Objective To evaluate the clinical efficacy and safety of intrapleural injection of recombinant human endostatin (endostar) and/or cisplatin in treatment of malignant hydrothorax and ascites. Methods The multicenter, randomized controlled study was conducted in 14 Chinese nationwide large hospitals from January 2011 to January 2014(Clinical Trials:NCT01327235.). A total of 317 patients with more than moderate amount of pericardial effusion malignant hydrothorax and ascites were randomly divided into group A (endostar group, n=105), group B (cisplatin group, n=104) and group C (endostar combined cisplatin group, n=108). After puncture and drainage, endostar 45 mg per time by intrathoracic injection or 60 mg per time by intraperitoneal injection was performed in group A. Cisplatin, 40 mg per time by intrapleural injection on day 1, 4 and 7, was administrated in group B. Endostar and cisplatin were administrated in group C. All three groups of patients were administrated on day 1, 4 and 7, and three times as a course. The main outcomes were objective response rate (ORR), time to disease progression (TTP), quality of life (QOL) and Karnofsky performance status (KPS), as well as the drug safety. Results A total of 317 patients were included in full analysis set (FAS) (group A with 105 cases, group B with 104 cases, group C with 108 cases), and 275 patients were included in perprotocol set (PPS) (group A with 98 cases, group B with 89 cases, group C with 88 cases). There were 298 cases and 273 cases qualified for evaluation on drug efficacy in FAS and PPS respectively. There was a significant difference in ORR among three groups (A: 48.51%, B: 46.39%, C: 63.00%, P=0.0373), and ORR was higher in group C than that in groups A and B (P=0.0189). Patients without intracavitary treatment history, with hydrothorax, female, without systemic chemotherapy, with initial treatment on effusion, with sufficient drainage, with hemorrhagic effusion and without diagnosis of gastric carcinoma had better outcome in ORR after treatment (P<0.05 or P<0.01). In those with hemorrhagic effusion, the ORRs in groups A and C were significantly higher than that of group B (A: 71.05%, B: 45.16%, C: 80.00%, P=0.0090). And in those with hemorrhagic pleural effusions specifically, the ORRs in groups A and C were significantly higher than that of group B (A: 71.42%, B: 40.00%, C: 88.88%, P=0.0013). The median TTP was 68.869 d, 44.951 d and 69.030 d in group A, B and C, respectively, with a significant difference (P=0.0121), and was shorter in group B than that in group A and C (P=0.0240, P=0.0046). The proportion of patients with improved QOL and KPS in group A was higher than that in group B and C after the third and sixth administration, respectively (P<0.05 or P<0.01). The incidence of adverse events was lower in group A than that in group B (P=0.0005), but no significant difference was shown between group B and C (P=0.2866).
Conclusion Intra-pleural injection of endostar is potentially effective in treatment of patients with malignant hydrothorax and ascites, especially those with hemorrhagic effusion. It shows a synergistic effect with cisplatin in improving the clinical efficacy, TTP and QOL, but without increasing the risk of adverse reactions. It is worth to be widely applied in clinical practice further.