INTRODUCTION

Ovarian diseases in children and adolescents rarely require surgical treatment, since most lesions are functional and do not require intervention ¹ . The ovary is responsible for puberty, reproductive function and also optimizes peak bone mass ² . Therefore, it is of great importance and, whenever possible, should be preserved. Usually, oophorectomy in young patients is indicated in cases of malignant ovarian tumors or adnexal torsion with necrosis.

Ovarian lesions are represented by neoplastic changes (benign and malignant tumors) or non-neoplastic changes (functional cysts ³ , adnexal torsion, endometriomas and infectious processes). Although ovarian tumors are the most common gynecological tumor in childhood and adolescence, they are still quite rare in women under 21 years of age ⁴ . Fortunately, when malignant and adequately treated (conservative surgery and combined chemotherapy), these tumors have low morbidity and a good fertility rate ⁵ . The greater availability of ultrasound has led to an increase in the number of cysts detected in children, most of which are small and functional (1-3 cm) ⁶ .

Ovarian torsion is an uncommon condition in pediatric patients ⁷ , but it should be diagnosed early. Initially, lymphatic and venous congestion occurs, which may be followed by arterial ischemia, often culminating in tissue necrosis ⁸ . However, it is important to note that the bluish appearance resulting from venous stasis should not be confused with necrosis, and in these cases the ovary can be preserved.

The objective of this study is to understand the causes of oophorectomies, oophoroplasties and ovarian biopsies performed in children and adolescents at the Santa Casa de Misericórdia Hospital Complex in Porto Alegre.

METHODS

A descriptive and retrospective cross-sectional study was carried out with female patients aged 0 to 19 years who underwent ovarian surgery between January 2006 and December 2016 at the Santa Casa de Misericórdia Hospital Complex in Porto Alegre. Data were obtained through review of the anatomopathological reports from the Pathology Service of the aforementioned complex and also through analysis of medical records.

Quantitative variables were described using mean and standard deviation or median and interquartile range, and qualitative variables were described using absolute and relative frequencies. Pearson’s chi-square or Fisher’s exact tests were used to assess associations between categorical variables. The Mann-Whitney or Kruskal-Wallis tests supplemented by Dunn were used to compare medians. The significance level adopted was 5% (p ≤ 0.05) and analyses were performed using SPSS version 21.0. The study was duly approved by the Research Ethics Committee of Irmandade Santa Casa de Misericórdia de Porto Alegre under Opinion No. 2,297,918.

According to the World Health Organization (WHO) classification, patients were classified as children (0 to 9 years) and adolescents (10 to 19 years).

RESULTS

The sample consisted of 117 patients, with a mean age of 11.78 years (SD ± 5.56 years) operated by different professionals. The age group with the highest prevalence of ovarian surgical procedures was that of newborns and after six years of age (Figure 1).

Regarding the procedure performed, the majority (75.2%) underwent oophorectomy, since many surgeons opt for this procedure when there is an adnexal mass, followed by oophoroplasty (24.8%).

There was no difference regarding the affected ovarian side: 59 patients (50.4%) underwent the procedure on the left ovary and 58 patients (49.6%) on the right ovary (p=0.926). Forty-two girls (35.9%) underwent salpingectomy at the same surgical time, 88.1% due to benign causes. Of these, only five (11.9%) patients had anatomopathological findings compatible with malignant lesions: two cases of sex cord tumor, one immature teratoma, one lymphoblastic leukemia and one choriocarcinoma.

Thirty patients (25.6%) presented ovarian torsion and all reported abdominal pain. There was no association between the age of the patients and the occurrence of ovarian torsion (p=0.48), nor association between the ovarian volume described in the imaging exams and the diagnosis of ovarian torsion (p=0.869). Of these 30 patients, 18 (60%) had a history of abdominal pain and vomiting prior to diagnosis. Of this same group, 16 (53.4%) patients presented torsion of the adnexa without underlying lesion, where the mean age of the patients was 10 years (SD±5.4 years). The remaining 14 (46.6%) patients presented torsion with ovarian disease and had an average age of 12.5 years (SD±4.8 years), presenting the following underlying ovarian lesions: six mature teratomas, three hemorrhagic cysts, two simple serous cysts, one follicular cyst, one cystadenofibroma and one fibrothecoma. There was no statistically significant difference between the two groups.

Regarding the anatomopathological results (Table 1), the majority (31/117 = 26.5%) presented mature cystic teratoma, followed by 16 cases (13.6%) of ovarian torsion without underlying lesion, 14 cases (12%) of hemorrhagic luteal cyst and 11 cases (9.4%) of cystadenomafibroma. Sixty-four patients (54.7%) presented ovarian tumor. Nine of the 117 cases (7.7%) were malignant lesions, three of which were immature teratoma.

Forty-six anatomopathological specimens had weights described in the pathology report, ranging from 4 to 3,000 grams. The lowest weight was equivalent to a cystadenofibroma (4 g), and the highest was a serous cystadenoma (3,000 g).

Seventy-two patients (61.5%) had imaging tests listed in their medical records. The majority (46.2%) underwent ultrasound, followed by tomography (11.1%) and magnetic resonance imaging (8.5%).

Tumor marker tests were performed on only 39 (33.3%) patients: 38 patients underwent alpha-fetoprotein (AFP) tests, 32 serum HCG (human chorionic gonadotropin), 17 LDH (lactic dehydrogenase), nine Ca-125 and CEA (carcinoembryonic antigen), and only five patients had alkaline phosphatase (AP) tests. Of these patients, none had altered serum HCG, CEA or ALP. Three patients had increased Ca-125, one due to mature teratoma and two due to endometrioma, and in a 19-year-old patient the Ca-125 was 787 U/mL (reference value = up to 35 U/mL). The highest value found for LDH was 357 U/L (reference value = 100-190 U/L) in a 10-year-old girl with anatomopathological findings compatible with mature cystic teratoma. The sensitivity of the tumor markers ranged from 0 to 75% and the specificity from 23.1 to 100% (Table 2).

DISCUSSION

Currently, the literature has suggested that most ovarian cancers originate in the fallopian tube, since a gradual transformation of the epithelium would have the capacity to invade and metastasize within the fallopian tube itself and reach the ovary ⁹ . However, it is still unknown whether salpingectomy will lead to a reduction in mortality from ovarian cancer ⁹ . In our study, 42 (35.9%) girls underwent salpingectomy at the same surgical time. Of these, only five (11.9%) had cancer and six of these (14.5%) were due to torsion. While prophylactic salpingectomy could contribute to the reduction of ovarian cancer, on the other hand, it would increase cases of infertility. Therefore, there is still a need for further discussion in the literature about the best approach in cases where it is necessary to remove the ovary in patients without complete offspring: whether or not to remove the ipsilateral fallopian tube at the same time.

According to the National Comprehensive Cancer Network (NCCN Guideline 2017) ¹⁰ , patients with tumors limited to one or both ovaries and who have an intact capsule (IA and IB respectively) do not necessarily need to undergo chemotherapy. However, even if the tumor is limited to the ovary, rupture of the capsule (as is possible in the case of an ovarian biopsy, for example) automatically changes the stage to IC (tumor limited to one or both ovaries with any of the following: rupture of the capsule, tumor on the surface of the ovary or malignant cells with ascites), which will result in the need for adjuvant chemotherapy and surgery ¹⁰ . Therefore, it is necessary to consider the advantages and risks of performing biopsies on ovaries that may contain a malignant tumor. Many services opt for oophorectomy in these cases, but prior discussion with the patient and her family is necessary regarding the risk/benefit of a biopsy vs. cystectomy vs. oophorectomy.

Ovarian torsion is a relatively rare event in the pediatric age group. ¹¹ It is an emergency condition that affects 4.9/100,000 women between 1 and 20 years of age, with most cases being accompanied by an ovarian mass or cyst ^1,11 . However, in childhood, torsion in an ovary without an underlying lesion is more common due to the greater length of the utero-ovarian ligament ¹² . In our study, the age difference between pure ovarian torsion and those with an underlying lesion was not statistically significant, although we observed a tendency for pure ovarian torsion to affect younger patients (10 years, SD ± 5.4 years). The same was true in the study by Karaman et al. ¹²where, of the 29 girls with ovarian torsion, eight (27.5%) had normal ovaries and a mean age of 13 years, and 21 (72.5%) had an adjacent mass or cyst and a mean age of 14 years. Furthermore, in the study by Jourjon et al. ¹³ , which evaluated 65 girls with ovarian torsion, 60.6% had torsion with an adjacent mass, which shows a higher proportion of torsion in ovaries with lesions when compared to a previous study that showed 27% ¹⁴ .

In our study, all patients with ovarian torsion (30/117) reported abdominal pain (p < 0.001). Of these, the majority (60%) presented vomiting, which was similar to the study by Karaman et al. ¹² where the most common symptom was abdominal pain with most cases presenting nausea and vomiting. Appelbaum et al. ¹⁵ state that vomiting is commonly associated with torsion in several studies of children and adolescents, agreeing with Jourjon et al. ¹³ who found nausea and vomiting as the only statistically significant clinical predictor of torsion.

There was no association between ovarian volume described in the imaging exams of our study and the diagnosis of ovarian torsion (p=0.869). This result contrasts with that found by Jourjon et al. ¹³ where the size of the mass predicts torsion because this occurred more frequently with lesions ≥ 5 cm. The finding of this study is supported by the literature since it was also found that tumors > 5 cm in patients older than 1 year are likely to twist, with a sensitivity of 83% ¹⁶ .

Ovarian tumors constitute 1% to 2% of all tumors reported in children and adolescents ⁵ . In this same age group, tumors of the female genital tract are rare, but among them, ovarian cancer is the most frequent, accounting for 1% of all childhood cancers ¹¹ . In our study, 64 (54.7%) patients had ovarian tumors, of which the most common type was germ cell tumors represented by mature teratomas in 31 cases (26.5%). Germ cell tumors are the most common ovarian neoplasm in childhood and adolescence, with mature cystic teratomas accounting for 55-70% of cases ¹⁷ . Al Jama et al ¹⁸ evaluated 52 patients, aged between 6 and 20 years, diagnosed with ovarian tumors and found that 87% were germ cell tumors. In addition, Akakpo et al ⁵ published a study with 706 patients diagnosed with ovarian tumors, of which 67 cases (9.5%) were in patients aged 0 to 19 years. Of these, 53.7% had mature teratomas.

In our sample, of the 64 patients with tumors, 45 (70.3%) were adolescents. This was similar to the study by Akakpo et al. ⁵ who reported that 48/67 patients (71.6%) were over nine years of age. In our study, nine (7.7%) patients had malignant tumors, a result similar to that of Piippo et al. ¹⁹ , who analyzed the treatment performed on 79 girls under 17 years of age who had ovarian masses and found that in seven (8.8%) of them the anatomopathological examination was compatible with malignant tumors.

The evaluation of tumor markers can increase the accuracy in the differential diagnosis of pediatric ovarian pathology, but their role is still controversial because they present false positive and false negative results ³ . Although they are not validated, tumor markers commonly evaluated in the pediatric population include AFP, HCG, LDH and inhibin A and B ²⁰ . When there is a high suspicion of malignancy, the use of these markers can facilitate preoperative planning and, in addition, assist in monitoring the patient with evidence of remission in the postoperative period ²⁰ .

In our study, 39 (33.3%) patients had tumor markers measured and they were elevated in 16 cases (41%): four with malignant ovarian tumors and 12 with benign tumors. Eleven girls with cancer had negative tumor markers. We can compare these results with those of Spinelli et al. who evaluated 130 children and adolescents with ovarian lesions, 110 of whom had serum levels of tumor markers (Ca-125, AFP, HCG and CA-19.9). Of these, 18 (16.4%) had elevated levels: five with malignant ovarian tumors and 13 with benign tumors. In the same study, Ca-125 was increased in seven cases of benign lesions: four mature teratomas, two serous cystadenomas and one fibroma. In our study, we had three cases of elevated Ca-125, all due to benign causes: two endometrioid cysts and one mature teratoma. Some markers (CEA, HCG, FA, AFP) were very specific (97-100%) but very insensitive (0-20%). On the other hand, in our sample, LDH had a sensitivity of 75%, but a specificity of 23%. Therefore, markers can help in decision-making, but have a greater value in follow-up when positive.

Although the vast majority of lesions in the patients analyzed were benign (92.3%) in the present study, oophorectomy was performed in 75% of cases, and this data deserves discussion. This finding is a warning for surgeons and gynecologists to discuss more precise criteria for ovarian removal in young patients, since this organ plays an important role in puberty and reproductive function. The risk of cystectomy (rupture of the capsule in a malignant lesion, bleeding) and oophorectomy for benign disease should be assessed together with the patient and her family prior to surgery.