Discussion
Teratomas present diverse biological behavior and continue to be the cause of many diagnostically and therapeutically challenging issues. While it is commonly accepted that mature teratomas of the ovary are benign neoplasms, the classification of the immature ones is still discussed. About 25% of all the pediatric GCTs present as tumors with more than one histologic type. In this situation, therapy and prognosis depend on the component with the highest malignancy [
12‐
17]. In accordance with this approach pure immature teratomas were excluded in the recent studies concerning pediatric malignant germ cell tumors. Conversely, patients with malignant germ cell tumors, even when these lesions contained mature teratoma or immature teratoma elements, were not eligible for studies concerning teratomas exclusively [
2,
12,
17,
18]. Nevertheless, there are also studies were a lesion was considered as immature teratoma although the pathology report revealed a component of another malignant tumor [
19]. Marks of the controversy regarding this topic are also reflected in the nomenclature of these lesions across the studies. The multiplicity of names describing immature teratomas (immature teratoma, malignant teratoma, teratoma with malignant elements, immature teratoma with malignant behavior) renders universally applicable classification of these lesions very difficult [
5,
12,
17‐
21].
All the girls in our study had at least stage III of the disease. In a study from 2012, Schneider et al. ascertain that immature teratomas behave in a malignant fashion only if foci of malignant germ cell elements are present and if they are resected incompletely. According to their paper, tumors containing clusters of yolk sac tumor are likely responsible for the reports that immature teratoma may metastasize [
2]. Nevertheless, in some recently published studies immature teratomas are still included in the evaluation of patients with malignant germ cell tumors and even at US National Cancer Institute website reads that “Immature teratomas can exhibit malignant behavior and metastasize” [
7,
14,
22,
23]. Therefore, can we definitively classify them as benign?
Another important aspect in attempts to estimate the real nature of these lesions is the importance of the grade of immaturity. In a study published by Malignant Germ Cell Tumor International Collaborative it was revealed that grade was the most important risk factor for relapse in ovarian immature teratoma [
6]. However, there is scarceness of papers reporting the use of a detailed grading system and its significance in children has been challenged by some former studies. Additionally, high immaturity itself is not associated with a poor prognosis if the tumor is completely resected [
2]. Defining real nature of the lesion is indispensable in choosing optimum management. Obtaining the correct pathology report is crucial here. Clusters of malignant tumor can be easily overlooked and yolk sac tumor components might be very small. Microfoci of yolk sac tumor (Heifetz lesions) are not precisely defined among the studies, thereby classifying a lesion as malignant yolk sac tumor is particularly difficult [
2,
5,
12]. Performing a central pathomorphological examination can be helpful in the case of these neoplasms. Perhaps new tumor markers, like the ones discovered by Feng et al. will help eliminate mistakes in pathomorphological diagnoses [
24]. Grading was available only in one patient in our study and no Heifetz lesions were observed. This is a retrospective study and we cannot verify the results of the pathology report. Taking into consideration that all immature teratoma cases were of stage III at least, it is questionable if no other malignant components were present.
Despite these unclear pathological aspects, commonly applied diagnostic work-up of ovarian teratomas include physical examination, ultrasound imaging and tumor markers evaluation. In most of the studies the peak incidence of these lesions is reported in early adolescence [
2,
12]. It was also confirmed by our study. Symptoms of ovarian teratoma do not differ from those observed in other ovarian masses. Some patients reveal acute symptoms rising suspicion of ovarian torsion. The risk of torsion complicating a case of ovarian teratoma is approximately 3 to 16% in children [
2,
25]. Those cases are rarely associated with malignancy. Tsai et al. revealed that young girls tend to have either torsion of a mature cystic teratoma or torsion without underlying condition, while older patients are more likely to present with torsion and a tumor [
2,
26]. In our study ovarian torsion was the most common intraoperative finding additional to the tumor in girls with acute presentation. The incidence of ovarian torsion was not higher in older girls (Kruskal-Wallis test,
p = 0,765).
As most of the mature teratomas are slow-growing cystic lesions and there are some common characteristic findings, they are easily recognized on US. Nevertheless, their typical features might be less clear in case of prepubertal girls and when the lesion is large [
13,
26,
27]. Our study poses limitations in that respect. The ultrasound description was often very short and we were unable to identify characteristic features for teratoma in each case. Diagnosis of immature type also poses problems. The appearance of the tumor on ultrasound is non-specific, although predominance of solid component is a differentiating factor. Identification of fat, mural nodules and calcified components within the lesion is typical for teratomas. However, there are reports suggesting that typical CT features of teratomas occur more often in mature than in immature lesions. Those findings suggest that CT or MRI imaging might be helpful, rather for staging and assessing of tumor respectability [
13,
28,
29].
Alphafetoprotein evaluation is an important prognostic factor in many malignant germ cell tumors treatment protocols and it is also used in follow-up of those patients. Additionally, it is a characteristic marker of yolk sac tumor. Regarding its levels in ovarian teratomas, it was revealed that they are rarely elevated [
2,
12]. In the recent Malignant Germ Cell Tumor International Collaborative study regarding ovarian immature teratomas, patients with AFP levels higher than 1000 ng/mL were excluded because this level was considered more likely to indicate malignant elements [
6]. However, this approach was questioned by Terenziani et al. who used an AFP level cutoff “high for age” in their study [
14]. Another limitation is caused by the vide variation in AFP levels at birth and the variability in its half-time within the first year of life [
2]. Multiple publications where cases of immature teratoma are analyzed without taking AFP levels into consideration makes it more difficult to form a credible assessment of this indicator [
14,
30,
31]. Only one girl in the immature teratoma group had AFP levels < 1000 ng/mL in our study. Given the results of the recent research this is another factor calling into question the result of pathology report in our cases.
According to the experience of many pediatric surgical centers, surgery remains the mainstay of treatment in ovarian teratomas. Under particular conditions ovarian-sparing surgery might be successfully applied. Preservation of ovarian tissue should be reserved for cases of localized mature teratoma, when there is a plane of dissection between the tumor and the normal ovary [
1,
2,
13,
14,
32]. The study by O’Neill et al. revealed that normal ovarian tissue was visualized on follow-up ultrasound after cystectomy even if there was no normal ovarian tissue visible preoperatively [
13]. Although incomplete resection is an important risk factor of recurrence, relapse is not inevitable in those cases. The use of laparoscopic techniques in the treatment of ovarian teratomas has its supporters and opponents. If the suspicion of malignancy is low and a surgeon is experienced in minimal invasive surgery, laparoscopic approach might be adequate [
2,
6,
7,
12,
15,
21,
33]. Children’s Cancer Study Group demonstrated in their study that cyst fluid aspiration or spillage during surgery are not associated with relapse, thereby this factor, claimed by laparoscopy opponents as more often associated with minimal invasive methods, might not have such importance for the final outcome [
12,
30]. Our results are promising in that respect. We revealed that the use of laparoscopy increased in time and the preservation rate was higher in the laparoscopy group. Laparotomy is the treatment of choice in large masses, suspicious for malignancy, if surgical staging is required [
13]. The preservation rate seemed no to be affected by the size of the lesion in our study.
An important aspect of the management of ovarian teratomas is their recurrence rate and the incidence of bilateral lesions. The recurrence rate after cystectomy vary between 3 to 13%. However, in the study were the recurrence rate was estimated to be of 10% the authors revealed that only 3% of the recurrent cases will require reoperation [
13,
34]. There were studies suggesting that women with bilateral or multiple dermoid cysts may include a subgroup of patients with a greater tendency to develop future ovarian germ cell neoplasms. Due to low risk of malignant transformation in case of mature cystic teratomas in children the treatment should be directed on the basis of age, fertility desire or presence of another pelvic pathology rather than the size or bilaterality [
35‐
37]. In the study by O’Neill et al. the biopsy of the contralateral macroscopically normal ovary revealed pathological lesions only in 1.1% of patients. Therefore, preoperative ultrasound in combination with careful inspection of the contralateral ovary at the time of surgery offer a safe alternative to biopsy [
13,
38]. Our results support this approach, none of the biopsies was positive in our study. Given the sensitivity of ultrasound in the detection of mature cystic teratomas, annual imaging seems appropriate as a postoperative surveillance. AFP monitoring is not recommended in case of completely resected ovarian teratomas without preoperatively elevated tumor markers [
12,
38].
The treatment of immature teratomas poses much more difficulties. Their possible malignant behavior requires appropriate risk classification. Among factors considered as those influencing the prognosis, except the previously mentioned, staging is the next unclear one. Since malignant tumors and benign ones cannot be distinguished based only on intraoperative appearance, there are studies recommending staging in all tumors [
2,
39]. However, Billimire et al. revealed that survival appears to have been unaffected by deviations from surgical guidelines when chemotherapy is administrated [
39]. Different staging systems across the surgical centers renders evaluation of their significance difficult [
2,
6]. Another controversial aspect is the use of chemotherapy. Number of therapeutic options might be found in the literature. To name some of them: chemotherapy for recurrence, only for malignant recurrence, in all cases of stage I grade 2 and higher or only in girls older than 15 [
2,
6,
7,
22,
40]. In the German MAKEI study more than half of the children with tumor recurrence after watch and wait strategy had yolk sac tumor in addition to teratoma. It might be one of the main counter-arguments to the results of the studies which did not find evidence that chemotherapy had any significant therapeutic effects in pure teratomas and that it does not decrease relapses in pediatric population [
40]. Recently created Malignant Germ Cell Tumor International Collaborative in its publications gathering data from four clinical trials recommends surgery alone for patients with stage I/II, grade 1/2 tumors and suggests conducting of a prospective trial of observation after surgery for patients with grade 2/3, stage II-IV tumors [
6]. This approach might be encouraged by the results of Park et al. as they revealed that most stage I malignant ovarian germ cell tumor recurrences can be successfully salvaged by surgery and BEP chemotherapy without compromising the overall survival [
22]. Adjuvant chemotherapy according to protocol TGM 95 was given to all girls with immature teratoma in our study group. They all responded well to the therapy but it remains questionable if these were real pure immature teratoma lesions.
Schneider et al. revealed that patients with cytogenetically abnormal immature teratomas often develop recurrence [
2]. An optimal consensus treatment strategy should be based on a detailed analysis of all possible factors. In this regard, the above mentioned study shows that we might be still not aware of a number of them. The future molecular research may allow distinguishing low-risk from high-risk patients accurately [
15].
Different staging systems, histopathologic classification systems and risk stratification are important factors impeding development of an appropriate treatment protocol [
2,
12]. Additional factors include the lack of consistent terminology in the literature and analyzing pediatric and adult patients together [
5,
20,
31]. International cooperation seems the only way to progress in the prospective knowledge of prognostic factors for these tumors. It will not be possible without high quality and meticulous diagnostic approach in all participating institutions.