Document resource
The treatment landscape for locoregionally advanced nasopharyngeal carcinoma (LA-NPC) is evolving, with adjuvant chemotherapy (AC) gaining renewed attention, driven by three recent pivotal randomised trials. Metronomic capecitabine for 1 year following chemoradiotherapy (CCRT) significantly improved failure-free survival (FFS) and overall survival (OS). Notably, 78% of the cohort also received induction chemotherapy (IC).1 Similarly, Miao et al reported that 6 months of adjuvant capecitabine improved FFS in LA-NPC with high-risk features, although no OS benefit was observed in the intention to treat cohort.2 Most recently, Liu et al compared two intravenous adjuvant regimens, demonstrating that gemcitabine plus cisplatin (GP) outperformed the traditional fluorouracil plus cisplatin (PF).3 Collectively, these studies affirmed that both oral and intravenous AC can improve outcomes in selected patients with LA-NPC—yet the optimal regimen remains to be defined.Against this evolving backdrop, Chen and colleagues4 now provide valuable real-world evidence from a large retrospective cohort at Sun Yat-sen University Cancer Center. The primary aim was to compare oral versus intravenous AC following CCRT in 470 patients with LA-NPC. After propensity score matching (n=154 per group), no significant difference was observed in 3 year progression free survival (PFS) between intravenous and oral AC (76.3% vs 73.9%; HR 0.803; p=0.316). OS was similarly comparable (p=0.817). As expected, intravenous AC was associated with higher rates of grade ≥3 haematological toxicities and electrolyte disturbances.In the unmatched cohort, multivariable analyses indicated that GP was associated with significantly improved PFS (HR 0.38, 95% CI 0.18 to 0.81; p=0.012) and distant metastasis-free survival (HR 0.42, 95% CI 0.19 to 0.96; p=0.039) compared with PF, aligning with findings from a prior trial.3 No significant differences were seen among PF, S-1 and capecitabine in adjusted analyses. Of note, the study was not designed to directly compare individual regimens.As with all retrospective studies, selection bias remains a major inherent limitation. For example, 28.7% (135/470) received IC, and among them, 98% received oral AC. This may reflect the ‘high-risk’ nature of this cohort, for example, persistent Epstein-Barr virus DNA5 or residual disease post-IC,6 whereby IC and CCRT alone may not suffice for treatment intensity in these patients. Additionally, patients with advanced nodal disease (N2–3) were more likely to receive intravenous rather than oral AC (93.4% vs 65.5%). These imbalances reflect real-world practice patterns—such as physicians’ perceptions of recurrence risk and tolerability—that may not be fully addressed by statistical adjustment.Tolerability of AC remains a clinical challenge, as patients often experience significant acute toxicity following CCRT. In clinical trial settings—where patients are typically highly selected and closely monitored—completion rates range from 50% to 70%3 7 8 for intravenous AC and 70% to 80% for oral capecitabine.1 2 This cohort offers valuable insights into real-world experience with AC: approximately 70% of patients completed three cycles of intravenous AC, while 85% completed eight cycles of capecitabine, although only 65% completed S-1. This result is reassuring of the tolerance of AC, which may be facilitated by advances in RT technique and cancer supportive care. In addition to the better tolerability, oral agents offer greater convenience, but require longer treatment durations and carry distinct toxicities, such as hand-foot syndrome and diarrhoea, which were not captured in the present analyses. Future studies should incorporate patient-reported outcomes to better reflect the lived experience of patients.In the broader treatment paradigm, the optimal sequencing of chemotherapy—induction, adjuvant or both—remains an area of active investigations. Guo et al directly compared IC (paclitaxel plus PF) versus AC (PF) and did not find a difference in PFS and OS. Furthermore, IC was associated with better distant control, while AC was associated with better loco-regional control.8 The Meta-analysis of chemotherapy in nasopharynx carcinoma (MAC-NPC) confirmed OS benefit of both IC followed by CCRT and CCRT followed by AC over CCRT, although claims of superiority are not appropriate based on p scores. However, it did not reveal inferior locoregional control of IC-CCRT versus CCRT.9 Notably, the recent pivotal AC trials were not included.9 Meanwhile, emerging phase II/ III data support the addition of PD-1 inhibitors during induction and adjuvant phases, demonstrating significant improvements in event-free survival, though mature OS data are pending.10 11 Importantly, AC was not permitted in these trials. In this new era of immunotherapy, the role of AC must be reconsidered in the context of evolving systemic strategies. The future of adjuvant therapy—and the broader treatment paradigm—must be guided by risk stratification. The open questions are: (1) Which patients will benefit from IC, and among those receiving upfront CCRT, who should receive AC? (2) For patients undergoing IC, how can we identify those who will gain additional benefit from subsequent AC? (3) What is the optimal IC/AC regimen balancing efficacy and toxicity? Only through answering these questions can we further individualise treatment through careful escalation and de-escalation.