PFS Improvement Seen With Cabozantinib in Previously Treated Advanced Neuroendocrine Tumors

Jennifer A. Chan, MD, MPH

Cabozantinib (Cabometyx) led to a progression-free survival improvement in patients with advanced neuroendocrine tumors (NET) who experienced progression following prior therapy vs with placebo, according to data from the phase 3 CABINET trial (NCT03375320) presented during the 2023 ESMO Congress.¹

At a median follow-up of 13.9 months, patients with extra-pancreatic NETs (epNETs) who received the TKI (n = 129) experienced a median PFS of 8.3 months vs 3.2 months with placebo (n = 68; stratified HR, 0.45; 95% CI, 0.30-0.66; P < .0001). Additionally, at a median follow-up of 16.7 months, patients with pancreatic NETs (pNETs) achieved a median PFS of 11.4 months vs 3.0 months in the cabozantinib (n = 62) and placebo (n = 31) arms, respectively (stratified HR, 0.27; 95% CI, 0.14-0.49; P < .0001).

CABINET enrolled patients with well- to moderately-differentiated, grade 1 to 3 NETs. Patients needed to have disease progression (PD) according to RECIST criteria within 12 months prior to random assignment and have PD after or intolerance of at least 1 prior FDA-approved systemic therapy, excluding somatostatin analogs. Concurrent use of somatostatin analogs was permitted if administered at a stable dose for at least 2 months.

Patients in both the epNETs and pNETs subgroups were randomly assigned 2:1 to receive either cabozantinib at 60 mg daily or daily placebo until PD. Patients in the placebo arm were allowed to crossover and receive cabozantinib 60 mg daily after PD. The primary end point was PFS by blinded independent central review, and secondary end points included overall survival (OS), objective response rate, safety, and tolerability.

In the epNETs group, at baseline, patients in the cabozantinib and placebo arms had a median age of 66 years (range, 28-86) vs 66 years (range, 30-82), respectively. Most patients in both arms had an ECOG performance status of 1 (64% vs 52%), well-differentiated disease (79% vs 84%), grade 2 disease (63% vs 66%), and received concurrent treatment with somatostatin analogs (62% vs 65%). The median number of prior lines of therapy received was 2 (range, 1-7) vs 2 (range, 1-6), respectively, and the median time from diagnosis to random assignment was 87 months (range, 10-489) vs 83 months (range, 13-340).

Similarly, in the pNETs group, patients in the cabozantinib and placebo arms had a median age of 60 years (range, 29-79) vs 64 years (range, 39-79), respectively. Most patients in both arms were men (56% vs 58%), had well-differentiated disease (90% vs 97%), had grade 2 disease (63% vs 63%), and received prior treatment with somatostatin analogs (97% vs 94%). The median number of prior lines of therapy received was 3 (range, 1-9) vs 2 (range, 0-7), respectively, and the median time from diagnosis to random assignment was 84 months (range, 1-213) vs 83 months (range, 18-197).

In the epNETs cohort, patients in the cabozantinib and placebo arms underwent a median number of 5 (range, 0-35) and 3 (range, 0-24) cycles of therapy, respectively. At the data cutoff, 22% and 16% remained on treatment, respectively. Reasons for treatment discontinuation included PD (36% vs 50%), adverse effects (AEs; 20% vs 15%), consent withdrawal (9% vs 4%), alternate therapy (3% vs 1%), physician decision (2% vs 6%) and death (5% vs 3%).

In the pNETs cohort, patients in the cabozantinib and placebo arms underwent a median number of 7 (range, 0-38) and 4 (range, 0-12) cycles of therapy, respectively. At the data cutoff, 24% and 13% of patients remained on treatment, respectively. Reasons for treatment discontinuation included PD (35% vs 71%), AEs (16% vs 0%), consent withdrawal (5% vs 10%), and intercurrent illness (5% vs 3%).

Additional data from the study revealed that the median OS in the epNETs cohort was 21.9 months vs 22.4 months in the cabozantinib and placebo arms, respectively (stratified HR, 0.90; 95% CI, 0.56-1.46; log-rank P = .34). In the pNETs cohort, the median OS was 43.5 months vs 31.0 months in the cabozantinib and placebo arms, respectively (stratified HR, 0.77; 95% CI, 0.34-1.73; log-rank P = .26).

In terms of safety, patients in the cabozantinib (n = 124) and placebo (n = 63) arms of the epNETs cohort experienced grade 3 (59.7% vs 33.3% respectively), grade 4 (7.3% vs 1.6%), and grade 5 (8.9% vs 7.9%) AEs. Commonly occurring grade 3 or higher AEs consisted of hypertension (27.4% vs 4.8%), fatigue (13.7% vs 7.9%), and diarrhea (10% vs 3%). Investigators noted that of the 11 grade 5 AEs that were observed in the cabozantinib arm, events were unrelated or unlikely to be related to treatment in 8 patients and possibly related in 3 patients.

In the pNETs cohort, patients in the cabozantinib arm (n = 60) experienced grade 3 (56.7%), grade 4 (8.3%), and grade 5 (3.3%) AEs; those in the placebo arm (n = 30) experienced grade 3 AEs (43.3%). The most common grade 3 or higher AEs included hypertension (26.7% vs 20.0% respectively), fatigue (13.3% vs 3.3%), and hyperglycemia (8.3% vs 10.0%). The grade 5 AEs in the cabozantinib arm were determined to be unrelated to treatment.

“The interim analyses and both of the cohorts showed a statistically significant and clinically meaningful difference in PFS in patients treated with cabozantinib,” Chan concluded. “Based on these results, the DSMB voted unanimously to stop accrual to the trial, unblind all patients, and all those receiving placebo to crossover to cabozantinib. The AEs that we observed were consistent with the known safety profile of cabozantinib. Based on these results, cabozantinib may become a new treatment option for patients with previously treated, progressive NETs.”

REFERENCE:

Chan J, Geyer S, Ou FS, et al. Alliance A021602: Phase III, double-blinded study of cabozantinib versus placebo for advanced neuroendocrine tumors (NET) after progression on prior therapy (CABINET). Ann Oncol. 2023;34(suppl 2):S192. doi:10.1016/j.annonc.2023.10.047