In the current retrospective study, we explored the treatment outcomes in patients with iGCTs with choriocarcinoma element or higher β-HCG level treated with a radiotherapy-based regimen. As indicated, the 5-year EFS and OS rates were > 80%. Furthermore, we identified chemotherapy cycles, β-HCG level, and intratumoral haemorrhage during the clinical course as independent risk factors for survival.
Choriocarcinoma is a rare subtype of NGGCTs. Based on our data, those with choriocarcinoma element or higher β-HCG levels accounted for 5.2% of patients with iGCTs diagnosed during the same period. Many studies have shown that radiotherapy is indispensable for the treatment of iGCTs, for either germinomas or NGGCTs [5–8]. In our cohort, only two patients did not receive radiotherapy due to severe complications. As a result, the 5-year survival rate for the entire series reached 80%, which is superior to that in historical reports. Although irradiation-based regimens have improved patients’ prognosis, the optimal radiation volume remains undetermined. In a study conducted by the International Society of Pediatric Oncology, patients with localised NGGCTs had a 5-year OS of 82% after FR [9]. Based on ACNS 0112 data, the 5-year OS was 93% after CSI [10]. In the ACNS 1123 study, patients who received WVI had a 3-year OS of 92% [11]. It seems that the efficacy is superior in the COG reports in which CSI or WVI was applied. However, the differences were comparable among different volumes for patients with NGGCTs with localised disease in the same cohort [11, 12]. Data from our series also showed that disease control and survival rates were similar between patients treated with CSI or non-CSI. Considering the potential side effects of extended-field irradiation, it is optional to choose non-CSI in this setting if there is no evidence of metastasis.
In terms of radiation dose, most patients in our series received a dose ≥ 54 Gy, as adopted by most groups for patients with NGGCTs. Simultaneously, there were a number of patients whose prescription dose was < 54 Gy. Survival was comparable between the two radiation dose groups, in either the single variate analysis or multivariate analysis. This may be attributed to the small sample size of this study. Interestingly, we also identified two patients who received 40 Gy and five cycles of chemotherapy. One patient underwent initial surgery, and the histology revealed germinoma with sporadic syncytiotrophoblastic element infiltration. The serum β-HCG level was 1032 IU/L. The other patient was diagnosed based on tumour marker level elevation, with a CSF β-HCG level of 997 IU/L. They remained disease-free at the last visit for 105 and 94 months, respectively (Fig. 3). The less aggressive treatment strategy may be decided based on the germinoma histology, homogeneous enhancement of the tumour on radiographic images, and satisfactory response after induction chemotherapy.
Generally, β-HCG is an important marker, with a remarkable elevation indicating the presence of choriocarcinoma element. However, a widely accepted cut-off value has not yet been established. It appears that it is more reliable when it is ≥ 10,000. Conversely, hundreds or even lower could occur, probably caused by the quantity of choriocarcinoma elements [13, 14]. Therefore, it is challenging to determine the presence of choriocarcinoma element based solely on β-HCG level elevation, especially when the elevation is mild to moderate. Otherwise, the individualised treatment strategy, which was determined by malignant elements, is far from established in patients with NGGCTs, even in those with homogeneous element. Thus, a dose of ≥ 54 Gy was unanimously recommended in all patients with NGGCTs. However, in combination with our data, some issues might be raised: Is there a possibility that the radiation dose could be reduced in patients with choriocarcinoma or in certain subgroups of patients, especially those with mild-to-moderate β-HCG level elevation and achieving CR after induction chemotherapy? Alternatively, in patients with germinoma with full histology evaluation, would it be necessary to be treated as NGGCTs even if β-HCG level is moderately elevated? Given that β-HCG > 3000 IU/L was identified as an adverse prognostic factor in our series, an aggressive treatment strategy, including a higher radiation dose, seems more justified in these patients. For those with β-HCG level < 3000 IU/L, more studies are needed to clarify these issues.
Intratumoral haemorrhage is a common complication in patients with choriocarcinoma, which could occur at any time during the clinical course. Some may occur spontaneously, and some may be intervention-related, such as haemorrhage after biopsy or radiotherapy [15–21]. Hence, as a major cause of early death, it is a major concern for clinicians that has a significant influence on decision-making regarding treatment modality arrangement. However, haemorrhage was not proven to be an independent factor for survival by far [1, 2]. In our study, patients with or without intratumoral haemorrhage at diagnosis showed a similar prognosis (data not shown). Although haemorrhage during the course was identified as an adverse factor in our series, only two patients were included in the model (patients with early death were excluded from the multivariate analysis), making the conclusion weak. Therefore, for the entire series, fatal haemorrhage events were not as common as those in the published reports. The possible reasons may lie in our careful surgical candidate selection and induction chemotherapy application.
Generally, choriocarcinoma elements are sensitive to antitumor therapy. As indicated in our series, 57.4% of patients achieved CR after induction chemotherapy. They showed similar survival rates as those receiving initial surgery. Even in patients with PR after induction chemotherapy, those undergoing second-look surgery had a good prognosis. Therefore, in patients with choriocarcinoma element, induction chemotherapy could be considered initially, which may provide some patients with the opportunity to avoid radical resection. After all, it is always challenging for surgeons to deal with this type of tumour, as it is characterised by spontaneous haemorrhage. Additionally, gross total resection may be easier to achieve during second-look surgery because of decreased tumour volume and/or blood supply and because the extent of resection is an independent factor for survival in patients with choriocarcinoma element [1, 2]. Probably because of the small sample size, the removal of residual disease was not found to be correlated with survival in our series.
In the multivariate analysis, we found that chemotherapy cycles were an independent prognostic factor. Both disease control and survival can be improved by additional cycles of chemotherapy. Given that choriocarcinoma is characterised by distant metastasis, even beyond the CNS, adequate chemotherapy may bring survival benefits by eliminating potential micrometastasis, especially in responders [2, 22, 23]. Thus, although the optimal number of chemotherapy cycles is still unknown in this setting, more than four cycles are warranted in patients with choriocarcinoma element or with higher β-HCG level. However, a small number of patients do not respond well to ICE-based regimens. Some authors have introduced high-dose chemotherapy with stem cell support in these patients and reported satisfactory results [22, 24]. Therefore, more aggressive regimens may be recommended, especially in patients with poorer prognosis, as indicated by our study.
Generally, in the current study, we reported better outcomes in patients with choriocarcinoma element or higher β-HCG levels treated with a radiotherapy-based regimen. However, the conclusions may be limited by the retrospective study design and number of cases included. To recruit a large number of cases for the study, β-HCG level > 500 IU/L was set as an inclusion criterion for our study. However, it is still difficult to establish a definite correlation between high β-HCG levels and presence of choriocarcinoma element. Even in those with choriocarcinoma element confirmed by histology, whether it occupied major or minor components was not further analysed. Moreover, the treatments were not consistent for the entire series, such as indications for surgical intervention, radiation dose, and chemotherapy cycles; this resultant bias could not be avoided. Therefore, multicentre collaboration is needed to initiate prospective studies, which is the best way to investigate this rare malignancy.