Neoadjuvant Carboplatin/Paclitaxel Showing Potential in Esophageal Cancer

Neoadjuvant treatment with carboplatin and paclitaxel-based chemotherapy produced a 27.9% pathologic complete response rate in patients with resectable esophageal cancer, according to results of the NEOSCOPE trial.

Somnath Mukherjee, MD

Neoadjuvant treatment with carboplatin and paclitaxel-based chemotherapy (CarPacRT) produced a 27.9% pathologic complete response (pCR) rate in patients with resectable esophageal cancer, according to results of the NEOSCOPE trial.

Data from the randomized phase II study, which were presented during the 2016 Gastrointestinal Cancers Symposium, also demonstrated good tolerability for the CarPacRT regimen, according to Somnath Mukherjee, MD. However, neoadjuvant oxaliplatin/capecitabine chemoradiotherapy (OxCapRT) failed to demonstrate activity to move forward, he said.

“Neoadjuvant chemoradiotherapy is a standard of care for operable esophageal cancer,” said Mukherjee, Department of Oncology, CRUK/MRC Institute for Radiation Oncology, University of Oxford, but he maintained that evidence suggests the current chemoradiotherapy standard practice may be improved upon by using a different chemoradiotherapy combination.

In NEOSCOPE, patients with operable cancer of the mid-lower esophagus/gastroesophageal junction (GEJ) received induction chemotherapy with 2 cycles of oxaliplatin and capecitabine. Eligible patients had tumor stage ≥T3 and/or ≥N1 and a total disease length <8 cm.

Eighty-five patients were randomized 1:1 to OxCapRT (oxaliplatin, 85 mg/m2 on alternate weeks; capecitabine, 625 mg/m2 on the days of radiotherapy) or CarPacRT (carboplatin area under the curve 2 weekly; paclitaxel 50 mg/m2 weekly). All patients underwent imaging to restage their disease, with surgery scheduled for 6 to 8 weeks following radiation therapy. The recommended surgery was a two-phase esophagectomy with two-field lymphadenectomy; minimally invasive esophagectomy was allowed.

Tumor characteristics were balanced between the 2 arms. Approximately 86% of patients, overall, had stage T3 disease and 57 of the 85 were node-positive. The predominant tumor site was the lower third of the gastroesophageal junction (76.2% in the OxCapRT arm; 90.7% in the CarPacRT arm); 14.3% in the OxCapRT arm had the middle third as the site of the predominant tumor compared with 4.7% in the CarPacRT arm. The maximal disease length was 5.85 cm and 5.7 cm in the 2 arms, respectively.

The rate of pCR was the primary endpoint, with a “pick a winner” design to the trial.

“A pCR of 15% was not considered large enough to warrant further investigation, whilst a pCR of 35% was considered worthwhile,” said Mukherjee. Either arm having at least 10 pCRs out of the first 38 patients would be considered for the phase III study.

The pCR rates were 11.9% in the OxCapRT arm and 27.9% in the CarPacRT arm by intent to treat. “Ten of the first 38 patients in the CarPacRT arm attained pCR, thereby meeting the prespecified criteria for success,” said Mukherjee.

Thirty-six of 42 patients randomized to OxCapRT proceeded to surgery compared with 41 of 43 randomized to CarPacRT. Principle reasons for not proceeding to surgery included disease progression and death during chemotherapy.

“Overall, the toxicity was similar across the arms, although the incidence of neutropenia was higher in the CarPacRT arm, 21.4% versus 2.6%,” said Mukherjee.

There were 3 deaths on the study; all occurred during induction chemotherapy.

Treatment was very well tolerated and treatment compliance was high. Median dose intensity was 96.9% for oxaliplatin, 97.5% for capecitabine, 85.3% for carboplatin, and 82.5% for paclitaxel. Similarly, radiotherapy was extremely well tolerated: 38 of 38 patients in the OxCapRT arm and 41 of 42 in the CarPacRT arm received full-dose radiotherapy.

About half of 3 patients in each arm experienced postoperative complications by 30 days on the treatment. Cardiac complications were slightly higher in the OxCapRT arm compared with the CarPacRT arm (25.0% vs 9.8%), but the difference was not statistically significant.

The R1 resection rate, defined as tumor within 1 mm of the circumferential resection margin, was 27.8% in the OxCapRT arm and 19.5% in the CarPacRT arm.

Further, there was overlap in the 95% confidence intervals in the 2 arms of NEOSCOPE, given its small size, he said, and “imbalances in poor prognostic factors disfavoring the oxaliplatin arm.”

Patients in the CarPacRT arm had a 10% lower rate of advanced disease and a lower rate of proximal disease as well, for which R0 resection is more difficult.

Both weekly carboplatin/paclitaxel-based neoadjuvant chemoradiotherapy and oxaliplatin/fluoropyrimidine-based neoadjuvant chemoradiotherapy have shown promise in multiple clinical trials. In a definitive trial of oxaliplatin/5-flurouracil (5-FU) versus cisplatin/5-FU-based chemoradiotherapy, oxaliplatin/5F-U had shown comparable efficacy but less toxicity.

Putting the NEOSCOPE data into context is challenging when considering the wide range of pCR rates, from 13% to 50%, in clinical trials of neoadjuvant chemoradiotherapy, with or without induction chemotherapy, according to invited discussant Daniel V.T. Catenacci, MD, assistant professor of Medicine, University of Chicago Medicine.

The CALGB 80803 trial, in which positron imaging/computed tomography is being used to determine whether or not a particular chemotherapy regimen is having an effect on tumor size in patients with esophageal cancer, may provide more clarity, according to Catenacci.

Mukherjee S, Hurt C, Gwynne S, et al. NEOSCOPE: A randomised phase II study of induction chemotherapy followed by either oxaliplatin/capecitabine (OXCAP) or carboplatin/paclitaxel (CarPac) based chemoradiation (CRT) as pre-operative regimen for resectable oesophageal adenocarcinoma. J Clin Oncol. 2016;34 (suppl 4S; abstr 3).


View more from the 2016 GI Cancer Symposium