In this study, most of the patients underwent previous surgical resection (86.8%, 33/38) and/or EBRT (71.1%, 27/38). Even 10 patients received two times of resection. It was a real difficulty to treat these patients with local recurrence after the multiple failures of previous treatment. These patients in the present study were deemed as inoperable as per the surgeon's consultation or personal refusal. Additionally, re-EBRT, due to the dose limitation of the normal structures or individual choice, was not recommended.
In theory, enough high radiation dose can be delivered to the tumor and the adjacent normal tissues are well spared with SABT due to the more rapid dose fall-off and lower dose rate of SABT compared with EBRT (24). The theory was validated by the facts soon that SABT had been recommended as one of the standard cares for prostate carcinoma by NCCN guidelines and for hepatocellular carcinoma by ESMO guideline for years (23, 25). Additionally, it was also recommended by expert consensus to treat most kinds of cancers including head and neck cancer, thoracic cancer, abdominal cancer, pelvic cancer and limb cancer (10, 11, 26). The SABT technique continued to advance over time, especially the invention of 3D-printing template (Patent No. ZL 2016 2 0414011.9 from our center) that significantly improved the accuracy, efficacy, security and productivity of the SABT. Therefore, SABT may be an appropriate alternative for the inoperable recurrent AF. So far, data regarding SABT for AF was little. As far as we know, this was the first multicenter study with the largest sample size, reporting the long-term outcomes of the SABT for inoperable recurrent AF. Moreover, this work firstly described the application of the advanced SABT method (3D-printing template assisted and CT-guided SABT) in recurrent AF.
In this study, the short-term efficacy was favorable with the ORR and DCR of 76.3% and 100.0% respectively. As to the long-term survival of the patients, the present study showed that the 3-, 5 and 9-year LCT rates were 81.0%, 68.6% and 62.3% respectively and the 3-, 5 and 9-year OS rates reached 94.5%, 84.8% and 60.4% respectively which indicated a favorable local control and survival of SABT for these inoperable recurrent AF patients. These favorable results were consistent with previous studies. Huang et al.(27) reported the efficacy of CT-guided SABT for AF in a small sample size study (14 patients) that the local control rate was 100% and the median OS reached 52.3 months. Zheng et al.(28) reconfirmed a good efficacy of SABT for AF in a study that the local control rate of sole brachytherapy was 100% in 5 pediatric patients. In the present study, the AEs of SABT for AF patients was mild and well acceptable. Previous studies aforementioned similarly reported that the AEs were in low probability and slight(27, 28). Collectively, our study echoed previous studies in the largest sample size indicating that SABT was characterized not only by good efficacy but also by low and accepted toxicity in recurrent AF. Ergen et al. (29) reported the efficacy and toxicity of radiotherapy for primary or recurrent AF patients with prior irradiation. In the study, the 2- and 5-year LCT rate of radiotherapy (18 patients received solo EBRT, 1 received EBRT combined with brachytherapy and 1 received solo brachytherapy) for AF was 80% and 69% respectively with a median LCT of 33 months. As to toxicity, in Ergen’s study, the most common radiation-related acute side effects were skin effects that the rate of Grade 1 and Grade 2 acute dermatitis was 55% and 35% respectively and the late side effects of radiotherapy was observed in 55% of the patients. What’ worse, one patient developed radiation-induced sarcoma after 14 years. It was worth noting that the efficacy of SABT for recurrent AF in our study was comparable with EBRT, but the AEs of SABT were significantly lower than EBRT, which indicated that SABT may be a good alternative treatment for the recurrent AF patients.
Multivariate analyses for LCT showed that the patients with smaller tumor volume and higher dose were significantly associated with a lower probability of local failure and longer LCT. Ji et al. (12) also reported that smaller tumor volume and higher radiation dose were significantly related with better local control in recurrent head and neck cancer. Qu et al. (17) released similar results in recurrent cervical carcinoma. These results were also consistent with our routine clinical experience. Enhancing local control can be achieved through dose escalation and careful selection of patients with smaller tumors.
Different from the previous studies in which the SABT was performed with free hand guided by CT, in our study, 9 patients were treated with SABT assisted by the 3D-printing template and guided by the CT. Wang et al. reported that the template assistance significantly improved the accuracy and efficacy of SABT for the recurrent rectal cancer (30). In this study, the 3D-template-assisted and CT-guided SABT showed a trend of better local control compared with CT-guided SABT, but the differences were insignificant (HR, 0.42, P=0.410) possibly due to that the sample size of the patients receiving advanced SABT was small. Moreover, the template assistance indeed increased the efficiency, shorten operation time and improve puncture accuracy of SABT.
Most of the patients in this study obtained symptom relief after SABT, especially the symptom of pain. Similar results were observed in previous studies (31-34). These consistent results suggested that SABT was effective not only on the local control of the tumor but also on the symptom relief.
As to limitations, the retrospective nature of this study lowered the evidence level of this study. Even so, it provided a new piece of evidence for SABT to treat the locally recurrent AF. More studies were warranted to further confirm the role of SABT in AF.