In Reply: Tumor Growth Rate as a New Predictor of Progression-Free Survival After Chordoma Surgery.

Abstract

To the Editor: We read with great interest the letter by Liu et al1 who commented on our article2 about the relationship between the preoperative tumor growth rate (TGR) and the progression-free survival (PFS) in patients with chordoma. In this study, we found that patients with preoperative TGR values ≥ 10.12%/m exhibited a poorer PFS than those with TGR < 10.12%/m (P < .0001). A correlation between TGR and the postoperative Ki-67 labeling index (LI) was also noted, suggesting that TGR could be considered as a preoperative radiological marker of tumor proliferation and a helpful tool permitting to identify more aggressive tumors and surgery decision making. As discussed in our article,2 we totally agree with the authors that patients with worsening symptoms are not eligible for TGR evaluation. Indeed, for these patients, the presence of rapidly progressive symptoms may already be a sign of tumor aggressiveness, suggesting a rapid preoperative tumor growth, and prompt radical surgery is recommended. We think that this situation of rapid tumor growth is more frequently observed in the following situations: de novo or secondary dedifferentiated chordoma, tumor recurrence with a history of radiation, or in case of tumor hemorrhage as described by Wang et al.3 However, primary chordomas are most often slow-growing tumors. A recent retrospective study3 of 19 patients with chordoma showed a mean tumor growth in volume of 0.8 cm3 or 8% per month. Accordingly, the use of TGR in clinical practice does not seem to add significant risks for the patient for tumor volume increase and symptoms worsening. In addition, the time needed to perform all the required preoperative workup (computed tomography scan, positron emission tomography scan, proton beam therapist consultation, etc) and to schedule the surgery is usually around 2 mo. Not considering, before treatment, this simple calculation of the difference in tumor volume between the diagnosis and the 24 h preoperative MRIs, which is an important information about the tumor behavior, would be pity. Surprisingly, Liu et al1 identified the inclusion of patients who underwent radiation therapy as a major limitation of our study. Nevertheless, as described in our exclusion criteria, patients with a history of radiation were, in fact, excluded from this study. Indeed, it is known that the effects of radiation on tumor biology can be prone to aggressive feature and risk of secondary dedifferentiated chordoma development.4 Considering the limited number of patients included in this study, the relation between TGR values and PFS has to be confirmed by a well-designed study with a larger sample size, taking into account other factors5-7 such as soft-tissue infiltration, initial size of the tumor, and radiological features. The underlying mechanisms explaining this variation in tumor growth remain unknown. As observed in other brain tumors, alterations in the genome, exome, and transcriptome may influence the preoperative TGRs in chordomas and therefore the clinical presentation, management, and outcomes. Worse genomic prognostic markers, such as PBRM1 gene alterations,8 have been described, and it would be interesting, for example, to correlate the frequency of this mutation with chordomas expressing a high preoperative TGR.