In Reply: Surgical Adverse Events for Primary Tumors of the Spine and Their Impact on Outcomes

To the Editor: Thank you for the opportunity to answer a letter to the editor1 about our article recently published in Neurosurgery.2 We are grateful that the authors appreciated the importance of this work and how it can positively affect shared decision making between patients and surgeons when it comes to surgical treatment of primary tumors of the spine. We are happy to provide some further clarification to some points raised by our colleagues. The authors mention the lack of stratification on tumor histology and surgical approach and how it limits the ability to draw deeper conclusions. There was indeed an effort to stratify our results according to those 2 variables. In Tables 7 and 8, we presented our results of patient-reported outcomes (PROs) stratified by histology (malignant, benign, intradural) and surgical plan (en bloc, intralesional). In Tables 12 and 13, we presented our clinical outcomes with the same stratification. We also considered the surgical approach (anterior/posterior) in our multivariate analyses of covariates associated with the risk of adverse events (AE) (Table 5) and PROs (Table 9). Although it would have been interesting to further stratify our results for histology and, for example, analyze separately our chordoma or chondrosarcoma patients and compare them, we think further stratification would have compromised our ability to find significant differences. Furthermore, these pathologies are relatively rare and one of the rationale of our study was to build a large cohort patients with primary tumors of the spine from multiple centers of expertise. In the second paragraph, the authors mention that our study did not show a survival advantage of en bloc resection. Indeed our study focused mainly on AE and their effect on outcomes and the survival analysis with the Kaplan-Meir curves shown in Figure 3 is stratified on AE status only. We stratified our PROs and clinical outcomes results according to surgical plan (en bloc vs intralesional) as a way to control for a potential confounding and clinically important factor. On the other hand, we agree that a time point of 1 year after surgery is probably too short to find a significant difference in survival according to the type of resection. We definitely plan to report our long-term results in the future, but we feel that the effect of perioperative AE was probably maximal in the first year. Another important point to mention is that the AE status did not affect the ability of the surgical team to perform their initial surgical plan of resection. The success rate was 68.8% in the group without intraoperative AE vs 64.5% in the group with an intraoperative AE (P = .538). This will most likely have the biggest effect in the long-term risk of local recurrence and survival.3 Finally, in the last paragraph, it is mentioned that future research should prioritize standardized AE monitoring and we cannot agree more. In our study, even with the use of Spinal Adverse Event Severity Sytem, version 2 (SAVES-V2), which is a validated system to record AE and specific to spine, variations in the way AE were collected was a challenge. The people and the timing for collecting AE varied across sites and that most likely explain the variability of AE rates within the study.4,5 The role of minimally invasive techniques to treat chordomas and chondrosarcomas was also mentioned. Although personalized care is paramount in treating primary spinal tumors and minimally invasive surgical techniques maybe has a role to play in some benign tumors,6 it is difficult to imagine minimally invasive techniques applied in most cases of malignant tumor requiring en bloc resection.3