Published online Aug 24, 2026. doi: 10.5306/wjco.124227
Revised: July 27, 2026
Accepted: August 20, 2026
Published online: August 24, 2026
Processing time: 72 Days and 1.5 Hours
Despite the wide utilization of intensity-modulated radiation therapy and other advanced treatment modalities, approximately 10% of patients with nasopharyn
To investigate the temporal distribution of site-specific recurrence and evaluate the prognostic impact of early vs late recurrence in NPC.
This study retrospectively included 330 recurrent NPC patients who received initial treatment of radical ra
Among the 330 eligible patients, 167 developed pure local recurrence, 86 had pure regional recurrence, and 77 had locoregional recurrence. With a median follow-up duration of 97.8 months, there was no significant difference in timing distribution across various patterns of failure. The peak incidence of progression was observed in the 2nd year, contributing 29.4% of all progression events, with a cumulative 5-year proportion of 82.8%. The optimal cutoff for differentiating early and late recurrence was identified as 24 months. Advanced T (P = 0.008) and N (P = 0.028) stages were predisposing factors for early recurrence. Survival analysis manifested that late locoregional recurrence was associated with markedly better post-recurrence OS [hazard ratio (HR) = 0.37, P = 0.006], whereas such scenario was not observed in entire cohort or sub-cohorts with pure local or regional failure. The complementary 24-month landmark analysis further unveiled that late recurrence was associated with superior post-landmark OS across the entire cohort (HR = 0.38, P < 0.001) and all failure subgroups.
Recurrence timing is a pivotal prognostic factor in recurrent NPC. Early locoregional recurrence within 24 months is indicative of significantly inferior post-recurrence OS, warranting more intensive surveillance and aggressive salvage strategies for this subset.
Core Tip: This is the core tip of our study focusing on recurrent nasopharyngeal carcinoma. In 330 recurrent patients, re
- Citation: Wang JB, Meng Y, He ML, Zhang SJ, Huang XD, Zhang Y, Wu RY, Wang K, Chen XS, Qu Y, Zhang JH, Liu QF, Luo JW, Gao L, Xiao JP, Li YX, Luo LH, Yi JL. Timing matters: Early vs late recurrence in nasopharyngeal carcinoma and its impact on survival. World J Clin Oncol 2026; 17(8): 124227
- URL: https://www.wjgnet.com/2218-4333/full/v17/i8/124227.htm
- DOI: https://dx.doi.org/10.5306/wjco.124227
In recent decades, the outcome of nasopharyngeal carcinoma (NPC) has improved remarkably due to the advancements in treatment modalities, especially the wide utilization of intensity-modulated radiation therapy as well as the integration of chemotherapy and immunotherapy[1]. However, still around 10% patients suffer from local or regional disease recurrence, leading to the deterioration in survival outcomes[2]. This highlights the necessity for a comprehensive understanding of the factors influencing survival after recurrence, which could ultimately optimize the management of recurrence. The therapeutic strategies for recurrent NPC partially depend on the recurrence site. For local recurrence, re-irradiation or surgical resection is typically recommended, whereas surgery is generally preferred for regional recurrence[3]. In cases where patients are unsuitable for radiotherapy (RT) or surgery, chemotherapy or best supportive treatment may be considered[4].
Beyond the recurrence site, the timing of recurrence also plays a critical role in shaping the overall prognosis and determining the treatment strategies. Generally, early recurrence is often associated with more aggressive biological behavior and greater resistance to the initial treatment, subsequently leading to poorer survival outcomes[5,6]. On the other hand, therapeutic approaches for recurrent NPC also vary between early and late recurrences[3]. For example, re-irradiation is generally not advised for individuals with a recurrence latency less than one year after completion of primary RT[4].
However, the temporal and locational features of recurrence and the following prognostic consequence remain poorly understood. This study sought to: (1) Depict the distribution pattern of recurrence timing and site in a large recurrent NPC cohort; (2) Further identify the predisposing factors for various recurrence timings and sites; and (3) Determine the prognostic factors for recurrent NPC.
Between January 1990 and December 2020, 330 NPC patients with complete medical records who received radical RT initially at our center and experienced local or regional recurrence were retrospectively reviewed and analyzed. All patients were histologically confirmed to have NPC with no evidence of distant metastases. All patients underwent comprehensive pre-treatment evaluation and were restaged using the 8th AJCC staging system[7]. This research was conducted in accordance with the principles outlined in the Declaration of Helsinki and obtained ethical approval from the Ethics Committee of our institution (NCC: 23/353-4095). Given its observational nature using retrospective clinical data, informed consent was waived.
Patients were advised to attend follow-up appointments at least every 3 months in the first 2 years, every 6 months during the third to fifth years, and subsequently annually until death. Assessment procedures included a thorough physical examination, fiberoptic nasopharyngoscopy, magnetic resonance for nasopharynx and neck, computed tomography (CT) of the chest, abdominal ultrasonography, as well as bone scans and PET/CT when indicated. Re
The optimum cutoff timing for the definition of early and late recurrence was determined by adopting the maximally selected rank method, facilitated by the survminer package in R[8]. To evaluate the stability of this cutoff, internal validation using 1000 bootstrap resamples was performed. After defining the “early” and “late” recurrence, patients were categorized into six groups: Early local recurrence (ETR), late local recurrence (LTR), early regional recurrence (ENR), late regional recurrence (LNR), early locoregional recurrence (ETNR), and late locoregional recurrence (LTNR).
Post-recurrence overall survival (OS) was evaluated as the duration between confirmation of specific recurrence and any cause of death or last follow up. Furthermore, 24-month landmark OS was also explored as a complementary assessment. χ2 test was employed to analyze clinical characteristics and treatment modalities among subgroups. A Kernel density function was generated to compare the timing distribution across various patterns of failure (POF). Kaplan-Meier approach was utilized to estimate survival, with log-rank test applied to assess differences between survival curves. Multivariable Cox regression analysis was conducted to evaluate prognostic factors affecting survival. The variables of interest included age, gender, initial TNM, recurrent TNM and recurrence timing. A significance threshold of P < 0.05 was applied to determine statistical significance. Data analysis was performed with R 3.6.2.
A total of 330 recurrent patients were eligible for final analysis, including 167 patients (50.6%) with pure local recurrence, 86 patients (26.1%) with pure regional recurrence, and 77 patients (23.3%) with locoregional recurrence. In the entire group, the median follow-up duration was 97.8 months [95% confidence interval (CI): 83.9-111.7], while the median time to recurrence was 24.7 months (95%CI: 22.7-27.4). Detailed characteristics and demographics are presented in Table 1. The median OS of the entire cohort was 88.4 months (95%CI: 76.7-100.1), while the median post-recurrence OS was 35.4 months (95%CI: 27.2-43.6).
| Variables | Total (n = 330) | Local (n = 167) | Regional (n = 86) | Locoregional (n = 77) | P value |
| Gender | 0.615 | ||||
| Male | 257 (77.88) | 131 (78.44) | 69 (80.23) | 57 (74.03) | |
| Female | 73 (22.12) | 36 (21.56) | 17 (19.77) | 20 (25.97) | |
| Age | 0.374 | ||||
| < 60 | 271 (82.12) | 133 (79.64) | 71 (82.56) | 67 (87.01) | |
| ≥ 60 | 59 (17.88) | 34 (20.36) | 15 (17.44) | 10 (12.99) | |
| Smoking status | 0.112 | ||||
| No | 202 (61.21) | 94 (56.29) | 54 (62.79) | 54 (70.13) | |
| Yes | 128 (38.79) | 73 (43.71) | 32 (37.21) | 23 (29.87) | |
| Drinking status | 0.663 | ||||
| No | 223 (67.58) | 109 (65.27) | 60 (69.77) | 54 (70.13) | |
| Yes | 107 (32.42) | 58 (34.73) | 26 (30.23) | 23 (29.87) | |
| KPS status | 0.783 | ||||
| 60-70 | 11 (3.33) | 7 (4.19) | 2 (2.33) | 2 (2.60) | |
| 80-90 | 319 (96.67) | 160 (95.81) | 84 (97.67) | 75 (97.40) | |
| Weight loss status | 0.044 | ||||
| < 10% | 308 (93.33) | 154 (92.22) | 85 (98.84) | 69 (89.61) | |
| ≥ 10% | 22 (6.67) | 13 (7.78) | 1 (1.16) | 8 (10.39) | |
| T | < 0.001 | ||||
| 1 or 2 | 150 (45.45) | 57 (34.13) | 58 (67.44) | 35 (45.45) | |
| 3 or 4 | 180 (54.55) | 110 (65.87) | 28 (32.56) | 42 (54.55) | |
| N | < 0.001 | ||||
| 0 or 1 | 150 (45.45) | 98 (58.68) | 24 (27.91) | 28 (36.36) | |
| 2 or 3 | 180 (54.55) | 69 (41.32) | 62 (72.09) | 49 (63.64) | |
| TNM | 0.853 | ||||
| I-II | 63 (19.09) | 33 (19.76) | 17 (19.77) | 13 (16.88) | |
| III-IV | 267 (80.91) | 134 (80.24) | 69 (80.23) | 64 (83.12) | |
| Radiotherapy technique | 0.863 | ||||
| Non-IMRT | 196 (59.39) | 101 (60.48) | 49 (56.98) | 46 (59.74) | |
| IMRT | 134 (40.61) | 66 (39.52) | 37 (43.02) | 31 (40.26) | |
| Time of treatment | 0.632 | ||||
| 1990-1999 | 113 (34.24) | 57 (34.13) | 25 (29.07) | 31 (40.26) | |
| 2000-2009 | 126 (38.18) | 65 (38.92) | 36 (41.86) | 25 (32.47) | |
| 2010-2020 | 91 (27.58) | 45 (26.95) | 25 (29.07) | 21 (27.27) | |
| Chemotherapy | 0.149 | ||||
| No | 206 (62.42) | 112 (67.07) | 47 (54.65) | 47 (61.04) | |
| Yes | 124 (37.58) | 55 (32.93) | 39 (45.35) | 30 (38.96) | |
| Early or late recurrence | 0.109 | ||||
| Early | 156 (47.27) | 74 (44.31) | 49 (56.98) | 33 (42.86) | |
| Late | 174 (52.73) | 93 (55.69) | 37 (43.02) | 44 (57.14) | |
| rT | < 0.001 | ||||
| 0 | 86 (26.06) | 0 (0.00) | 86 (100.00) | 0 (0.00) | |
| 1 or 2 | 119 (36.06) | 74 (44.31) | 0 (0.00) | 45 (58.44) | |
| 3 or 4 | 125 (37.88) | 93 (55.69) | 0 (0.00) | 32 (41.56) | |
| rN | < 0.001 | ||||
| 0 | 167 (50.61) | 167 (100.00) | 0 (0.00) | 0 (0.00) | |
| 1 or 2 | 150 (45.45) | 0 (0.00) | 77 (89.53) | 73 (94.81) | |
| 3 | 13 (3.94) | 0 (0.00) | 9 (10.47) | 4 (5.19) | |
| rTNM | < 0.001 | ||||
| I-II | 158 (47.88) | 74 (44.31) | 57 (66.28) | 27 (35.06) | |
| III-IV | 172 (52.12) | 93 (55.69) | 29 (33.72) | 50 (64.94) | |
| Post recurrence treatment | < 0.001 | ||||
| Chemotherapy | 36 (10.91) | 19 (11.38) | 7 (8.14) | 10 (12.99) | |
| RT | 150 (45.45) | 90 (53.89) | 19 (22.09) | 41 (53.25) | |
| Surgery | 107 (32.42) | 33 (19.76) | 56 (65.12) | 18 (23.38) | |
| BST | 37 (11.21) | 25 (14.97) | 4 (4.65) | 8 (10.39) |
Table 1 compares the baseline characteristics of patients among the local, regional, and locoregional recurrence groups. Not surprisingly, the local recurrence group harbored more aggressive T stage at the initial diagnosis, and the regional recurrence group exhibited a more advanced initial N stage. There were also significant distinctions with regard to the post-recurrence management. More patients in the local and locoregional recurrence groups underwent re-irradiation, whereas the regional recurrence group received surgical resections more frequently. No differences between groups were found regarding gender, age, smoking habits, drinking habits, RT techniques or treatment era.
The detailed distribution of time to recurrence is illustrated in Figure 1A-C. In the overall cohort, the peak incidence of recurrence was observed in the 2nd year, contributing 29.4% of all recurrence events. Afterwards, a gradual decline in the annual number of recurrence was observed, resulting in the scenario that recurrences occurring within 5 years accounted for 82.8% (Figure 1A). Specifically for local recurrence, the highest recurrence (27.9%) was also noted in the 2nd year, with cumulative rates reaching 83.7% at the fifth year (Figure 1B). In cases of regional recurrence, the peak period was almost evenly distributed within the 1st and 2nd year, achieving a cumulative percentage of 81.0% by the 5th year (Figure 1C). There was no significant difference in timing distribution across various POF (Supplementary Figure 1).
The 24-month interval of time to recurrence was identified as the optimal timing cutoff to define early and late recurrence with most significantly distinct post-recurrence OS. Internal bootstrap validation (1000 resamples) confirmed the robu
Table 2 compares the characteristics of patients among 4 pairs of groups. More advanced T stage was significantly linked with overall early recurrence and ETR, but not with ENR or ETNR. Accordingly, more advanced N stage was signi
| Variables | Total | Local | Regional (n = 86) | Locoregional (n = 77) | ||||||||
| Early (n = 156) | Late (n = 174) | P value | ETR | LTR | P value | ENR | LNR | P value | ETNR | LTNR | P value | |
| Gender | 0.689 | 0.718 | 0.708 | 0.822 | ||||||||
| Male | 123 (78.85) | 134 (77.01) | 59 (79.73) | 72 (77.42) | 40 (81.63) | 29 (78.38) | 24 (72.73) | 33 (75.00) | ||||
| Female | 33 (21.15) | 40 (22.99) | 15 (20.27) | 21 (22.58) | 9 (18.37) | 8 (21.62) | 9 (27.27) | 11 (25.00) | ||||
| Age | 0.975 | 0.424 | 0.754 | 0.129 | ||||||||
| < 60 | 128 (82.05) | 143 (82.18) | 61 (82.43) | 72 (77.42) | 41 (83.67) | 30 (81.08) | 26 (78.79) | 41 (93.18) | ||||
| ≥ 60 | 28 (17.95) | 31 (17.82) | 13 (17.57) | 21 (22.58) | 8 (16.33) | 7 (18.92) | 7 (21.21) | 3 (6.82) | ||||
| Smoking status | 0.090 | 0.144 | 0.729 | 0.281 | ||||||||
| No | 88 (56.41) | 114 (65.52) | 37 (50.00) | 57 (61.29) | 30 (61.22) | 24 (64.86) | 21 (63.64) | 33 (75.00) | ||||
| Yes | 68 (43.59) | 60 (34.48) | 37 (50.00) | 36 (38.71) | 19 (38.78) | 13 (35.14) | 12 (36.36) | 11 (25.00) | ||||
| Drinking status | 0.922 | 0.922 | 0.300 | 0.350 | ||||||||
| No | 105 (67.31) | 118 (67.82) | 48 (64.86) | 61 (65.59) | 32 (65.31) | 28 (75.68) | 25 (75.76) | 29 (65.91) | ||||
| Yes | 51 (32.69) | 56 (32.18) | 26 (35.14) | 32 (34.41) | 17 (34.69) | 9 (24.32) | 8 (24.24) | 15 (34.09) | ||||
| KPS status | 0.177 | 0.213 | 0.504 | 0.504 | ||||||||
| 60-70 | 3 (1.92) | 8 (4.60) | 1 (1.35) | 6 (6.45) | 2 (4.08) | 0 (0.00) | 0 (0.00) | 2 (4.55) | ||||
| 80-90 | 153 (98.08) | 166 (95.40) | 73 (98.65) | 87 (93.55) | 47 (95.92) | 37 (100.00) | 33 (100.00) | 42 (95.45) | ||||
| Weight loss status | 0.250 | 0.471 | 1.000 | 0.419 | ||||||||
| < 10% | 143 (91.67) | 165 (94.83) | 67 (90.54) | 87 (93.55) | 48 (97.96) | 37 (100.00) | 28 (84.85) | 41 (93.18) | ||||
| ≥ 10% | 13 (8.33) | 9 (5.17) | 7 (9.46) | 6 (6.45) | 1 (2.04) | 0 (0.00) | 5 (15.15) | 3 (6.82) | ||||
| T | 0.008 | 0.002 | 0.627 | 0.064 | ||||||||
| 1 or 2 | 59 (37.82) | 91 (52.30) | 16 (21.62) | 41 (44.09) | 32 (65.31) | 26 (70.27) | 11 (33.33) | 24 (54.55) | ||||
| 3 or 4 | 97 (62.18) | 83 (47.70) | 58 (78.38) | 52 (55.91) | 17 (34.69) | 11 (29.73) | 22 (66.67) | 20 (45.45) | ||||
| N | 0.028 | 0.162 | 0.416 | 0.338 | ||||||||
| 0 or 1 | 61 (39.10) | 89 (51.15) | 39 (52.70) | 59 (63.44) | 12 (24.49) | 12 (32.43) | 10 (30.30) | 18 (40.91) | ||||
| 2 or 3 | 95 (60.90) | 85 (48.85) | 35 (47.30) | 34 (36.56) | 37 (75.51) | 25 (67.57) | 23 (69.70) | 26 (59.09) | ||||
| TNM | 0.002 | 0.003 | 0.708 | 0.114 | ||||||||
| I-II | 19 (12.18) | 44 (25.29) | 7 (9.46) | 26 (27.96) | 9 (18.37) | 8 (21.62) | 3 (9.09) | 10 (22.73) | ||||
| III-IV | 137 (87.82) | 130 (74.71) | 67 (90.54) | 67 (72.04) | 40 (81.63) | 29 (78.38) | 30 (90.91) | 34 (77.27) | ||||
| Radiotherapy technique | 0.351 | 0.473 | 1.000 | 0.569 | ||||||||
| Non-IMRT | 88 (56.41) | 108 (62.07) | 42 (56.76) | 59 (63.44) | 28 (57.14) | 21 (56.76) | 18 (54.55) | 28 (63.64) | ||||
| IMRT | 68 (43.59) | 66 (37.93) | 32 (43.24) | 34 (36.56) | 21 (42.86) | 16 (43.24) | 15 (45.45) | 16 (36.36) | ||||
| Time of treatment | 0.008 | 0.205 | 0.032 | 0.531 | ||||||||
| 1990-1999 | 41 (26.28) | 72 (41.38) | 20 (27.03) | 37 (39.78) | 9 (18.37) | 16 (43.24) | 12 (36.36) | 19 (43.18) | ||||
| 2000-2009 | 71 (45.51) | 55 (31.61) | 33 (44.59) | 32 (34.41) | 25 (51.02) | 11 (29.73) | 13 (39.39) | 12 (27.27) | ||||
| 2010-2020 | 44 (28.21) | 47 (27.01) | 21 (28.38) | 24 (25.81) | 15 (30.61) | 10 (27.03) | 8 (24.24) | 13 (29.55) | ||||
| Chemotherapy | 0.753 | 0.902 | 0.733 | 0.946 | ||||||||
| No | 96 (61.54) | 110 (63.22) | 50 (67.57) | 62 (66.67) | 26 (53.06) | 21 (56.76) | 20 (60.61) | 27 (61.36) | ||||
| Yes | 60 (38.46) | 64 (36.78) | 24 (32.43) | 31 (33.33) | 23 (46.94) | 16 (43.24) | 13 (39.39) | 17 (38.64) | ||||
| rT | 0.069 | 0.488 | - | 0.423 | ||||||||
| 0 | 49 (31.41) | 37 (21.26) | 0 (0.00) | 0 (0.00) | 49 (100.00) | 37 (100.00) | 0 (0.00) | 0 (0.00) | ||||
| 1 or 2 | 56 (35.90) | 63 (36.21) | 35 (47.30) | 39 (41.94) | 0 (0.00) | 0 (0.00) | 21 (63.64) | 24 (54.55) | ||||
| 3 or 4 | 51 (32.69) | 74 (42.53) | 39 (52.70) | 54 (58.06) | 0 (0.00) | 0 (0.00) | 12 (36.36) | 20 (45.45) | ||||
| rN | 0.532 | - | 0.791 | 0.824 | ||||||||
| 0 | 74 (47.44) | 93 (53.45) | 74 (100.00) | 93 (100.00) | 0 (0.00) | 0 (0.00) | 0 (0.00) | 0 (0.00) | ||||
| 1 or 2 | 75 (48.08) | 75 (43.10) | 0 (0.00) | 0 (0.00) | 43 (87.76) | 34 (91.89) | 32 (96.97) | 41 (93.18) | ||||
| 3 | 7 (4.49) | 6 (3.45) | 0 (0.00) | 0 (0.00) | 6 (12.24) | 3 (8.11) | 1 (3.03) | 3 (6.82) | ||||
| rTNM | 0.342 | 0.488 | 0.826 | 0.836 | ||||||||
| I-II | 79 (50.64) | 79 (45.40) | 35 (47.30) | 39 (41.94) | 32 (65.31) | 25 (67.57) | 12 (36.36) | 15 (34.09) | ||||
| III-IV | 77 (49.36) | 95 (54.60) | 39 (52.70) | 54 (58.06) | 17 (34.69) | 12 (32.43) | 21 (63.64) | 29 (65.91) | ||||
| Post recurrence treatment | 0.054 | 0.283 | 0.516 | 0.573 | ||||||||
| Chemotherapy | 18 (11.54) | 18 (10.34) | 11 (14.86) | 8 (8.60) | 3 (6.12) | 4 (10.81) | 4 (12.12) | 6 (13.64) | ||||
| RT | 64 (41.03) | 86 (49.43) | 38 (51.35) | 52 (55.91) | 11 (22.45) | 8 (21.62) | 15 (45.45) | 26 (59.09) | ||||
| Surgery | 61 (39.10) | 46 (26.44) | 17 (22.97) | 16 (17.20) | 34 (69.39) | 22 (59.46) | 10 (30.30) | 8 (18.18) | ||||
| BST | 13 (8.33) | 24 (13.79) | 8 (10.81) | 17 (18.28) | 1 (2.04) | 3 (8.11) | 4 (12.12) | 4 (9.09) | ||||
Multivariable analysis of post-recurrence OS is exhibited in Table 3. Age over 60 was associated with an unfavorable prognosis in the entire group [hazard ratio (HR) = 1.73, 95%CI: 1.21-2.48, P = 0.003] and regional recurrence group (HR = 2.53, 95%CI: 1.20-5.31, P = 0.014). In the regional failure group, female gender was an independent factor associated with more favorable post-recurrence OS (HR = 0.29, 95%CI: 0.11-0.76, P = 0.012). More advanced rTNM stage was related to worse post-recurrence OS in the entire group (HR = 1.88, 95%CI: 1.38-2.56, P < 0.001) and the regional recurrence group (HR = 3.26, 95%CI: 1.68-6.34, P < 0.001). In terms of recurrence timing, late recurrence was linked with improved post-recurrence OS in the locoregional recurrence subgroup (HR = 0.37, 95%CI: 0.19-0.76, P = 0.006), whereas no similar trend was observed in other subgroups. Accordingly, patients with late recurrence exhibited notably better post-recurrence OS in the locoregional recurrence cohort (median: 45.4 months vs 17.2 months, P = 0.004) (Figure 2). No significant differences were found between early and late recurrence subgroups among the entire cohort (median: 45.4 months vs 29.7 months, P = 0.401), LTR and ETR (median: 42.1 months vs 29.7 months, P = 0.783), LNR and ENR (median: 60.0 months vs 59.2 months, P = 0.817), respectively.
| Variables | Total (n = 330) | Local (n = 167) | Regional (n = 86) | Locoregional (n = 77) | ||||
| P value | HR (95%CI) | P value | HR (95%CI) | P value | HR (95%CI) | P value | HR (95%CI) | |
| Gender | ||||||||
| Male | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | ||||
| Female | 0.135 | 0.76 (0.53-1.09) | 0.358 | 0.79 (0.47-1.31) | 0.012 | 0.29 (0.11-0.76) | 0.394 | 1.35 (0.68-2.66) |
| Age | ||||||||
| < 60 | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | ||||
| ≥ 60 | 0.003 | 1.73 (1.21-2.48) | 0.056 | 1.61 (0.99-2.61) | 0.014 | 2.53 (1.20-5.31) | 0.385 | 1.50 (0.60-3.78) |
| TNM | ||||||||
| I-II | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | ||||
| III-IV | 0.269 | 1.26 (0.84-1.88) | 0.053 | 1.77 (0.99-3.17) | 0.553 | 0.79 (0.36-1.74) | 0.422 | 0.65 (0.23-1.85) |
| Early or late recurrence | ||||||||
| Early | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | ||||
| Late | 0.461 | 0.89 (0.66-1.21) | 0.872 | 1.04 (0.67-1.59) | 0.561 | 1.21 (0.64-2.28) | 0.006 | 0.37 (0.19-0.76) |
| rTNM | ||||||||
| I-II | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | 1.00 (reference) | ||||
| III-IV | < 0.001 | 1.88 (1.38-2.56) | 0.132 | 1.39 (0.91-2.13) | < 0.001 | 3.26 (1.68-6.34) | 0.451 | 1.32 (0.64-2.74) |
After excluding 26 patients who died or were censored prior to the 24-month timepoint, 304 patients remained eligible for the landmark OS analysis. Multivariable Cox regression suggested that late recurrence independently predicted superior post-landmark survival in the entire cohort (HR = 0.38, 95%CI: 0.27-0.53, P < 0.001) and across all failure subgroups: Local (HR = 0.45, 95%CI: 0.29-0.72, P < 0.001), regional (HR = 0.44, 95%CI: 0.22-0.88, P = 0.021), and locoregional recurrence (HR = 0.11, 95%CI: 0.04-0.26, P < 0.001) (Supplementary Table 1). Relevant survival comparisons are exhibited in Supplementary Figure 2.
Based on 330 recurrent NPC patients, we observed that the peak recurrence period was in the second year after the completion of primary treatment. The time to recurrence of 24 months was identified as the optimal cutoff for differentiating early from late recurrence. Patients experiencing early recurrence were prone to harbor more advanced T, N and TNM stage of disease. Early recurrence demonstrated markedly worse post-recurrence OS in the cohort with locoregional recurrence whereas no significantly inferior survival was found in patients with early recurrence in exclusive local or regional site.
In the present study, we first illustrated the chronological distribution of various recurrence patterns in NPC. The first 2 year was revealed to be the peak period of local and regional recurrence, which was consistent with corresponding reports from endemic centers[9-11] and our previous report regarding the dynamic hazard of progression[12]. In addition, we further unveiled a basically consistent distribution of the failure time-frame among various recurrence pattern. This result validates the paradigm that primary tumor and cervical regions could share a uniform post-treatment monitoring schedule. These findings also highlight the importance of dynamic risk-adapted surveillance strategies. For instance, intensified follow-up during the first 2-3 years is highly recommended to timely detect recurrence and provide salvage therapies to mitigate the devastating effect of tumor relapse on OS.
To understand the predisposing factors associated with the site-specific recurrence, we compared the clinical characteristics between different recurrence sites. Patients suffering from local recurrence tended to have more advanced initial T stage, while those experiencing regional recurrence had more unfavorable initial N stage. Additionally, we also analyzed the predictors that distinguish early and late recurrence. Patients with more advanced T stage or overall TN stage were more likely to suffer from early recurrence and ETR. More advanced N stage was more frequently found in entire early recurrence cohort whereas no such trend was observed in other three sub-cohorts. This indicates that the ETR may result from the initial primary tumor burden. On the contrary, LNR may be more related to historical treatment modalities or the host factors, such as immune status or genetic polymorphisms[13], while less related to the gross nodal burden.
Although studies variably defined “early” recurrence in NPC, the cutoff of 24 months was most frequently rendered[9,14,15]. However, most studies identified the early recurrence as ≤ 24 months mainly on the basis of arbitrary or empirical definition. In our study, we attempted to identify the optimal cutoff timing by using a more rational approach, the maximally selected rank method, to maximize the post-recurrence OS difference. Coincidentally, it was still the 24-month that was considered the most suitable cutoff to distinguish early and late recurrence. The underlying mechanisms of early or late recurrences remain inconclusive. Biologically, the early progression indicated resistance to RT and chemotherapy, inadequate intensity of therapy, or accelerated re-population after initial treatment[16]. Instead, decreased anti-tumor immune response or enhanced tumor stem cell properties may, in part, explain the occurrence of late relapse[17-20].
By rendering 24-month as the cutoff for defining early and late recurrence, several studies consistently reported significantly inferior OS in patients who had early recurrence within 2 years after treatment[9,14,15]. In general, the reported OS for early recurrence was around 30%-40% and that for late recurrence was about 40%-70%. However, when survival is calculated from initial diagnosis, this observation is mathematically confounded by time bias, as late-relapsing patients are guaranteed to survive through their longer recurrence-free latency period. To diminish the guarantee-time bias, we conducted a complementary 24-month landmark OS analysis and revealed that late recurrence was indepen
Nevertheless, post-recurrence OS represents a clinically more meaningful assessment since it could better guide sal
This intriguing disparity may be explained by underlying tumor biology, particularly intratumor heterogeneity and clonal evolution[21,22]. Simultaneous locoregional recurrence represents a highly aggressive phenotype characterized by widespread spatial dissemination and multiclonal seeding. Under the intense selection pressure of definitive chemoradiotherapy, early relapsing tumors likely emerge through the rapid clonal selection of intrinsically radioresistant and che
Current study has several limitations. Firstly, its retrospective design introduces unavoidable inherent biases. Se
In conclusion, the peak recurrence period of NPC was within the first two years after the completion of primary treat
We thank all the patients and their families for their contribution to this study.
| 1. | Lai SZ, Li WF, Chen L, Luo W, Chen YY, Liu LZ, Sun Y, Lin AH, Liu MZ, Ma J. How does intensity-modulated radiotherapy versus conventional two-dimensional radiotherapy influence the treatment results in nasopharyngeal carcinoma patients? Int J Radiat Oncol Biol Phys. 2011;80:661-668. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 436] [Cited by in RCA: 538] [Article Influence: 35.9] [Reference Citation Analysis (1)] |
| 2. | Wu LR, Liu YT, Jiang N, Fan YX, Wen J, Huang SF, Guo WJ, Bian XH, Wang FJ, Li F, Song D, Wu JF, Jiang XS, Liu JY, He X. Ten-year survival outcomes for patients with nasopharyngeal carcinoma receiving intensity-modulated radiotherapy: An analysis of 614 patients from a single center. Oral Oncol. 2017;69:26-32. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 80] [Cited by in RCA: 88] [Article Influence: 9.8] [Reference Citation Analysis (0)] |
| 3. | Lee AWM, Ng WT, Chan JYW, Corry J, Mäkitie A, Mendenhall WM, Rinaldo A, Rodrigo JP, Saba NF, Strojan P, Suárez C, Vermorken JB, Yom SS, Ferlito A. Management of locally recurrent nasopharyngeal carcinoma. Cancer Treat Rev. 2019;79:101890. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 90] [Cited by in RCA: 267] [Article Influence: 38.1] [Reference Citation Analysis (0)] |
| 4. | Ng WT, Soong YL, Ahn YC, AlHussain H, Choi HCW, Corry J, Grégoire V, Harrington KJ, Hu CS, Jensen K, Kwong DL, Langendijk JA, Le QT, Lee NY, Lin JC, Lu TX, Mendenhall WM, O'Sullivan B, Ozyar E, Pan JJ, Peters LJ, Poh SS, Rosenthal DI, Sanguineti G, Tao Y, Wee JT, Yom SS, Chua MLK, Lee AWM. International Recommendations on Reirradiation by Intensity Modulated Radiation Therapy for Locally Recurrent Nasopharyngeal Carcinoma. Int J Radiat Oncol Biol Phys. 2021;110:682-695. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 50] [Cited by in RCA: 77] [Article Influence: 15.4] [Reference Citation Analysis (0)] |
| 5. | Mase S, Shinjo K, Totani H, Katsushima K, Arakawa A, Takahashi S, Lai HC, Lin RI, Chan MWY, Sugiura-Ogasawara M, Kondo Y. ZNF671 DNA methylation as a molecular predictor for the early recurrence of serous ovarian cancer. Cancer Sci. 2019;110:1105-1116. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 19] [Cited by in RCA: 32] [Article Influence: 4.6] [Reference Citation Analysis (0)] |
| 6. | Lee DY, Abraham J, Ross E, Ridge JA, Lango MN, Liu JC, Bauman JR, Avkshtol V, Galloway TJ. Rapid recurrence in head and neck cancer: Underappreciated problem with poor outcome. Head Neck. 2021;43:212-222. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 7] [Cited by in RCA: 16] [Article Influence: 2.7] [Reference Citation Analysis (0)] |
| 7. | Tang LL, Chen YP, Mao YP, Wang ZX, Guo R, Chen L, Tian L, Lin AH, Li L, Sun Y, Ma J. Validation of the 8th Edition of the UICC/AJCC Staging System for Nasopharyngeal Carcinoma From Endemic Areas in the Intensity-Modulated Radiotherapy Era. J Natl Compr Canc Netw. 2017;15:913-919. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 86] [Cited by in RCA: 155] [Article Influence: 17.2] [Reference Citation Analysis (0)] |
| 8. | Lausen B, Schumacher M. Maximally Selected Rank Statistics. Biometrics. 1992;48:73. [DOI] [Full Text] |
| 9. | Li F, Chen FP, Chen YP, Chen Y, He XJ, Huang XD, Zheng ZQ, Zheng WH, Liu X, Sun Y, Zhou GQ. Clinical Characteristics and Prognostic Factors of Early and Late Recurrence After Definitive Radiotherapy for Nasopharyngeal Carcinoma. Front Oncol. 2020;10:1469. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 3] [Cited by in RCA: 12] [Article Influence: 2.0] [Reference Citation Analysis (0)] |
| 10. | Li Y, Huang Z, Zeng X, Pan Y, Wu L, Wang J, Chen R, Xie Y, Lai J, Lin D, Qiu S. Early recurrence as a pivotal event in nasopharyngeal carcinoma: identifying predictors and key molecular signals for survivors. Head Face Med. 2024;20:55. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 7] [Reference Citation Analysis (0)] |
| 11. | Zhang LL, Zheng WH, Zhu WJ, Deng QL, Peng JL, Li YY, Sun Y, Lin L. Prognostic models for early and late tumor progression prediction in nasopharyngeal carcinoma: An analysis of 8292 endemic cases. Cancer Med. 2023;12:5384-5396. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 3] [Cited by in RCA: 5] [Article Influence: 1.7] [Reference Citation Analysis (0)] |
| 12. | Wang J, Huang X, Sun S, Wang K, Qu Y, Chen X, Wu R, Zhang Y, Liu Q, Zhang J, Luo J, Xiao J, Gao L, Xu G, Hu C, Li YX, Yi J. Stage-dependent conditional survival and failure hazard of non-metastatic nasopharyngeal carcinoma after intensity-modulated radiation therapy: Clinical implications for treatment strategies and surveillance. Cancer Med. 2021;10:3613-3621. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 20] [Reference Citation Analysis (0)] |
| 13. | Chen YP, Chan ATC, Le QT, Blanchard P, Sun Y, Ma J. Nasopharyngeal carcinoma. Lancet. 2019;394:64-80. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1109] [Cited by in RCA: 2408] [Article Influence: 344.0] [Reference Citation Analysis (4)] |
| 14. | Zhang LL, Xu F, He WT, Huang MY, Song D, Li YY, Deng QL, Huang YS, Wang T, Shao JY. Development and validation of a prognostic nomogram for the pre-treatment prediction of early metachronous metastasis in endemic nasopharyngeal carcinoma: a big-data intelligence platform-based analysis. Ther Adv Med Oncol. 2020;12:1758835920978132. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 6] [Cited by in RCA: 8] [Article Influence: 1.3] [Reference Citation Analysis (0)] |
| 15. | Chee J, Ting Y, Ong YK, Chao SS, Loh KS, Lim CM. Relapse status as a prognostic factor in patients receiving salvage surgery for recurrent or residual nasopharyngeal cancer after definitive treatment. Head Neck. 2016;38:1393-1400. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 11] [Cited by in RCA: 24] [Article Influence: 2.4] [Reference Citation Analysis (0)] |
| 16. | Liu Y, Yan W, Qi X, Zhang Y, Wang K, Qu Y, Chen X, Zhang J, Luo J, Li YX, Huang X, Wu R, Wang J, Yi J. Significance of longitudinal Epstein-Barr virus DNA combined with multipoint tumor response for dynamic risk stratification and treatment adaptation in nasopharyngeal carcinoma. Cancer Lett. 2024;605:217276. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 3] [Cited by in RCA: 10] [Article Influence: 5.0] [Reference Citation Analysis (0)] |
| 17. | Chen Y, Feng Y, Lin Y, Zhou X, Wang L, Zhou Y, Lin K, Cai L. GSTM3 enhances radiosensitivity of nasopharyngeal carcinoma by promoting radiation-induced ferroptosis through USP14/FASN axis and GPX4. Br J Cancer. 2024;130:755-768. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 20] [Cited by in RCA: 43] [Article Influence: 21.5] [Reference Citation Analysis (0)] |
| 18. | Li JY, Zhao Y, Gong S, Wang MM, Liu X, He QM, Li YQ, Huang SY, Qiao H, Tan XR, Ye ML, Zhu XH, He SW, Li Q, Liang YL, Chen KL, Huang SW, Li QJ, Ma J, Liu N. TRIM21 inhibits irradiation-induced mitochondrial DNA release and impairs antitumour immunity in nasopharyngeal carcinoma tumour models. Nat Commun. 2023;14:865. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 110] [Reference Citation Analysis (4)] |
| 19. | Lu Y, Liang Y, Zheng X, Deng X, Huang W, Zhang G. EVI1 promotes epithelial-to-mesenchymal transition, cancer stem cell features and chemo-/radioresistance in nasopharyngeal carcinoma. J Exp Clin Cancer Res. 2019;38:82. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 22] [Cited by in RCA: 41] [Article Influence: 5.9] [Reference Citation Analysis (0)] |
| 20. | Zhou F, Shayan G, Sun S, Huang X, Chen X, Wang K, Qu Y, Wu R, Zhang Y, Liu Q, Zhang J, Luo J, Shi X, Liu Y, Liang B, Li YX, Wang J, Yi J. Spatial architecture of regulatory T-cells correlates with disease progression in patients with nasopharyngeal cancer. Front Immunol. 2022;13:1015283. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 18] [Reference Citation Analysis (0)] |
| 21. | McGranahan N, Swanton C. Clonal Heterogeneity and Tumor Evolution: Past, Present, and the Future. Cell. 2017;168:613-628. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1799] [Cited by in RCA: 2090] [Article Influence: 232.2] [Reference Citation Analysis (5)] |
| 22. | Dagogo-Jack I, Shaw AT. Tumour heterogeneity and resistance to cancer therapies. Nat Rev Clin Oncol. 2018;15:81-94. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 3236] [Cited by in RCA: 2733] [Article Influence: 341.6] [Reference Citation Analysis (10)] |