Hoskins et al (1) reported the relationship between the diameter of residual tumors and prognosis in primary debulking surgery (PDS) in ovarian cancer. Subsequently, the importance of PDS has been examined. Although platinum preparations are usually employed, it has been concluded that tumor reduction surgery is a strong determinant factor influencing survival periods in stage III/IV patients. This conclusion was reached following a meta-analysis conducted by Bristow et al (2). The study by these authors is the only available evidence of the effectiveness of surgical therapy with respect to advanced ovarian cancer. In particular, surgical therapy was considered to be optimal when residual tumors were less than 1 cm since the prognosis following surgery was significantly more satisfactory, making PDS the main objective.

Patients with advanced ovarian cancer often have metastatic disease in the upper abdominal region. Nevertheless, it is difficult to achieve optimal surgery using only standard surgical techniques such as hysterectomy, oophorectomy and omentectomy, without a surgical approach that involves upper abdominal surgery (3). Using upper abdominal surgery, the prognosis after optimal surgery is significantly more satisfactory compared to suboptimal cases (3,4,5). In particular, metastases to the diaphragm are common in the latter patients (6). Thus, diaphragmatic surgery is an operational procedure essential in achieving optimal debulking surgery in advanced mullerian carcinomas such as ovarian, primary peritoneal and fallopian tube carcinomas. There are two significant procedures for diaphragmatic surgery: one involves the stripping of disseminated lesions, and the other involves full-thickness resections of nodal lesions which penetrate the diaphragm. The two procedures are often performed simultaneously. Neither of these procedures is frequently performed by gynecological oncologists. In current medical practice, it appears that practitioners lacking experience in such procedures hesitate to perform them due to the possibility of incidental events. This study investigated incidental events retrospectively to evaluate the safety of the two procedures.

A total of 82 patients who underwent surgery at the Tokyo Medical University Hospital between 2006 and 2009 were analyzed in this study. As noted in Table I, the primary sites were the ovaries in 71 patients, the peritoneum in 9 and the fallopian tubes in 2 patients. The histological types were serous adenocarcinomas including serous papillary in 72 patients, endometrioid adenocarcinomas in 5, mucinous adenocarcinoma in 1 patient and clear cell carcinomas in 4 patients. The majority of cases were serous types.

With regard to FIGO clinical staging, 2 patients were in stage IIIb, 74 were in stage IIIc and 6 patients were in stage IV. Diaphragmatic surgery was performed during PDS in 63 patients, interval debulking surgery (IDS) in 13 and secondary debulking surgery (SDS) in 6 patients. The patients in the IDS group received 2–6 courses of paclitaxel/carboplatin combination chemotherapy prior to undergoing surgery.

Out of the 82 patients, stripping and full-thickness resections were performed in 56 and 12 patients, respectively, and the two procedures were performed in 14 patients. Diaphragmatic tumors were removed completely in 71 patients but incompletely in 11 patients (13.4%). Regarding intra-operative incidental events, unexpected open chest surgery occurred in 5 patients (8.9%) in the group that underwent stripping (Table II). On the other hand, post-operative incidental events including pleural effusion were observed in 43 patients (52.4%). Pleural effusion was observed in 25 patients (44.6%) of the group that underwent stripping, 7 (58.3%) of the group that underwent full-thickness resection and 11 patients (78.6%) of the group that underwent both procedures (Table II). No significant differences among the groups were noted.

In terms of the timing of the operations, unexpected open chest surgery occurred in 1 out of 63 patients (1.6%) in the PDS group, 4 out of 13 patients (30.8%) in the IDS group and 0 out of 6 patients (0.0%) in the SDS group. Incidences were significantly high in the IDS group (Table III).

Of note is that none of the patients with incidental events required chest drainage or experienced problems during the natural course of the observations.

Complete cytoreduction is the principle goal in mullerian carcinomas such as ovarian, tubal and peritoneal carcinomas. However, if this poses obstacles, efforts should be made to the greatest extent possible to make the residual disease less than 1 cm in order to improve prognosis (5). It has also been reported that, following the peritoneum and colon, lesions were frequently observed on the diaphragm. This hypothesis was posited by an analysis of tumor localizations in surgical reports of approximately 200 ovarian cancer cases (7). An analysis of bulky diseases in the upper abdominal region, including the diaphragm cranial to the omentum, has shown that there is a correlation between large-volume ascites and disease sites. The same study also revealed that the percentage of cases resulting in suboptimal cytoreductions increased (8). Thus, diaphragmatic disease is a crucial factor in the treatment of advanced ovarian cancer. Complete and optimal debulking rates are thought to increase if diaphragmatic lesions are removed using surgical procedures. Moreover, a better prognosis is likely to be achieved in comparison with cases which do not undergo diaphragmatic surgery (9,10).

However, diaphragmatic metastases are a common obstacle to optimal cytoreduction for patients with ovarian cancer (10). Care with regard to incidental events associated with diaphragmatic surgery must therefore be exercised at all times. The most common incidental events are increased postoperative pleural effusion. It has been reported that incidences increase particularly when open chest occurs (10). With regard to postoperative incidental events, especially pleural effusion and pneumothorax, Devolder et al investigated 69 ovarian cancer patients who underwent diaphragmatic surgery (11). Of the 69 patients, 17 underwent stripping, 22 coagulation and 30 patients underwent both procedures. Postoperative pleural effusion was reported to have been observed in 41 patients, 3 of whom underwent chest drainage, 7 pleural punctures and 1 patient underwent both procedures. In addition, pneumothorax was observed in 4 out of 69 patients, and 1 of these patients required chest drainage. Moreover, Eisenhauer et al reported that ipsilateral pleural effusion was observed in more than half of the patients (an overall rate of 58%) that underwent diaphragmatic peritonectomies or resections. These authors concluded that conservative management is possible in the majority of cases without the need for chest tubes or thorcentesis (12). Postoperative pleural effusion, which may pose clinical obstacles, was not observed in the present study of 82 patients, nor was routine chest tube placement performed at the time of stripping or full-thickness resection. Postoperative pneumothorax was observed at an overall rate of 2.4% (2/82 patients). However, it was possible to prevent pneumothorax from occurring through the use of adequate vacuuming prior to closing the chest (13). Attention should therefore be paid when performing IDS since, as previously analyzed, the incidence of postoperative incidental events was significantly high in the case of diaphragmatic surgery following IDS. Disseminated tumors on the diaphragm, observed prior to neoadjuvant chemotherapy, are thought to induce changes such as necrosis and weakening of the layer structure of the diaphragm.

As observed in attempts at using EndoGIA stapling devices or Gore-tex grafts in diaphragmatic resections or diaphragmatic reconstructions (14), diaphragmatic surgery is likely to be actively performed in the future with respect to advanced ovarian cancer (with peritoneal or fallopian tube cancer). However, the prevention of postoperative complications and incidental events should be further investigated.