Discussion
This large nationwide study showed that a conservative approach is possible in OMT, with 60% of patients who kept at least one ovary. The incidence of recurrence was low at 8.3% and was not influenced by the type of initial surgical approach or the type of tumor resection. Risk factors for recurrence were intrinsic to the tumor features, such as a large palpable mass and bilateral synchronous forms at diagnosis.
The clinical presentation of OMT in our series did not differ from the literature (1, 3, 4, 14), with the most frequent revealing symptom being abdominal pain in 61% of cases. One-third of patients underwent emergency surgery and revealed adnexal torsion in 40% of cases, which represents 25% of the whole cohort. This incidence is relatively high compared to the 3% to 16% of cases in cohorts including both adults and children in the literature (6-8, 15, 20, 21). A palpable mass was present in 27% of cases in our cohort. Data on the exact size of the lesion was not available for all patients, and the retrospective nature of our study precludes any analysis of the risk of recurrence according to tumor size. Nevertheless, the association of a palpable mass with a risk of recurrence as an independent variable suggests that the size of the teratoma may influence the risk of recurrence.
Elevations in AFP and HCG are considered features of a malignant GCT, but in a case of OMT, these tumor markers are usually normal (14, 15). An elevation in AFP was observed in five cases in our series and may be related to the presence of a small immature teratoma or a differentiated liver component (3, 23). This possibility prompts avoidance of any spillage during teratoma surgery because of the possible association with an immature component, which was not found at the preoperative staging.
Our experience showed, as others, that laparoscopy and laparotomy can both be safely practiced for OMT surgery without any influence on OMT recurrence (19, 24). Laparoscopy has the advantages of decreasing postoperative pain, complications, and length of hospitalization (6, 8, 20, 22, 25). It also minimizes postoperative adhesions, which may have an impact on fertility (6, 7, 24). All of these benefits promote laparoscopy as the first surgical approach in adults (6, 20, 21). However, more tumor rupture and spillage are described during laparoscopy (6, 26), and this risk is increased when OSS is considered.
In our study, laparotomy with or without preceding laparoscopy was performed in 83% of cases and comprised OSS in 58% of these cases. Laparoscopy was performed only in 15% of cases, precluding any analysis of the influence of the type of surgery on the second event in this group. In the whole cohort, however, OSS did not increase the risk of a second event. In the literature, unilateral oophorectomy seems to have no impact on puberty development (27) but could advance menopause by 1.1 to 1.8 years with the health consequences of estrogen deprivation (16, 17, 28, 29). Bilateral forms and second events, reported in 5% and 8.3% of cases in our cohort, may lead to additional amputation of the ovarian tissue. All these features favor conservative surgery in OMT to preserve the ovarian tissue and reserve as much as possible (8, 11, 14, 25, 30). In our series, all patients maintained at least one ovary even after recurrence. Additionally, fertility preservation should be contemplated in cases of bilateral forms at diagnosis or of a second event after surgery (31). In this case of benign indications at the postpubertal stage, oocyte cryopreservation is still the highest-yield strategy to offer (31). Cryopreservation of ovarian tissue is also an option and remains the only one before the onset of puberty (31-33).
The incidence of metachronous OMT was 8.3% in our cohort, slightly lower than the 10 to 20% reported in the literature (6, 8, 17, 19). Recurrence was observed mainly on the contralateral ovary (62%) or on both ovaries (10%). The second tumor occurred on the same ovary in only 28% of recurrent cases, an observation already made in other reports (6, 8, 17). This suggests that metachronous OMT is more a new event than a local recurrence of the primary tumor. This hypothesis is supported by DNA profile analysis of OMT at diagnosis and after relapse (34) and PCR analysis of a case with synchronous bilateral OMT (2). Multiple synchronous forms and the observation in our study that bilateral forms are at risk of a second event may suggest a genetic origin in OMT (6). Data on the presence of a familial history was retrieved in only three percent of cases in our cohort. This incidence is certainly higher, as the questioning for familial history is not systematic and information on this is difficult to retrieve. To date, no specific germinal gene pathogenic variant background has ever been documented in such families, as in our cohort (15, 35, 36).
Risk factors for a second event related in our study were also the occurrence of a perioperative rupture. As most of the recurrences were on the contralateral ovary, it is difficult to interpret and could be biased by confounding factors such as the size of the lesion, although no significant association was observed between perioperative rupture and palpable mass. Other studies did not find perioperative rupture to be a risk factor for recurrence in an adult population (7, 37).
As the EFS curve did not show any plateau, it suggests the need for prolonged surveillance after surgery (18). Considering the median time to a second event of 30.5 months and the slow growth rate of this benign tumor, we propose a pelvic and abdominal US exam every 6 months during the five years following initial surgery. Patients and family should be informed of the risk of recurrence and of the importance of consulting in cases of pelvic symptoms. These patients have amputation of their ovarian capital and may also present adnexal torsion on a functional cyst or tumor recurrence, even years after the initial event. They must therefore be considered at risk for premature ovarian failure.
In conclusion, despite the limitations linked to the retrospective nature of our analysis, these results support a conservative approach for OMTs in children. The rationale for surgery is the risk of adnexal torsion and the natural growth of the teratoma, which may preclude any possibility of OSS (6, 17, 20). In an emergency, laparoscopy is a good option, particularly when adnexal torsion is observed; simple detorsion is recommended with a second-step resection surgery. When planned resection is performed, OSS should be preferred, and laparoscopy should be considered only if there is no doubt on the benign nature of the lesion. If this is not the case, TO should be performed preferentially with a laparotomy preceded by a laparoscopic exploration of the peritoneum, and peritoneal fluid sending for analysis. Laparoscopy should be proposed only if the removal of the tumor can be performed without rupture, avoiding morcellation of the specimen to allow proper pathological analysis of the tumor (38). Finally, the observation of multiple bilateral lesions and/or familial history as well as the contralateral side of most metachronous lesions suggest a genetic predisposing background and invite wider genetic investigations in pediatric OMTs.