To evaluate the feasibility and pharmacology of intrapleural (IPL) administration of paclitaxel, 18 patients with malignant pleural effusions were entered onto a phase I clinical study, 13 were caused by lung cancer. Following an effusion drainage rate of less than 100 ml/24 h and full expansion of the lung, patients were treated with a single instillation of pacilitaxel administered IPL in 500 ml of normal saline and retained for a maximum of 96 h when tolerated. No systemic chemotherapy or ipsilateral thoracic irradiation was given for 4 weeks before and after the IPL treatment. The starting dose was 82.5 mg/m(2) with the dose escalation schedule of 125, 175, 225 and 300 mg/m(2). There were minimal local or systemic toxicities, such as local chest pain or myelosuppression, even when the paclitaxel dose reached 225 mg/m(2). The pharmacological advantages of the IPL administration of paclitaxel were demonstrated by the mean exposure of the pleural cavity (area under the concentration-time curve) to paclitaxel after IPL delivery exceeding that of the plasma by approximately 370-fold (range 55-684) and by the extraordinarily slow IPL clearance of paclitaxel (mean +/- SE 0.49 +/- 0.07 l/m(2)/day; range 0.08-1.16 l/m(2)/day) with significant concentrations of paclitaxel persisting within the cavity for more than 48-96 h after a single IPL instillation. In patients with detectable plasma paclitaxel levels, the plasma levels achieved exceed the minimal concentrations that are required to induce cytotoxic effects in vitro. Four patients had progressive dyspnea during IPL retention of paclitaxel solution because of treatment failure and needed drainage of effusion. One of these patients who was at the dose level of 225 mg/m(2) originally had severely chronic obstructive lung disease, developed acute respiratory failure, refused mechanical ventilation support and succumbed to respiratory failure. No further patients were included after this event. Antitumor effect was shown by four of the 15 evaluable patients having no recurrence of effusion on chest radiograph at 1 month. Most of these responders had a good performance status, normal pretreatment pleural pH and/or glucose compared with the non-responders. We conclude that paclitaxel at a dose level of 175 or 225 mg/m(2) is feasible for use intrapleurally. It could be considered for incorporation into treatment programs for patients with less advanced thoracic tumors with carcinomatous pleuritis or with IPL tumors following surgical debulking.
