Phase III Study of Concurrent Chemoradiotherapy Versus Radiotherapy Alone for Advanced Nasopharyngeal Carcinoma: Positive Effect on Overall and Progression-Free Survival

2003 ◽  
Vol 21 (4) ◽  
pp. 631-637 ◽  
Author(s):  
Jin-Ching Lin ◽  
Jian-Sheng Jan ◽  
Chen-Yi Hsu ◽  
Wen-Miin Liang ◽  
Rong-San Jiang ◽  
...  
Keyword(s):  
Nasopharyngeal Carcinoma ◽  
Concurrent Chemoradiotherapy ◽  
Combined Treatment ◽  
Survival Rates ◽  
Progression Free Survival ◽  
Patient Characteristics ◽  
Concurrent Chemotherapy ◽  
Phase Iii ◽  
Rank Test ◽  
Free Survival

Purpose: Nasopharyngeal carcinoma (NPC) is a radiosensitive and chemosensitive tumor. This randomized phase III trial compared concurrent chemoradiotherapy (CCRT) versus radiotherapy (RT) alone in patients with advanced NPC. Patients and Methods: From December 1993 to April 1999, 284 patients with 1992 American Joint Committee on Cancer stage III to IV (M0) NPC were randomly allocated into two arms. Similar dosage and fractionation of RT was administered in both arms. The investigational arm received two cycles of concurrent chemotherapy with cisplatin 20 mg/m2/d plus fluorouracil 400 mg/m2/d by 96-hour continuous infusion during the weeks 1 and 5 of RT. Survival analysis was estimated by the Kaplan-Meier method and compared by the log-rank test. Results: Baseline patient characteristics were comparable in both arms. After a median follow-up of 65 months, 26.2% (37 of 141) and 46.2% (66 of 143) of patients developed tumor relapse in the CCRT and RT-alone groups, respectively. The 5-year overall survival rates were 72.3% for the CCRT arm and 54.2% for the RT-only arm (P = .0022). The 5-year progression-free survival rates were 71.6% for the CCRT group compared with 53.0% for the RT-only group (P = .0012). Although significantly more toxicity was noted in the CCRT arm, including leukopenia and emesis, compliance with the combined treatment was good. The second cycle of concurrent chemotherapy was refused by nine patients and was delayed for ≥ 1 week for another nine patients. There were no treatment-related deaths in either arm. Conclusion: We conclude that CCRT is superior to RT alone for patients with advanced NPC in endemic areas.

Concurrent Chemotherapy-Radiotherapy Compared With Radiotherapy Alone in Locoregionally Advanced Nasopharyngeal Carcinoma: Progression-Free Survival Analysis of a Phase III Randomized Trial

2002 ◽  
Vol 20 (8) ◽  
pp. 2038-2044 ◽  
Author(s):  
A.T.C. Chan ◽  
P.M.L. Teo ◽  
R.K. Ngan ◽  
T.W. Leung ◽  
W.H. Lau ◽  
...  
Keyword(s):  
Nasopharyngeal Carcinoma ◽  
Progression Free Survival ◽  
Patient Characteristics ◽  
Tumor Stage ◽  
Concurrent Chemotherapy ◽  
Phase Iii ◽  
Free Survival ◽  
Advanced Tumor ◽  
Nodal Size ◽  
Significant Difference

PURPOSE: Nasopharyngeal carcinoma (NPC) is highly sensitive to both radiotherapy (RT) and chemotherapy. This randomized phase III trial compared concurrent cisplatin-RT (CRT) with RT alone in patients with locoregionally advanced NPC. PATIENTS AND METHODS: Patients with Ho’s N2 or N3 stage or N1 stage with nodal size ≥ 4 cm were randomized to receive cisplatin 40 mg/m2 weekly up to 8 weeks concurrently with radical RT (CRT) or RT alone. The primary end point was progression-free survival (PFS). RESULTS: Three hundred fifty eligible patients were randomized. Baseline patient characteristics were comparable in both arms. There were significantly more toxicities, including mucositis, myelosuppression, and weight loss in the CRT arm. There were no treatment-related deaths in the CRT arm, and one patient died during treatment in the RT-alone arm. At a median follow-up of 2.71 years, the 2-year PFS was 76% in the CRT arm and 69% in the RT-alone arm (P = .10) with a hazards ratio of 1.367 (95% confidence interval [CI], 0.93 to 2.00). The treatment effect had a significant covariate interaction with tumor stage, and a subgroup analysis demonstrated a highly significant difference in favor of the CRT arm in Ho’s stage T3 (P = .0075) with a hazards ratio of 2.328 (95% CI, 1.26 to 4.28). For T3 stage, the time to first distant failure was statistically significantly different in favor of the CRT arm (P = .016). CONCLUSION: Concurrent CRT is well tolerated in patients with advanced NPC in endemic areas. Although PFS was not significantly different between the concurrent CRT arm and the RT-alone arm in the overall comparison, PFS was significantly prolonged in patients with advanced tumor and node stages.

Randomized Phase III Trial of Fluorouracil Alone Versus Fluorouracil Plus Cisplatin Versus Uracil and Tegafur Plus Mitomycin in Patients With Unresectable, Advanced Gastric Cancer: The Japan Clinical Oncology Group Study (JCOG9205)

2003 ◽  
Vol 21 (1) ◽  
pp. 54-59 ◽  
Author(s):  
Atsushi Ohtsu ◽  
Yasuhiro Shimada ◽  
Kuniaki Shirao ◽  
Narikazu Boku ◽  
Ichinosuke Hyodo ◽  
...  
Keyword(s):  
Gastric Cancer ◽  
Advanced Gastric Cancer ◽  
Controlled Trial ◽  
Survival Rates ◽  
Progression Free Survival ◽  
Toxic Effects ◽  
Phase Iii ◽  
Rank Test ◽  
Survival Times ◽  
Free Survival

Purpose: To compare fluorouracil (FU) alone with FU plus cisplatin (FP) and with uracil and tegafur plus mitomycin (UFTM) for patients with advanced gastric cancer in a prospective, randomized, controlled trial. Patients and Methods: A total of 280 patients with advanced gastric cancer were randomly allocated and analyzed for survival, response, and toxicity. The survival curves were compared between groups by log-rank test on an intent-to-treat basis. Results: At the interim analysis, the UFTM arm showed a significantly inferior survival with higher incidences of hematologic toxic effects than did control arm FU alone, and the registration to UFTM was terminated. Both investigational regimens, FP and UFTM, had a significantly higher incidence of hematologic toxic effects than FU alone, although the effects were manageable. The overall response rates of the FU-alone, FP, and UFTM arms were 11%, 34%, and 9%, respectively. The median progression-free survival was 1.9 months with FU alone, 3.9 months with FP, and 2.4 months with UFTM, respectively. Although FP demonstrated a higher response rate (P < .001) and longer progression-free survival than did FU alone (P < .001), no differences in overall survival were observed between the arms. The median survival times and 1-year survival rates were 7.1 months and 28% with FU, 7.3 months and 29% with FP, and 6.0 months and 16% with UFTM, respectively. Conclusion: Neither investigational regimen, FP nor UFTM, showed a survival advantage as compared with FU alone. FU alone will remain a reference arm in our future trial for advanced gastric cancer.

scholarly journals Efficacy of Concurrent Chemoradiotherapy In Subgroups of Stage III Nasopharyngeal Carcinoma: An Analysis Based On 10-Year Follow-Up

2021 ◽  
Author(s):  
Lei Wang ◽  
Wu Zheng ◽  
Wanqin Cheng ◽  
Dehuan Xie ◽  
Feifei Lin ◽  
...  
Keyword(s):  
Nasopharyngeal Carcinoma ◽  
Concurrent Chemoradiotherapy ◽  
Cox Regression ◽  
Disease Free Survival ◽  
Concurrent Chemotherapy ◽  
Stage Iii ◽  
Rank Test ◽  
Free Survival ◽  
N2 Disease

Abstract PurposeTo evaluate the efficacy of concurrent chemoradiotherapy (CCRT) in subgroups of stage III nasopharyngeal carcinoma (NPC) in the context of intensity-modulated radiotherapy (IMRT).Methods272 patients with stage III NPC who underwent IMRT with or without concurrent chemotherapy (CCT) were retrospectively reviewed. Clinicopathological features were evaluated by a Cox regression model to identify independent prognostic factors. Survival outcomes were assessed using Kaplan-Meier method and log-rank test.ResultsThe median follow-up time was 108 months. The 10-year locoregional-free survival (LRFS), distant metastasis-free survival (DMFS), disease-free survival (DFS), and overall survival (OS) rates were 87.8%, 80.7%, 68.8%, and 74.9%, respectively. Multivariate analysis showed that the N classification was significantly associated with DMFS (hazard ratio [HR] 3.616, 95% confidence interval [CI] 1.387-9.428, P=0.009), DFS (HR 2.417, 95% CI 1.291-4.423, P=0.006), and OS (HR 3.024, 95% CI 1.385-6.602, P=0.005). In patients with T1-3N2 disease, CCRT was associated with improved 10-year LRFS (89.6% vs. 65.4%, P=0.005), DFS (71.9% vs. 39.4%, P=0.001) and OS (80.0% vs. 50.5%, P=0.004) compared with IMRT alone. However, in patients with T3N0-1 disease, no significant survival differences were observed between patients treated with IMRT alone and CCRT (P>0.05). ConclusionCCRT is an effective therapy in stage III NPC, especially for patients with N2 disease but N0-1 disease. Individualized treatment strategies are essential for patients with varying disease risks.

scholarly journals Efficacy of concurrent chemoradiotherapy in subgroups of stage III nasopharyngeal carcinoma: an analysis based on 10-year follow-up

2021 ◽  
Vol 16 (1) ◽  
Author(s):  
Lei Wang ◽  
Zheng Wu ◽  
Wanqin Cheng ◽  
Dehuan Xie ◽  
Feifei Lin ◽  
...  
Keyword(s):  
Nasopharyngeal Carcinoma ◽  
Concurrent Chemoradiotherapy ◽  
Cox Regression ◽  
Disease Free Survival ◽  
Concurrent Chemotherapy ◽  
Stage Iii ◽  
Rank Test ◽  
Free Survival ◽  
N2 Disease

Abstract Purpose To evaluate the efficacy of concurrent chemoradiotherapy (CCRT) in subgroups of stage III nasopharyngeal carcinoma (NPC) in the context of intensity-modulated radiotherapy (IMRT). Methods A total of 272 patients with stage III NPC who underwent IMRT with or without concurrent chemotherapy were retrospectively reviewed. Clinicopathological features were evaluated by a Cox regression model to identify independent prognostic factors. Survival outcomes were assessed using the Kaplan–Meier method and log-rank test. Results The median follow-up time was 108 months. The 10-year locoregional-free survival (LRFS), distant metastasis-free survival (DMFS), disease-free survival (DFS), and overall survival (OS) rates were 87.8%, 80.7%, 68.8%, and 74.9%, respectively. Multivariate analysis showed that the N classification was significantly associated with DMFS (hazard ratio [HR] 3.616, 95% confidence interval [CI] 1.387–9.428, P = 0.009), DFS (HR 2.417, 95% CI 1.291–4.423, P = 0.006), and OS (HR 3.024, 95% CI 1.385–6.602, P = 0.005). In patients with T1-3N2 disease, CCRT was associated with improved 10-year LRFS (89.6% vs. 65.4%, P = 0.005), DFS (71.9% vs. 39.4% P = 0.001) and OS (80.0% vs. 50.5%, P = 0.004) compared with IMRT alone. However, in patients with T3N0-1 disease, no significant survival differences were observed between patients treated with IMRT alone and CCRT (P > 0.05). Conclusions CCRT is an effective therapy in stage III NPC, especially for patients with N2 disease, but IMRT alone may be adequate for N0-1 disease. Individualized treatment strategies are essential for patients with varying disease risks.

Phase III Study Comparing Standard Radiotherapy With or Without Weekly Oxaliplatin in Treatment of Locoregionally Advanced Nasopharyngeal Carcinoma: Preliminary Results

2005 ◽  
Vol 23 (33) ◽  
pp. 8461-8468 ◽  
Author(s):  
Li Zhang ◽  
Chong Zhao ◽  
Pei-Jian Peng ◽  
Li-Xia Lu ◽  
Pei-Yu Huang ◽  
...  
Keyword(s):  
Nasopharyngeal Carcinoma ◽  
Concurrent Chemoradiotherapy ◽  
Patient Characteristics ◽  
Phase Iii ◽  
Response Analysis ◽  
Relapse Free Survival ◽  
Free Survival ◽  
Significant Difference ◽  
Phase Iii Study

Purpose A prospective, randomized, phase III study was performed to evaluate the feasibility and efficacy of concurrent weekly oxaliplatin with radiotherapy in patients with locoregionally advanced nasopharyngeal carcinoma (NPC). Patients and Methods From January 2001 to January 2003, 115 patients with locoregionally advanced NPC were randomly assigned to either radiotherapy (RT) alone (56 patients) or concurrent chemoradiotherapy (CCRT; 59 patients). All patient characteristics were well balanced in both arms. CCRT with oxaliplatin 70 mg/m2 weekly was administered for six doses from the first day of RT. Results All patients were eligible for toxicity and response analysis. Compliance with the protocol treatment was excellent, with 97% of patients completing all planned doses of oxaliplatin, and a lack of high-grade toxicity was observed. After a median follow-up time of 24 months, there was a significant difference in overall survival (OS), relapse-free survival (RFS), and metastasis-free survival (MFS) in favor of the CCRT arm. The 2-year OS rates were 100% for the CCRT arm and 77% for the RT arm (P = .01). The 2-year MFS rates were 92% for the CCRT arm and 80% for the RT arm (P = .02). The 2-year RFS rates were 96% for the CCRT arm and 83% for the RT arm (P = .02). Conclusion CCRT with weekly oxaliplatin is feasible and improves OS, MFS, and RFS rates in patients with locoregionally advanced NPC. Therefore, further randomized trials including oxaliplatin are warranted.

scholarly journals Induction Chemotherapy Followed by Cetuximab Radiotherapy Is Not Superior to Concurrent Chemoradiotherapy for Head and Neck Carcinomas: Results of the GORTEC 2007-02 Phase III Randomized Trial

2018 ◽  
Vol 36 (31) ◽  
pp. 3077-3083 ◽  
Author(s):  
Lionnel Geoffrois ◽  
Laurent Martin ◽  
Dominique De Raucourt ◽  
Xu Shan Sun ◽  
Yungan Tao ◽  
...  
Keyword(s):  
Squamous Cell Carcinoma ◽  
Cell Carcinoma ◽  
Head And Neck ◽  
Induction Chemotherapy ◽  
Squamous Cell ◽  
Concurrent Chemoradiotherapy ◽  
Progression Free Survival ◽  
Standard Of Care ◽  
Phase Iii ◽  
Free Survival

Purpose Both concurrent chemoradiotherapy (CT-RT) and cetuximab radiotherapy (cetux-RT) have been established as the standard of care for the treatment of locally advanced squamous cell carcinoma of the head and neck. It was not known whether the addition of induction chemotherapy before cetux-RT could improve outcomes compared with standard of care CT-RT. Patients and Methods The current trial was restricted to patients with nonmetastatic N2b, N2c, or N3 squamous cell carcinoma of the head and neck and fit for taxotere, cisplatin, fluorouracil (TPF). Patients were randomly assigned to receive three cycles of TPF followed by cetux-RT versus concurrent carboplatin fluorouracil and RT as recommended in National Comprehensive Cancer Network guidelines. The trial was powered to detect a hazard ratio (HR) of 0.66 in favor of TPF plus cetux-RT for progression-free survival at 2 years. The inclusion of 180 patients per arm was needed to achieve 80% power at a two-sided significance level of .05. Results Between 2009 and 2013, 370 patients were included. All patients and tumors characteristics were well balanced between arms. There were more cases of grade 3 and 4 neutropenia in the induction arm, and the induction TPF was associated with 6.6% treatment-related deaths. With a median follow-up of 2.8 years, 2-year progression-free survival was not different between both arms (CT-RT, 0.38 v TPF + cetux-RT, 0.36; HR, 0.93 [95% CI, 0.73 to 1.20]; P = .58). HR was 0.98 (95% CI, 0.74 to 1.3; P = .90) for locoregional control and 1.12 (95% CI, 0.86 to 1.46; P = .39) for overall survival. These effects were observed regardless of p16 status. The rate of distant metastases was lower in the TPF arm (HR, 0.54 [95% CI, 0.30 to 0.99]; P = .05). Conclusion Induction TPF followed by cetux-RT did not improve outcomes compared with CT-RT in a population of patients with advanced cervical lymphadenopathy.

scholarly journals Induction chemotherapy followed by radiotherapy versus concurrent chemoradiotherapy in the treatment of different risk locoregionally advanced nasopharyngeal carcinoma

2020 ◽  
Vol 12 ◽  
pp. 175883592092821
Author(s):  
Li-Ting Liu ◽  
Yu-Jing Liang ◽  
Shan-Shan Guo ◽  
Hao-Yuan Mo ◽  
Ling Guo ◽  
...  
Keyword(s):  
High Risk ◽  
Nasopharyngeal Carcinoma ◽  
Induction Chemotherapy ◽  
Concurrent Chemoradiotherapy ◽  
Risk Group ◽  
Risk Groups ◽  
Rank Test ◽  
Intermediate Risk ◽  
Free Survival ◽  
Alternative Treatment Strategy

Background: This study aimed to investigate the efficiency and toxicities of concurrent chemoradiotherapy (CCRT) and induction chemotherapy (IC) followed by radiotherapy (RT) in different risk locoregionally advanced nasopharyngeal carcinoma (NPC). Methods: A total of 1814 eligible patients with stage II–IVB disease treated with CCRT or IC plus RT were included. The overall survival (OS), progression-free survival (PFS) and distant metastasis-free survival (DMFS) were calculated using the Kaplan–Meier method, and the differences were compared using the log-rank test. Results: Nomograms were developed to predict OS, PFS and DMFS (C-index: 0.71, 0.70 and 0.71, respectively). Patients were then divided into three different risk groups based on the scores calculated by the nomogram for OS. In the low and intermediate-risk group, no significant survival differences were observed between patients treated with IC plus RT alone and CCRT (5-year OS, 97.3% versus 95.6%, p = 0.642 and 87.6% versus 89.7%, p = 0.381, respectively; PFS, 95.9% versus 95.6%, p = 0.325 and 87.6% versus 89.0%, p = 0.160, respectively; DMFS, 97.2% versus 94.8%, p = 0.339 and 87.2% versus 89.3%, p = 0.628, respectively). However, in the high-risk group, IC plus RT displayed an unfavorable 5-year OS (71.0% versus 77.2%, p = 0.022) and PFS (69.4.0% versus 75.4%, p = 0.019) compared with CCRT. A significantly higher incidence of grade 3 and 4 adverse events was documented in patients treated with CCRT than in those treated with IC plus RT in all risk groups ( p = 0.040). Conclusion: IC followed by RT represents an alternative treatment strategy to CCRT for patients with low and intermediate-risk NPC, but it is not recommended for patients with high-risk NPC.

Phase III Trial of Gemcitabine Plus Tipifarnib Compared With Gemcitabine Plus Placebo in Advanced Pancreatic Cancer

2004 ◽  
Vol 22 (8) ◽  
pp. 1430-1438 ◽  
Author(s):  
E. Van Cutsem ◽  
H. van de Velde ◽  
P. Karasek ◽  
H. Oettle ◽  
W.L. Vervenne ◽  
...  
Keyword(s):  
Pancreatic Cancer ◽  
Overall Survival ◽  
Single Agent ◽  
Survival Rates ◽  
Advanced Pancreatic Cancer ◽  
Progression Free Survival ◽  
Phase Iii ◽  
Controlled Study ◽  
Double Blind ◽  
Free Survival

Purpose To determine whether addition of the farnesyltransferase inhibitor tipifarnib (Zarnestra, R115777; Johnson and Johnson Pharmaceutical Research and Development, Beerse, Belgium) to standard gemcitabine therapy improves overall survival in advanced pancreatic cancer. Patients and Methods This randomized, double-blind, placebo-controlled study compared gemcitabine + tipifarnib versus gemcitabine + placebo in patients with advanced pancreatic adenocarcinoma previously untreated with systemic therapy. Tipifarnib was given at 200 mg bid orally continuously; gemcitabine was given at 1,000 mg/m2 intravenously weekly × 7 for 8 weeks, then weekly × 3 every 4 weeks. The primary end point was overall survival; secondary end points included 6-month and 1-year survival rates, progression-free survival, response rate, safety, and quality of life. Results Six hundred eighty-eight patients were enrolled. Baseline characteristics were well balanced between the two treatment arms. No statistically significant differences in survival parameters were observed. The median overall survival for the experimental arm was 193 v 182 days for the control arm (P = .75); 6-month and 1-year survival rates were 53% and 27% v 49% and 24% for the control arm, respectively; median progression-free survival was 112 v 109 days for the control arm. Ten drug-related deaths were reported for the experimental arm and seven for the control arm. Neutropenia and thrombocytopenia grade ≥ 3 were observed in 40% and 15% in the experimental arm versus 30% and 12% in the control arm. Incidences of nonhematologic adverse events were similar in two groups. Conclusion The combination of gemcitabine and tipifarnib has an acceptable toxicity profile but does not prolong overall survival in advanced pancreatic cancer compared with single-agent gemcitabine.

Is RECIST-defined progression free-survival a meaningful endpoint in the era of immunotherapy?

2017 ◽  
Vol 35 (6_suppl) ◽  
pp. 488-488 ◽  
Author(s):  
Sukhvinder Johal ◽  
Irene Santi ◽  
Justin Doan ◽  
Saby George
Keyword(s):  
Overall Survival ◽  
Disease Progression ◽  
Tumor Regression ◽  
Progression Free Survival ◽  
Phase Iii ◽  
Rank Test ◽  
Apparent Lack ◽  
Free Survival ◽  
Treatment Beyond Progression ◽  
Significant Difference

488 Background: Progression-free survival (PFS) is often used as a primary endpoint in oncology clinical trials as a surrogate for overall survival. Traditionally, the Response Evaluation Criteria in Solid Tumors (RECIST) have defined disease progression as a significant increase in the size of tumor lesions and the development of new lesions. However, some patients starting immunotherapy have shown initial increased size of tumor lesions followed by tumor regression, due to the unique mechanism of action of immunotherapies. This initial “pseudo-progression” could be classified inaccurately as disease progression, as evidenced by benefit from the treatment beyond progression approach ( JAMA Oncol 2016). The phase III CheckMate 025 trial of nivolumab versus everolimus in patients with advanced renal cell carcinoma allowed treatment beyond progression if there was investigator-assessed clinical benefit and tolerability. The purpose of our study was to test if treatment duration for an immunotherapy was different from RECIST-defined PFS, and as such, could potentially explain the apparent lack of correlation between RECIST progression and overall survival shown in CheckMate 025. Methods: Using 1-year data from CheckMate 025, Kaplan–Meier methodology was used to estimate the median duration of PFS and time to treatment discontinuation (TTD). Stratified log-rank test was used to assess the difference in treatments. Results: For all patients, the median PFS with nivolumab was 4.6 months (95% CI, 3.7–5.4 months) and median TTD was 6.2 months (95% CI, 5.6–7.7 months). For everolimus, the median PFS was 4.4 months (95% CI, 3.7–5.5 months) and median TTD was 3.9 months (95% CI, 3.7–4.6 months). Conclusions: Patients in CheckMate 025 had significantly longer survival with nivolumab than with everolimus, but with similar PFS. Our analysis demonstrated that while PFS was similar to TTD with everolimus, there was a significant difference between the 2 measures for nivolumab, suggesting that RECIST-defined PFS may not be the proper endpoint to define progression for immunotherapies. Further evaluation of the association of TTD and other immune-related progression endpoints with overall survival is warranted. Clinical trial information: NCT01668784.

scholarly journals Progression-Free Survival Prediction in Patients with Nasopharyngeal Carcinoma after Intensity-Modulated Radiotherapy: Machine Learning versus Traditional Statistics

2021 ◽  
Vol 11 (8) ◽  
pp. 787
Author(s):  
Ronald Wihal Oei ◽  
Yingchen Lyu ◽  
Lulu Ye ◽  
Fangfang Kong ◽  
Chengrun Du ◽  
...  
Keyword(s):  
Machine Learning ◽  
Nasopharyngeal Carcinoma ◽  
Proportional Hazards ◽  
Progression Free Survival ◽  
Cox Proportional Hazards ◽  
Brier Score ◽  
Rank Test ◽  
Survival Prediction ◽  
Free Survival ◽  
Log Rank Test

Background: The Cox proportional hazards (CPH) model is the most commonly used statistical method for nasopharyngeal carcinoma (NPC) prognostication. Recently, machine learning (ML) models are increasingly adopted for this purpose. However, only a few studies have compared the performances between CPH and ML models. This study aimed at comparing CPH with two state-of-the-art ML algorithms, namely, conditional survival forest (CSF) and DeepSurv for disease progression prediction in NPC. Methods: From January 2010 to March 2013, 412 eligible NPC patients were reviewed. The entire dataset was split into training cohort and testing cohort in a ratio of 90%:10%. Ten features from patient-related, disease-related, and treatment-related data were used to train the models for progression-free survival (PFS) prediction. The model performance was compared using the concordance index (c-index), Brier score, and log-rank test based on the risk stratification results. Results: DeepSurv (c-index = 0.68, Brier score = 0.13, log-rank test p = 0.02) achieved the best performance compared to CSF (c-index = 0.63, Brier score = 0.14, log-rank test p = 0.38) and CPH (c-index = 0.57, Brier score = 0.15, log-rank test p = 0.81). Conclusions: Both CSF and DeepSurv outperformed CPH in our relatively small dataset. ML-based survival prediction may guide physicians in choosing the most suitable treatment strategy for NPC patients.

Sign in / Sign up

Export Citation Format

Share Document