Phase II study of dasatinib in patients with previously treated malignant mesothelioma (Cancer and Leukemia Group B 30601): a brief report

Abstract

Introduction

We conducted a phase II trial of dasatinib in malignant mesothelioma (MM) patients to evaluate its toxicity and efficacy as a second-line treatment.

Material and Methods

Patients with unresectable MM and no symptomatic effusions were given dasatinib 70 mg twice daily as part of a 28-day cycle. We also measured plasma VEGF and PDGFβ and serum CSF-1 and mesothelin-related protein at baseline and during therapy.

Results

Forty-six patients were enrolled in this study. Fifty percent of the first 12 patients enrolled experienced ≥ grade 3 treatment-related adverse events, and therefore, the starting dose was reduced to 50 mg twice daily. Grade 3 and 4 toxicities included fatigue (11%) and pleural effusion (9%). The overall disease control rate was 32.6%, and PFS at 24 weeks was 23% (95% CI: 13.5%, 40.0%). Survival was markedly longer in patients with lower pre-treatment CSF-1 levels and in patients whose CSF-1 levels decreased from baseline during therapy.

Discussion

Single-agent dasatinib has no activity in MM and is associated with pulmonary toxicities that prohibit its use in an unselected MM population.

Introduction

Cisplatin in combination with pemetrexed is now a standard first-line treatment for malignant mesothelioma (MM) (1). However, once patients fail treatment with a pemetrexed-containing regimen, there is no standard of care available. Dasatinib, an aminothiazole analogue, is an orally administered protein tyrosine kinase (PTK) inhibitor with activity against several dysfunctional signaling pathways in MM, including the SRC family kinases (SFKs), vascular endothelial growth factor (VEGF), platelet-derived growth factor β (PDGFβ) receptor, and several other PTKs (2–3). The present study was based on the hypothesis that dasatinib would have direct and indirect anti-proliferative effects on mesothelioma and would slow disease progression no longer controlled with first-line platinum and pemetrexed therapy. As secondary endpoints, we analyzed the effect of dasatinib on plasma VEGF and PDGF levels. In addition, we followed levels of serum colony stimulating factor-1 because it activates MAPK in myeloid progenitors via a SRC-dependent mechanism (4) and serum mesothelin-related protein because it is known to be decreased in patients responding to mesothelioma therapy (5).

Materials and Methods

Patient Selection

Eligible patients were over 18 years of age, had histologically confirmed MM not amenable to curative surgery, had unidimensionally measurable disease, and an ECOG performance status of 0–1. Patients were excluded if they were pregnant, nursing, on antithrombotic or antiplatelet agents, or had documented brain metastases, symptomatic pleural effusion, or serious cardiac disease or bleeding disorders. Patients were required to fail one and only one pemetrexed-containing chemotherapy regimen. Patients were allowed to have received prior intrapleural therapy with sclerosing agents or bleomycin. At least 4 weeks were required since prior major surgery and 4 weeks since first-line chemotherapy or radiotherapy to lesions other than those used for measurement. Patients were required to have adequate hematologic, hepatic, and renal function and an electrocardiogram showing a QTc interval of <450 msec. All patients provided informed consent prior to receiving study treatment, and the study was approved by the Institutional Review Boards of participating institutions.

Treatment Schedule

Dasatinib was given orally at an initial fixed dose of 70 mg twice daily. After 23 patients had enrolled, however, the dose was reduced to 50 mg twice daily because of poor tolerance, including fatigue and pleural effusion. One cycle was defined as 28 days.

Study Endpoints and Analysis

The primary objective of this study was to determine the rate of progression-free survival (PFS) at 24 weeks. Secondary objectives included description of: (1) response rate (CR+PR); (2) response duration; (3) overall survival (OS); (4) toxicity; and (5) analysis of plasma VEGF and PDGF and serum CSF-1 and mesothelin-related protein levels. Response assessment was performed using modified RECIST criteria as described before (6). A CT scan was performed at baseline and every 8 weeks until the patient had documented progression. Plasma VEGF was analyzed using Human VEGF QuantikineR^® ELISA kit (R&D Systems, Minneapolis, MN ), and plasma PDGF was analyzed using Human PDGF-AB ELISA kit (Insight Genomics, Falls Church, VA). Serum CSF-1 levels were detected using Human M-CSF Immunoassay QuantikineR kit (R&D Systems, Minneapolis, MN ), and serum mesothelin-1 related protein levels were analyzed using MESOMARKR assay (FujirebioR Diagnostics, Inc, Malvern, PA).

Sample Size Considerations

Sample size was determined as before (7) by applying the following considerations: if the true 24-week PFS rate for the single agent was ≥54%, we would conclude that the agent was worthy of further investigation. However, if the true 24-week PFS rate was ≤34%, we would conclude that the agent was not worthy of further investigation. Under constant hazards, a 24-week PFS of 34% and 54% corresponded to a median PFS of 3.55 months and 6.22 months or a hazard ratio of 1.75. Forty three eligible patients were to be enrolled in the study. The probability of erroneously concluding that the treatment was worthy of further investigation (p ≥ 0.54) when the success rate was truly ≤34% (p ≤ 0.34) was 0.06. The probability of erroneously concluding that the treatment was not worthy of further investigation (p ≤ 0.34) when the success rate was truly ≥54% (p ≥ 0.54) was 0.13. If 19 or fewer (42.9%) eligible patients were successful in their treatment, we would conclude that the treatment was not worthy of additional investigation.

Statistical Analysis

OS was measured from the day of registration until date of death; living patients were censored at the date of last follow-up. PFS was measured from the day of registration until disease progression or death, whichever came first; living patients who did not progress were censored at the date of last follow-up. Progression was defined as at least a 20% increase in the sum of the longest diameter (LD) of target lesions or the appearance of one or more new lesions. Kaplan-Meier estimates were used to illustrate PFS and OS curves. The correlation of response rate and biomarkers was examined using Wilcoxon rank sum test. In multivariate analysis, a logistic regression model was used to explore the relationship of response rate and biomarkers after adjusting for other significant prognostic factors. To examine the relationship between these biomarkers and PFS and OS, a log rank test and a Cox’s proportional hazard model was used, after adjusting for other prognostic factors.

Results

Drug Safety

Forty-six patients were enrolled in this study from 09/15/2007 to 08/31/2009 (Table 1). Three patients withdrew from the study before treatment initiation, leaving 43 patients eligible for evaluation. Six of the first 12 patients experienced grade ≥3 adverse events. After 23 patients had enrolled in the study, the starting dose of dasatinib was decreased from 70 mg to 50 mg twice daily. The most common grade 3 and 4 non-hematologic events were fatigue (11%), pleural effusion (9%), and dyspnea (7.5%). One patient developed grade 3 pericardial effusion. The most common grade 3 or 4 hematologic events (7% of patients) were grade 3 anemia (2%) and lymphopenia (4%). A greater percentage of grade ≥3 toxicities was seen in patients receiving 70 mg of dasatinib (52%) compared to 50 mg of dasatinib (39%).

Table 1.

Baseline Characteristics

Variable	N=46*
Age
Median, years (range)	68 (35, 81)
Gender
Male	31 (72%)
Female	12 (28%)
ECOG performance status
0	19 (44%)
1	24 (56%)
Site of origin
Pleura	36 (84.0%)
Peritoneum	6 (14.0%)
Other	1 (2.0%)
Histology
Epithelial	33 (77%)
Biphasic	5 (12%)
Sarcomatoid	2 (5%)
Missing	3 (7%)
Prior therapy
Chemotherapy	43 (100%)
Radiation	6 (14.0%)
Surgery	23 (53%)

Abbreviations: ECOG = Eastern Cooperative Oncology Group; N = Number of patients

^*3 patients were excluded because they withdrew from study before starting therapy.

Treatment Response

Most patients stopped therapy due to disease progression (56%) or adverse events (12%). The median follow-up time was 84 weeks as of April 29, 2011. No complete responses were observed (Table 2). Although partial responses were seen in 4.7% of patients (95% CI: 0.6%, 15.8%) and stable disease in 28% of patients, progressive disease was seen in 42% of patients. The 24-week PFS rate was 23% (95% CI: 13.5%, 40.0%). The 48-week PFS rate was 2.3% (95% CI: 0.3%, 16.1%). The clinical benefit rate (CR + PR + SD) was 32.6% (95% CI: 19.1%, 48.5%). Median PFS was 9.1 weeks (95% CI: 7.4, 16.9). Median overall survival was 26.1 weeks (95% CI: 18.6, 36.7), with a 1-year survival rate of 25.6%. Thirty-eight deaths had occurred at the time of analysis. No difference in overall survival or PFS was observed in log-rank analysis between patients treated with the initial or reduced dose of dasatinib (p=0.68 and p=0.82, respectively).

Table 2.

Best Responses to treatment^*

Best Response	N	%
Partial response	2	5
Stable disease	12	28
Progressive disease	18	42
Unevaluable#	6	14
Early deaths**	4	9
Missing	1	2
Total	43	100

^*Excludes canceled patients.

^# 6 'unevaluable' patients had off treatment dates before Day 1 of the 3^rd cycle, when the first on-treatment CT scan was scheduled. Four had patient refusal as an off-treatment reason; 1 withdrew; 1 due to AE, and 1 as 'other'.

^**4 patients had 'early deaths' (i.e., patient did not survive to Day 1 of the 3^rd cycle) when the first on-treatment CT scan was scheduled.

Correlative Data

Biomarker levels were measured before therapy and at the end of the first and second cycle of therapy, and hazard ratios were then estimated for risk of death (Table 3). A higher level of CSF-1 before treatment was associated with shorter overall survival (Figure 1) (HR=1.63, 95% CI 1.01, 2.64; p-value = 0.046; n=36). A decrease in CSF-1 levels from baseline to the end of cycle 2 was associated with longer PFS (HR=6.67, 95% CI: 1.13, 39.4; p-value=0.037; n=11), and the median PFS for these patients was 25 weeks.

Table 3.

Hazard ratios for risk of death according to biomarker levels

	Overall Survival		Progression-free Survival
	P-value	Hazard Ratio (95% CI)	P-value	Hazard Ratio (95% CI)
Pre-VEGF* (n=38)	0.96	1.00(0.92,1.1)	0.19	1.06(0.97,1.15)
VEGF: log(cycle 1/pre) (n=20)	0.71	1.08(0.72,1.62)	0.41	0.83(0.58,1.30)
VEGF: log(cycle 2/pre) (n=13)	0.95	1.01(0.69,1.48)	0.84	1.04(0.74,1.45)
Pre-PDGF (n=38)	0.92	0.97(0.93,1.07)	0.61	1.02(0.95,1.09)
PDGF: log(cycle 1/pre) (n=20)	0.61	0.86(0.47,1.56)	0.10	0.60(0.32,1.11)
PDGF: log(cycle 2/pre) (n=13)	0.60	1.14(0.70.1.85)	0.51	1.15(0.76,1.74)
Pre- CSF-1** (n=36)	0.05	1.63(1.01,2.64)	0.27	1.27 (0.83,1.96)
CSF-1: log(cycle 1/pre) (n=17)	0.15	0.37(0.10,1.41	0.27	0.53(0.17,1.64)
CSF-1: log(cycle2 /pre) (n=11)	0.11	5.47(0.67,44.7)	0.04*	6.67(1.13,39.4)
MRP (n=36)	0.52	0.99 (0.96,1.02)	0.53	1.01(0.98,1.04)
MRP: log(cycle1/pre) (n=17)	0.23	1.95(0.66,5.75)	0.14	2.24 (0.77,6.49)
MRP: log(cycle2/pre) (n=11)	0.63	1.32 (0.42,4.12)	0.11	2.67 (0.80,8.91)

^*Pre-Treatment VEGF values in the model were divided by 100 for more interpretable Hazard Ratios.

^**M-CSF values in the model were divided by 1000 for more interpretable Hazard Ratios.

Abbreviation: MRP, mesothelin related protein

Figure 1.

Overall Survival time in relation to baseline CSF-1 level.

Discussion

This phase II study of single-agent dasatinib in patients with previously treated MM demonstrated that 23% of subjects were progression free at 24 weeks. This percentage is less than the pre-specified 42% needed to declare the regimen worthy of further investigation. Additionally, both median PFS (9.1 weeks) and median OS (26.1 weeks) are comparable to those seen with best supportive care in a recent randomized study (8). The only biomarker associations observed were that OS was markedly longer in patients with lower pre-treatment CSF-1 levels and PFS was nearly 3 times longer (25 weeks) in patients whose CSF-1 levels decreased from baseline during therapy. Both of these observations support the notion that CSF-1 may play a role in more aggressive mesothelioma biology and are consistent with other studies predicting poor survival in patients with Hodgkin’s lymphoma (9), metastatic colorectal carcinoma (10), prostate cancer (11), and NSCLC (12). SRC participates in regulation of CSF-1 activation of MAPK in myeloid progenitors (13), and SRC inhibition has been reported to impair downstream signaling from the CSF-1 receptor (14). Identification of tumors with molecular markers of sensitivity to SRC inhibition or dasatinib may be a more successful strategy in future clinical studies testing inhibition of this pathway in mesothelioma.

Despite promising preclinical data suggesting that SRC plays a critical role in advanced MM, our clinical study suggests that dasatinib is inactive in unselected patients with previously treated unresectable mesothelioma. We therefore conclude that the lack of activity and unfavorable toxicity of dasatinib should preclude its further development in an unselected mesothelioma patient population.