Discussion:
Ovarian cancer is the leading cause of death from gynecological cancer [1]. Early-stage ovarian carcinoma seldom produces symptoms, and consequently more than half of all ovarian carcinomas are at an advanced stage[2]. 70% of patients are diagnosed at an advanced stage [1]. Although serous adenocarcinoma is the most common histological subtype [1], an exceptional complication has been described. According to a recent literature review by a Japanese team [2], regarding colo-adnexal fistula, only a case was noted with serous carcinoma. Fistulae usually occur after previous surgery and chemotherapy, in the setting of relapsed disease rather than as an initial presentation [3], spontaneous fistulization of ovarian cancer into digestive tract with a fistula tract is a different and rare phenomenon [4]. Direct invasion of ovarian cancer to the walls of the colon, rectum, and other organs is not unusual, the extension of ovarian cancer to invade the colon and rectum with resultant fistulas is extremely unusual [2]. Enterogenital fistulas occupy the dominant cause of gynecological fistulas, according to a retrospective national cohort study of women treated for gynecological fistula, whereas urogenital fistulas are the prerogative of post-hysterectomy women [5]. Formation of a fistula to the recto-sigmoid colon, is the most common site [2], accounting for 40% of all involved organs as described in KIZAKI and al. review [6]. This complication affects the prognosis. Indeed, the fistulous communication in the digestive lumen leads to the overflow of it microbial deposit. The tumor therefore becomes superinfected and may result in pelvic peritonitis in case of secondary rupture. Surgical treatment therefore becomes unavoidable. On the other hand, the patient is deprived of the benefit of undergoing neoadjuvant chemotherapy, which will decrease the chances of a complete macroscopic cytoreduction. In addition, the operative procedure becomes more laborious: the patient will require segmental colonic resection during the same narcosis. Trimming then suturing the fistula orifice by the colonic side seems to us inadequate because it is necessary to perform a monobloc resection of the entire conglomerate mass in order to comply with the carcinological constraints. Moreover, the disconnection of the fistula may induce a spillover of tumor clones into the large peritoneal cavity which will inevitably evolve into a peritoneal miliary. The other issue is the potential to leak stercoral and bowel flora contents into the peritoneal cavity. A primary anastomosis is required in clement conditions. In case of uncompensated hemodynamic state or peritoneal pyo-stercoral diffusion, transformation into a terminal stoma is the rule. Spontaneous resolution can be awaited with bowel discharge of pus as reported by KAZUOIMAMURA and al. [7], the patient reconsulted only after 46 years after a recrudescence of her symptoms. We cannot count on this eventuality while risking the expansion of suppuration and peritoneal neoplastic graft, without taking into account the overestimation of hospital costs dilapidated in the control. The case reported by SHAI et al. [8] demonstrates that a conservative treatment could be undertaken in order to avoid an extensive resecting surgery and to permit a neoadjuvant treatment but this eventuality will be validated only if the fistulous track has closed spontaneously thus requiring a simple separation of the two abutting viscera. But evidence of fistulous communication may be difficult to obtain with the usual imaging methods. Rectal contrast is not recommended during the period of acuitization, as it can repermeabilize or enlarge the visceral communication. Thus, it seems that this attitude could bring more risk than it offers safety. It is true that fistulization does not always imply a tumor invasion, the chronic inflammation inflicted by the iterative twisting attacks or chronic pressure causes this fistulous communication as it was described by [2]. This is explained by the fact that the dermoid cyst, which for a long time was proclaimed to be at risk of torsion because of its large size [9] occupies the first rank of the histological type involved in oophoro-sigmoid fistula. But, considering the high probability of cancerous invasion causing fistulization, reaching up to 38% according to the review of KIZAKI and al. [6], en-bloc resection is imperative in order not to leave a residual neoplastic niche. Finally, considering that concomitant colo-rectal resection doesn’t worsen the prognosis, an optimal resection is the mainstray of available therapeutic tools. This is supported by the review by FOURNIER [10] who recommend extensive resection which has little influence on post-operative morbidity but significantly improves survival.