Abstract
Two phase 3 trials, COMET-1 and COMET-2, have reported that cabozantinib did not significantly extend overall survival (OS) compared to prednisone and prednisone plus mitoxantrone, respectively, in post-docetaxel patients with metastatic castration-resistant prostate cancer (mCRPC). We conducted a retrospective analysis of a combined data set from these trials to identify a benefit in subsets of patients according to prognostic risk factors. The prognostic ability of factors to predict survival was evaluated using Cox proportional hazards regression models. Evaluation of potential beneficial subsets was performed using interaction terms between factors and cabozantinib. All tests were two-sided and p ≤ 0.05 was considered statistically significant. A total of 1147 post-docetaxel patients with mCRPC were available (1028 from COMET-1 and 119 from COMET-2). The following factors were prognostic for OS: age, disease-free interval, hemoglobin, prostate-specific antigen, alkaline phosphatase, albumin, bone scan lesion area, lactate dehydrogenase, Eastern Cooperative Oncology Group performance status, and pain (all p < 0.05). There was no interaction effect on survival between cabozantinib versus comparator arms and any prognostic group. After adjusting for prognostic factors, cabozantinib was associated with better OS (hazard ratio 0.80, 95% confidence interval 0.67–0.95; p = 0.012). Further investigation of cabozantinib in a better-powered trial or a rational patient population based on a molecular biomarker may be warranted.
Keywords: Cabozantinib, Metastatic, Post-docetaxel, Castration-resistant, Prostate cancer, Survival, Prognosis
Patient summary:
Two phase 3 trials have reported no survival benefit for cabozantinib, a multitarget oral drug, in metastatic castration-resistant prostate cancer. This analysis pooled 1147 patients from these trials to identify a survival benefit for cabozantinib. This study suggests that further rational development may be justified.
Cabozantinib is an orally administered tyrosine kinase inhibitor of MET and VEGF receptors that has been approved by the US Food and Drug Administration for metastatic renal cell carcinoma and medullary thyroid cancer. Two phase 3 trials, COMET-1 and COMET-2, have reported that cabozantinib did not extend overall survival (OS) among post-docetaxel patients with metastatic castration-resistant prostate cancer (mCRPC) compared to prednisone and prednisone plus mitoxantrone, respectively [1,2]. However, there were multiple signals of activity and benefit when examining secondary endpoints. In the COMET-1 trial (n = 1028), cabozantinib significantly improved radiographic progression-free survival (median 5.6 vs 2.8 mo; p < 0.001) but not OS (median 11.0 vs 9.8 mo; hazard ratio [HR] 0.91, 95% confidence interval [CI] 0.77–1.07; p = 0.213), which was the primary endpoint.
Cabozantinib was also associated with improvements in circulating tumor cell conversion, bone biomarkers, and the incidence of symptomatic skeletal events.
After 119 patients were randomized, the COMET-2 trial was discontinued after COMET-1 reported a lack of survival extension, and although the primary endpoint of pain response was not significantly different (15% vs 17%; p = 0.773), there was a trend for better OS (9 vs 7.9 mo; HR 0.71, 95% CI 0.45–1.12; p = 0.121) and a significantly better bone scan response with cabozantinib (31% vs 5.2%; p < 0.001). Previous phase 2 trials in mCRPC also demonstrated robust bone-targeted activity in terms of bone scan improvements [3,4].
Notably, analysis of the phase 3 METEOR trial comparing cabozantinib and everolimus in metastatic renal cell carcinoma demonstrated greater relative benefit among those with bone disease [5]. We conducted a retrospective analysis of a combined data set from the COMET-1 and COMET-2 phase 3 trials with the hypothesis that a benefit would be observed in subsets of high-risk groups and those with a higher burden of bone metastasis.
Deidentified patient-level data were obtained from the COMET-1 and COMET-2 trials. Cox proportional hazards regression models were constructed to explore the OS prognostic ability of clinical and laboratory factors [6–8]. The trial (COMET-1 or COMET-2) was included as a stratification factor for all models. Subsets were identified by testing for an interaction term between treatment and selected factors. Patients were grouped into risk quartiles according to the previously described post-docetaxel nomogram [6–8]. Analyses were performed in SAS v.9.0 (SAS Institute, Cary, NC, USA) or R v.3.2.2 (R Foundation for Statistical Computing, Vienne, Austria).
Patient characteristics have been described previously and are summarized in Supplementary Table 1 [1,2]. The following factors were significantly predictive for OS on univariable analyses: age, disease-free interval (post-docetaxel and post-androgen deprivation therapy initiation), hemoglobin, prostate-specific antigen, albumin, bone scan lesion area (BSLA), lactate dehydrogenase, Eastern Cooperative Oncology Group performance status, and pain (all p < 0.05; Table 1). No interaction was statistically significant, indicating no differential impact of cabozantinib on OS for any prognostic factor evaluated or risk group based on quartiles (Fig. 1). After adjusting for all potential prognostic factors in a multivariable model (Table 1), treatment with cabozantinib versus the comparator was associated with better OS (HR 0.80, 95% CI 0.67–0.95; p = 0.012).
Table 1 –
Univariable and multivariable Cox regression analyses for association of factors with overall survival
| Parameter | HR (95% CI) | p value | Interaction p value |
|---|---|---|---|
| Univariable | |||
| Age per decade | 1.12 (1.02–1.23) | 0.019 | 0.59 |
| DFI post-docetaxel per year | 0.88 (0.82–0.94) | <0.001 | 0.20 |
| DFI post-ADT per year | 0.95 (0.93–0.97) | <0.001 | 0.20 |
| Hemoglobin per 10 U | 0.73 (0.69–0.77) | <0.001 | 0.72 |
| PSA log-transform | 1.25 (1.20–1.31) | <0.001 | 0.14 |
| BSLA log-transform | 1.31 (1.23 –1.40) | <0.001 | 0.23 |
| LDH log-transform | 2.54 (2.26–2.86) | <0.001 | 0.61 |
| LDH ≥ ULN (280 U/l) | 3.10 (2.67–3.59) | <0.001 | 0.47 |
| ECOG performance status | |||
| 0 | 0.28 (0.22–0.35) | <0.001 | 0.59 |
| 1 | 0.52 (0.43–0.64) | ||
| ≥2 | Reference | ||
| Lymph node metastases | 1.27 (1.10–1.46) | 0.001 | 0.79 |
| Liver metastases | 3.03 (2.50–3.68) | <0.001 | 0.20 |
| Lung metastases | 1.34 (1.07–1.69) | 0.012 | 0.95 |
| BPI score ≥4 | 1.79 (1.54–2.09) | <0.001 | 0.93 |
| Prior abiraterone | 1.09 (0.83–1.43) | 0.54 | 0.48 |
| Prior enzalutamide | 1.17 (1.00–1.38) | 0.055 | 0.43 |
| Site of disease Visceral | 2.32 (1.93–2.79) | <0.001 | 0.30 |
| Lymph node ± bone | 1.24 (1.05–1.46) | ||
| Bone only | Reference | ||
| Cabozantinib treatment | 0.87 (0.75–1.01) | 0.062 | |
| Multivariable | |||
| Age per decade | 1.39 (1.23–1.56) | <0.001 | |
| DFI post-docetaxel per year | 0.84 (0.77–0.91) | <0.001 | |
| DFI post-ADT per year | 0.99 (0.97–1.02) | 0.53 | |
| Hemoglobin per 10 U | 0.90 (0.94–0.96) | 0.002 | |
| PSA log-transform | 1.10 (1.03–1.17) | 0.002 | |
| LDH log-transform | 1.63 (1.38–1.91) | <0.001 | |
| ECOG performance status | |||
| 0 | 0.50 (0.37–0.68) | ||
| 1 | 0.68 (0.54–0.86) | ||
| ≥2 | Reference | ||
| Lymph node metastases | 1.18 (0.99–1.40) | 0.069 | |
| Liver metastases | 1.89 (1.48–2.42) | <0.001 | |
| Lung metastases | 1.03 (0.77–1.37) | 0.84 | |
| BPI score ≥4 | 1.32 (1.10–1.58) | 0.003 | |
| Prior abiraterone | 0.67 (0.48–0.94) | 0.019 | |
| Prior enzalutamide | 1.09 (0.89–1.34) | 0.41 | |
| Prior cabazitaxel | 1.66 (1.37–2.00) | <0.001 | |
| BSLA log-transform | 1.14 (1.05–1.24) | 0.003 | |
| Cabozantinib treatment | 0.80 (0.67–0.95) | 0.012 |
HR = hazard ratio; CI = confidence interval; DFI = disease-free interval; ADT = androgen deprivation therapy; PSA = prostate-specific antigen; BSLA= bone scan lesion area; LDH = lactate dehydrogenase; ULN = upper limit of normal; ECOG = Eastern Cooperative Oncology Group; BPI = Brief Pain Inventory.
Fig. 1 –
Survival according to risk group quartile. There was no significant difference in survival between the cabozantinib and control arms stratified by prognostic risk group quartile.
These data suggest that cabozantinib may confer benefits in a rationally selected mCRPC patient population. Our hypothesis that those with higher prognostic risk and greater bone tumor burden (measured as BSLA in the COMET-1 and −2 trials) or higher risk groups might experience greater benefits was not demonstrated by the analysis. Surprisingly, a modest survival benefit was observed in the overall population. These data suggest that an undefined subgroup of patients, potentially selected on the basis of molecular factors, may derive clinically relevant benefits from cabozantinib. In this context, studies in other malignancies suggest that MET alterations or expression levels may be a potential predictive biomarker for benefit [9,10]. Given the trend for better OS in COMET-2, which enrolled only symptomatic patients, a larger trial targeting this subgroup could also be considered. Indeed, given the elderly mCRPC population and the potential toxicities of cabozantinib, better patient selection might improve the therapeutic index.
Systemic therapy for mCRPC has witnessed several advances with the emergence of second-generation androgen inhibitors (enzalutamide, apalutamide, abiraterone acetate) and a potential role for PARP inhibitors and T-cell checkpoint inhibitors in appropriately selected patients. However, most of these agents are expected to provide incremental advances and are not expected to be curative. Evaluation of cabozantinib in a better-powered trial or optimal patient population guided by the discovery of a potential predictive molecular biomarker of activity (eg, MET expression) could resurrect a role for this drug in patients with mCRPC and should be considered.
Supplementary Material
Acknowledgments
Financial disclosures: Guru P. Sonpavde certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: Guru P. Sonpavde is a consultant for Bayer, Sanofi, Pfizer, Novartis, Exelixis, Eisai, Janssen, Amgen, AstraZeneca, Merck, Genentech, EMD Serono, and Astellas; has received institutional research support from Sanofi, Bayer, Boehringer-Ingelheim, Merck, AstraZeneca, BMS, and Pfizer; serves on steering committees for trials sponsored by AstraZeneca, BMS, Bavarian-Nordic, and Astellas; has been an author for UpToDate; and has received speaker fees from Clinical Care Options, Physicians Education Resource (PER), Research to Practice (RTP), and OncLive. Gregory R. Pond has nothing to disclose. Karim Fizazi has acted in a consulting or advisory role for Janssen Oncology, Astellas Pharma, Sanofi, Orion Pharma GmbH, Curevac, AstraZeneca, ESSA, Roche/Genentech, Clovis Oncology, and Amgen; has received travel or accommodation expenses from Amgen and Janssen; and has received honoraria from Janssen, Sanofi, Astellas Pharma, and Merck. Johann S. de Bono has a commercial interest in abiraterone, AKT and PI3K inhibitors, and PARP inhibitors in BRCA cancers; and has served as an advisor for AstraZeneca, Sanofi-Aventis, Astellas, Janssen, Pfizer, Genentech, GSK, MSD, Merck Serono, and Bayer. Ethan M. Basch is employed by the University of North Carolina; has received research funding from the National Cancer Institute and the Patient-Centered Outcomes Research Institute; is on the editorial board for the Journal of the American Medical Association; has been a consultant on research projects for Memorial Sloan Kettering Cancer Center, Dana-Farber Cancer Institute, Centers for Medicare & Medicaid Services, and the Research Triangle Institute; and is a scientific advisor for Noona Healthcare and Sivan Healthcare. Howard I. Scher is a consultant for BIND Therapeutics Inc., Astellas, Clovis Oncology, Ferring Pharmaceuticals, Janssen R&D, LLC, Merck, Sanofi Aventis, and WCG Oncology; is a member of the board of directors for Asterias Biotherapeutics; and is participating in an unpaid industry collaboration with EPIC Sciences, Inc. Matthew R. Smith is a consultant for Amgen, Bayer, Exelixis, Janssen, and Lilly.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Smith M, de Bono J, Sternberg C, et al. Phase III study of cabozantinib in previously treated metastatic castration-resistant prostate cancer: COMET-1. J Clin Oncol 2016;34:3005–13. [DOI] [PubMed] [Google Scholar]
- 2.Basch EM, Scholz MC, de Bono JS, et al. Final analysis of COMET-2: cabozantinib versus mitoxantrone/prednisone (MP) in metastatic castration-resistant prostate cancer patients with moderate to severe pain who were previously treated with docetaxel and abiraterone and/or enzalutamide. J Clin Oncol 2015;33(Suppl 7):141.25185099 [Google Scholar]
- 3.Smith MR, Sweeney CJ, Corn PG, et al. Cabozantinib in chemotherapy-pretreated metastatic castration-resistant prostate cancer: results of a phase II nonrandomized expansion study. J Clin Oncol 2014;32:3391–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Smith DC, Smith MR, Sweeney C, et al. Cabozantinib in patients with advanced prostate cancer: results of a phase II randomized discontinuation trial. J Clin Oncol 2013;31:412–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Escudier B, Powles T, Motzer RJ, et al. Cabozantinib, a new standard of care for patients with advanced renal cell carcinoma and bone metastases? Subgroup analysis of the METEOR trial. J Clin Oncol 2018;36:765–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Halabi S, Lin CY, Small EJ, et al. Prognostic model predicting metastatic castration-resistant prostate cancer survival in men treated with second-line chemotherapy. J Natl Cancer Inst 2013;105:1729–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Halabi S, Lin CY, Kelly WK, et al. Updated prognostic model for predicting overall survival in first-line chemotherapy for patients with metastatic castration-resistant prostate cancer. J Clin Oncol 2014;32:671–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chi KN, Kheoh T, Ryan CJ, et al. A prognostic index model for predicting overall survival in patients with metastatic castration-resistant prostate cancer treated with abiraterone acetate after docetaxel. Ann Oncol 2016;27:454–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Goyal L, Zheng H, Yurgelun MB, et al. A phase 2 and biomarker study of cabozantinib in patients with advanced cholangiocarcinoma. Cancer 2017;123:1979–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Klempner SJ, Borghei A, Hakimian B, Ali SM, Ou SI. Intracranial activity of cabozantinib in MET exon 14-positive NSCLC with brain metastases. J Thorac Oncol 2017;12:152–6. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.