Study Design

This is a phase I nonrandomized and uncontrolled single-arm clinical trial without dose escalation to establish the safety of oxaliplatin PIPAC in cycle 1, and the combination of oxaliplatin PIPAC with 5-FU/LV in cycles 2–3 given within 24 h of the PIPAC. The rules for accrual were based on limiting the risk to that of the traditional 3+3 phase I trial design with modifications that allow for reduced time to complete the study without compromising patient risk.¹³ These rules are based on not exceeding the risks associated with the traditional 3+3 phase I design, while better accounting for the patient queue including the potential for inevaluable patients (patients who do not get fully treated or followed for reasons other than unacceptable treatment-related adverse event such as rapidly progressive disease). This approach has been used in several NCI studies (NCI CTEP protocols nos. 9091 and 9924, to reduce study duration without adversely impacting operating characteristics or patient safety). Unless the trial stops accrual for unacceptable AEs, the expected number of patients starting treatment is eight during the safety evaluation, which allows for inevaluable patients and a maximum of eight evaluable patients. Additional patients are permitted as noted above after the safety evaluation, up to 14 patients. The study required no more than one unacceptable treatment related AE (DLT) in the first six patients and no more than two in the first seven or eight patients. Once initial safety was established without unacceptable adverse events on the basis of those rules or any treatment death, the study planned for accrual of up to 14 patients. The PIPAC procedure was performed as previously described,¹⁴ with intraperitoneal drug delivered using a high-pressure injection (Medrad Stellant injector, Bayer Corporation) and Capnopen nebulizer (Capnomed Corporation, Tubingen, Germany) at a maximum of 200 psi and 30 mL/min, followed by a 30-min pneumoperitoneum at 12 mmHg containing the aerosolized at room temperature prior to release of the pneumoperitoneum. Oxaliplatin 90 mg/m² was administered intraperitoneally (IP) via PIPAC at 6-week intervals for a total of three treatments provided no adverse event (AE), dose-limiting toxicity, disease progression, or patient withdrawal occurred. Prior to the second and third treatments, 5-FU 400 mg/m² IV with LV 20 mg/m² IV were administered for sensitization. There were no planned dose modifications for this study. If a patient was unable to receive the next scheduled PIPAC due to toxicity, a delay of up to 21 days was permissible, with any delay greater than 21 days considered a DLT. Other toxicities considered a DLT were any grade 3 or higher nonhematologic toxicity, excluding grade 3 nausea, vomiting, abdominal pain or diarrhea adequately treated to grade 2 or lower within 48 h; grade 3 fatigue that returns to grade 2 or less within 7 days; grade 3 laboratory/metabolic abnormalities that are not considered clinically significant and are easily correctable to grade 2 or lower within 72 h; grade 3 infusion-related reaction (first occurrence and in the absence of steroid prophylaxis) that resolves within 6 h with appropriate clinical management, and grade 3 peripheral neuropathy. Additionally, surgical complications of Clavien–Dindo grade IIIB or higher; Grade 4 thrombocytopenia, grade 4 neutropenia lasting more than 7 days or associated with fever or infection were also considered a DLT. The DLT rules were applied for cycle 2 to assess the safety of PIPAC oxaliplatin, and separately in cycle 2 to assess the addition of 5-FU/LV preceding PIPAC oxaliplatin.