Choudhury Discusses Sequencing of Therapy in Metastatic Castration-Resistant Prostate Cancer

Atish D. Choudhury, MD, PhD

Codirector of the Prostate Cancer Center

Dana-Farber Cancer Institute

Assistant Professor of Medicine, Harvard Medical School

Boston, MA

Targeted Oncology^TM: Please discuss the design and results of the CARD trial (NCT02485691).

CHOUDHURY: This was a randomized open-label study, where patients with metastatic castration-resistant prostate cancer [mCRPC] who progressed in less than or equal to 12 months on a prior alternative androgen receptor [AR]–targeted agent—so abiraterone [Zytiga] or enzalutamide [Xtandi]—before or after docetaxel, who were randomized to get cabazitaxel [Jevtana] at 25 mg/m2 or the other agent.¹

The primary end point was imaging-based progression-free survival [PFS], with important secondary end points around clinical factors, [as well as] other secondary end points around quality of life, pain response, time to skeletal-related event, etc.

The baseline characteristics—on the cabazitaxel arm, if anything—had more patients [above the age of] 75. Many patients in both arms of the study had pain, [approximately] 70%, and a good number of patients had metastatic disease at diagnosis. However, this is a trial that enrolled before patients were routinely getting docetaxel and abiraterone in mHSPC [metastatic hormone-sensitive prostate cancer], so it was only [approximately] 11% to 14%.

It was an even split on whether they [received] abiraterone or enzalutamide before or after docetaxel.¹ Imaging-based PFS dramatically favored cabazitaxel over the alternative androgen-signaling targeted inhibitor. It was 8 months for cabazitaxel vs 3.7 months for the other agents, and it was statistically significant [HR, 0.54; 95% CI, 0.40-0.73; P < .001].

What were the results of the subgroup analyses in the CARD trial?

There were many planned subgroups. However, there was no subgroup that they tested where there was not favorable benefit for cabazitaxel compared [with] the other agent. So even if your performance status is poor, you’re elderly, or [you] have visceral metastases, none of those features would suggest you’re better off switching to the other oral agent. Then looking at PFS, cabazitaxel is superior to abiraterone or enzalutamide in a statistically significant way [HR, 0.52; 95% CI, 0.40-0.68; P < .0001].

The overall survival [OS] also favored cabazitaxel over the other agents, even though there was crossover allowed [for] cabazitaxel in patients in the other arm of the study. The median OS was 13.6 months for cabazitaxel and 11 months for the [androgen]-signaling targeted inhibitor [HR, 0.64; 95% CI, 0.46-0.89; P = .0008].

Looking at all the secondary end points, PSA response, objective tumor response, and pain response favored cabazitaxel.¹ There was a preplanned analysis for time to SSE [symptomatic skeletal event] in the cabazitaxel arm, and the median time to SSE was not reached. It was 17 months to SSE for the alternative AR-targeting agent, which was [also] a statistically significant P-value [HR, 0.59; 95% CI, 0.35-1.01; P = .05].

What were the safety results of this trial?

The adverse events [AEs] leading to treatment discontinuation were more frequent in cabazitaxel than the other AR-targeted agent [19.8% with cabazitaxel vs 8.9% for abiraterone or enzalutamide].¹ However, the frequency of any grade 3 AEs was similar and only slightly higher for cabazitaxel than the other agents [56.3% with cabazitaxel vs 52.4% for abiraterone or enzalutamide].

Serious AEs were very similar in frequency between groups, and it was only treatment discontinuation that seemed to be different because AEs leading to death were somewhat better with cabazitaxel [5.6% with cabazitaxel vs 11.3% for abiraterone or enzalutamide].

Looking at probability of deterioration of quality of life, the Functional Assessment of Cancer Therapy- Prostate]² questionnaire [was used], [which surrounds] many features of quality of life that are most relevant for our patients with prostate cancer. There are several questions related to well-being [and] pain. Looking at the well-being and pain-related subscales, the pain-related subscale for well-being favored cabazitaxel in a quite significant way [P < .001].

Even the prostate-specific concerns of well-being favored cabazitaxel slightly, [although] not statistically significant, even though you might think well-being would be worse for patients who are getting chemotherapy compared with an oral drug.

How do these safety data impact your own decision-making?

[Again, for the baseline characteristics], 67% and 71% of these patients had pain.¹ I would say, now that we are being a little more proactive about considering next treatments for our patients in the context of so many options, I’m not sure that [up to] 70% of my patients, who have previously received abiraterone and docetaxel, have [meaningful] pain.

This is an advanced population. If they have visceral metastases, would that lead me to chemotherapy rather than a second AR-targeted therapy? [Approximately] 16% to 20% [of patients had] visceral metastases, and [only 8%] of these patients had PSA-only progression. So, if you have a patient with PSA-only progression, [do these] data apply to them? That part is also hard to say.

This doesn’t mean every single patient who has gotten abiraterone and docetaxel must get cabazitaxel as their very next line of therapy. But people who think patients with symptoms related to cancer and poor performance status related to their symptoms would be better off with an alternative AR-targeted therapy rather than cabazitaxel are probably wrong because of the pain-related quality-of-life data we’re seeing here.

How does the VISION study (NCT03511664) differ from the CARD trial?

These are patients who had mCRPC who received prior treatment with both—[or at least] 1—AR pathway inhibitors and [1 to 3] taxane regimens. To enroll a patient on this study—and we enrolled several patients on this study at Dana-Farber [Cancer Institute]—you had to have declared the protocol-permitted standard of care you’re planning for prior to the randomization. You could not choose chemotherapy, radium-223, or immunotherapy. Basically, [the only treatments allowed were] a hormonal agent, steroid, [or a] bone-protective agent—but not chemotherapy or a radiopharmaceutical. These patients were randomized to receive the protocol-permitted standard of care with lutetium-177 every 6 weeks for 4 cycles, but that could be increased to 6, or the protocol-permitted standard of care alone.³ There were alternate primary end points of radiographic PFS and OS.

To qualify for VISION, you had to have had a Gallium-68 prostate-specific membrane antigen [PSMA] PET/CT, which was complicated because not every site had that capability.³ So you had to send patients somewhere they could get it, and the patients who had the PET/CT [also] had to have PSMA-expressing disease without a lesion that was not PSMA expressing. Eighty-three percent of the patients who had the PSMA PET/CT scan were eligible, based on the PSMA criteria.

They started enrollment in June 2018. Initially, they had trouble enrolling patients onto the standard of care–alone arm. There was a huge amount of early dropout, because investigators were using this trial to try to get the patient lutetium-177 PSMA, and if they didn’t, then they didn’t want to follow through on the trial procedures, which violates the spirit of the trial.

What were the efficacy results of this trial?

The investigators implemented several measures to increase adherence to the trial procedures, then they enrolled the rest of the patients. When you look at the different end points, the radiographic PFS [rPFS] end point per trial was only after the measures were implemented, whereas the OS was analyzed in all randomized patients.³

When you look at the subgroup analyses, it’s important to know that the race-based subgroups enrolled very poorly in terms of the number of Black, African American, or Asian patients. [Therefore, it is difficult] to draw any conclusions about those patients in this study. Approximately 40% to 50% of patients received more than 1 AR inhibitor and/or taxane.

The primary analysis was all randomized patients.3 Again, this is even before the measures were implemented, and there was a 4-month improvement in OS in this heavily pretreated population—from 11 months to 15 months. So the addition prolonged the median OS and the rPFS analysis set. And again, the OS benefit was observed across most prespecified subgroups. In certain subgroups, the difference did not meet statistical significance, but in no subgroups did it seem that standard of care alone benefited patients more than the lutetium-177 PSMA.

The rPFS primary end point, which was after the measures were implemented, was prolonged from 3.4 months to 8.7 months [HR, 0.40; 95% CI, 0.29-0.57; P < .001]. Moreover, the rPFS benefit was observed across most prespecified subgroups. The AEs with standard of care alone vs lutetium-177 PSMA—the number of grade 3 events with lutetium- 177 PSMA was modest.³ What was seen primarily was myelosuppression, and there were some patients who did have renal effects more than in the standard of care alone. [There was] a little more fatigue as well, but overall, the lutetium-177 PSMA was well tolerated. There was certainly a much higher rate of dry mouth—grades 1 and 2—with the lutetium-177 PSMA compared [with] standard of care.

_REFERENCES

_{1. de Wit R, de Bono J, Sternberg CN, et al. Cabazitaxel versus abiraterone or enzalutamide in metastatic prostate cancer. N Engl J Med. 2019;381(26):2506- 2518. doi:10.1056/NEJMoa1911206}

_{2. Fizazi K, Kramer G, Eymard JC, et al. Quality of life in patients with metastatic prostate cancer following treatment with cabazitaxel versus abiraterone or enzalutamide (CARD): an analysis of a randomised, multicentre, open-label, phase 4 study. Lancet Oncol. 2020;21(11):1513-1525. doi:10.1016/S1470-2045(20)30449-6}

_{3. Sartor O, de Bono J, Chi KN, et al. Lutetium-177-PSMA-617 for metastatic castration-resistant prostate cancer. N Engl J Med. 2021;385(12):1091-1103. doi:10.1056/NEJMoa2107322}