Enhancement of local control in locally advanced node-positive nasopharyngeal carcinoma by adjunctive chemotherapy

Purpose: To determine the efficacy of chemotherapy adjunctive to radical ra diotherapy (neoadjuvant +/- adjuvant) in patients with node-positive nasoph aryngeal carcinoma (NPC). Methods and Materials: All the node-positive patients given adjunctive chem otherapy between 1984-1989 (n = 209, CHEMO) were compared with all the node -positive patients treated by radical radiotherapy alone during the same pe riod (n = 409, NCHEMO). The CHEMO group had significantly more bulky nodes, lower cervical/supraclavicular nodes, and more advanced overall stages tha n the NCHEMO group because nodal size (greater than or equal to 4 cm) was u sed as a selection criterion for chemotherapy (1984-1988 departmental proto col and 1988-1989 prospective randomized trial). The chemotherapy consisted of two courses of neoadjuvant cisplatin (100 mg/m(2) D1) and 5-fluorouraci l (5-FU) (1 gm/m(2) D1-D3) in 191 patients. In addition to the two courses of neoadjuvant, four courses of adjuvant chemotherapy, of the same combinat ion, were given after radical radiotherapy in a further 18 patients. Radica l radiotherapy delivered a nasopharyngeal dose of 60-62.5 Gy. In addition, parapharyngeal booster external radiotherapy (20 Gy) was given in the prese nce of parapharyngeal involvement, and intracavitary brachytherapy (24 Gy) was used to treat any local residual tumor diagnosed at 4-6 weeks after ext ernal radiotherapy. Both crude and actuarial rates were compared (survival, distant metastases, and local failures) between CHEMO and NCHEMO for all p atients, for individual Ho's overall stage, for patients with nodes of diff erent sizes (less than or equal to 3 cm, >3-less than or equal to 6 cm, >6 cm), for individual T-stage and individual N-stage, and for patients belong ing to different gender and different age groups (<40 years, greater than o r equal to 40 years). Multivariate analyses using the Cox Regression Model were performed to identify significant prognostic factors. Results: With a median follow-up of 5.5 years (range 0.7 to 10 years), CHEM O had significantly less local failures overall than NCHEMO; this was espec ially true for patients with advanced stages (III + IV). Additionally, in a ll nodal-size subgroups, in all node-positive T3, and in node-positive T3-S tage IV, there was a significant reduction in local failures after chemothe rapy. There was a trend toward fewer local failures in favor of chemotherap y in Stage III, Stage IV, and T3-Stage III (0.05 < p less than or equal to 0.1). There was no difference in local failures between CHEMO and NCHEMO in Stage II or in T1 and T2. The multivariate analyses identified the adminis tration of adjunctive chemotherapy to be of independent significance in det ermining the local failure rate for all patients, the T3 (node-positive), a nd the advanced overall stages (III and IV combined). There was no differen ce in overall survival, relapse-free survival, and distant metastasis rates between CHEMO and NCHEMO among patients belonging to Stages III and IV des pite the presence of more advanced nodal diseases in CHEMO. There were very few late local relapses in patients given adjunctive chemotherapy, in cont radistinction to the well-known predisposition of NPC to late local relapse s after radical radiotherapy. Conclusion: Adjunctive chemotherapy enhanced local control in node-positive NPC in general, and node positive-T3 and -T3-Stage IV in particular with r eduction of late local relapses. The enhancement in local control of the lo cally advanced NPC could be explained by the significant shrinkage of the p rimary tumor by the neoadjuvant chemotherapy, leading to an increased safet y margin between the tumor volume and the radiation volume. We recommend th at adjunctive chemotherapy (neoadjuvant +/- adjuvant) should become an inte gral part of the multimodality curative treatment for patients with node-po sitive T3 NPC. (C) 1999 Elsevier Science Inc.