Neoadjuvant Cabazitaxel plus Abiraterone/Leuprolide Acetate in Patients with High-Risk Prostate Cancer: ACDC-RP Phase II Trial

Definitive local therapy provides excellent oncologic outcomes for most patients with clinically localized prostate cancer; however, patients with high-risk, localized disease treated with radical prostatectomy experience high treatment failure rates approaching 65% (1–4). Fifteen-year mortality rates remain in excess of 30% despite treatment intensification with adjuvant/salvage radiation and/or chemohormonal therapy (2). These poor outcomes strongly suggest the presence of micrometastatic disease at the time of initial diagnosis (5, 6). Early treatment intensification with neoadjuvant therapy has thus been proposed to improve oncologic outcomes of patients with high-risk prostate cancer.

Although previous neoadjuvant trials/meta-analyses in this setting have shown improved pathologic outcomes in the form of pathologic complete response (pCR), tumor downstaging, and decreased positive surgical margins, such approaches have focused on short-term pathologic outcomes and have not demonstrated clinically significant survival benefits (7–13). Over the past decade, novel agents targeting the androgen receptor (AR) axis have emerged (14, 15). Although standard hormonal therapy using luteinizing hormone–releasing hormone (LHRH) agonists/antagonists suppresses serum testosterone levels by 90%, tumor tissue androgen levels remain at 25% to 35% of the original levels (14, 15). Continued intratumoral androgen synthesis, along with AR amplification, contributes to castration resistance (14–17). Abiraterone acetate, a cytochrome P450c17 (CYP17) inhibitor, suppresses additional sources of androgen production, including intratumor and adrenal sources, and is currently approved for the treatment of metastatic prostate cancer in both the castrate sensitive and resistant settings (18). The taxanes, docetaxel, and cabazitaxel, inhibit microtubule polymerization and may interfere with AR nuclear translocation, thus potentially having synergistic mechanisms of action with androgen synthesis inhibitors (19, 20). We hypothesized that the combination of abiraterone and a taxane might overcome early resistance mechanisms characteristic of androgen deprivation therapy (ADT) exposure, with this treatment strategy showing survival benefits in the first-line metastatic hormone-sensitive setting (PEACE-1; ref. 21). In this study, we compared the pathologic, early oncologic, and safety outcomes of neoadjuvant leuprolide acetate and abiraterone acetate with or without cabazitaxel in patients with localized, high-risk prostate cancer prior to planned radical prostatectomy.

This was a randomized, open-label, multicenter, parallel group phase II trial with a planned safety run-in (trial registration ID: NCT02543255). All procedures were approved by the University Health Network (Toronto, Ontario, Canada) and London Health Sciences Centre (London, Ontario, Canada) research ethics boards in accordance with the Declaration of Helsinki principles and the International Council for Harmonization Good Clinical Practice guidelines. All patients provided written informed consent. This was an investigator-initiated trial with funding support by the Ontario Institute for Cancer Research. Study drugs were provided as follows: abiraterone acetate (by Janssen, Inc.), cabazitaxel and leuprolide (Sanofi-Aventis Canada, Inc.), and pegfilgrastim (Amgen Canada, Inc.). A data and safety monitoring board reviewed unblinded safety and efficacy data throughout the trial. The data were collected, analyzed, and interpreted by the investigator-led team. The trial was conducted in accordance with CONSORT guidelines. Study rigor was assessed using SciScore. No research resource identifier was available for this study.

This study included men ≥18 years of age with histologically or cytologically confirmed D'Amico high-risk prostatic adenocarcinoma without neuroendocrine differentiation or small cell features. D'Amico high-risk disease was defined as: prostate-specific antigen (PSA) >20 ng/mL, Gleason score (GS) ≥8 with minimum two cores positive for cancer, or GS 4+3 = 7 with minimum three cores positive and PSA >10 ng/mL. All patients had Eastern Cooperative Oncology Group (ECOG) performance status 0–1, life expectancy >6 months, and had surgically resectable disease per the treating physician. Included patients had no evidence of nodal/metastatic disease as determined by conventional imaging (performed at the participating center): radionuclide bone scans and CT/ MRI. Nonpathologic lymph nodes were less than 15 mm in the short/transverse axis. Clinical stage was evaluated using a combination of clinical examination/radiologic evaluation. Prostate-specific membrane antigen (PSMA) PET/CT became available midway through this study and was permitted at the treating physician's discretion. PSMA-PET/CT findings did not affect patient study eligibility.

Exclusion criteria included prior chemotherapy, surgery, or radiotherapy for prostate cancer. Patients were permitted 3 months of ADT (LHRH agonists/antagonists, first generation antiandrogens, or both) prior to study enrollment, but not administered within 6 months prior to randomization. Further exclusion criteria included significant abnormalities on screening laboratory exams or active, uncontrolled medical comorbidities (Supplementary Materials and Methods).

Consenting, eligible participants were randomized 1:1 to neoadjuvant treatment with abiraterone acetate (1,000 mg/day) plus prednisone (5 mg twice daily), leuprolide acetate (22.5 mg subcutaneously every 12 weeks), and cabazitaxel (25 mg/m² body surface area; arm A) or abiraterone acetate plus prednisone and leuprolide acetate only (arm B). Randomization was centrally performed using permuted blocks. The planned treatment duration was 24 weeks in each arm, followed by a radical prostatectomy with standard template pelvic lymph node dissection to the level of the aortic bifurcation. Metastasis-directed therapy via a targeted lymph node dissection based on findings of preoperative PSMA PET/CT (if performed) was permitted. Abiraterone acetate was started on day 1 and stopped one day prior to prostatectomy. Patients randomized to arm A were planned to receive a maximum of six cycles of cabazitaxel, administered at 25 mg/m² per body surface area on day 1 of each cycle, every 21 days. Cabazitaxel was started 14 days after the initiation of abiraterone with the last cycle planned by week 17 to allow for adequate recovery time prior to prostatectomy. Pegfilgrastim (6 mg subcutaneously daily) was administered 24 hours post-cabazitaxel. Assigned treatments were continued until participants could no longer tolerate treatment due to an adverse event (AE), another systemic antineoplastic drug or investigational drug was initiated, or at the investigator's discretion. Following surgery, patients underwent PSA/testosterone checks every 3 months for the first 2 years, followed by every 6 months thereafter, or as clinically indicated per the treating clinician.

The primary endpoint was presence of a pCR or minimal residual disease (MRD) in the radical prostatectomy specimen. pCR was defined as the absence of morphologically identifiable carcinoma in the prostatectomy specimen. MRD was defined as residual cancer burden ≤0.25 cm (ref. 3; total tumor volume ≤0.5 cm³ × tumor cellularity ≤50%). Pathologists were blinded to treatment arm. Specimens were evaluated by the respective institutional genitourinary pathologist(s), and there was no central pathology review. All radical prostatectomy specimens were processed within 30 minutes of excision. The specimens were painted as per institutional protocol, and the apex and bladder neck were removed, sectioned, and blocked for paraffin embedding. The prostate was sliced in serial transverse sections (5 mm in thickness) from apex to base, paraffin processed (formalin fixation), and evaluated accordingly (Supplementary Material section).

Secondary outcomes included other pathologic outcomes at time of prostatectomy: positive surgical margins, tumor size, extraprostatic extension, seminal vesicle invasion, and nodal involvement. PSA failure data were limited to men with recovered serum testosterone levels (defined as: >1.7 nmol/L), to minimize potential immortal time bias during castrate states, and were defined as two consecutive PSA levels >0.2 ng/mL or use of adjuvant/salvage radiation or ADT from date of radical prostatectomy. Post hoc analysis evaluating demographic and oncologic predictors of pCR/MRD was performed. Oncologic variables included baseline serum PSA level, biopsy grade, percent positive biopsy cores, maximum percent core involvement, and number of chemotherapy cycles received.

A safety run-in after administration of two cycles of cabazitaxel was preplanned to evaluate the first six participants assigned to arm A for toxicity prior to further enrollment. Full details on the safety run-in procedures are included in the appended protocol.

Safety and tolerability were determined by frequent assessment of AEs, physical examinations, vital signs, and safety laboratory assessments as detailed in the study schedule of events (Supplementary Materials and Methods). A safety follow-up visit was conducted approximately 30 days after the last dose of abiraterone acetate.

The sample size was calculated on the basis of the primary endpoint (pCR/MRD). A pCR/MRD of 18% following 6 months of neoadjuvant leuprolide/abiraterone acetate was anticipated (10). This study was designed so that a two-sided z test had 70% power to detect a 25% difference between the two groups, assuming a type I error rate of 5%. The calculated sample size was 76 patients with 38 participants in each arm.

Statistical analysis was performed by treatment arm using all patients with evaluable data. For safety analyses, all patients who received at least one dose of safety drug were evaluated. χ² and Fisher exact testing were used for comparison of categorical variables, including the primary outcome of pCR/MRD. The Mann–Whitney U test was used for nonparametric comparisons of continuous measures. Survival analysis was performed using Kaplan–Meier curves with the log-rank test used for group comparisons. Predictors of pCR/MRD were evaluated using univariable logistic regression. All study participants who were randomized and underwent subsequent radical prostatectomy were evaluated using the intent to treat analysis. Per protocol analysis excluded noncompliant patients (defined as missing 14 or more days of oral dosing in a 28-day period), those who received less than 90 days of abiraterone or less than three cycles of cabazitaxel. As this was a surgical endpoint-based trial, those who did not undergo radical prostatectomy were excluded. Hypotheses testing was two-sided with P ≤ 0.05 denoting statistical significance. All statistical analysis was performed using R version 3.6.1.

We have full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The data generated in this study are only available upon request from the corresponding author, due to the protected nature of patient health information.

Between December 2016 and April 2021, 78 patients were enrolled. Forty patients were randomly allocated to arm A (cabazitaxel) and 38 to arm B (no cabazitaxel). One participant in arm A opted against chemotherapy (analyzed in the chemotherapy arm as per the intent-to-treat principle). During follow-up, 3 patients in arm A and 5 patients in arm B discontinued intervention prior to radical prostatectomy. The final cohort included 70 patients, with 37 in arm A and 33 in arm B (CONSORT diagram in Fig. 1). Safety run-in for the first 6 patients did not demonstrate any dose limiting toxicities and the cabazitaxel dose of 25 mg/m² was continued.

Baseline patient characteristics were well balanced between the two arms (Table 1; Supplementary Table S1). Median patient age was 63.5 years [interquartile range (IQR), 58.0–68.0] and median PSA at time of randomization was 21.9 ng/mL (IQR, 14.6–42.8). GS 8 disease or worse at biopsy was present in 25 (67.7%) and 17 (51.5%) patients in arms A and B, respectively (P = 0.42). Ten patients (12.8%) received 3 months ADT prior to study entry (5 per arm). Of the 37 patients in arm A, 30 (81.1%) completed six cycles of chemotherapy. Seven patients (9.0%) underwent preoperative PSMA PET/CT (arm A: 4; arm B: 3). Of these 7 patients, locoregional involvement of the pelvic lymph nodes was noted in 4 patients (57.1%). None of the patients had evidence of metastatic disease (i.e., all M0). None of the patients underwent an extended pelvic lymph node dissection beyond the level of the common iliac vessels or metastasis-directed therapy in the form of targeted nodal excision or stereotactic body radiotherapy given that the positive nodes were within the standard nodal dissection template region.

A pCR or MRD was observed in 16 patients (43.2%) in arm A versus 15 patients (45.5%) in arm B (P = 0.85). pCR was observed in two patients (5.4%) in arm A versus three patients (9.1%) in arm B (P = 0.66). Median tumor sizes in arms A and B were 1.34 cm³ (IQR, 0.42–8.9 cm³) and 1.77 cm³ (0.35–5.4 cm³), respectively (P = 0.49). Surgical margins were positive in 5 patients (13.5%) in arm A versus 10 (30.3%) in arm B (P = 0.16). ypT3 disease or worse was present in 22 patients (59.5%) in arm A compared with 19 (57.6%) in arm B (P = 0.95). Positive lymph nodes were detected in 12 patients (32.4%) in arm A versus 6 patients (18.2%) in arm B (P = 0.28). Seven (18.9%) and 6 (18.2%) patients in arms A and B, respectively, received radiotherapy in the adjuvant or salvage settings (P = 0.74, Table 2).

On univariable analysis, the percentage of positive biopsy cores predicted pCR/MRD. Compared with patients with ≤25% positive biopsy cores, patients with 25.1% to 50% (OR, 0.45; P = 0.21), 50.1% to 75.0% (OR, 0.24; P = 0.061), and 75.1% to 100% positive biopsy cores (OR, 0.18; P = 0.035) had progressively decreasing odds of pCR/MRD. Ad hoc logistic regression analysis adjusting for baseline differences in median biopsy percent total (50% in arm A vs. 35.0% in arm B) demonstrated no significant differences in odds of pCR/MRD by treatment arm [OR, 0.91; 95% confidence interval (CI), 0.32–2.57; P = 0.85].

Patient age, baseline PSA, biopsy GS, maximum percent core involvement, and number of chemotherapy cycles did not significantly predict pCR/MRD (Supplementary Table S2).

Testosterone recovery was noted in 37 of 40 (92.5%) patients in arm A versus 33/38 (86.8%) patients in arm B (P = 0.48). Median serum testosterone levels at last follow-up were 7.60 nmol/L (IQR, 3.1–12.1) and 5.50 nmol/L (IQR, 0.3–12.6), respectively (P = 0.45).

Median follow-up was 13.1 (range = 1.4–48.8) and 11.2 months (range = 1.6–49.6) in arms A and B, respectively. The 12-month PSA relapse-free survival rate was 78.0% (95% CI, 64.0–94.0%) in arm A versus 76.0% (95% CI, 62.0–93.0%) in arm B. The corresponding 24-month rates were 48.0% (95% CI, 29.0%–80.0%) and 50.0% (95% CI, 31.0%–82.0%) in arms A and B, respectively (log-rank test P = 0.75; Fig. 2).

The 12-month PSA relapse-free survival rate was superior in those with pCR/MRD: 96.0% (95% CI, 89.0%–100.0%) versus those without pCR/MRD: 62.0% (95% CI, 47.0%–81.0%, P = 0.03). At 24 months, however, PSA relapse-free survival rates were non-significantly different (P = 0.096; Fig. 3).

Two of the first 6 patients experienced thromboembolic events postoperatively. The protocol was amended to include 28 days of low molecular weight heparin for both arms, in addition to standard preoperative subcutaneous heparin prophylaxis. No further thromboembolic events occurred.

Grade 3 or worse AEs were observed in 17 of 40 (42.5%) patients in arm A and 9 of 38 (23.7%) patients in arm B (P = 0.078). Notably, 4 of 38 patients in arm B discontinued treatment due to hepatotoxicity (none discontinued due to hepatoxicity in arm A). Febrile neutropenia occurred in 3 of 40 (7.5%) patients in arm A (Table 3).

In this randomized, open-label, multicenter, phase II trial, addition of cabazitaxel to combination abiraterone and leuprolide acetate in the neoadjuvant setting prior to radical prostatectomy was not associated with significant improvement in pCR/MRD (43.2% vs. 45.5%) or pCR (5.4% vs. 9.1%). Furthermore, no significant difference was seen in biochemical recurrence rates post-radical prostatectomy. Patients with pCR/MRD, however, had improved early PSA relapse-free survival rates (96.0% vs. 62.0% at 12 months, P = 0.04), with longer-term comparisons underpowered secondary to the limited follow-up period and low number of patients at risk. Treatment was reasonably well tolerated in both arms, with grade 3 or worse AEs observed in in 42.5% and 23.7% of patients in arms A and B, respectively (P = 0.078).

In the phase III PUNCH trial of neoadjuvant chemohormonal therapy, 788 patients with clinically localized, high-risk prostate cancer were randomized to radical prostatectomy alone or ADT plus docetaxel prior to radical prostatectomy. This trial failed to meet its primary endpoint of 3-year biochemical progression-free survival (BPFS; 0.89 vs. 0.84, P = 0.11); however, overall BPFS [HR, 0.69; 95% CI, 0.48–0.99) and metastasis-free survival (HR, 0.70; 95% CI, 0.51–0.95) were improved with neoadjuvant chemohormonal therapy. Although both the PUNCH trial and our trial studied chemohormonal therapy in the neoadjuvant setting, there are important differences between the studies. The control arm in the PUNCH trial consisted of radical prostatectomy alone, whereas the current trial used an “active” control with hormonal therapy. This may have attenuated any observed pathologic/outcome benefits seen with combined chemohormonal therapy in our trial. Furthermore, hormonal therapy in the PUNCH trial consisted of LHRH agonist with/without an antiandrogen alone, as compared with combined LHRH agonist and abiraterone acetate/prednisone in this trial. This was designed to maximize suppression of intratumoral androgen production, an important cause of hormone resistance. This “maximal androgen blockade” may explain why pCR was seen in 5 of 70 (7.1%) patients in this trial versus none in the Alliance trial (22).

The importance of maximal androgen blockade with an AR-axis targeted therapy in the neoadjuvant setting is reinforced by results of the ARNEO trial. In this trial, apalutamide addition to degarelix was associated with significant improvements in radical prostatectomy MRD rates (38% vs. 9.1%, P = 0.002). Notably, patients in the intervention arm of the ARNEO trial had lower MRD responses (38%) as compared with patients in both the intervention (43.2%) and control arms (45.5%) of our trial. Furthermore, no pCR was seen in either arm of the ARNEO trial (23). These differences are likely related to differences in patient selection and use of 3 months of neoadjuvant therapy in the ARNEO trial versus 6 months in this trial. The utility of AR signaling inhibitors use in the neoadjuvant setting was initially highlighted by Taplin and colleagues, who demonstrated that 24 weeks of neoadjuvant abiraterone/LHRH agonist was associated with MRD in 52% of prostatectomy specimens (10). A pooled analysis by McKay and colleagues in 2021 further demonstrated that neoadjuvant AR signaling inhibitor use was associated with a pCR rate of 9.4% with a 3-year biochemical recurrence-free rate of almost 60% (12).

The results of this trial have important implications for future treatment/study planning. Neoadjuvant chemotherapy is standard of care for many cancers, including subsets of bladder, colorectal, and breast cancer. It is conceivable that the future of high-risk, localized prostate cancer follows such a paradigm, with neoadjuvant utilization of second-generation androgen axis–targeting agents. As such, data from our trial will serve as a baseline measure of pCR/MRD by which additional agents can be compared. Future research should explore biologic determinants of treatment response and resistance. Translational analysis within this trial evaluating genomic predictors of pathologic response is currently underway, as are novel neoadjuvant approaches utilizing genomic risk stratification [Genomic Biomarker-Selected Umbrella Neoadjuvant Study for High Risk Localized Prostate Cancer (GUNS) trial, NCT04812366].

Important limitations to this trial include limited clinical follow-up, lack of metastatic/survival outcomes data, and use of pathologic, as opposed to survival, outcomes as the primary endpoint. PSA/testosterone follow-up data were available for only 70/78 patients. Although study treatment allocation was randomized, imbalances in baseline patient characteristics were present, likely due to the relatively small sample size of this trial (e.g., 57% of patients in arm A with GS 4+5 disease or worse vs. 33% in arm B; Percent Gleason 4 or 5 disease on biopsy: 87.5% in arm A vs. 70.0% in arm B). These imbalances may have theoretically contributed to the overall negative trial results. The observed pCR/MRD in both the control and experimental arms was significantly higher than that anticipated during the sample size calculations, with low, erroneous initial estimates during sample size calculations given the study inclusion criteria and baseline patient characteristics. Between study differences in the rates of pCR/MRD may also be secondary to differences in defining/evaluating pCR/MRD (10, 12, 23). This study also had limited power (70%) for the detection of between group differences. It is also important to note that PSMA PET/CT-based stratification for all patients would have demonstrated evidence of radiographic distant disease spread in a subset of trial patients and perhaps should reignite investigation into utilizing neoadjuvant regimens among patients without evidence of metastasis on PSMA PET/CT-based imaging. PSMA PET/CT in this trial was performed at the investigator's discretion and this may have introduced an element of a detection/measurement bias in the 7 patients who underwent this imaging modality. We note that patients with diffusely metastatic disease were discouraged from surgical trial enrollment. The lack of central pathology review for the primary outcome of pCR/MRD likely introduced inter-observer variability between the genitourinary pathologists. Toxicity profile, including risk of venous thromboembolic events, was increased in the chemohormonal therapy arm. Correlative, translational analyses to evaluate potential subsets of patients that may benefit from neoadjuvant treatment intensification with cabazitaxel addition to abiraterone/LHRH agonists are currently underway and not available at the time of reporting.

Abiraterone addition to neoadjuvant regimens appears to improve pCR compared with prior LHRH-based neoadjuvant trials. Neoadjuvant cabazitaxel addition to abiraterone acetate plus leuprolide acetate prior to radical prostatectomy did not improve pCR/MRD in clinically localized patients with high-risk prostate cancer.

N. Fleshner reports nonfinancial support from Amgen and Sanofi; grants and personal fees from Janssen, Astellas, Bayer, and Sanofi; grants from Nucleix, Progenix, and OICR during the conduct of the study; and personal fees from Amgen, Abbvie, Ferring, Verity Pharmaceuticals, and POINT Biopharma outside the submitted work. R.K. Sayyid reports grants from Astellas Canada outside the submitted work. A.R. Hansen reports other support from Bristol Myers Squibb (BMS), Pfizer, AstraZeneca, MSD, GSK, Janssen, Astellas, Genentech, Novartis, and Roche outside the submitted work. R. Fernandes reports other support from Bayer, Janssen, Pfizer, Ipsen, Merck, and EMD Serono outside the submitted work. E. Winquist reports personal fees from Advanced Accelerator Applications, Astellas, and Bayer outside the submitted work; in addition, E. Winquist is a clinical expert (ad hoc) for the Canadian Agency for Drug Technologies and Health. S.S. Sridhar reports personal fees from Janssen during the conduct of the study. A. Finelli reports grants from OICR, Sanofi, Amgen, and Janssen during the conduct of the study. G. Kulkarni reports grants and personal fees from Janssen as well as personal fees from Astellas, BMS, Pfizer, Ferring, EMD Serono, Merck, Photocure, TerSera, Knight Therapeutics, Theralase, Bayer, Astra Zeneca, Novartis, and Verity Pharmaceuticals outside the submitted work. A.M. Joshua reports grants from Ontario Institute for Cancer Research; nonfinancial support from Sanofi and Janssen during the conduct of the study; and other support from Astellas and Janssen outside the submitted work. No disclosures were reported by the other authors.

N.E. Fleshner: Conceptualization, resources, data curation, supervision, funding acquisition, investigation, methodology, writing–original draft, writing–review and editing. R.K. Sayyid: Resources, data curation, software, formal analysis, investigation, methodology, writing–original draft, writing–review and editing. A.R. Hansen: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, visualization, methodology, writing–review and editing. J.L. Chin: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, methodology, writing–review and editing. R. Fernandes: Conceptualization, resources, data curation, supervision, investigation, writing–review and editing. E. Winquist: Conceptualization, resources, data curation, formal analysis, supervision, investigation, writing–review and editing. T. van der Kwast: Conceptualization, resources, data curation, supervision, methodology, writing–review and editing. J. Sweet: Resources, data curation, formal analysis, supervision, investigation, methodology, writing–review and editing. K. Lajkosz: Resources, data curation, software, formal analysis, supervision, methodology, writing–review and editing. M. Kenk: Conceptualization, resources, data curation, supervision, funding acquisition, investigation, methodology, writing–review and editing. K. Hersey: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, methodology, writing–review and editing. R. Veloso: Resources, data curation, supervision, funding acquisition, methodology, project administration, writing–review and editing. D. Berlin: Resources, data curation, supervision, investigation, methodology, project administration, writing–review and editing. J.O. Herrera-Caceres: Resources, data curation, funding acquisition, investigation, methodology, project administration, writing–review and editing. S. Sridhar: Conceptualization, resources, data curation, supervision, investigation, methodology, project administration, writing–review and editing. M. Moussa: Resources, data curation, formal analysis, methodology, writing–review and editing. A. Finelli: Resources, data curation, formal analysis, supervision, investigation, methodology, writing–review and editing. R.J. Hamilton: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, methodology, writing–review and editing. G.S. Kulkarni: Conceptualization, resources, data curation, formal analysis, supervision, investigation, methodology, writing–review and editing. A.R. Zlotta: Conceptualization, resources, data curation, formal analysis, funding acquisition, validation, methodology, writing–review and editing. A.M. Joshua: Conceptualization, resources, data curation, formal analysis, supervision, funding acquisition, investigation, methodology, project administration, writing–review and editing.

This work was supported by: Ontario Institute for Cancer Research, Janssen, Sanofi, and Amgen. We would like to acknowledge the Ontario Institute for Cancer Research, Janssen, Sanofi, and Amgen for their research funding support.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.