Summary of Main Results
71.4% of the patients are candidates for IDS in this study, while 38(88.4%) achieved optimally debulking and 24(55.8%) had no residual disease after IDS. We did not find a significant difference in intraoperative blood loss between the treatment pattens receiving the regimen containing bevacizumab or not in the last preoperative chemotherapy. The only factor affecting PFS was the dosage of bevacizumab. The safety profile was acceptable in this study as only 5.2% (3/58) of the patients discontinued bevacizumab due to toxicity.
Results in the Context of Published Literature
NACT is a treatment strategy for patients with advanced ovarian cancer with the aim of reducing the morbidity of IDS and increasing the probability of optimal debulking. ANTHALYA trial[10]is a multicenter open-label noncomparative Phase II study aimed to investigate whether adding bevacizumab to neoadjuvant carboplatin-paclitaxel helps achieve optimal debulking. 95 patients with stage III or IV ovarian cancer were randomized 2:1 to receive four cycles of neoadjuvant carboplatin-paclitaxel with or without bevacizumab during cycles 1-3 before IDS. More patients in the bevacizumab group were candidates for IDS(69 vs. 60% ) and the proportion is similar to this study(71.4%). The complete resection rate was 58.6% compared with 46.3% in the Phase III EORTC study[11]. GEICO1205 trial[12]is a randomized Phase II study to evaluate neoadjuvant bevacizumab in newly diagnosed stage III or IV ovarian cancer. More patients receiving bevacizumab than chemotherapy alone underwent IDS(89 vs 67%, p=0.029). However, the complete resection rates were very similar(66 vs 64%, p=0.858). In this study, for stage IV and stage III patients with extensive tumor spread, doctors are more inclined to choose the regimen containing bevacizumab at the beginning of NACT, so as to enhance the efficacy of NACT and surgical resection rate. For patients with SD or unsatisfactory CA125 reduction assessed after NACT, the addition of bevacizumab in the chemotherapy regimen enhanced the efficacy, and likewise enabled more patients to obtain the opportunity of surgery. Nevertheless, the complete resection rate in this study was lower than that of the ANTHALYA and GEICO1205 studies, which may be caused by the lower dose of bevacizumab. Of the 43 surgical patients, 26 (60.5%) patients received 7.5 mg/kg dose and 17 (39.5%) received 15 mg/kg dose while the dosage of the above two studies was 15 mg/kg. At the same time, the number of courses of bevacizumab varies widely, as 20 of the 43 surgical patients received only 1-2 courses of bevacizumab in NACT. It also exacerbated the shortage of bevacizumab. Whether increasing the dosage of bevacizumab can improve the complete resection rate needs further study.
To reduce the risk of bleeding and wound healing complications, the drug label for bevacizumab recommends discontinuation of the drug at least 28 days before surgery. In this study, doctors commonly performed surgery 6 weeks after withdrawal of bevacizumab without increasing the amount of intraoperative bleeding. We did not find a significant difference in intraoperative blood loss between the treatment pattens receiving the regimen containing bevacizumab or not in the last preoperative chemotherapy. As well, prophylactic anticoagulant therapy with low molecular heparin was performed in 39.5%(17/43) of the patients after surgery, and no significant postoperative bleeding was observed. The incidence of impaired wound healing(2.3%) was also similar to that of the NACT cohort in the MITO16A study (3%)[13]. Adding bevacizumab to neoadjuvant chemotherapy is feasible and tolerable.
The safety profile was acceptable in this study as only 5.2% (3/58) of the patients discontinued bevacizumab due to toxicity. No ≥grade 3 AEs was observed, demonstrating that bevacizumab was well tolerated. Most AEs occurred during the combination chemotherapy phase of bevacizumab administration. Similar to some clinical trials[4-5,14-16], hypertension was the most common AE and the percentage of hypertension in this study was 10.3%. Other more serious AEs, such as gastrointestinal perforation, fistula and central nervous system bleeding, were not observed in this study. In the real world, doctors may avoid bevacizumab therapy in patients with bleeding tendencies and gastrointestinal invasion of the tumor. Meanwhile, 62.1%(36/58) of patients received a dosage of ≤7.5 mg/kg and 37.9%(22/58) received 15 mg/kg. The lower dose may result in a lower incidence of adverse reactions.
Implications for Practice and Future Research
Although there was no significant difference between the two groups in the median PFS. The NT+ FL group had a better 12-month PFS. We found that the median number of bevacizumab courses in the NT+ FL group was eight and four patients completed 22 treatment courses. While the median number of bevacizumab courses was three in the NT group. A more adequate number of courses may lead to a better PFS. Patients in the NT group did not continue to use bevacizumab after IDS, which may be due to: the high proportion of no residual surgery (63.6%) leading to the decrease of doctors' willingness to add bevacizumab in postoperative treatment; six patients received maintenance therapy with a poly-ADP-Ribose polymerase(PARP) inhibitor after chemotherapy, and the choice of another maintenance therapy led to the abandonment of bevacizumab. Bevacizumab was discontinued in one patient due to impaired wound healing and in two patients due to deep vein thrombosis; and economic issues. Inadequate treatment of bevacizumab, premature discontinuation and the use of PARP inhibitors may have an impact on survival.
Currently, with the development of PARP inhibitors in the treatment of ovarian cancer, patients have more options for maintenance therapy. According to NCCN guideline, maintenance therapy with PARP inhibitors may benefit for newly diagnosed stage II-IV high-grade serous carcinoma, G2/3 ovarian endometrioid carcinoma, clear cell carcinoma with BRCA1/2 mutation, and carcinosarcoma after CR or PR is achieved following surgery and platinum-based first-line therapy[6]. In this study, six patients(10.3%, 6/58) received maintenance therapy with PARP inhibitors instead of bevacizumab after surgery and chemotherapy and BRCA mutations were present in three patients. Nevertheless, PARP inhibitors have limited efficacy for patients with an effective Homologous Recombination proficient(HRP) mechanism. Whether bevacizumab maintenance therapy is better than PARP inhibitors for HRP patients needs further verification. In addition, bevacizumab requires intravenous administration every 3 weeks and is less convenient than PARP inhibitors, which also reduces patients' willingness to take it.
Economic factors were another crucial factor in discontinuing bevacizumab, with 11 patients discontinuing it because they could not afford it. It may improve in the future as the price of bevacizumab in China has gradually decreased recently.
The univariate analysis indicated that age(P=0.047), dosage(P=0.020) and complete resection(P=0.075) had an impact on the PFS. However, according to the multivariate regression analysis, we found that the only factor affecting PFS was the dosage of bevacizumab. The patients receiving the dosage of 15 mg/kg had a better PFS than those receiving the dosage of ≤7.5 mg/kg(HR=0.352, 95% CI: 0.143-0.864, P = 0.023). To date, studies of neoadjuvant bevacizumab have been limited and all the studies[10,12,13] used the dosage of 15 mg/kg. There is a lack of clinical studies using 7.5 mg/kg dose in NACT or comparing 15 mg/kg to 7.5mg/kg. In the real world, doses of 7.5 mg / kg are usually used because of concerns about side effects such as hypertension and increased risk of bleeding during surgery, as well as economic factors. More prospective studies are needed to compare the different dosages and survival of NACT with bevacizumab.
Strengths and Weaknesses
To the best of our knowledge, no studies of NACT with bevacizumab treatment in ovarian cancer patients in China have been published. We studied the prognostic factors and outcomes of 58 patients receiving the NACT treatment regimen containing bevacizumab. However, our study has several limitations, including its single-center, retrospective design. The sample of each treatment group was small, particularly for subgroup analysis. The patients' follow-up time was not long enough to obtain more survival data and analysis. Retrospective follow-up data may result in a lower incidence of AEs than actual.