastic astrocytoma, is an oral alternative to dacarbazine that is orally bioavailable, has higher cerebrospinal fluid levels, (approximately 40% of plasma concentrations), compared to dacarbazine and has clinical activity against melanoma equivalent to dacarbazine.4 Thalidomide is an orally bioavailable agent that possesses sedative, antiemetic, antiangiogenic, and immunomodulatory activities and has clinical activity against several cancers including malignant melanoma.5–8 In a small dose-finding trial, the combination of temozolomide and thalidomide in the treatment of advanced melanoma was shown to be well tolerated and active.9 A follow up single institution Phase II trial, reported by the same investigators, using the combination of thalidomide and temozolomide reported an overall response rate of 32% with one durable complete remission in 38 patients.10 In a 181 patient, two-institution, randomized Phase II trial reported from England, the combination of temozolomide and thalidomide produced a small yet not statistically significant improvement in response and survival duration over temozolomide alone and a temozolomide plus interferon-α combination.11 Despite a lack of prospective confirmation, this oral combination regimen rapidly became a community standard for patients with unresectable metastatic melanoma. Our intent was to expand on these results to better define the tolerability and efficacy of a set dose and schedule of the combination of temozolomide with thalidomide9,10 in the treatment of metastatic malignant melanoma in the cooperative group setting.
The effectiveness of thalidomide historically has been attributed to its ability to modulate tumor necrosis factor alpha (TNFα) production, though several studies now indicate that thalidomide also co-stimulates differential subsets of T-cells resulting in regulation of additional cytokines involved in immune responses.12–15 In an attempt to better understand the immune modulatory effects of thalidomide (in combination with temozolomide) in patients with metastatic malignant melanoma, a number of laboratory immune correlates were measured. CD4+/CD25+ regulatory T-cells are known to be important modulators of immune response.16–20 Increased levels of these circulating CD4+/CD25+ regulatory T-cells have been described in patients with melanoma, reflecting a potentially immune suppressive environment in vivo.21 With the patient’s consent, blood samples were to be obtained at four time points each: prior to and at 5, 9 and 13 weeks following initiation of therapy, in order to assess treatment-induced immune responsiveness.
PATIENTS AND METHODS
Patients
Eligible patients must have had a biopsy-proven diagnosis of unresectable stage IV melanoma of cutaneous origin, and they must have been without active brain metastases as documented by baseline CT or MRI of the brain. Patients may have received up to one prior systemic therapy for Stage IV disease excluding prior thalidomide, temozolomide or dacarbazine. Prior adjuvant alpha interferon was allowed. A Zubrod performance status of 0 or 1 was required and patients must have had adequate hepatic, hematologic and renal function. Patients must have been fully aware of the teratogenic potential of thalidomide and agree to fully comply with the FDA-mandated S.T.E.P.S.® program. Women of childbearing potential must have had a negative pregnancy test, with a sensitivity of at least 50 mIU/ml, performed within 24 hours prior to the start of treatment with thalidomide. Local institutional review board approval of the trial was required at each participating institution and all patients gave written informed consent.
Treatment
All patients were to receive both agents by oral dosing on a daily basis using the same doses and schedule as described by Hwu et al.9,10 Thalidomide was administered at bedtime 30 – 60 minutes prior to temozolomide. For patients less than 70 years of age, thalidomide was started at 200 mg per day and titrated up at 100 mg/day increments at 2-week intervals, to a maximum of 400 mg/day during the first cycle and maintained at that dose as long as therapy was tolerated. For patients ≥70 years of age, the starting dose of thalidomide was 100 mg/day and escalated in 50 mg/day increments at 2-week intervals, to a maximum of 250 mg/day. Temozolomide was administered at a dose of 75 mg/m2/day daily for 6 weeks followed by a 2-week break with eight week cycles repeated until progression or unacceptable toxicity. No prophylactic anticoagulation was required.
Dose modifications were performed using the CI Common Terminology Criteria for Adverse Events (CTCAE) Version 3.0. For Grade ≥2 rash/desquamation or for Grade 3 or 4 somnolence/depressed level of consciousness, sensory neuropathy or constipation, thalidomide was to be held and restarted at the next lowest dose level previously tolerated by the patient (or at 50 mg or 100 mg if the initial dose of 100 mg or 200 mg was not tolerated) when toxicity resolved to Grade 1 for at least 2 weeks. For all other Grade 3 or 4 non-hematologic toxicity, temozolomide was to be held until the toxicity resolved to ≤Grade 1 and restarted at a dose of 60 mg/m2/day. If the Grade 3/4 toxicity recurred, temozolomide was again held until the toxicity resolved to ≤Grade 1 and restarted at a dose of 45 mg/m2/day. If the Grade 3/4 toxicity again recurred, the patient was removed from protocol treatment. For Grade 4 hematologic toxicity, including neutropenia or thrombocytopenia, temozolomide was withheld until the toxicity resolved to ≤Grade 1 then resumed at the same dose. If Grade 4 neutropenia or thrombocytopenia recurred, temozolomide was again held until toxicity resolved to ≤Grade 1 then resumed at a dose of 60 mg/m2/day. Again, if Grade 4 neutropenia or thrombocytopenia recurred, temozolomide was again held until toxicity resolved to ≤Grade 1 then resumed at a dose of 45 mg/m2/day. If again Grade 4 neutropenia or thrombocytopenia recurred or if temozolomide was withheld for ≥4 consecutive weeks due to toxicity, patients were removed from protocol treatment. No dose reductions were to be performed for Grade 3/4 anemia.
Tumor response was evaluated by physical examination, chest x-ray, computed tomography scan, or other diagnostic tests, as appropriate, after every 8-week cycle of therapy. Standard RECIST criteria were used for response.
Statistical Methods
The primary objective of this trial was to assess the six-month progression-free survival (PFS) in patients with Stage IV melanoma treated with the combination of oral thalidomide and temozolomide. Secondary objectives included assessment of objective tumor response, overall survival, qualitative and quantitative toxicities of this combination, and a preliminary investigation of specific circulating biomarkers and the relationship of these biomarkers with clinical outcomes. Based on data from previous Southwest Oncology Group Phase II trials in Stage IV melanoma, it was assumed that this regimen would not be of interest if the true 6-month PFS probability was less than 10% and that a true 6-month PFS of 25% or more would be of considerable interest. The accrual goal was 55 evaluable patients. If 10 or more of the 55 patients survived at least 6 months with no evidence of disease progression, this would be considered sufficient evidence to warrant further study of this regimen. This design had a significance level of 4%, (probability of falsely declaring an agent with a 10% 6-month PFS probability to warrant further study), and a power of 92%, (correctly declaring an agent with a 25% 6-month PFS probability as warranting further study). Assuming complete follow-up, 55 patients would be sufficient to estimate the 6-month PFS probability, the 1-year overall survival probability, and toxicity rates to within at least ±13% (95% confidence interval). Furthermore, with 55 patients any toxicity occurring with at least 5% probability would likely be observed at least once (94% probability). Response was to be evaluated only among the subset of patients with measurable disease at baseline. Assuming 80% of the 55 eligible patients would have measurable disease, this would be sufficient to estimate the response rate (confirmed and unconfirmed, complete and partial responses) to within at least ±15% (95% confidence interval).
Progression-free survival was defined as the time from study enrollment to the date of progression, symptomatic deterioration, or death. Patients last known to be alive and progression-free were censored at the date of last contact. Overall survival was defined as the time from study enrollment to the date of death or last follow up. Progression-free and overall survival distributions were estimated by the Kaplan Meier method.22
An exploratory analysis was performed to investigate the relationship between clinical outcomes and immunologic biomarker values. The relationship between biomarker values and response was analyzed using logistic regression methods or by a two-sided Fisher’s Exact test. Cox proportional hazard models were fit to assess the relationship between the biomarkers and PFS and OS.23 In addition, the change in biomarker value from baseline to week 5, week 9 and week 13 were analyzed using a paired two-sided nonparametric sign test. Due to the limited sample size, the objective of this analysis was limited to generating hypotheses to be further investigated in larger cohorts as there would only be sufficient power to detect very strong relationships. No adjustments for multiple comparisons were made.
Biomarkers
Serum samples were obtained at baseline, and at 5, 9, and 13 weeks following initiation of therapy. Unprocessed venipuncture tubes were shipped overnight at ambient temperature to a central processing lab where immunomonitoring protocols were carried out in standard, previously published fashion and included cytokines (IL-2, IL-4, IL-5, IL-6, IL-10, TNFα, and IFNγ) measured in serum using cytokine immunoassays, IFNγ ELISPOT against peptides comprised of a pool of 32 8–12-mer peptides representing immunodominant CD8+ T cell epitopes from human cytomegalovirus, Epstein Barr virus and influenza virus (CEF peptide pool), and phytohemagglutinin (PHA) as a polyclonal T cell activator.24,25 Flow cytometry was performed to enumerate NK (CD16+/CD56+), CD4 (CD4+/CD3+), Treg (CD4+/CD25hi), CD8 (CD8+/CD3+), activated CD8 (CD8+/CD25+), CD25- Treg (CD4+/CD25-/FoxP3+), and FoxP3+ Treg (CD4+/CD25hi/FoxP3+) cells.
RESULTS
Patients
A total of 64 patients were enrolled in this multi-institutional phase II trial from June 2005 to January 2007. Two eligible patients who withdrew consent never received any protocol treatment, thus are not evaluable for any of the study endpoints. Patient characteristics are presented in Table 1. Two-thirds of the patients were male. Three patients had a history of previously treated brain metastases. Over half the patients had M1c disease. Of the 11 patients who received prior systemic therapy, 7 had received cytotoxic chemotherapy and 5 patients received immunotherapy including one patient who received a combination biochemotherapy regimen.
Table 1.
Patient Characteristics (n = 62)
| Variable | No. patients |
|---|---|
| Median age (range) | 62 (33 – 82) |
| Sex M:F | 42:20 |
| Performance Status 0/1 | 40/22 |
| Metastatic site: | |
| Bone | 14 |
| Brain | 3 |
| Liver | 19 |
| Lung | 42 |
| Distant nodes | 42 |
| Other non-visceral | 10 |
| Other visceral | 12 |
| Serum LDH: | |
| Elevated | 22 |
| Normal | 40 |
| M stage: | |
| M1a | 6 |
| M1b | 21 |
| M1c | 35 |
| Prior systemic therapy | 11 |
Efficacy
All 62 patients had documented disease progression, including 8 patients with new brain metastases. The 6-month progression-free survival estimate was 15% (6–23% 95%CI); median PFS was 2 months (2–4 months 95%CI), (Figure 1). The estimated 1-year overall survival was 35% (23–47% 95%CI); median OS was 8 months (6–12 months 95%CI), (Figure 2). For the 6 surviving patients, the median follow-up is 21 months (range: 20–32 months).
Figure 1.
Progression Free Survival
Figure 2.
Overall Survival
Fifty-four patients had measurable disease at baseline and are included in the response analysis. There were 7 responses observed, all partial, for an estimated overall response rate of 13% (5–25% 95%CI).
Toxicity
There was one treatment-related death, due to cardiac ischemia/infarction along with grade 4 thromboembolism in the same patient, (Table 2). An additional two patients experienced Grade 4 adverse events, including one case each of neutropenia and central nervous system ischemia. Grade 3 fatigue was the most common significant toxicity, occurring in 6 patients. Ninety percent of patients experienced some degree of toxicity.
Table 2.
Grade 2 5 Toxicity (n = 62) # of patients with Given Type and Grade
| Grade (#) | ||||
|---|---|---|---|---|
| Adverse Event | 2 | 3 | 4 | 5 |
| Abdominal/stomach pain | 1 | |||
| Anemia | 3 | 1 | ||
| Anorexia | 6 | 1 | ||
| Cardiac ischemia/infarction | 1 | |||
| CNS ischemia | 1 | |||
| Confusion | 1 | 1 | ||
| Constipation | 21 | 1 | ||
| Creatinine | 1 | |||
| Depression | 1 | 1 | ||
| Diarrhea | 1 | |||
| Dizziness | 3 | |||
| Dry skin | 3 | |||
| Dyspnea | 1 | 1 | ||
| Edema limb | 1 | 1 | ||
| Fatigue | 18 | 6 | ||
| Headache | 2 | |||
| Heartburn | 1 | 1 | ||
| Hypokalemia | 1 | |||
| Hypotension | 2 | |||
| Infection | 1 | |||
| Lymphopenia | 3 | |||
| Musculoskeletal pain limb | 1 | |||
| Nausea | 6 | 1 | ||
| Neuropathy motor | 1 | 1 | ||
| Neuropathy sensory | 4 | 1 | ||
| Neutropenia | 4 | 2 | 1 | |
| Pulmonary embolism | 1 | |||
| Pruritus | 2 | |||
| Rash | 6 | |||
| Somnolence | 10 | 2 | ||
| Syncope | 2 | |||
| Taste alteration | 1 | |||
| Thrombocytopenia | 1 | 1 | ||
| Thromboembolism | 1 | |||
| Vomiting | 3 | 1 | ||
| Weight loss | 2 | |||
Abbreviation: CNS, central nervous system
Biomarker Analysis
Assay results were received representing 52 of the patients enrolled. This included two patients with no assay results from baseline specimens. Summary statistics for each of the biomarkers at baseline, week 5, week 9, and week 13 are displayed in Table 3. For this analysis, assay values extrapolated beyond the normal range of the assay and with questionable reproducibility were discarded. The values for CEF and PHA represent the number of spot forming units, (SFU), per million minus the background, (SFU for no peptide), plus two standard deviations.
Table 3.
Biomarker results
| Baseline | Week 5 | Week 9 | Week 13 | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Marker | N | Median | Min | Max | N | Median | Min | Max | N | Median | Min | Max | N | Median | Min | Max |
| IL-2 (pg/ml) | 7 | 11.25 | 1.53 | 317.94 | 7 | 11.21 | 1.75 | 80.26 | 6 | 6.35 | 0.96 | 851.57 | 1 | 35.58 | 35.58 | 35.58 |
| IL-4 (pg/ml) | 28 | 0.67 | 0.15 | 20.67 | 15 | 1.11 | 0.32 | 4.74 | 12 | 0.57 | 0.12 | 1.62 | 4 | 1.49 | 0.91 | 2.70 |
| IL-5 (pg/ml) | 7 | 2.93 | 1.5 | 76.25 | 7 | 4.90 | 1.92 | 16.08 | 6 | 3.79 | 1.10 | 34.49 | 2 | 17.68 | 3.60 | 31.75 |
| IL-6 (pg/ml) | 48 | 169.965 | 1.64 | 9056.3 | 31 | 288.05 | 2.58 | 6084.79 | 29 | 191.11 | 2.46 | 4503.34 | 16 | 82.70 | 7.51 | 7642.37 |
| IL-10 (pg/ml) | 36 | 7.42 | 1.62 | 178.34 | 27 | 7.28 | 1.70 | 306.00 | 23 | 5.46 | 1.69 | 45.89 | 14 | 4.91 | 1.72 | 25.30 |
| TNF-α (pg/ml) | 40 | 32.405 | 2.08 | 1874.92 | 25 | 52.62 | 3.56 | 2515.99 | 23 | 36.03 | 2.87 | 909.96 | 13 | 54.10 | 1.82 | 762.86 |
| IFN-γ (pg/ml) | 39 | 38.74 | 0.63 | 5222 | 21 | 44.46 | 0.81 | 366.30 | 21 | 11.16 | 1.10 | 1135.26 | 11 | 17.07 | 0.81 | 276.64 |
| CEF* | 31 | 603.17 | 0 | 4673.75 | 15 | 631.56 | 0.00 | 3630.00 | 12 | 251.19 | 0.00 | 2596.42 | 9 | 405.25 | 0.00 | 797.57 |
| PHA* | 28 | 2891.28 | 0 | 9878.58 | 15 | 3272.39 | 565.00 | 10460.12 | 11 | 1490.86 | 0.00 | 11851.98 | 9 | 2198.58 | 0.00 | 12863.33 |
| %CD16+/CD56+ | 24 | 8.3 | 1 | 27.1 | 15 | 4.80 | 0.70 | 12.60 | 14 | 5.20 | 1.40 | 9.80 | 11 | 4.70 | 1.20 | 19.10 |
| %CD4+/CD3+ | 33 | 38.1 | 9.8 | 68.5 | 22 | 30.05 | 13.90 | 48.70 | 21 | 33.90 | 16.40 | 61.80 | 10 | 31.45 | 10.10 | 38.30 |
| %CD4+/CD25+ | 33 | 8.9 | 1.1 | 37.3 | 22 | 5.40 | 0.50 | 16.60 | 21 | 4.40 | 0.20 | 30.50 | 10 | 3.35 | 1.00 | 10.20 |
| %CD8+/CD3+ | 28 | 15.3 | 5.9 | 32.7 | 18 | 13.55 | 2.10 | 21.90 | 18 | 16.55 | 4.90 | 32.10 | 11 | 17.20 | 3.10 | 24.40 |
| %CD8+/CD25+ | 28 | 0.6 | 0.1 | 2.5 | 18 | 0.50 | 0.10 | 2.10 | 18 | 0.45 | 0.00 | 2.60 | 11 | 0.70 | 0.10 | 17.50 |
| %CD4+/CD25−/FoxP3+1 | 23 | 1.6 | 0.2 | 9.7 | 12 | 2.65 | 0.40 | 13.40 | 14 | 2.75 | 0.20 | 12.80 | 8 | 3.65 | 1.60 | 13.20 |
| %CD4+/CD25+/FoxP3+2 | 19 | 52.6 | 24.6 | 79.5 | 10 | 66.45 | 49.30 | 83.20 | 12 | 65.75 | 0.50 | 84.30 | 7 | 61.30 | 46.70 | 82.40 |
SFU – spot forming units per million (background SFU +2SD)
CEF – cytomegalovirus, Epstein-Barr virus, influenza virus
PHA – phytohemagglutinin
%FoxP3+ cells only within CD4+CD25− population
%FoxP3+ cells only within CD4+CD25+ population
The change in biomarker value from baseline to week 5, week 9, and week 13 were analyzed using a paired two-sided nonparametric sign test. Only those patients with data at both time points were included. The results are shown in Table 4. Of these 42 comparisons, significant differences were associated with: an increase in IL-10 from prestudy to week 5, (median difference = 3.17), (p = 0.001), and decreases in %CD16+/CD56+ from prestudy to week 9, (median difference = −2.85), (p = 0.03), %CD4+/CD3+ from prestudy to week 5, (median difference = −4.4), (p = 0.01), and %CD4+/CD25+ from prestudy to week 13, (median difference = −5), (p= 0.04).
Table 4.
Biomarker results compared to baseline
| week 5 - baseline | week 9 - baseline | week 13 - baseline | ||||
|---|---|---|---|---|---|---|
| Marker | N | p-value* | N | p-value* | N | p-value* |
| IL-4 | 11 | 0.55 | 11 | 1.00 | 4 | 0.63 |
| IL-6 | 24 | 0.84 | 26 | 0.85 | 14 | 0.79 |
| IL-101 | 19 | 0.001 | 19 | 0.17 | 13 | 0.27 |
| TNF-α | 19 | 0.36 | 19 | 0.36 | 12 | 0.77 |
| IFN-γ | 17 | 0.14 | 18 | 0.48 | 10 | 0.34 |
| CEF | 13 | 0.27 | 9 | 1.00 | 8 | 1.00 |
| PHA | 13 | 0.58 | 8 | 0.73 | 7 | 1.00 |
| %CD16+/CD56+2 | 9 | 0.51 | 6 | 0.03 | 7 | 0.69 |
| %CD4+/CD3+3 | 18 | 0.01 | 15 | 1.00 | 9 | 0.51 |
| %CD4+/CD25+4 | 18 | 1.00 | 15 | 1.00 | 9 | 0.04 |
| %CD8+/CD3+ | 15 | 0.61 | 12 | 1.00 | 10 | 0.75 |
| %CD8+/CD25+ | 15 | 0.27 | 12 | 0.39 | 10 | 0.75 |
| %CD4+/CD25−/FoxP3+ | 9 | 0.51 | 7 | 1.00 | 5 | 1.00 |
| %CD4+/CD25+/FoxP3+ | 8 | 0.07 | 6 | 0.69 | 5 | 1.00 |
| Table 4a. Relationship between baseline serum cytokine levels and response | ||||
|---|---|---|---|---|
| IL4 | response | Total | p-value | |
| INC/STA | CR/PR | |||
| low (IL4 < 1.2) | 15 | 1 | 16 | 1.000 |
| normal (IL4 > 1.2) | 7 | 1 | 8 | |
| Total | 22 | 2 | 24 | |
| IL6 | response | Total | p-value | |
| INC/STA | CR/PR | |||
| normal (IL6 ≤ 9.3) | 5 | 0 | 5 | 1.000 |
| high (IL6 > 9.3) | 30 | 6 | 36 | |
| Total | 35 | 6 | 41 | |
| IL10 | response | Total | p-value | |
| INC/STA | CR/PR | |||
| normal (IL10 ≤ 15.5) | 18 | 4 | 22 | 0.645 |
| high (IL10 > 15.5) | 6 | 2 | 8 | |
| Total | 24 | 6 | 30 | |
| IFN-γ | response | Total | p-value | |
| INC/STA | CR/PR | |||
| normal (IFN-γ ≤ 48) | 15 | 4 | 19 | 0.347 |
| high (IFN-γ > 48) | 15 | 1 | 16 | |
| Total | 30 | 5 | 35 | |
| TNF-α | response | Total | p-value | |
| INC/STA | CR/PR | |||
| normal (TNF-α ≤ 7.8) | 9 | 0 | 9 | 0.303 |
| high (TNF-α > 7.8) | 21 | 6 | 27 | |
| Total | 30 | 6 | 36 | |
| Table 4b. Relationship between baseline ELISPOT and response | ||||
|---|---|---|---|---|
| CEF | response | Total | p-value | |
| INC/STA | CR/PR | |||
| Weak/Intermediate (≤ 1000) | 17 | 3 | 20 | 0.536 |
| Strong (> 1000) | 8 | 0 | 8 | |
| Total | 25 | 3 | 28 | |
| PHA | response | Total | p-value | |
| INC/STA | CR/PR | |||
| Weak/Intermediate (≤ 1000) | 4 | 1 | 5 | 0.504 |
| Strong (> 1000) | 18 | 2 | 20 | |
| Total | 22 | 3 | 25 | |
| Table 4c. Relationship between baseline circulating cells and response | ||||
|---|---|---|---|---|
| %CD16+/CD56+ | response | Total | p-value | |
| INC/STA | CR/PR | |||
| Low (≤ 8.3) | 10 | 1 | 11 | 0.586 |
| High (> 8.3) | 8 | 2 | 10 | |
| Total | 18 | 3 | 21 | |
| %CD4+/CD3+ | response | Total | p-value | |
| INC/STA | CR/PR | |||
| Low (≤ 38.1) | 14 | 2 | 16 | 0.642 |
| High (> 38.1) | 11 | 3 | 14 | |
| Total | 25 | 5 | 30 | |
| %CD4+/CD25+ | response | Total | p-value | |
| INC/STA | CR/PR | |||
| Low (≤ 8.9) | 11 | 5 | 16 | 0.045 |
| High (> 8.9) | 14 | 0 | 14 | |
| Total | 25 | 5 | 30 | |
| %CD8+/CD3+ | response | Total | p-value | |
| INC/STA | CR/PR | |||
| Low (≤ 15.3) | 9 | 4 | 13 | 0.096 |
| High (> 15.3) | 12 | 0 | 12 | |
| Total | 21 | 4 | 25 | |
| %CD8+/CD25+ | response | Total | p-value | |
| INC/STA | CR/PR | |||
| Low (≤ 0.6) | 12 | 3 | 15 | 0.626 |
| High (> 0.6) | 9 | 1 | 10 | |
| Total | 21 | 4 | 25 | |
| %CD4+/CD25−/FoxP3+ | response | Total | p-value | |
| INC/STA | CR/PR | |||
| Low (≤ 1.6) | 10 | 1 | 11 | 1.000 |
| High (> 1.6) | 9 | 0 | 9 | |
| Total | 19 | 1 | 20 | |
| %CD4+/CD25+/FoxP3+ | response | Total | p-value | |
| INC/STA | CR/PR | |||
| Low (≤ 52.6) | 8 | 0 | 8 | 1.000 |
| High (> 52.6) | 8 | 1 | 9 | |
| Total | 16 | 1 | 17 | |
N = # of patients with data at both baseline and the respective time point
Sign test
There was a significant increase in IL-10 from prestudy to week 5 (median difference = 3.17)
There was a significant decrease in %CD16+/CD56+ from prestudy to week 9 (median difference = −2.85)
There was a significant decrease in %CD4+/CD3+ from prestudy to week 5 (median difference = −4.4)
There was a significant decrease in %CD4+/CD25+ from prestudy to week 13 (median difference = −5)
Abbreviations: INC, increased; STA, stable; CR, complete response; PR, partial response
The relationship between clinical outcomes and baseline biomarker values was then explored. Due to the highly skewed distributions of the baseline data, the values for each assay were dichotomized into two groups as follows. For the serum cytokine assays, results were categorized as falling within the normal range, (as specified by one of the assay manufacturers), vs. above normal. For ELISPOT, results were categorized as strong, (>1000 SFU/million), vs. weak or intermediate, (≤1000 SFU/million). The flow cytometry results were split along the observed sample medians. The relationship between each biomarker and response (confirmed and unconfirmed complete and partial), was analyzed using a two-sided Fisher’s Exact test. These results are shown in Tables 4a – c (available on-line). The only statistically significant relationship was with low circulating levels of %CD4+/CD25+ cells at baseline, (p = 0.045). However, this assumes no adjustment for multiple comparisons. Cox proportional hazard models were fit to assess the relationship between the biomarkers and PFS and OS. Tables 5 and 6, (available on-line), display the p-values, hazard ratios, and 95% confidence intervals for the hazard ratios for each. None of the biomarkers were found to be related with PFS or OS at a significance level of 0.05.
Table 5.
Relationship between baseline biomarker and progression-free survival
| Marker | p-value | Hazard Ratio | 95% CI | |
|---|---|---|---|---|
| IL-4: normal (1.2– 6.8) vs. Low (< 1.2) | 0.594 | 0.79 | 0.33 | 1.88 |
| IL-6: high (> 9.3) vs. normal (≤ 9.3) | 0.464 | 0.72 | 0.29 | 1.75 |
| IL-10: high (> 15.5) vs. normal (≤ 15.5) | 0.965 | 0.98 | 0.44 | 2.18 |
| TNF-α: high (> 7.8) vs. normal (≤ 7.8) | 0.672 | 0.87 | 0.46 | 1.66 |
| IFN-γ: high (> 48) vs. normal (≤ 48) | 0.240 | 0.64 | 0.31 | 1.34 |
| CEF: strong (> 1000) vs. weak/intermediate (≤ 1000) | 0.389 | 1.41 | 0.65 | 3.06 |
| PHA: strong (> 1000) vs. weak/intermediate (≤ 1000) | 0.220 | 1.95 | 0.67 | 5.70 |
| %CD16+/CD56+: high (> 8.3) vs. low (≤ 8.3) | 0.336 | 1.50 | 0.66 | 3.44 |
| %CD4+/CD3+: high (> 38.1) vs. low (≤ 38.1) | 0.300 | 0.69 | 0.34 | 1.39 |
| %CD4+/CD25+: high (> 8.9) vs. low (≤ 8.9) | 0.129 | 1.77 | 0.85 | 3.71 |
| %CD8+/CD3+: high (> 15.3) vs. low (≤ 15.3) | 0.412 | 1.39 | 0.63 | 3.04 |
| %CD8+/CD25+: high (> 0.6) vs. low (≤ 0.6) | 0.528 | 1.28 | 0.59 | 2.78 |
| %CD4+/CD25−/FoxP3+: high (> 1.6) vs. low (≤ 1.6) | 0.415 | 1.43 | 0.61 | 3.36 |
| %CD4+/CD25+/FoxP3+: high (> 52.6) vs. low (≤ 52.6) | 0.057 | 0.36 | 0.13 | 1.03 |
Table 6.
Relationship between baseline biomarker and overall survival
| Marker | p-value | Hazard Ratio | 95% CI | |
|---|---|---|---|---|
| IL-4: normal (1.2–6.8) vs. Low (< 1.2) | 0.771 | 1.14 | 0.47 | 2.81 |
| IL-6: high (> 9.3) vs. normal (≤ 9.3) | 0.223 | 1.94 | 0.67 | 5.65 |
| IL-10: high (> 15.5) vs. normal (≤ 15.5) | 0.361 | 0.65 | 0.26 | 1.62 |
| TNF-α: high (> 7.8) vs. normal (≤ 7.8) | 0.061 | 0.52 | 0.26 | 1.03 |
| IFN-γ: high (> 48) vs. normal (≤ 48) | 0.581 | 1.25 | 0.56 | 2.78 |
| CEF: strong (> 1000) vs. weak/intermediate (≤ 1000) | 0.359 | 1.47 | 0.65 | 3.33 |
| PHA: strong (> 1000) vs. weak/intermediate (≤ 1000) | 0.453 | 1.52 | 0.51 | 4.51 |
| %CD16+/CD56+: high (> 8.3) vs. low (≤ 8.3) | 0.449 | 0.70 | 0.27 | 1.78 |
| %CD4+/CD3+: high (>38.1) vs. low (≤ 38.1) | 0.673 | 1.17 | 0.56 | 2.44 |
| %CD4+/CD25+: high (> 8.9) vs. low (≤ 8.9) | 0.947 | 0.98 | 0.46 | 2.08 |
| %CD8+/CD3+: high (> 15.3) vs. low (≤ 15.3) | 0.301 | 0.66 | 0.30 | 1.45 |
| %CD8+/CD25+: high (> 0.6) vs. low (≤ 0.6) | 0.654 | 1.20 | 0.55 | 2.63 |
| %CD4+/CD25−/FoxP3+: high (> 1.6) vs. low (≤ 1.6) | 0.301 | 0.62 | 0.25 | 1.54 |
| %CD4+/CD25+/FoxP3+: high (> 52.6) vs. low (≤ 52.6) | 0.577 | 1.34 | 0.48 | 3.76 |
DISCUSSION
When compared with a meta-analysis on systemic therapy of advanced melanoma, this multicenter phase II trial of thalidomide plus temozolomide yielded results no better than 70 negative phase II trials.26 Based on the model presented in this meta-analysis, the predicted 6-month PFS and 1-year survival rates for the patients enrolled on this trial are 16% and 34% respectively. The observed 6-month PFS of 15% (6–23%, 95%CI), and the 1-year survival of 35% (23–47%, 95%CI), are virtually identical to what was predicted based on observations from 2001 stage IV melanoma patients treated on cooperative group phase II trials. The overall response rate of 13% (5–25%, 95%CI) for this combination is also equivalent to what would be expected from dacarbazine or temozolomide alone.4 Based on these results, use of this combination regimen is not recommended nor is further evaluation in a phase III trial appropriate.
The toxicity of thalidomide with temozolomide was acceptable in this study, despite nearly all patients experiencing some adverse event. No prophylactic anticoagulation was mandated, yet only 4 thromboembolic phenomena were observed, albeit one led to treatment related death from cardiac ischemia/infarction. Only three grade 4 events were observed. Expected toxicities of thalidomide including constipation and fatigue were common. Myelosuppression and neuropathy were uncommon.
Immunomodulatory effects of thalidomide have been well described.12–15,27 The effect on immunomodulation of this agent when combined with temozolomide, an inhibitor of DNA replication, in the treatment of advanced melanoma has not been previously reported. We evaluated a number of circulating cytokines reported to be modulated by thalidomide treatment either in vitro or in vivo, frequency of immune suppressive regulatory T-cells and other immune cell types, and functional CD8 T-cell responses against a pool of recall antigens at various time points throughout the course of therapy in an exploratory attempt to correlate immune response biomarkers with clinical outcome. CD8+ T-cell responses were evaluated to assess the overall immune status of patients over the course of treatment. The results of these efforts were inconclusive and reflect the difficulty in finding significant correlations that are biologically relevant given the relatively small number of patients whose samples were evaluable. That is, baseline biomarker samples were submitted or results reproducible in often less than 50% of patients enrolled and in subsequent weeks of therapy, consistently less than 30% of patients had samples submitted for testing. Given the large number of statistical comparisons necessary, the number of blood samples tested was far too small to allow for any definitive conclusions. Despite this, the most interesting of the biomarker analyses was the decrease in %CD4+CD25hi (Treg) cells over the treatment period, suggesting that the treatment regimen may have a beneficial, though not clinically significant, effect on decreasing immune suppressive T-cells. This observation may be related to the reported effects of thalidomide on the preferential costimulation of CD8+ T-cells and type 1 helper T-cells.12,13 Thalidomide-induced reduction of the immune suppressive environment may improve natural immunity against melanoma and consequently improve the prospect of clinical responses, as is the goal of active immunotherapy for melanoma.28
The treatment for stage IV melanoma remains unsatisfactory, and despite the introduction of a number of different therapeutic agents in other areas of oncology, there have been no new drugs approved for use in melanoma for many years. Even approaches that appear to have definite activity against melanoma, such as temozolomide, biochemotherapy and anti-CTLA4 antibodies, have failed in industry-sponsored or cooperative group phase III trials.3,4,29 It is increasingly apparent that efforts to identify active agents in melanoma solely by objective response rates are inadequate. Better endpoints need to be selected, such as 6-month PFS and one-year OS. Equally importantly, the role of patient selection in influencing results of metastatic melanoma trials has been under appreciated; promising results of single-institution studies repeatedly fail to survive the scrutiny of multi-institutional trials. Despite this, results of single-institution trials with commercially available agents frequently end up in widespread use in the community. Identification of predictive markers for clinical benefit is one avenue worth pursuit as is confirmation of limited institution phase II results in larger multi-institutional studies before alterations in the standard of care are made. It is clear that more innovative clinical trials in the treatment of malignant melanoma are required, with enrollment of a higher percentage of such patients since the standard of care in this disease has such poor results.
Acknowledgments
This investigation was supported in part by the following PHS Cooperative Agreement grant numbers awarded by the National Cancer Institute, DHHS: CA32102, CA38926, CA46282, CA20319, CA22433, CA35178, CA45807, CA35261, CA35176, CA35431, CA37981, CA74647, CA45808, CA58861, CA45461, CA04919, CA58416, CA67575, CA45560, CA76426, CA11083 and supported in part by the Celgene Corporation
We wish to acknowledge Dr. W. Martin Kast, Ph.D., Beckman Center for Immune Monitoring, University of Southern California, for his significant contribution in providing biomarker analyses for this study.
Footnotes
Presented at the ASCO 2008 Annual Meeting in Chicago, Illinois.
References
- 1.Chapman PB, Einhorn LH, Meyers ML, et al. Phase III multicenter randomized trial of the Dartmouth regimen versus dacarbazine in patients with metastatic melanoma. J Clin Oncol. 1999;17:2745–2751. doi: 10.1200/JCO.1999.17.9.2745. [DOI] [PubMed] [Google Scholar]
- 2.Atkins MB, Lotze MT, Dutcher JP, et al. High-dose recombinant interleukin 2 therapy for patients with metastatic melanoma: Analysis of 270 patients treated between 1985 and 1993. J Clin Oncol. 1999;17:2105–2116. doi: 10.1200/JCO.1999.17.7.2105. [DOI] [PubMed] [Google Scholar]
- 3.Atkins MB, Hsu J, Lee S, et al. Phase III trial comparing concurrent biochemotherapy with cisplatin, vinblastine, dacarbazine, interleukin-2, and interferon alfa-2b with cisplatin, vinblastine, and dacarbazine alone in patients with metastatic malignant melanoma (E3695): A trial coordinated by the Eastern Cooperative Oncology Group. J Clin Oncol. 2008;26:5748–5754. doi: 10.1200/JCO.2008.17.5448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Middleton MR, Grob JJ, Aaronson N, et al. Randomized phase III study of temozolomide versus dacarbazine in the treatment of patients with advanced metastatic melanoma. J Clin Oncol. 2000;18:158–166. doi: 10.1200/JCO.2000.18.1.158. [DOI] [PubMed] [Google Scholar]
- 5.Rajkumar SV. Current status of thalidomide in the treatment of cancer. Oncology. 2001;15:867–874. [PubMed] [Google Scholar]
- 6.Eisen T, Boshoff C, Mak I, et al. Continuous low dose thalidomide: A phase II study in advanced melanoma, renal cell, ovarian and breast cancer. Br J Cancer. 2000;82:812–817. doi: 10.1054/bjoc.1999.1004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kudva GC, Collins BT, Dunphy FR. Thalidomide for malignant melanoma. N Engl J Med. 2001;345:1214–1215. doi: 10.1056/NEJM200110183451617. [DOI] [PubMed] [Google Scholar]
- 8.Reiriz AB, Richter MF, Fernandes S, et al. Phase II study of thalidomide in patients with metastatic malignant melanoma. Mel Res. 2004;14:527–531. doi: 10.1097/00008390-200412000-00014. [DOI] [PubMed] [Google Scholar]
- 9.Hwu WJ, Krown SE, Panageas KS, et al. Temozolomide plus thalidomide in patients with advanced melanoma: Results of a dose-finding trial. J Clin Oncol. 2002;20:2610–2615. doi: 10.1200/JCO.2002.09.034. [DOI] [PubMed] [Google Scholar]
- 10.Hwu WJ, Krown SE, Menell JH, et al. Phase II study of temozolomide plus thalidomide for the treatment of metastatic melanoma. J Clin Oncol. 2003;21:3351–3356. doi: 10.1200/JCO.2003.02.061. [DOI] [PubMed] [Google Scholar]
- 11.Danson S, Lorigan P, Arance A, et al. Randomized phase II study of temozolomide given every 8 hours or daily with either interferon alfa-2b or thalidomide in metastatic malignant melanoma. J Clin Oncol. 2003;21:2551–2557. doi: 10.1200/JCO.2003.10.039. [DOI] [PubMed] [Google Scholar]
- 12.McHugh SM, Rifkin IR, Deighton J, et al. The immunosuppressive drug thalidomide induces T helper cell type 2 (Th2) and concomitantly inhibits Th1 cytokine production in mitogen- and antigen-stimulated human peripheral blood mononuclear cell cultures. Clin Exp Immunol. 1995;99:160–167. doi: 10.1111/j.1365-2249.1995.tb05527.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Haslett PAJ, Corral LG, Albert M, et al. Thalidomide costimulates primary human T lymphocytes, preferentially inducing proliferation, cytokine production, and cytotoxic responses in the CD8+ subset. J Exp Med. 1998;187:1885–1892. doi: 10.1084/jem.187.11.1885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Corral LG, Haslett PAJ, Muller GW, et al. Differential cytokine modulation and T cell activation by two distinct classes of thalidomide analogues that are potent inhibitors of TNF-α. J Immunol. 1999;163:380–386. [PubMed] [Google Scholar]
- 15.Verbon A, Juffermans NP, Speelman P, et al. A single oral dose of thalidomide enhances the capacity of lymphocytes to secrete gamma interferon in healthy humans. Antimicrob Agents Chemother. 2000;44:2286–2290. doi: 10.1128/aac.44.9.2286-2290.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ng WF, Duggan PJ, Ponchel F, et al. Human CD4+CD25+ cells: A naturally occurring population of regulatory T cells. Blood. 2001;98:2736–2744. doi: 10.1182/blood.v98.9.2736. [DOI] [PubMed] [Google Scholar]
- 17.Annunziato F, Cosmi L, Liotta F, et al. Phenotype, localization, and mechanism of suppression of CD4+CD25+ human thymocytes. J Exp Med. 2002;196:379–387. doi: 10.1084/jem.20020110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bach JF. Regulatory T cells under scrutiny. Nature Rev Immunol. 2003;3:189–198. doi: 10.1038/nri1026. [DOI] [PubMed] [Google Scholar]
- 19.Belkald Y, Piccirillo GA, Mendez S, et al. CD4+CD25+ regulatory T cells control Leishmania major persistence and immunity. Nature. 2002;420:502–507. doi: 10.1038/nature01152. [DOI] [PubMed] [Google Scholar]
- 20.Shevach EM. CD4+CD25+ suppressor T cells: More questions than answers. Nature Rev Immunol. 2002;2:389–400. doi: 10.1038/nri821. [DOI] [PubMed] [Google Scholar]
- 21.Gray CP, Arosio P, Hersey P. Association of increased levels of heavy-chain ferritin with increased CD4+ CD25+ regulatory T-cell levels in patients with melanoma. Clin Cancer Res. 2003;9:2551–2559. [PubMed] [Google Scholar]
- 22.Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. JASA. 1958;53:457–481. [Google Scholar]
- 23.Cox DR. JR Stat Soc Ser B. 1972;34:187–202. [Google Scholar]
- 24.Lambeck AJ, Crijns AP, Leffers N, et al. Serum cytokine profiling as a diagnostic and prognostic tool in ovarian cancer: a potential role for interleukin 7. Clin Cancer Res. 2007;13:2385–2391. doi: 10.1158/1078-0432.CCR-06-1828. [DOI] [PubMed] [Google Scholar]
- 25.Matijevic M, Uban RG. Use of interferon-gamma ELISPOT in monitoring immune responses in humans. Methods Mol Biol. 2005;302:237–252. doi: 10.1385/1-59259-903-6:237. [DOI] [PubMed] [Google Scholar]
- 26.Korn EL, Liu PY, Lee SJ, et al. Meta-analysis of phase II cooperative group trials in metastatic stage IV melanoma to determine progression-free and overall survival benchmarks for future phase II trials. J Clin Oncol. 2008;26:527–534. doi: 10.1200/JCO.2007.12.7837. [DOI] [PubMed] [Google Scholar]
- 27.Calabrese L, Fleischer AB. Thalidomide: Current and potential clinical applications. Am J Med. 2000;108:487–495. doi: 10.1016/s0002-9343(99)00408-8. [DOI] [PubMed] [Google Scholar]
- 28.Rosenberg SA. Overcoming obstacles to the effective immunotherapy of human cancer. Proc Natl Acad Sci USA. 2008;105:12643–12644. doi: 10.1073/pnas.0806877105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kirkwood JM, Lorigan P, Hersey P, et al. A phase II trial of tremilimumab (CP-675,206) in patients with advanced refractory or relapsed melanoma. J Clin Oncol. 2008;26:488s. doi: 10.1158/1078-0432.CCR-09-2033. (suppl; abstr 9023) [DOI] [PubMed] [Google Scholar]