In this study on the effectiveness of salvage HT for patients with persistent cervical cancer after definitive RT/CCRT, we found the following two major points. First, the adoption of salvage HT reduced mortality rate up to about 60%, compared to systemic CT. Second, the relative resectability of a persistent tumor obviously affected survival, and tumor size alone was a significant predictive factor regarding resectability.
We showed that, when compared to systemic CT, salvage HT provided a significant PFS and OS improvement for patients with persistent cervical cancer after definitive RT/CCRT. These patients have an extremely poor prognosis due to their chemoresistance [5, 17–18]. Therefore, our findings have important clinical implications, as they may provide a strategy to address an unmet medical need. Salvage HT has been suggested to impart benefits for treatment of persistent disease after RT since the 1990s [9–14]. However, all of the reports to support this conclusion were based on retrospective studies from single institutions with small sample sizes. Furthermore, no randomized trials have directly compared salvage HT and systemic CT, and thus the clinical role of surgical treatment has remained unclear. Specifically, there has been a lack of Level 1 evidence to guide management for patients with persistent tumors after RT/CCRT. To overcome the limitations of published noncomparative retrospective studies, we used propensity score matching with patients who underwent contemporary salvage HT to generate comparable subjects in a systemic CT group. To the best of our knowledge, this is the first study to employ propensity-matched analysis to look at survival benefits in this patient population.
We showed that the resectability of persistent tumor obviously affected survival, and that tumor size alone is a significant predictive factor regarding resectability. Therefore, the goal of salvage HT should be R0 resection, and we suggest that stratification based on residual tumor size could be used to identify appropriate candidates for salvage HT. Indeed, the heterogeneity of residual tumor size might explain why surgical treatment has not been applied in clinical practice more successfully, even though there have been some reports of the benefits of surgery [9–14]. The opportunity to identify patients who would not benefit from a morbid surgery is a key consideration for patients who are deemed to have resectable disease. Based on our results, we suggest that patients with persistent cervical tumors less than 40 mm, which do not affect bladder and/or rectum, could be potential candidates for salvage HT. A more reliable cut-off value for complete resection would be 11 mm, although our results did not show a significant survival difference between pathological and non-pathological R0 groups. Houvenaeghel et al. showed relatively good results of 3-year and 5-year survival rates (64.9% and 55.6%, respectively) after curative surgery for patients with residual tumors of 20 mm or more [14]. However, Azria et al. reported that the therapeutic effect of HT was disappointing in a small series of 10 patients who developed residual disease of 20 mm or more after CCRT [13].
Intensity of RT is the most important factor that affects treatment outcomes of cervical cancer patients treated with RT/CCRT. In the current study, RT doses were evaluated as EQD2, which quantifies the biological effect of any RT method and also takes account of dose-per-fraction or dose-rate [19]. Our study is the first to take the treatment intensity of RT into account when evaluating the efficacy of surgical treatment for cervical cancer after definitive RT. Indeed, previous studies, including prospective trials or meta-analysis, did not contain details of RT intensity [20–22], which might preclude the adoption of salvage HT after definitive RT/CCRT in the clinic. In the current study, the median dose of EQD2 in the HT cohort was 68.1 Gy, and this dose would ensure that RT/CCRT leads to a complete cure. Thus, we have shown that salvage HT could be a valuable curative intervention for persistent cervical tumors, even after definitively invasive treatment has been performed.
PE could be a curative option for some of those patients [6, 23]. Total PE offers a 5-year survival of 23–50%, however, high rate of severe postoperative complications, such as infection, injury of the urinary and gastrointestinal tracts, and small-bowel obstruction, occurs, in addition to a 4–14% surgery-related mortality rate [24–27]. The major clinical issue associated with surgical treatment after RT/CCRT is the high incidence of post-operative complications, and this must be balanced against the potential benefit of treatment. In this sense, PE has been employed sparingly, and salvage HT could therefore be an acceptable alternative to PE. The current study, in which patients who underwent PE were excluded, showed that early and late severe adverse events occurred in 4.2% and 9.9% of patients, respectively (Table 2). In a recent retrospective analysis of 362 locally advanced cervical cancer patients undergoing adjuvant HT after CCRT, grade 3 or 4 post-operative complications occurred in 5.8% of cases [28].
The current study has some limitations that should be taken into account when interpreting our results. First, there were some unbalanced variables present in the data, although these variables were not statistically significant between the two groups, even after propensity score matching was performed (Table 1). This may be explained by significant differences in the characteristics of treated populations in clinical practice, which would lead to large biases between the groups (supplementary Figure. A1A). The retrospective nature of the study meant that our ability to control for differences was limited to variables for which data were available. In spite of this limitation, we suggest that our current results, and our interpretation of them, are valid, and provide important information. Second, the salvage HT was not based on random assignment, and we could not control the eligibility criteria for salvage HT in the study. The result of the current study may be confounded by other unobserved variables. Although we used a rigorous statistical method to adjust for baseline differences between treatment groups, including propensity score matching, unobserved variables could be unbalanced across treatment groups and partly explain the difference in survival. Furthermore, safety records of the salvage HT could be changed depending on the eligibility criteria. Despite the recent development of vessel-sealing and cutting devices and advances in surgical techniques, salvage HT in a previously irradiated field is usually challenging, even in patients with seemingly resectable disease. Third, we were unable to determine the specific time at which the diagnosis of persistent tumor was made based on the available data. In the current study, the median time at which diagnosis was made in the salvage HT and systemic CT groups was 51 days and 50 days after completion of RT/CCRT, respectively. Some tumors continue to regress several months after radiation treatment, meaning that a complete response may eventually be achieved, even if the tumor existed at the time when RT was terminated. This selection bias is fully acknowledged. The opportunity to identify the appropriate time of diagnosis and understand tumor biology could make salvage HT a more appealing treatment strategy.
In conclusion, we have shown that the adoption of salvage HT for patients with persistent cervical cancer after definitive RT/CCRT significantly reduces mortality rate when compared to systemic CT. Furthermore, complete resection obviously affected survival compared to incomplete resection and systemic CT, and increased size of the persistent tumor is negatively correlated with successful resection rate. Further prospective clinical trials with regard to salvage HT after RT/CCRT are now warranted. The other survey to address appropriate patients for salvage HT after definitive RT/CCRT was conducted and several factors were identified as selection criteria for salvage HT (no publication). Based on the results of that study, the Japanese Clinical Oncology Group is now planning to conduct a multicenter prospective trial to address this issue.