Pediatric ovarian neoplastic tumors: incidence, age at presentation, tumor markers and outcome

Transcription

1 A C TA Obstetricia et Gynecologica AOGS MAIN RESEARCH ARTICLE Pediatric ovarian neoplastic tumors: incidence, age at presentation, tumor markers and outcome SEPPO TASKINEN 1, RIITTA FAGERHOLM 1, JOUKO LOHI 2 & MERVI TASKINEN 3 1 Department of Pediatric Surgery, Children s Hospital, University of Helsinki, 2 Department of Pathology, Haartman Institute and HUSLAB, University of Helsinki, and 3 Department of Hematology, Oncology and Stem Cell Transplantation, Children s Hospital, University of Helsinki, Helsinki, Finland Key words Ovary, tumor, pediatric, children, teratoma, tumor marker, incidence Correspondence Seppo Taskinen, Children s Hospital, Stenb ackinkatu 11, Helsinki, Finland. seppo.taskinen@hus.fi Conflict of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article. Please cite this article as: Taskinen S, Fagerholm R, Lohi J, Taskinen M. Pediatric ovarian neoplastic tumors: incidence, age at presentation, tumor markers and outcome. Acta Obstet Gynecol Scand 2015; 94: Received: 7 August 2014 Accepted: 20 January 2015 DOI: /aogs Abstract Objective. To investigate the incidence, age of onset and tumor marker levels in benign and malignant pediatric ovarian neoplastic tumors. Design. Retrospective database study. Setting. Single-center study. Population. Forty-five 0 15-year-old girls operated on for ovarian neoplastic tumors from the beginning of 1999 to the end of Methods. Serum alpha-fetoprotein, human chorionic gonadotropin and CA 125 levels as well as follow-up data were recorded from patient charts and tumor histology was re-evaluated. Main outcome measures. Incidence of ovarian neoplastic tumors in the pediatric population. Differences in patient characteristics and tumor marker levels between those with benign and malignant tumors. Results. The annual incidence of ovarian tumors was 2.2/ females. Median age at presentation was 13.0 years (range ), similar in both the 33 (73%) girls with a benign tumor and the 12 (27%) girls with a malignant tumor. The tumors with the highest propensity to metastasize (yolk sac tumors, mixed germ cell tumors, small cell carcinoma) were only found in girls > 9 years. Elevated serum alpha-fetoprotein and CA 125 values associated more often with malignant tumors (p < and p < 0.031, respectively). There were no deaths or local recurrences. Four girls with a mature teratoma developed a contralateral benign ovarian tumor during follow up. Conclusions. Both benign and malignant ovarian tumors are rare in the pediatric population, but the incidence increases with age. High alpha-fetoprotein and CA 125 levels were associated with malignant tumors. The prognosis of the pediatric ovarian tumors seems to be favorable. Abbreviation: AFP, alpha-fetoprotein. Introduction Ovarian neoplastic tumors are rare in children. More than 40 years ago the annual incidence was reported to be 2.6 cases/ girls aged 0 14 years in a study from our metropolitan area (1). Approximately half of the ovarian masses in children are neoplastic in nature and the proportion of malignant ovarian tumors is variable (16 55%) in different series, probably because of differences in the referral population, age cut-offs and inclusion of non-neoplastic cysts (2,3). The histology of Key Message The annual incidence of ovarian neoplastic tumors was 2.2/ in females younger than 16 years. One-fourth of the tumors were malignant and especially high serum alpha-fetoprotein and CA 125 levels were associated with malignancy. However, the outcome of pediatric ovarian tumors seems to be good. ª 2015 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 94 (2015)

2 Pediatric ovarian neoplastic tumors S. Taskinen et al. ovarian tumors is very variable and the most common tumors in adults and in children are different (2). The incidence of malignant tumors is known to increase with patient age and tends to be associated with higher levels of tumor markers (4,5). In this study we analyzed the incidence of ovarian neoplastic tumors in our population-based pediatric material. In addition, we evaluated differences in age of onset, levels of tumor markers and outcome between benign and malignant tumors. Material and methods Girls operated on for ovarian tumors during in the Children s Hospital of Helsinki University Central Hospital were identified from the operative database. The referral area for the hospital is Helsinki and Uusimaa districts with a total population of around 1.5 million. Only girls with neoplastic ovarian tumors were included and they were classified according to the WHO classification. Patients with non-neoplastic cysts were excluded. In principle, all girls aged 0 15 years with ovarian tumors among those living in our referral area were operated on at this hospital during the study period. The number of annual births and of girls <16 years of age living in the referral area were provided by Statistics Finland. The mean annual birth rate of female neonates was 8687 during the study period and the mean number of 0 15 yearold girls was (range ) for the inclusive whole years For the incidence calculations we used the number of girls living in the referral area during each individual year of the study period (altogether girls). The clinical characteristics were recorded from the patient files, including the preoperative serum tumor markers alpha-fetoprotein (AFP) (39 patients), human chorionic gonadotropin (37 patients), CA 125 (39 patients) and CA 19-9 (35 patients). All patients with tumor marker measurements were older than 1 year and normal laboratory ranges for tumor markers were: AFP 0 9 ku/l for patients >1 year, human chorionic gonadotropin 0 9 IU/ L, CA ku/l and CA ku/l. The histology of the tumors was re-evaluated by one of the authors (JL). A stromal tumor with uncertain malignant potential was included in the malignant group. Types of treatment are summarized in Table 1. Only girls with tumors that had a clear tendency for metastases received chemotherapy. The ethics committee of the hospital approved the study. Mann Whitney U-test was used to compare continuous variables and Fisher s exact test was used to compare categorical variables (STATVIEW â 5.0.1, SAS Institute Inc., Cary, NC, USA). Table 1. Treatment of the patients with ovarian neoplastic tumors. Type of treatment Results Benign tumors (n = 33) Malignant tumors (n = 12) Immature teratomas, sex cord stromal tumor Open surgery Laparoscopic surgery 3 a 0 0 Oophorectomy Ovary-preserving surgery b Chemotherapy Yolc sac tumors, mixed germ cell tumors, small cell carcinoma a Laparoscopic surgery was carried out only during the last 8 years in the study. In two additional cases initial laparoscopy was converted to laparotomy. Two mucinous cystadenomas had rupture during the laparoscopy. b Ovary-preserving surgery was carried out only during the last 2 years of the study. The annual incidence was 2.2 cases/ Thirty-three girls (73%) had a benign tumor and 12 (27%) had a malignant tumor. Accordingly, the annual incidence of malignant ovarian tumors was 0.6/ females. The prevalence of both benign and malignant tumors increased with patient age. All the high-grade malignant tumors (yolk sac tumors, mixed germ cell tumors and small cell carcinoma) were diagnosed in girls >9 years. Two girls (4%) had an underlying tumor-predisposing condition. Both had Turner syndrome with chromosomal 46XY/45XO mosaicism. These two girls were found to have gonadoblastomas at gonadectomy performed prophylactically at the age of 0.9 and 15.1 years without preoperative evidence of a tumor. The most prevalent tumor was benign teratoma, which was found in all age groups except for infants. Second in frequency was immature teratoma, found throughout all age groups except the infants. Mucinous and serous cystadenomas and mixed germ cell tumors were found in girls aged >13 years, as also was the single case of small cell carcinoma (Table 2). Elevated serum AFP and CA 125 levels were detected more often in malignant than in benign tumors (p < and p < 0.031, respectively, Table 3). All eight girls who had a serum AFP level > 35 U/L (3.5 times the normal upper limit) and all eight girls who had CA 125 > 93 ku/l (2.7 times the upper normal limit) had a malignant tumor (p < for both tumor markers). Either a serum AFP > 35 IU/L or CA 125 > 93 ku/l was observed in nine of 12 girls with a malignant tumor. The girls with yolk sac tumor or mixed germ cell tumor with yolk sac component had the highest serum AFP levels. CA 19-9 levels were not different between girls with 426 ª 2015 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 94 (2015)

3 S. Taskinen et al. Pediatric ovarian neoplastic tumors Table 2. Ovarian tumor histology and patient age. Tumor histology Patients n % Patient age (years) median (range) Benign tumors 33 73% 13.0 ( ) Germ cell tumors 27 60% 12.2 ( ) Mature teratoma 25 56% 12.2 ( ) Gonadoblastoma 2 4% 0.9 and 15.1 Epithelial tumors 6 13% 14.0 ( ) Mucinous cystadenoma 4 9% 13.3 ( ) Serous cystadenoma 2 4% 14.5 and 15.6 Malignant tumors or tumors 12 27% 12.9 ( ) with malignant potential Germ cell tumors 10 22% 12.9 ( Immature teratoma 5 11% 11.9 ( ) Yolk sac tumor 3 7% 14.0 ( ) Mixed germ cell tumor 2 4% 14 and 14.6 Sex cord stromal tumors Sex cord stromal tumor 1 2% 8.1 with annular tubules Epithelial tumors Small cell carcinoma 1 2% 14.2 benign or malignant tumors. Serum human chorionic gonadotropin was within the normal range in all girls. The girls with sex cord stromal tumor with annular tubules had virilizing symptoms. One girl with serum hypercalcemia at the time of diagnosis had an ovarian small cell carcinoma. All the girls are alive after a median follow-up time of 5.1 (range ) years. Metastases or local recurrences were not observed during follow up. However, four girls with mature teratoma were later diagnosed with a new benign tumor in the contralateral ovary. These cases have been published previously (6) and one of these girls has since had a second recurrence in the remaining ovary. Discussion The main finding of our study was that both benign and malignant ovarian tumors are rare in the pediatric population, but the incidence increases with age. The prognosis of these pediatric ovarian tumors was favorable and no recurrences of malignant tumors were observed during the follow up. Our results also emphasize the role of high serum tumor marker levels as an indicator of malignant tumors. We report an annual incidence of 2.2 ovarian tumors in females aged 0 15 years. A quarter of the tumors were malignant, which is in line with previous studies (1,2,7 9). The annual incidence for malignant ovarian tumors in our study (0.6/ females) was similar to the incidence reported previously in a Finnish study (0.7/ in 0 14-year-old females) (1) and an Israeli study (0.52/ in 0 19-year-old females) (10). Recently an increase in the incidence of adolescent testicular germ cell tumors has been observed in our country (11). Interestingly, the annual incidence of ovarian tumors has not changed at all in our metropolitan area over the last 40 years (1). This may suggest that environmental factors do not have a role in the development of these tumors. We observed that all the tumors that were considered high-grade malignant (yolk sac tumors, mixed germ cell tumors and small cell carcinoma) were detected in girls >9 years. Overall, the incidence of both malignant and benign ovarian tumors increased with age, while the proportion of malignant tumors remained similar in different age groups, which is comparable to a previous study (3). So far, there are no data to explain this phenomenon. Only two of the girls in this study had a predisposing condition, i.e. Turner syndrome. Table 3. Tumor marker levels in different ovarian tumors. AFP (ku/l) CA 125 (ku/l) CA 19-9 (ku/l) Tumors Patients (n) Median (range) Elev. Median (range) Elev. Median (range) Elev. Benign tumors 31 1 (1 28) 2/28 (7%) 25 (8 80) 7/27 (26%) 8 (2 84) 5/24 (21%) Mature teratoma 25 1 (1 28) 2/23 25 (8 50) 5/23 8 (2 84) 4/21 Cystadenoma 6 1 (1 2) 0/5 36 (15 80) 2/4 9 (6 58) 1/3 Malignant tumors or tumor ( ) 8/11 (73%) 135 (9 1315) 8/12 (75%) 12 (2 359) 4/11 (36%) with malignant potential Immature teratoma 5 59 (3 110) 3/4 35 (24 268) 2/5 16 (7 200) 1/4 Sex cord stromal tumors a 1 2 0/1 15 0/1 5 0/1 Yolk sac tumor ( ) 3/3 140 ( ) 3/3 6 (3 32) 1/3 Mixed germ cell tumor and /2 93 and 597 2/2 76 and 359 2/2 Small cell carcinoma b 1 1 0/ /1 2 0/1 AFP, alpha-fetoprotein; Elev., elevated. a The patient had virilizing symptoms. b The patient had hypercalcemia. ª 2015 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 94 (2015)

4 Pediatric ovarian neoplastic tumors S. Taskinen et al. Germ cell, epithelial and sex cord tumors were observed in 82, 15 and 2% of the girls, respectively. Germ cell tumors represent about 60 80% of all neoplastic ovarian tumors (2,9,12), benign teratoma being the most frequently diagnosed type in children (3,7 9). In contrast, epithelial tumors are the most common type in adults, and in our series all the girls with epithelial tumor were older than 13 years. Sex cord stromal tumors form a heterogeneous group of tumors with varying malignant potential (2,13). Virilizing symptoms, as was seen in our patient, have been reported in about half of the patients with sex cord stromal tumors (13). Seventy-five percent of the girls in this study with a malignant ovarian tumor had highly elevated serum levels of either AFP or CA 125, or both. If serum AFP or CA 125 was more than 3.5 or 2.7 times the highest reference value, respectively, the tumor was malignant. AFP values, in particular, are increased in patients with tumors including yolk sac components (5). In a recent study, high serum AFP separated benign from malignant tumors in children and adolescents with a specificity of 89% and with sensitivity of 50% (5). Specificity increased if the AFP cut-off value was increased, but did not reach 100% as in our study. Because teratomas are usually very heterogeneous in histological composition, it is also possible that immature or even yolk sac elements are missed (14). Re-evaluation of originally benign pathology specimens should be considered if very high serum AFP values are found. On re-evaluation one of our mature teratoma cases with a high serum AFP concentration turned out to be an immature teratoma. Because tumor marker levels can be high in both benign and malignant tumors, it has been recommended to perform frozen-section analysis before radical surgery and not to rely solely on tumor marker levels in decision-making (15,16). However, frozen section biopsies have limitations; as stated above, teratomas are often heterogeneous in tissue composition and immature elements or even yolk sac foci may be missed in frozen section. The outcome for our young patients was good, as all girls, both with malignant and benign tumors, are alive after a median follow up of 5 years. No local recurrences of the malignant tumors have been observed, but all six girls with high-grade malignancy received chemotherapy. In a recent report by the Children s Oncology Group from USA, half of the 25 patients with stage I malignant ovarian germ cell tumors had persistent or recurrent disease without primary chemotherapy, but the patients were rescued with chemotherapy, resulting in an overall survival of 95% (17). More studies are needed to establish the characteristics of patients with malignant germ cell tumors that appear cured with only surgery. However, four of the girls in our study with a benign tumor developed a contralateral benign neoplasm (6). Hence, we feel that pediatric patients undergoing operations for mature teratoma need follow up up to a potential child-bearing age to detect early possible contralateral recurrence in order to make ovary-preserving surgery possible. Ovarian tumors in children are a heterogeneous group and especially the incidence of the rarest tumors can vary from decade to decade in a small population. However, we consider our incidence data of the more common tumors, such as mature and immature teratomas, cystadenomas as well as yolk sac tumors and mixed germ cell tumors, to be reliable, due to the presumably close to 100% referral rate of pediatric tumors to our hospital. Excellent outcome of malignant tumors may be related to the small sample size; however, it is encouraging for the patients and the hospital personnel. Funding No specific funding. References 1. Lindfors O. Primary ovarian neoplasms in infants and children. A study of 81 cases diagnosed in Finland and Sweden. Ann Chir Gynaecol Fenn Suppl. 1971;177: von Allmen D. Malignant lesions of the ovary in childhood. Semin Pediatr Surg. 2005;14: Cass DL, Hawkins E, Brandt ML, Chintagumpala M, Bloss RS, Milewicz AL, et al. Surgery for ovarian masses in infants, children, and adolescents: 102 consecutive patients treated in a 15-year period. J Pediatr Surg. 2001;36: Brookfield KF, Cheung MC, Koniaris LG, Sola JE, Fischer AC. A population-based analysis of 1037 malignant ovarian tumors in the pediatric population. J Surg Res. 2009;156: Loh AH, Gee KW, Chua JH. Diagnostic accuracy of preoperative alpha-fetoprotein as an ovarian tumor marker in children and adolescents: not as good as we thought? Pediatr Surg Int. 2013;29: Taskinen S, Urtane A, Fagerholm R, Lohi J, Taskinen M. Metachronous benign ovarian tumors are not uncommon in children. J Pediatr Surg. 2014;49: Al Jama FE, Al Ghamdi AA, Gasim T, Al Dakhiel SA, Rahman J, Rahman MS. Ovarian tumors in children and adolescents a clinical study of 52 patients in a university hospital. J Pediatr Adolesc Gynecol. 2011;24: Islam S, Yamout SZ, Gosche JR. Management and outcomes of ovarian masses in children and adolescents. Am Surg. 2008;74: Liu H, Wang X, Lu D, Liu Z, Shi G. Ovarian masses in children and adolescents in China: analysis of 203 cases. J Ovarian Res. 2013;6:47. ecollection ª 2015 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 94 (2015)

5 S. Taskinen et al. Pediatric ovarian neoplastic tumors 10. Menczer J, Sadetzki S, Murad H, Barda G, Andreev H, Barchana M. Childhood and adolescent ovarian malignant tumors in Israel. A nationwide study. Acta Obstet Gynecol Scand. 1999;78: Pauniaho SL, Salonen J, Helminen M, Heikinheimo O, Vettenranta K, Heikinheimo M. Germ cell tumors in children and adolescents in Finland: trends over Cancer Causes Control. 2014;25: Massicot R, Rousseau V, Darwish AA, Sauvat F, Jaubert F, Nihoul-Fekete C. Serous and seromucinous infantile ovarian cystadenomas a study of 42 cases. Eur J Obstet Gynecol Reprod Biol. 2009;142: Cecchetto G, Ferrari A, Bernini G, Alaggio R, Collini P, Virgone C, et al. Sex cord stromal tumors of the ovary in children: a clinicopathological report from the Italian TREP project. Pediatr Blood Cancer. 2011;56: Heifetz SA, Cushing B, Giller R, Shuster JJ, Stolar CJ, Vinocur CD, et al. Immature teratomas in children: pathologic considerations: a report from the combined pediatric oncology Group/Children s cancer group. Am J Surg Pathol. 1998;22: Abdel-Hady el-s, Abdel-Hady Hemida R, Gamal A, El- Shamey M. Fertility sparing surgery for ovarian tumors in children and young adults. Arch Gynecol Obstet. 2012;285: Spinelli C, Pucci V, Buti I, Liserre J, Messineo A, Bianco F, et al. The role of tumor markers in the surgical approach of ovarian masses in pediatric age: a 10-year study and a literature review. Ann Surg Oncol. 2012;19: Billmire DF, Cullen JW, Rescorla FJ, Davis M, Schlatter MG, Olson TA, et al. Surveillance after initial surgery for pediatric and adolescent girls with stage I ovarian germ cell tumors: report from the children s oncology group. J Clin Oncol. 2014;32: ª 2015 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 94 (2015)