In this study, we observed a median latency time of 17 years, with a range of 9 to 30 years, suggesting the importance of a long follow-up period. Among the 11 patients with RIG, ReRT combined with chemotherapy was performed in 5 patients at the initial treatment of RIG and for 6 patients at the time of recurrence; the median PFS and MST were 11.3 months and 28.3 months, respectively. The local recurrences were observed less frequently in patients receiving ReRT combined with chemotherapy than in those receiving chemotherapy alone. We identified no genetic alterations in the IDH1/2 and TERT promoters or in the H3F3A, and BRAF genes; these are commonly mutated in spontaneous high-grade gliomas. In addition, we found that evaluation of the IDH1/2 mutational status was helpful for establishing the diagnosis of RIG in cases whose IDH1/2 mutational states were different between the primary and secondary glioma.
The median latency time from the primary cancer treatment to the development of RIG was 17 years and varied from 9 to 30 years in our cohort. Previous studies reported that the median latency period was 9–11 years, and the incidence of RIG largely disappeared after 15–20 years [4, 5]. However, in our series, 4 out of 11 patients (36.4%) had a latency period of 20 years or more. Nakao et al. also reported that the latency period was more than 20 years in 4 patients . These results indicate that pediatric patients with primary diseases that were treated successfully with radiation therapy have a risk of developing a secondary neoplasm more than 20 years after the initial treatment. From the clinical point of view, these patients should be carefully monitored much longer than we have expected; based on our results, we conclude that the follow-up time should be extended to 30 years.
One of the controversies in the treatment of RIG is whether ReRT has a therapeutic effect on tumors that were induced by previous radiation. In our study, 5 patients underwent ReRT during the initial treatment of RIG, and 6 patients had ReRT at the time of RIG recurrence. The median PFS and MST were 11.3 and 28.3 months, respectively, which were better than are those reported in previous studies [5, 6, 23]. In addition, patients initially receiving ReRT combined with chemotherapy tended to have a longer PFS than those receiving chemotherapy alone did, but MST was similar between the groups. The use of ReRT at the time of recurrence might prolong the OS in patients who have not undergone initial ReRT. Moreover, we found the local recurrences were observed less frequently in patients receiving ReRT combined with chemotherapy than in those receiving chemotherapy alone, suggesting a therapeutic effect on local control from ReRT. These observations are key findings, as they emphasize the importance of ReRT in the treatment of RIG, regardless of the timing. Paulino et al. reported that patients who underwent ReRT for RIG showed better survival rates than those who did not (13 vs. 8 months; p=0.0009), suggesting that ReRT was efficacious in treating these tumors . Yamanaka et al. reported that the MST of patients who received surgery, chemotherapy, and ReRT was 18 months; those who did not undergo chemotherapy had an MST of 9 months (p=0.0006), suggesting that the combination of ReRT and chemotherapy is a potential treatment option for RIG . Recent molecular analyses suggest that RIGs harbor a largely homogeneous genetic and epigenetic profile of sporadic pediatric receptor tyrosine kinase I in GBM, which might be particularly vulnerable to ionizing radiation [14, 15, 24]. While clinical and genetic evidence support that ReRT is a potential therapeutic option for the treatment of RIG, further studies using large cohorts are needed to confirm the efficacy of ReRT.
Another concern in the treatment of RIG is ReRT-related radiation necrosis. There was no symptomatic radiation necrosis observed in our patients. According to literature reviews, the risk of developing necrosis is less than 10% in the patients who underwent ReRT treatment for RIG . The period between the first and second radiation sessions is usually more than 10 years, and most patients with RIG die within 3 years; therefore, late complications related to ReRT might not be clinically relevant. Two patients (Cases 10 and 11) in our cohort received Bev treatment combined with postoperative ReRT and TMZ, and they did not develop symptomatic radiation necrosis during the follow-up period. In Case 11, ReRT combined with TMZ/Bev after biopsy showed significant therapeutic efficacy without any sign of radiation necrosis (Figure 2). These cases suggest that the use of Bev combined with ReRT may be a promising therapeutic option for minimizing radiation-related risks with favorable therapeutic efficacy. It will be of interest to investigate the efficacy of combined ReRT and TMZ/Bev therapy in larger patient cohort.
We investigated genetic alterations in 8 patients whose tumor samples were available. We found no alterations in the IDH1/2 or TERT promoters or in the H3F3A or BRAF genes; these are commonly altered in spontaneous high-grade gliomas. Two patients had a hypermethylated MGMT promoter; the other 6 patients had a hypomethylated MGMT promoter. These results are consistent with those of previous reports and confirm the genetic characteristics of RIG [10–15, 24]. In Case 9, the difference in the IDH1/2 mutational status between the primary and secondary tumors was helpful for establishing the diagnosis of RIG. This case highlights the importance of determining the IDH1/2 mutational status in both primary and secondary gliomas. In addition, it is a reminder of the potential risk of developing RIG after radiation therapy for diffuse gliomas and the importance of long-term follow-up. Prior studies consistently report that RIGs hardly harbor IDH1/2 mutations [12–14]; therefore, identifying a wild-type IDH1/2 is diagnostically relevant. We recommend evaluating the IDH1/2 mutational status between the primary and secondary gliomas, especially in cases with recurrent glioma occurring a long time after a successful initial treatment.
Our study had certain limitations. First, this was a retrospective study, and the indications or dose/fraction regimens of ReRT were heterogeneous. They might have depended on the previous radiation field or dose/fraction regimen or period from the previous radiation; thus, heterogeneity was inevitable. Second, we did not investigate the genetic status in three patients because tissue samples were unavailable. Thus, further studies are needed to elucidate the genetic characteristics of RIG. Third, our cohort was too small to draw definitive conclusions. We acknowledge that the power of the survival analysis regarding the usefulness of ReRT, and the ReRT-related toxicity, was limited by the sample size; therefore, our results need to be confirmed in larger cohort studies.