A Phase II Study of Pazopanib in Patients with Malignant Pleural Mesothelioma: NCCTG N0623 (Alliance)

Abstract

Purpose

Preclinical and clinical data have shown promise in using antiangiogenic agents to treat malignant pleural mesothelioma (MPM). We conducted this phase II study to evaluate the efficacy and toxicity of single‐agent pazopanib in patients with MPM.

Materials and Methods

Patients with MPM who had received 0–1 prior chemotherapy regimens were eligible to receive pazopanib at a dose of 800 mg daily. The primary endpoint was progression‐free survival rate at 6 months (PFS6), with a preplanned interim analysis for futility. Secondary endpoints included overall survival (OS), PFS, adverse events assessment and clinical benefit (complete response, partial response [PR], and stable disease [SD]).

Results

Thirty‐four evaluable patients were enrolled, with a median age of 73 years (49–84). The trial was closed early because of lack of efficacy at the preplanned interim analysis. Only 8 patients (28.6%; 95% confidence interval [CI], 13.2–48.7%) in the first 28 evaluable were progression‐free at 6 months. PFS6 was 32.4% (95% CI, 17.4–50.5). There were 2 PR (5.9%) and 16 SD (47.1%). The overall median PFS and OS were 4.2 months (95% CI, 2.0–6.0) and 11.5 months (95% CI: 5.3–18.2), respectively. The median PFS and OS for the previously untreated patients was 5.4 months (95% CI, 2.7–8.5) and 16.6 months (95% CI, 6.6–30.6), respectively; and 2.0 months (95% CI, 1.3–4.2) and 5.0 months (95% CI: 3.0–11.9), respectively, for the previously treated patients. Grade 3 or higher adverse events were observed in 23 patients (67.6%).

Conclusion

Single‐agent pazopanib was poorly tolerated in patients with MPM. The primary endpoint of PFS6 was not achieved in the current study. ClinicalTrials.gov identification number. NCT00459862.

Implications for Practice

Single‐agent pazopanib did not meet its endpoint in this phase II trial in malignant mesothelioma. Pazopanib is well tolerated in mesothelioma patients with a manageable toxicity profile. There is a need to better identify signals of angiogenesis that can be targeted in mesothelioma. Encouraging findings in frontline treatment warrant further investigations in combination with chemotherapy or immunotherapy.

Short abstract

This article presents results of a phase II study evaluating the efficacy and toxicity of single agent pazopanib in patients with malignant pleural mesothelioma.

Introduction

With increasing occurrence in the past few decades, malignant pleural mesothelioma (MPM) remains a serious problem worldwide because of its refractory nature in response to current therapy and its highly mortality rate. Although the use of the principal etiological agent, asbestos, has declined significantly, it has been estimated that the incidence of MPM will increase in the next decades mainly because of the previously wide use of asbestos and the long latency period from the first exposure to asbestos to the disease onset 1. MPM is a highly aggressive tumor with a median survival of no more than 1 year without treatment 1. It is difficult to diagnose at an early stage, is refractory to conventional chemotherapy and radiotherapy, and is seldom cured by radical resection 2. Antifolates, specifically pemetrexed and raltitrexed, combined with a platinum regimen have been the cornerstone of treatment of MPM in recent years, resulting in improvement of survival by a few months and enhancement quality of life 3. However, relapse rates remain high and long‐term survival is still very poor. Most patients with advanced‐stage MPM progress during or shortly after the first‐line chemotherapy and no standard treatment has yet been defined 3. Therefore, new approaches are urgent needed to improve the treatment activity as well as the tolerability in advanced‐stage MPM.

Angiogenesis, one of the hallmarks of cancer including mesothelioma, plays a key role in tumor growth, invasion, and metastasis 4, 5. Improved understanding of the molecular biology underlying MPM has led to the discovery of an essential role of angiogenesis in this disease. Both vascular endothelial growth factor (VEGF) and platelet‐derived growth factor (PDGF) are autocrine growth factors in MPM 6. Previous studies demonstrated that VEGF is overexpressed in MPM cell lines, tumor tissues, and pleural effusion as well as patients’ serum 7, 8, 9. In addition, VEGF is an independent prognostic marker and is closely associated with tumor stage 8, 9. Compared with a population who had a history of asbestos exposure but without MPM development, patients diagnosed with MPM had significantly higher serum VEGF levels that were positively correlated with tumor stage 10. Recombinant VEGF stimulated proliferation of MPM cells, whereas the use of neutralizing antibodies against VEGF or VEGF receptors (VEGFRs) inhibited tumor growth 8. In addition to promoting tumor angiogenesis and stimulating tumor growth, VEGF is also involved in the development of pleural effusions by enhancing vascular permeability in MPM 11. Moreover, the other critical factor in angiogenesis, PDGF receptor (PDGFR), has been shown to be differentially expressed in MPM, and blocking PDGFR in MPM cells leads to tumor inhibition and enhanced sensitivity to chemotherapy 12, 13.

Antiangiogenic drugs such as monoclonal antibodies directed against VEGF and small molecule tyrosine kinase inhibitors (TKIs) have demonstrated antitumor activities in various solid tumors. The high microvascular density found in MPM make this disease an attractive target for antiangiogenic therapies. Pazopanib, a second generation multikinase inhibitor against VEGFR‐1, ‐2, and ‐3; platelet‐derived growth factor receptor‐α; platelet‐derived growth factor receptor‐β; and c‐kit, was approved by the U.S. Food and Drug Administration in 2009 for the treatment of advanced renal cell carcinoma 14. In addition, pazopanib has also shown great antitumor activity and good tolerability in many other solid tumors including non‐small cell lung cancer, cervical cancer, breast cancer, thyroid cancer, soft tissue sarcoma, and nasopharyngeal carcinoma 15, 16, 17, 18, 19. Pazopanib seems to have a lower incidence of adverse events (e.g., hand‐foot syndrome, diarrhea, and myelosuppression) compared with other VEGFR TKI agents 19. This makes it more attractive for use in MPM, because an increasing fraction of patients with MPM are older adults with comorbidities 1 who may not be able to tolerate common treatment‐related toxicities.

Based on emerging data with antiangiogenic agents demonstrating promising activities in multiple cancers as well as the critical role of angiogenesis in MPM, we conducted North Central Cancer Treatment Group (NCCTG) N0623, a single‐arm phase II clinical trial of pazopanib for patients with MPM with the goal of evaluating its efficacy in this cohort. NCCTG is now part of the Alliance for Clinical Trials in Oncology.

Materials and Methods

Patients

Patients aged 18 years or older with histologically or cytologically confirmed malignant pleural mesothelioma with unresectable disease were eligible. Patients were either chemo‐naive or had only one prior chemotherapy regimen, with Eastern Cooperative Oncology Group performance score 0–2, and were required to have at least one measurable lesion to be used to assess response as defined by RECIST 20. Patients were required to have a life expectancy of more than 12 weeks. Additional eligibility criteria included adequate organ function. A negative pregnancy test must have been performed within 7 days prior to registration for women of childbearing potential.

Patients who were pregnant or nursing were not eligible, and birth control methods were required, although oral contraceptives were not considered reliable because of potential drug‐drug interaction. Patients with second primary malignancy within the past 5 years (except for carcinoma in situ of the cervix, nonmelanomatous skin cancer, or low‐grade [Gleason score ≤ 6] localized prostate cancer) were ineligible. Patients with any of the following conditions were excluded: uncontrolled infection; uncontrolled hypertension; any condition that impaired the ability to swallow and retain pazopanib tablets; any other severe underlying diseases; history of allergic reactions attributed to compounds of similar chemical or biologic composition to pazopanib or other agents used in the study; >+1 proteinuria on two consecutive dipsticks taken ≥1 week apart; medications that act through the CYP450 system; or other concurrent chemotherapy, immunotherapy, hormonal therapy, or radiotherapy. Other exclusion criteria included significant cardiac history; serious or nonhealing wound, ulcer, or bone fracture; history of abdominal fistula, diverticulosis, gastrointestinal perforation, or intra‐abdominal abscess ≤28 days prior to registration; history of cerebrovascular accident ≤6 months; current use of therapeutic warfarin; uncontrolled diabetes; interstitial pneumonia or extensive and symptomatic interstitial fibrosis of the lung; QTc prolongation or other significantly electrocardiographic abnormalities; and untreated or uncontrolled central nervous system metastases or seizure disorder. Also excluded were patients who were HIV positive and on combination antiretroviral therapy or had experienced major surgery, open biopsy, or significant traumatic injury ≤4 weeks, or concurrent radiation therapy to chest. The study was approved by the institution review board at Mayo Clinic, and all patients were required to sign the written informed consent (NCT00459862).

Study Design and Treatment

Pazopanib was administered orally at a fixed dose of 800 mg daily for a maximum of 2 years from registration or until disease progression, unacceptable toxicity, investigator's decision to remove patient, patient refusal, or alternative treatment. Twenty‐one days of treatment constituted one cycle. Dose modifications included dose level −1 (600 mg daily), dose level −2 (400 mg daily), and dose level −3 (200 mg daily). Adverse events were graded according to the National Cancer Institute's Common Terminology Criteria Version 3.0. Doses were delayed until toxicity resolved to grade ≤1 (grade ≤2 for blood and/or bone marrow). Dose modifications were made for patients with grade ≥2 hemorrhage, coagulation, and metabolic (aspartate aminotransferase [AST]/alanine aminotransferase [ALT]); grade ≥3 nonhematologic; grade ≥4 bone marrow toxicities; and moderate hypertension and proteinuria.

Blood pressure monitoring (twice daily) and blood pressure diaries were required for all enrolled patients for the duration of the study. Patients’ history, physical examination, and laboratory tests were assessed before each subsequent treatment cycle. Tumor assessments were performed at least every other cycle after completing cycle 2 (after cycles 2, 4, 6, etc.). Tumor response was assessed by RECIST 1.0 criteria.

Statistical Design and Analysis

The primary goal of this single‐arm phase II study was to evaluate the efficacy and biological effects and the safety profile of pazopanib treatment in patients with MPM. The primary endpoint of this study was the proportion of patients who remain progression free at 6 months based on the RECIST 1.0 criteria 21. Secondary endpoints included response rate, overall survival, progression‐free survival, adverse events assessment, and clinical benefit (complete response [CR], partial response [PR], and stable disease [SD]). Evaluation of predictive and serologic markers to better understand pazopanib activity and its target inhibition was an exploratory endpoint.

A success for the primary endpoint was defined as an evaluable patient who was alive and progression free at 6 months and had not gone off treatment due to adverse reactions, refusal, or alternate therapy. Based on a three‐outcome design with one preplanned interim analysis 21, the trial was designed to test the hypothesis that the true 6‐month progression‐free survival (PFS) rate is at most 50% versus the alternative hypothesis that the 6‐month PFS (success) rate is at least 65% using a one‐sided type I error rate of 10%. The pemetrexed plus cisplatin first‐line study showed a median time to progression of 5.7 (∼6) months 22. In the current study, patients who had at most one prior systemic therapy were also eligible. However, based on NCCTG N0021, a trial in a similar patient population, <18% of patients had received one prior chemotherapy for MPM 23. At the time of designing this trial, options for MPM upon progression on frontline therapy were limited. Moreover, trial enrollment and accrual were often by low number of eligible patients. These considerations led to including treatment‐naive patients as well as those who progressed on one prior line of therapy in this trial. Given the modest accrual rate of patients with MPM, we based our design and primary analysis on all patients enrolled in the trial. With a sample size of 50 evaluable patients, the treatment was to be considered effective, ineffective, or inconclusive if at least 30 successes, at most 28 successes, or 29 successes are observed respectively. If the true success rate was at least 65%, this design has 80% power to detect an effective treatment and a 5% chance of declaring the results to be inconclusive. The goal of this design was to more accurately reflect the clinical reality that a single success should not make the difference between a determination of “ineffective” and “promising.”

A preplanned interim analysis based on the 6‐month PFS rate was to take place after 28 evaluable patients were accrued and followed for 6‐months for assessment of progression status. The interim analysis would declare the regimen ineffective and accrual to the trial terminated if only at most 14 successes were observed. If 15 or more successes were observed in the first 28 evaluable patients, then accrual would continue and additional 22 patients will be enrolled.

PFS was defined as the time from registration to the first date of disease progression or death as a result of any cause. PFS was censored at the date of the last contact for patients alive and progression free at the time of this analysis. OS was defined as the time from registration to death due to any cause. Exact binomial confidence intervals for the proportion of successes (6‐month PFS rate) were constructed. The distribution of PFS and OS time was estimated using the Kaplan‐Meier method 24. The proportion of patients with confirmed responses (complete and partial) as well as disease control rate (confirmed CR, confirmed PR, and stable disease lasting ≥4 weeks) was summarized using exact binomial confidence intervals. Adverse events, regardless of attribution to treatment, were summarized similarly, using the maximum grade for each type of adverse event experienced by a patient across all cycles of treatment. All analyses were performed in the overall cohort, as well as by line of therapy (first‐line versus second‐line treatment for MPM).

Data collection and statistical analyses were conducted by the Alliance Statistics and Data Center. Data quality was ensured by review of data by the Alliance Statistics and Data Center and by the study chairperson following Alliance policies. All analyses were based on the study database frozen on May 6, 2014. This phase II therapeutic trial was monitored at least twice annually by the Mayo Clinic or Alliance Data and Safety Monitoring Boards.

Results

Patient Characteristics

Patient recruitment was undertaken from March 23, 2007 through December 28, 2008. Thirty‐four patients were enrolled before the trial was permanently closed because of lack of efficacy after the preplanned interim analysis. All patients had received study treatment and were eligible (Fig. 1). Table 1 summarizes the patient characteristics. Of the 34 patients, 82.3% were men, and median age was 73 (49–84). Fifteen (44.1%) patients had received previous systemic therapy, all with pemetrexed‐containing regimen (nine of whom had a response of SD or better to therapy), and 19 (55.9%) patients were treatment‐naive. Race composition was predominantly white (97.1%) with 2.9% black patients. Performance status was 0 in 44.1%, 1 in 50%, and 2 in 5.9% of patients.

Figure 1.

Consort diagram.

Table 1.

Patients characteristics (n = 34)

Characteristics	Prior treatment (n = 15), n (%)	No prior treatment (n = 19), n (%)	Overall (n = 34), n (%)
Median age (range), years	69 (57‐84)	74 (49‐83)	73 (49‐84)
Gender, male	12 (80.0)	16 (84.2)	28 (82.3)
Race
White	15 (100.0)	18 (94.7)	33 (97.1)
Black	0 (0)	1 (5.3)	1 (2.9)
Performance score
0	8 (53.3)	7 (36.8)	15 (44.1)
1	6 (40.0)	11 (57.9)	17 (50.0)
2	1 (6.7)	1 (5.3)	2 (5.9)
Histologic type
Epithelial	10 (66.7)	16 (84.2)	26 (76.5)
Sarcomatous	1 (6.7)	1 (5.3)	2 (5.9)
Mixed (biphasic)	1 (6.7)	2 (10.5)	3 (8.8)
Desmoplastic	1 (6.7)	0 (0)	1 (2.9)
Un‐ or poorly differentiated	1 (6.7)	0 (0)	1 (2.9)
Unknown or unavailable	1 (6.7)	0 (0)	1 (2.9)
Any previous cancer	3 (20.0)	3 (15.8)	6 (17.7)
Prior systemic cancer therapy
Alimta/cisplatin	8 (53.3)
Alimta/carboplatin	5 (33.3)
Alimta/cisplatin/taxol	1 (6.7)
Alimta/cisplatin/Avastin	1 (6.7)
Subsequent treatment received after off protocol treatment
No subsequent treatment	6 (40.0)	9 (47.4)	15 (44.1)
Subsequent treatment with Alimta	3 (20.0)	9 (47.4)	12 (35.3)
Subsequent treatment without Alimta	6 (40.0)	1 (5.3)	7 (20.6)

The median (range) of follow‐up on the three surviving patients is 28 (25.9–54.7) months. The median number of treatment cycles administered was 6 (1–25) cycles in the group with no prior treatment and 2 (1–16) cycles in the group with prior treatment. All patients were off treatment at the time of this analysis. Twenty‐three of 34 (67.7%) patients ended treatment for disease progression.

Primary Endpoint

Based on a preplanned interim analysis, only 8 (28.6%; 95% confidence interval [CI], 13.2%–48.7%) patients in the first 28 evaluable patients were progression free at 6 months. Thus, the study did not meet the protocol defined success criteria for the interim analysis. The 6‐month PFS rate was 32.4% (11/34; 95% CI, 17.4%–50.5%) in all 34 registered patients, 42.1% (8/19; 95% CI, 20.3%–66.5%) in the group with no prior treatment, and 20% (3/15; 95% CI, 4.3%–48.1%) in the group with prior treatment.

Secondary Efficacy Endpoints

The overall clinical benefit rate was 52.9% (95% CI, 35.1%–70.2%); 63.2% (95% CI, 38.4%–83.7%) in the group with no prior treatment and 40% in the group with prior (95% CI, 16.3%–67.7%) treatment. No complete responses were reported and there were two patients with partial response in the group with no prior treatment (10.5%). The median PFS was 4.2 months (95% CI, 2.0–6.0) in the overall group, 5.4 months (95% CI, 2.7–8.5) in treatment‐naive group (no prior treatment group), and 2.0 months (95% CI, 1.3–4.2) in previously treated patients (Fig. 2). The median OS was 11.5 months (95% CI, 5.3–18.2) in the overall group: 16.6 months (95% CI, 6.6–30.6) in treatment‐naive and 5.0 months (95% CI, 3.0–11.9) in previously treated patients (Fig. 3).

Figure 2.

Kaplan‐Meier estimates of progression‐free survival in the entire study population and by prior treatment (n = 34).Abbreviation: CI, confidence interval.

Figure 3.

Kaplan‐Meier estimates of overall survival in the entire study population and by prior treatment (n = 34).Abbreviation: CI, confidence interval.

Adverse Events

All 34 patients were evaluated for adverse events (AEs). Three patients had grade 5 events (none were treatment related). One patient died of acute respiratory distress syndrome (ARDS) secondary to disease progression and another had grade 5 event of depression, but events were deemed unrelated to pazopanib. Four patients reported grade 4 events: chest pain, fatigue, dyspnea, pleuritic pain, and adult respiratory distress syndrome. Grade 3 and higher (3+) adverse events were reported in 23 (67.6%) patients, and grade 4+ adverse events were reported in 5 (14.7%) patients. Grade 3 hematologic events were reported in two patients (5.9%): one patient with neutrophil count decreased and the other one with lymphopenia. There were no grade 4 hematologic AEs reported.

The most common (occurring in 10% [n = 4] or more of patients) grade 3 nonhematologic AEs were hypertension (14.7%), ALT increase (14.7%), and AST increase (11.8%). The common AEs related with other TKIs, such as skin rash, diarrhea and hand‐foot syndrome, were less frequent in this trial, specifically grade 1 (32.4%) and grade 3 (2.9%) rash and grade 1 (32.4%) and grade 2 (11.8%) diarrhea. Only one patient had grade 1 hand‐foot syndrome (2.9%).

Eight patients (23.5%) discontinued treatment for adverse events, seven of whom reported specific adverse events at the time of treatment discontinuation. One patient had fatigue, ascites, and dyspnea; one patient had ARDS; one patient had hypertension, and dose was held for over 14 days; one patient had grade 3 pain; one patient had persistent or recurrent liver toxicity (elevation of AST and ALT); and one patient had grade 3 ALT and AST.

Thirty‐four patients started the treatment for cycle 1, and four (11.8%) patients had dose reduction or adjustment (defined as not taking the assigned dose for 21 days) during cycle 1. Of the 30 patients who completed cycle 1 with no dose reductions or adjustments, 2 (6.7%) patients did not start cycle 2, and 5 (16.7%) had dose reduction or adjustment during cycle 2. Of the 23 patients who completed cycle 2 with no dose reduction or adjustments, 5 patients (21.7%) did not start cycle 3, and 10 (43.5%) had at least one dose reduction or adjustment in cycle 3 and beyond. Details on dose reduction are summarized in Table 2.

Table 2.

Summary of dose reduction

Dose reduction	Prior treatment, n (%)	No prior treatment, n (%)	Overall, n (%)
During first cycle	15	19	34
No reduction or adjustmenta (at 800 mg daily)	13 (86.7)	17 (89.5)	30 (88.2)
Reduction to 600 mg daily	0 (0)	1 (5.3)	1 (2.9)
Reduction to 400 mg daily	0 (0)	1 (5.3)	1 (2.9)
Adjustmenta	2 (13.3)	0 (0)	2 (5.9)
During cycle 2b	13	17	30
No reduction or adjustmenta (at 800 mg daily)	9 (69.2)	14 (82.4)	23 (76.7)
Reduction to 600 mg daily	1 (7.7)	1 (5.9)	2 (6.7)
Reduction to 400 mg daily	0 (0)	1 (5.9)	1 (3.3)
Adjustmenta	1 (7.7)	1 (5.9)	2 (6.7)
Did not start cycle 2	2 (15.4)	0 (0	2 (6.7)
Cycle 3 and beyondc	9	14	23
No reduction or adjustmenta	2 (22.2)	6 (42.9)	8 (34.8)
At least one reduction or adjustmenta	3 (33.3)	7 (50.0)	10 (43.5)
Did not start cycle 3	4 (44.4)	1 (7.1)	5 (21.7)

^aDid not take assigned dose on all 21 days in a given cycle.

^bPatients who completed cycle 1 with no dose reductions or adjustments.

^cPatients who completed cycle 1 and cycle 2 with no dose reductions or adjustments.

Discussion

MPM is a rare but highly aggressive tumor that remains a significant challenge in modern oncology because of the lack of significant response to systematic therapy. This is especially true for advanced‐stage patients, who have less opportunity to benefit from surgery combined with adjuvant therapy. Although the current first‐line chemotherapy, pemetrexed combined with cisplatin, has a response rate of over 40%, the median survival is only 2.8 months superior to the cisplatin alone 22. Carboplatin has also shown similar efficacy as cisplatin in advanced mesothelioma and is routinely used instead 25. In addition, many patients are not medically suitable for platinum‐based treatment or they declined the option of chemotherapy. Furthermore, after failure to the first‐line therapy, no standard second‐line chemotherapy is available 4, leaving many advanced‐stage patients with MPM suffering from symptoms such as shortness of breath and chest pain. Clearly, both the highly lethality and the low efficacy of current treatment for MPM makes it urgent to identify and test novel systemic compounds in prospective clinical trials. Molecular targeted therapy is one method that is under investigation.

The growing evidence that MPM has the highest VEGF level of any solid tumor, the role of angiogenesis as an important growth signaling pathway in MPM, and confirmation of the overexpression of PDGFR‐β in MPM 26, 27 provided a rationale for targeting these growth factor signaling pathways in MPM. Pazopanib, an oral, small‐molecule tyrosine kinase inhibitor of VEGF, PDGF, and c‐KIT, had been approved for the treatment of advanced renal cell carcinoma (RCC) by the U.S. Food and Drug Administration and shown activity across a range of solid tumors. Our study showed that single‐agent 800‐mg pazopanib daily was not well tolerated in the cohort of patients with MPM, with a 67.6% grade 3 adverse event incidence.

Treatment with single‐agent pazopanib resulted in a 6‐month progression‐free survival (PFS6) of 32.4%, lower than what would be expected according to the trial design. Although our study failed to meet the primary endpoint, the efficacy data, particularly the data on OS, are clinically interesting. The data suggest a superior trend associated with pazopanib treatment (median overall survival [mOS], 11.5 months), especially in the setting of patients with no prior treatment (mOS, 16.6 months), which compared favorably to previously reported mOS 12.1 months for pemetrexed plus cisplatin in the chemotherapy‐naive patients 22. Almost half of all treatment‐naive patients (9/19) went on to receive no further therapy upon progression on pazopanib. Another nine patients receive pemetrexed‐based chemotherapy upon progression. The differences in survival between treatment‐naive and pretreated patients may be explained by different disease biology and natural history of the disease. The discrepancy observed in our study on two survival endpoints might be related, at least partly, to the radiological evaluation method that was based on morphological imaging. In fact, the unique growth pattern of MPM presents challenges for the current linear method of assessing tumor progression or treatment efficacy 28, 29. Even with the modified RECIST criteria, there is still high interobserver variability and the possibility of overclassification of PD 28. In addition, antiangiogenic agents lead to tumor inhibition rather than tumor regression; hence, there is difficulty with determination of PFS based on change of tumor size, which cannot reflect direct vascular effects, possible reduced permeability, and inhibition of tumor growth. More accurate assessment for efficacy of antiangiogenic agents based on the functional and molecular imaging is still needed in future clinical trials of MPM 30.

No complete response and only two partial responses were found in the current study. While there is no clear association between response rate and survival in MPM 31, interestingly, in treatment‐naive patients superior overall survival although limited response rate was also observed in our study, which is consistent with previous results. Early trials of small molecule inhibitors (i.e., gefitinib, erlotinib, imatinib, sorafenib, sunitinib, vetalanib, cediranib) in malignant mesothelioma as first‐ and/or second‐line therapy met with unsatisfactory results, with median PFS (2–3.6 months), median OS (6.7–10 months), and stable disease (34%–72%) 14, 32, 33, 34, 35, 36, 37, 38. The current trial of single pazopanib in MPM yielded similar stable disease rate of 47.1%, median PFS of 4.2 months, and better median OS (11.5 months) in the combined first‐ and second‐line group. However, intertrial comparisons must be interpreted with great caution given the diversities in trial design, sample size, and patients’ characteristics. Also, our study is limited by lacking a control arm and relying on historical values as a comparison for treatment efficacy.

Although our study showed promising results of single‐agent pazopanib on overall survival in treatment‐naive patients with MPM, to date, phase II clinical trials based on antiangiogenic TKI therapy in MPM have failed to meet the primary endpoint of the trials. There may be underlying biological reasons for the nonsignificant activity. First, other signal pathways may exist that are involved in angiogenesis. Thus, inhibition of predominant VEGF/PDGF channels might not be sufficient to completely block signaling pathways in angiogenesis; tumor cells could escape from antiangiogenic therapy through activated alternative pathways 39. In addition, it has been suggested that the main reason for the low activity of EGFR inhibitor (gefitinib, erlotinib) in MPM may be related to the absence of EGFR‐activating mutations in MPM cells 40. Recently, mutations in target kinases have been found to be correlated with resistance to sunitinib 40. Because antiangiogenic TKI agents, like pazopanib, may also inhibit tumor cells directly 41, it is reasonable to extrapolate that the gene mutations in MPM cells might be related to resistance to pazopanib. Furthermore, because most clinical trials in MPM enrolled patients with advanced‐stage disease, earlier intervention might be important for better efficacy of antiangiogenic agents. Considering the central role of VEGF in the formation of neovasculature in the early stages of tumorigenesis, Li and colleagues 11 demonstrated that the most effective inhibition of tumor growth by bevacizumab in the implanted human MPM model occurs when the antiangiogenic drug is administrated in the early stage of tumor implantation.

Theoretically, improved antitumor efficacy of pazopanib in MPM might be achieved through its combination with chemotherapy or other drugs targeting different signal pathways. Imatinib, which showed no significant activity as a single agent in MPM, increased the antitumor efficacy of gemcitabine in mesothelioma xenografts 2. Preclinical data have shown great synergetic or addictive activity by the combination of pazopanib with cytotoxic drugs such as paclitaxel, docetaxel, and topotecan, in a variety of tumor models including pediatric solid tumor, bladder cancer, ovarian cancer, and chronic lymphocytic leukemia 42, 43, 44, 45, 46. It is believed that tumor vascular normalization by anti‐VEGF agents may improve the delivery of cytotoxic drugs by changing tumor interstitial pressure 47. Meanwhile, cytotoxic drugs may enhance the efficacy of antiangiogenic agents by directly damaging the endothelial cells 39. For instance, the MAPS study showed that addition of bevacizumab to cisplatin and pemetrexed significantly improved OS by 2.7 months compared with cisplatin and pemetrexed (18.8 months vs 16.1 months; hazard ratio [HR], 0.77; p = .0167) 48. However, this may come at an added cost of increased toxicity, as has been the case with multiple chemotherapy and TKI combination trials. In a phase II trial of pazopanib and paclitaxel in melanoma, 70% of patients needed dose interruptions due to adverse events 49. Nindetanib, another multikinase inhibitor targeting VEGFR 1, 2, and 3 also showed promising activity in the phase II LUME‐Meso trial when combined with cisplatin and pemetrexed, with PFS of 7.8 months compared with 5.3 months for chemotherapy with placebo (HR, 0.56; 95% CI, 0.34–0.91; p = .017) 50. However, the phase III part of the LUME‐Meso study yielded negative results with no difference in PFS between the two groups 51. Furthermore, tumor vascular normalization, being the result of adaptive resistance to antiangiogenic agents, might provide better delivery of cytotoxic drugs to tumor growth areas 47, thereby providing the rationale for combining pazopanib with chemotherapy to overcome acquired resistance of antiangiogenic agents. In addition to the modification of administration method, the other way to maximize the efficacy of pazopanib in MPM may be with the exploration of predictive biomarkers. In a previous study, monoclonal antibodies of VEGF (e.g., bevacizumab) did not show the meaningful clinical activity when combined with gemcitabine and cisplatin chemotherapy. However, in subgroup analysis, patients with lower circulating levels of VEGF had longer PFS and OS compared with those who had higher VEGF 52. Nikolinakos and colleagues 53 found that serum cytokine and angiogenic factors, such as soluble vascular endothelial growth receptor 2, may be useful biomarkers for pazopanib efficacy in patients with non‐small cell lung cancer. Xu and colleagues demonstrated that polymorphisms in angiogenesis‐ and drug exposure‐ related genes may predict response of pazopanib in patients with RCC 54. Moreover, in terms of decreasing the possibility of pazopanib‐related toxicities, the study in the patients with RCC suggested that UGT1A1 polymorphism was related to pazopanib‐induced hyperbilirubinemia 55.

Conclusion

In the current phase II clinical trial, single‐agent 800‐mg daily pazopanib was well tolerated and feasible in patients with MPM. The trial did not meet the predefined primary endpoint of PFS6. However, based on the intriguingly long OS in this trial, the biologic rationale behind prolonged OS with pazopanib in treatment‐naive patients needs to be further explored. In addition, the identification of effective predictive biomarkers will still be critical for personalizing treatment strategies with pazopanib to achieve maximal efficacy with minimal toxicity. However, given the lackluster results with other antiangiogenic TKIs in mesothelioma, further development of pazopanib for this indication is unlikely to be fruitful.

Author Contributions

Conception/design: Kaushal Parikh, Julian R. Molina

Provision of study material or patients: Benjamin Marchello, Alex A. Adjei, Julian R. Molina

Collection and/or assembly of data: Kaushal Parikh, Brandt Esplin, Benjamin Marchello, Alex A. Adjei, Julian R. Molina

Data analysis and interpretation: Sumithra J. Mandrekar, Katie Allen‐Ziegler, Angelina D. Tan

Manuscript writing: Kaushal Parikh, Sumithra J. Mandrekar, Julian R. Molina

Final approval of manuscript: Kaushal Parikh, Sumithra J. Mandrekar, Katie Allen‐Ziegler, Brandt Esplin, Angelina D. Tan, Benjamin Marchello, Alex A. Adjei, Julian R. Molina

Disclosures

The authors indicated no financial relationships.