Laparoscopy for Ovary-Sparing Tumorectomy in Children with Ovarian Tumors: A Clinical Retrospective Analysis

Introduction

Ovarian tumors (OTs) in children are rare, with 60%–80% of them being germ cell tumors.^1–4 In contrast, 60%–80% of OTs in adults are epithelial cell tumors. The incidence of OTs in the pediatric population is 2.2–2.6:100,000 girls per year.^2–7 Fortunately, malignant OTs compose only 1%–2% of OTs in children.^3,4,6,8,9 Given the heterogeneity of biological behavior, excellent prognosis, and physiological developmental characteristics, management of OTs in children differs from that in adults. Importantly, preservation of reproductive and ovarian reserve capacity and future potential contralateral ovarian lesions must be considered as much as possible.

Although the benign and malignant tendencies of tumors can be preliminarily judged by preoperative clinical characteristics, there are still many staging factors that are only based on intraoperative findings and frozen pathological sections. These factors affect the decision to perform ovary-sparing tumorectomy (OST), which has become increasingly standardized over the past decade and proven to be safe for children with OTs. Notably, laparoscopic OST has become increasingly common in the recent decades. However, there are still some knowledge gaps in the literature.

This study aimed to report our experience by retrospectively analyzing the clinicopathological characteristics of OTs in children as well as the outcomes of different surgical management strategies.

Materials and Methods

All cases in this retrospective study were obtained from our hospital database, from which all cases of ovarian lesions in hospital admission records were screened from October 2011 to December 2021. All non-neoplastic ovarian lesions and nonsurgical cases were excluded. Children under 18 years of age with OTs were retrospectively identified and analyzed. The data extracted from computer-based patient records included demographic data, clinical characteristics, serum tumor markers, OT size, surgical approach, ovarian surgical procedure, intraoperative tumor characteristics, postoperative complications, pathological diagnosis, and postoperative chemotherapy data.

The onset time refers to the time from initial symptoms to surgical intervention. Asymptomatic refers to incidental findings of a palpable abdominal mass or B-ultrasonography with slight abdominal discomfort. OT markers include alpha-fetoprotein (AFP), cancer antigen (CA) 125, carcinoembryonic antigen (CEA), CA19-9, CA15-3, human epididymal protein 4 (HE4), and β-human chorionic gonadotropin (β-HCG).

Some studies found that tumor size was a useful preoperative predictor of malignancy in pediatric patients; a diameter of 9.0 cm or greater was a risk factor for malignancy of a solid tumor, as well as a cutoff of 184 mL with 100% sensitivity and 54% specificity. ⁹ Therefore, tumor size was selected as a variable for analysis. In addition, Lawrence et al ⁷ reported that an ovarian mass size >5 cm had 2.1 times the odds of developing torsion relative to a smaller ovarian mass size.

Considering the irregular ellipsoid shape of ovarian tumors, we introduced a volume variable that had not been previously reported. The tumor size was extracted from B-ultrasonography reports, sometimes followed by computed tomography (CT) or magnetic resonance imaging (MRI) data. These data were compared with the surgical records to ensure accuracy. The volume of OTs was calculated using the formula for a prolate ellipsoid (0.52 × maximal longitudinal × anteroposterior × transverse diameters) according to three-dimensional data. ⁹ The surgical approaches included laparotomy and laparoscopy.

Clinical staging was based on the Pediatric Oncology Group (POG)/Children's Cancer Group (CCG) guidelines. Ovarian surgical procedures included ovary-sparing tumorectomy, oophorectomy, and salpingo-oophorectomy. The endpoint and follow-up time were obtained by comparison with outpatient and inpatient databases. Patients were considered lost to follow-up if no relevant information was available.

All ovarian procedures were performed by 6 veteran surgeons, including 5 gynecologists and 1 pediatric surgeon. The earliest laparoscopic OST for a child with a benign tumor was reported by Garcia et al^10,11 in 1996. Four-port and puncture decompression techniques were used. Monopolar electrocoagulation was used to stop bleeding. The edge of the capsule was not sewn or closed using a clip. We usually use the three- to four-port techniques. For younger children or those with larger tumors, a trocar was implanted through the umbilicus under direct visualization.

Puncture with a trocar was used for cystic decompression. Larger tumors accompanied by abdominal wall tension were decompressed through a small incision in the lower abdomen or gradually brought out of the abdominal cavity for OST with the assistance of laparoscopy. The remaining ovarian edge after laparotomy is usually sutured, but the ovarian edge after laparoscopic surgery may or may not be sutured, which is preferred by the operator, except for borderline and malignant tumors.

The study protocol was approved by the institutional review board of Weihai Municipal Hospital. The need for informed consent was waived owing to the retrospective nature of the study.

All data handling and statistical analyses were performed using SPSS 23.0 (Statistical Package for the Social Sciences). Continuous measurements are presented as mean ± standard deviation, and an independent Student's t-test was used for continuous variables. Categorical variables are summarized as numbers and percentages and compared using chi-squared and Fisher's exact tests. The cutoff was calculated using the receiver operating characteristic (ROC) curve and Excel. The statistical significance level was set at P ≤ .05.

Results

A total of 1658 patients (age range: 6–84 years) were managed for OTs between October 2011 and December 2021. Of these, 1239 (74.7%) had benign tumors, 98 (5.9%) had borderline tumors, and 321 (19.4%) had malignant tumors. Of the 1658 patients, 61 (3.7%) patients totaling 73 lesions and under 18 years of age with OTs were screened out. Of these, 54 (88.5%) patients had benign tumors, 2 (3.3%) had borderline tumors, and 5 (8.2%) had malignant tumors.

Clinical characteristics of patients are summarized in Table 1. The median age at the time of surgery was 14.8 ± 3.0 years. The primary presenting symptoms were abdominal pain (62.5%), nausea/vomiting (20.3%), continuous abdominal pain with nausea/vomiting (20.3%), and asymptomatic symptoms (37.5%). Among the tumor markers, there was no positive expression of HE4 or β-HCG.

There were significant differences in AFP (P = .001), CEA (P = .006), and CA125 (P = .002) levels between benign and malignant tumors. In total, 33 patients underwent serum CA199 tests, for which there were no significant differences between benign and malignant tumors. The median longest diameter and median volume of borderline and malignant OTs were larger than those of benign OTs (P < .001 and P = .05, respectively).

The area under the ROC curve for the longest diameter was 0.86 (95% confidence interval [CI]: 0.74–0.98, P = .002). A cutoff of 10 cm (100% sensitivity and 54% specificity) ruled out malignancy and a cutoff of 12 cm had 86% sensitivity and 79% specificity. The area under the ROC curve for volume was 0.85 (95% CI: 0.71–0.99, P = .002). A cutoff of 174 cm³ (100% sensitivity and 55% specificity) ruled out malignancy and a cutoff of 755 cm³ had 71% sensitivity and 88% specificity.

OT torsion occurred in 18 (24.7%) of the 73 lesions, of which 16 were treated with OST. A total of 8 lesions (44.4%) were treated with laparoscopic OST. The median volume of torsion OTs was 327.2 ± 218.9 cm³ and that of nontorsion OTs was 703.9 ± 1257.3 cm³. There was a significant difference in the median OT volume between torsion and nontorsion OTs (P = .04). The median length of torsion OTs was 9.7 ± 2.6 cm and that of nontorsion OTs was 10.7 ± 7.6 cm. There was no significant difference in the median length between torsion and nontorsion OTs (P = .39). No OT torsion was caused by borderline or malignant OTs.

Overall, 37 (57.8%) patients underwent laparoscopy, of which 1 patient was subsequently converted to laparotomy (Table 2). No significant difference was observed in the choice of emergency surgery between the laparoscopy and laparotomy groups. The overall OST rate was 91.8% (67/73). Among all laparoscopic surgeries, 95.1% (39/41) of the lesions were treated with OST and 4.9% (2/41) with salpingo-oophorectomy; of these, one case had a borderline tumor complicated by suppurative infection and rupture (stage II) and the other had serous cystadenoma necrosis due to torsion.

Among all laparotomy surgeries, 87.6% (28/32) of the lesions were treated with OST and four lesions with salpingo-oophorectomy or oophorectomy; of these, three cases were malignant tumors and the fourth case was mature teratoma necrosis due to torsion. There was no difference in the OST success rates between laparotomy and laparoscopy (P = .45). There was a significant difference in the clinical type, with cystic tumors accounting for 68.4% (26/38) in laparoscopic surgery and 20.0% (7/35) in laparotomy surgery (P < .001).

The probability of intraoperative tumor rupture or spillage was higher with laparoscopy than with laparotomy (P = .02). The OT volume was significantly smaller in patients who underwent laparoscopy than in those who underwent laparotomy (P = .04). The mean operation time was significantly shorter in patients who underwent laparoscopy than in those who underwent laparotomy (P = .04).

Laparoscopic procedures for children accounted for 42.3% (11/26) of ovarian surgeries 6 years ago and have accounted for 68.4% (26/38) over the last 5 years. None of the patients experienced serious postoperative complications.

Seven patients had borderline and malignant tumors, 3 of whom had stage IA tumors and underwent OST through laparotomy at the request of the parents. One patient had a mucinous borderline tumor without rupture; one had an immature teratoma without rupture, which was treated with postoperative chemotherapy; and one had a granulosa cell tumor without rupture, which was treated with postoperative chemotherapy.

The other four cases were serous borderline tumor with spontaneous rupture (II), immature teratoma without rupture (III), yolk sac tumor without rupture (IA), and yolk sac tumor with teratoma with rupture (III). All 4 patients underwent oophorectomy or salpingo-oophorectomy and postoperative chemotherapy, except for the patient with serous borderline tumor (II). No wedge biopsy of the contralateral ovary was performed in any patients. None of the patients relapsed according to the study endpoints. The median follow-up time was 61.7 ± 54.3 months (range: 8–125 months).

All patients were followed up, with an average follow-up time of 39.2 ± 33.4 months. The recurrence-free survival rate was 95%. Only one of these patients with mature teratoma experienced three recurrences bilaterally within 10 years, twice on the right side and once on the left side, and OST was performed each time.

The first surgery was a laparoscopic OST for a right mature ovarian teratoma. The second surgery was an open OST for a left metachronous mature ovarian teratoma and right recurrent mature ovarian teratoma 41 months after the first surgery. The third surgery was an open OST for a left recurrent mature ovarian teratoma 12 months after the second surgery. The fourth surgery was a laparoscopic OST for a right recurrent mature ovarian teratoma 71 months after the third surgery. The patient could also be regarded as having a metachronous tumor (1.6%).

Discussion

OTs are uncommon in children. In our study, OTs in children under 18 years of age accounted for 3.7% of all OTs, among which borderline and malignant tumors were rare. Previous studies have reported that the rate of mature teratomas is approximately 55–80%,^4,6,11–13 which is consistent with our finding that mature teratomas are common among the benign tumors in children (60.7%).

It is important for children to undergo OST because ovariectomy increases the risk of occasional castration and hormonal health. The risk factors include subsequent tumor lesions of the contralateral ovary or ovarian torsion with no lesions.^12,14,15 Furthermore, the risk of developing a second neoplasm (benign or malignant) in the contralateral ovary in children with OTs is 10%–23%.^9,16,17

Losing one ovary results in a 32% decrease in the odds of successful assisted reproduction and a more than threefold increase in the risk of premature ovarian failure as well as early menopause.^14,17 The first step in increasing the success rate of OST is to perform an accurate preoperative risk assessment based on medical history, imaging, and tumor markers.

Abdominal pain in children with OTs has been found to account for 49%–75% of cases,^3,4,18–21 whereas nausea and vomiting have been found to be less frequent presenting symptoms, except in OT cases with torsion. ³ Asymptomatic presentation accounts for 25%–60% of cases.^2,19 However, in our study, abdominal pain and vomiting were poor predictors for distinguishing between benign and malignant lesions.

While multiple tumor markers are useful adjuncts in the preoperative diagnosis of malignancy, the absence of elevated tumor markers does not exclude malignancy as only approximately 50% of malignant tumors present with elevated values. Additionally, the levels of serum tumor markers may be affected by age, anatomical site, histological type, and clinical stage of tumors. ²² Nevertheless, markers associated with OTs are still highly specific; at least one marker is positive in up to 54%–83% of malignant lesions and varies from 3% to 21% in benign lesions.^{6,16,18,19,23} In this study, tumor markers (AFP, CEA, and CA125) were significant in the differential diagnosis of benign and malignant tumors.

Tumor size is a helpful preoperative predictor of malignancy in pediatric OTs, especially in those with solid components and elevated tumor markers.^16,19,23 Tumor size greater than 10 cm is considered a high-risk factor for malignant OTs.^9,19,24 For example, Abbas et al ²⁵ reported OT volume as a predictor of malignancy, in which a threshold volume of 184 cm³ (100% sensitivity and 54% specificity) ruled out malignancy.

In our study, tumor size was a significant factor in the differential diagnosis of benign and malignant tumors in terms of length and volume. The length and volume thresholds of OTs were consistent with those reported previously. Therefore, more attention should be paid to the differentiation tendency of malignant OTs with length >10 cm and volume >180 cm³ before choosing the laparoscopic approach.

Ovarian torsion is common in children with OTs, occurring in 22%–51% of cases.^{6–8,12,18–20} Notably, 86% of OTs with acute symptoms have ovarian torsion. ²⁰ Previous studies have reported that OTs with torsion are significantly smaller than those without torsion.^26,27 Torsion often occurs in tumors ranging from 5 to 10 cm in size. Importantly, torsion associated with malignancy is rare, with the ratio varying from 1% to 6%.^7,8,18,19,27

The American College of Obstetricians and Gynecologists ²⁸ recommends that a surgeon should not remove a torsion ovary unless oophorectomy is unavoidable, such as when a severely necrotic ovary falls apart. As such, timely intervention with diagnostic laparoscopy is indicated to preserve ovarian function. In our study, ovarian torsion accounted for 24.7% of cases, and while the volume of the torsion tumor was smaller than that of the nontorsion tumor, the maximum length did not reflect this difference.

The reason for this is unclear; however, the shape and weight of the OT may be relevant factors. Notably, the rate of ovarian preservation with torsion was 88.9% and only 44.4% of the lesions were treated laparoscopically. Considering that the size of a typical torsion OT is usually about 10 cm and it is rarely a malignant tumor, laparoscopic OST is safe in an emergency.

In recent decades, OST has been recommended as the first-line treatment for pediatric OTs, whenever possible, and has been proven safe for children.^13,26,29,30 While it is generally agreed upon that OST should be performed as the primary treatment for benign ovarian lesions that have been identified preoperatively,^6,9,16,26 there are still disputes regarding the best primary treatment for borderline and malignant stage I tumors.^2,9,18,26

Fears of OT recurrence and upstaging have prevented the widespread acceptance of OST in their initial management. However, several recent studies have demonstrated that OST results in acceptable recurrence rates for both benign and malignant lesions, which differ from the high recurrence rate in adults.^16,29,31 The rate of OST has been found to vary from 22% to 77%.^{7,15,16,18,23,27} Although the risk of recurrence is higher after OST for borderline OTs, recurrence may be salvaged without any effect on overall survival.

In fact, evidence-based medicine recommends that OST should be considered under close follow-up. ⁹ However, there is limited literature evaluating OST for immature or malignant germ cell tumors and it is not considered standard management, even when utilizing platinum-based chemotherapy. ⁹ In contrast, unilateral salpingo-oophorectomy or oophorectomy without postoperative chemotherapy, regardless of the grade for stage I, has been the standard surgical treatment for immature teratomas.^9,26,32

Yet, many studies have reported that OST has been performed for immature teratomas with or without adjuvant chemotherapy and has an acceptable outcome.^2,9,18,26,30 For example, OST has been performed on stage I ovarian juvenile granulosa cell tumors without chemotherapy and no recurrence has been observed. ³³ In our study, the rate of OST was up to 91.8%.

Furthermore, three patients with borderline and malignant OTs (stage IA) underwent OST, of whom two patients with malignant OTs received postoperative chemotherapy and had no recurrence. Nonetheless, OST for malignant tumors still lacks support from evidence-based medicine, therefore it should be strictly staged and carefully selected.

Laparoscopic OST has become increasingly common in recent decades, but it cannot completely replace laparotomy. Laparoscopy is the standard approach for benign OTs and its use in early-stage malignant tumors is currently increasing.^4,20,29,33 However, laparoscopy has limitations in terms of tumor size, which is smaller when treated with laparoscopy than with laparotomy.^6,12,34 For example, larger tumors limit the operative space within the abdominal cavity.

Additionally, tumor rupture and intraoperative spillage are the main complications of laparoscopy, which raises the question of whether they increase the rate of tumor recurrence. Some reports have demonstrated that intraoperative spillage was not associated with recurrence.^{11,13,29,31,34,35} In our study, OT size, histopathological type, and cystic and solid lesions were related factors for the laparoscopic approach. Notably, laparoscopic surgery is more suitable for cystic lesions.

Although we found that the intraoperative rupture or spillage rate of OTs in laparoscopy was higher than that in laparotomy, it did not affect the outcomes of OST. Among the 7 patients with borderline and malignant tumors, only 1 underwent laparoscopic exploration due to suppurative infection and rupture of the OT. Therefore, laparotomy remains the first choice of treatment for OTs with suspected malignancy.

Regardless of grade, children with stage I disease (POG/CCG) can be cured only by complete ovarian resection without chemotherapy.^2,24,32,33 It remains unclear whether adjuvant chemotherapy is needed after OST for stage I tumors, although OST for stage I tumors without chemotherapy has been reported to have an excellent outcome.^2,24,33

Notably, OST is sometimes performed for borderline or malignant tumors because of an uncertain, intraoperative, frozen pathological condition or at the request of parents for ovarian sparing. In these cases, it is important to decide whether to perform a secondary surgery, close observation, or adjuvant chemotherapy according to the pathology type, grade, and stage of the OTs. In our study, two patients with malignant OTs were administered postoperative adjuvant chemotherapy after OST and neither relapsed.

Good outcomes have been reported in patients with both malignant and benign tumors who underwent ovary-sparing tumor resection or oophorectomy for pediatric OTs.^5,21,30 In fact, the 5-year, event-free survival rate has been reported to be 91%, ³³ and the recurrence rate after OST ranges from 3% to 13%.^13,20,21,34 Furthermore, previous studies have reported no significant differences between oophorectomy and OST in terms of the recurrence rate.^31,34 In our study, all patients were followed up after surgery, and only 1 patient with a mature teratoma had multiple recurrences that were treated with OST each time.

This study was limited by the nature of its retrospective design, single-center study, and few malignant patients, which may have an impact on the statistical results. Finally, some patients lacked follow-up data beyond 3 years.

Conclusions

OT size is a useful reference factor for differential diagnosis and choosing laparoscopic surgery. Intraoperative tumor rupture and spillage of benign tumors during laparoscopy and laparotomy did not seem to be associated with recurrence, and laparoscopic OST was considered safe. Further prospective studies are required to confirm these conclusions.