Eribulin Demonstrates Limited Activity in Advanced or Recurrent Cervical Cancer

Eribulin (Halaven) induced limited activity with a favorable toxicity profile in patients with recurrent or advanced cervical cancer, and prior exposure to paclitaxel was associated with decreased response to the agent, according to data from a phase 2 trial (NCT01676818) presented during the 2021 SGO Virtual Annual Meeting on Women’s Cancer.

Among 30 evaluable patients, 1 (3%) had a complete response (CR) with eribulin, 5 (17%) had a partial response (PR), and 13 (43%) had stable disease (SD). The clinical benefit achieved with the agent was 63%. Moreover, the median duration of response for patients who achieved a CR or PR (n = 6) with the agent was 4.3 months (95% CI, 2.7-51.1). The median overall survival (OS) was 6.5 months (95% CI, 4.5-12.7), while the median progression-free survival was 2.5 months (95% CI, 1.2-4.2).

Patients who had previously been treated with paclitaxel had less favorable responses to the agent (P = .002) and a significantly shorter PFS vs those who did not, at 1.2 months and 3.5 months, respectively (P = .008). The OS was also shorter for patients who had prior paclitaxel vs those who did not receive it, at 4.2 months vs 7.7 months, respectively (P = .14).

“While the primary end point [of 6-month PFS] was not met, the eribulin may be an option for women with paclitaxel-naïve, recurrent or advanced cervical cancer,” the authors of the poster wrote.

Eribulin, an analog of Halichondrion B, isolated from the Japanese marine sponge Halichondria okadai, induces anticancer activity by way of a tubulin-based antimitotic mechanism. The agent has a unique mechanism of microtubule inhibition that suppresses microtubular growth without impacting microtubular shortening.

Previously, eribulin has demonstrated improved response rates, as well as OS benefit, in patients with pretreated metastatic breast cancer. Moreover, the agent has shown improved OS in patients with previously treated, advanced liposarcoma. Preclinical antitumor activity with eribulin has also been reported in patients with squamous cell carcinoma.

The primary end point for the study was 6-month PFS, while secondary end points included best overall response per RECIST v1.1 criteria, toxicity, and OS. Exploratory end points comprised associations of apoptosis and proliferation markers, unfolded protein response markers, and tubulin subtypes with clinical activity.

A prespecified futility analysis gating stage 2 was established if 0 of 15 patients achieved at least stable disease at 6 months. Thirty evaluable patients would ensure 80% power when the true 6-month PFS was 26% with a 1-sided a of 0.1 or lower. Archival tumor samples were examined via immunohistochemistry for t-protein, tubulin subtypes, pAKT, EGFR, GRP78, p53, Bax and Ki67. Additionally, intensity and percent distribution were assigned by clinical pathologists, and overexpression was determined to be a score of 2 or higher in both parameters.

To be eligible for enrollment, patients had to have advanced or recurrent cervical cancer who had received 1 or fewer prior chemotherapy treatments, had measurable disease, and an ECOG performance status of 0 to 2. From November 2014 through May 2017, a total of 32 patients were enrolled to the study.

Participants were treated with intravenous eribulin at a dose of 1.4 mg/m2 on day 1 and day 8 of a 21-day cycle. Those who achieved a CR, PR, or who had stable disease, were able to continue on therapy; those who experienced peripheral neuropathy of grade 2 or higher that did not improve with dose reduction, disease progression, or a delay of treatment for at least 2 weeks were removed from treatment.

Two patients were unevaluable after having received less than 2 cycles of treatment. The 30 evaluable patients received a median of 4 cycles of treatment (range, 1-74), and the median time on study was 2.7 months (range, 0.6-52.1). The median follow-up time was 52.1 months (range, 29.6-77.3).

The median age of the study participants was 51 years (range, 29-76) and 63% had an ECOG performance status of 1. Additionally, 25% of patients had adenocarcinoma, 3% had adenosquamous carcinoma, 3% had small cell carcinoma, and 69% had squamous cell carcinoma. The majority, or 91%, had received prior chemotherapy treatment. Forty-one percent of patients had previously received cisplatin plus paclitaxel and bevacizumab (Avastin), while 41% had received cisplatin/gemcitabine, 7% had prior carboplatin/paclitaxel, and 10% had cisplatin with or without pemetrexed or capecitabine only. Moreover, 66% of patients had recurrent disease, and 56% had not received prior treatment with paclitaxel. Most patients had not previously received radiation treatment.

Correlative studies showed that βII-tubulin and βIII-tubulin, as well as BAX expression were predictors of response to eribulin for this patient population. Specifically, patients who did not overexpress βII-tubulin and BAX were more likely to respond to the agent (P = .40 and P = .002, respectively). Univariate analyses revealed that PFS was shorter in patients with βII-tubulin overexpression, at 1.2 months (95% CI, 0.9-2.6), and in patients with BAX overexpression, at 1.2 months (95% CI, 1.1-2.6). OS was shorter in those with βIII-tubulin and BAX overexpression, at 6.4 months (95% CI, 3.1-12.7) and 4.3 months (95% CI, 2.7-6.5), respectively.

In terms of safety, 2 patients were taken off of the study after experiencing grade 3 paresthesia after 7 cycles and 74 cycles of treatment each. Additionally, 5 patients required a dose reduction while on treatment, while 3 patients required 2 dose reductions.

The most common grade 3 or higher adverse effects were anemia (38%), decreased neutrophil count (22%), and decreased white blood cell count (19%).

Reference

Garcia-Sayre J, Lin Y, Matuso K, et al. Phase II trial of eribulin in advanced/recurrent cervical cancer. Presented at: 2021 SGO Virtual Annual Meeting on Women’s Cancer; March 19-25, 2021; Virtual. Accessed March 19, 2021.