Session Item

Intensity modulated radiation (IMRT) technique, associated with induction chemotherapy (IC) and/or concomitant chemotherapy (CC) is actually the recommended treatment modality for nasopharyngeal carcinomas (NPC). The aim of this study was to evaluate the loco regional relapse (LRR) rates and their patterns of relapse with this treatment protocol.

A retrospective study of 145 patients with NPC treated between June 2016 and July 2021. All patients received IMRT with integrated simultaneous boost (SIB) of 33 daily fractions at a dose of 69.96 Gy for high-risk volume, 60 Gy for intermediate risk volume and 54 Gy for low-risk volume. High risk volume dose was 66.5 Gy in children.
For patients presenting a LRR, the CT scan showing the recurrence was merged with the initial dosimetric CT scan and the macroscopic volume of the recurrence was contoured (GTVr). An evaluation of the V95% received by the GTVr was made by reproducing the initial dosimetry on the new scanner. Thus, we have defined 3 situations: a recurrence “on the primary site” (V95% ≥ 95%), a “marginal recurrence” (20% ≤ V95% <95%) and an “out of field” recurrence (V95% <20%).
Survival analysis was performed according to Kaplan-Meier method and Log-rank test was used to compare factors that may influence loco regional free survival (LRFS). Cox regression method was used for multivariate analyses.

Median age was 48 years (11-80) with a sex ratio of 2.9. One hundred-twenty tumors (82.7%) were classified as stages III-IV according to the 2017 UICC TNM classification. Ten patients (6.9%) were metastatic at diagnosis. One hundred-thirty-five patients (93.1%) received IC and 138 (95.2%) received CC.
After a median follow up of 48 months (24-83), 17 patients (11.7%) experienced LRR after a median of 18 months (6-43). Three years LRFS was 88.1%. A metastatic relapse was also noted for 5 of these patients (29.4%). Ten relapses (58.8%) were on the primary site, five (29.4%) were marginal and two pretragian relapses (11.7%) were “out of field”.
Factors predicting LRR were a consultation delay of 6 months or more (p=0.042), histologic type other than undifferentiated (UCNT) (p<0.0001), a delay more than 14 days starting IC (p=0.029),  number of CC courses (less than 4, p=0.013; less than 5, p=0.032, less than 6, p=0.05) and a total cumulative cisplatin dose < 380 mg/m² (p=0.004). UICC TNM classification did not impact LRR. In multivariate analyses, histology other than UCNT (p=0.003) and total cumulative cisplatin dose (p=0.002) were significant prognosis factors of LRR.

For nasopharyngeal carcinoma, tumors with histologic type other than UCNT and that receive a low cumulative cisplatin dose, have a higher risk of LRR. Therefore, they require a more aggressive therapeutic approaches and a suitable monitoring protocol.