Efficacy and Safety Evaluation of the Various Therapeutic Options in Locally Advanced Cervix Cancer: A Systematic Review and Network Meta-Analysis of Randomized Clinical Trials

Int J Radiat Oncol Biol Phys. 2019 Feb 1;103(2):411-437. doi: 10.1016/j.ijrobp.2018.09.037. Epub 2018 Nov 1.


Treatment options in locally advanced cervix cancer (LACC) have evolved around radiation therapy (RT) and/or chemotherapy (CT), hypoxic cell sensitizers, immunomodulators (Imm), and locoregional moderate hyperthermia (HT). A systematic review and network meta-analysis was conducted to synthesize the evidence for efficacy and safety in terms of long-term locoregional control (LRC), overall survival (OS), and grade ≥3 acute morbidity (AM) and late morbidity (LM). Five databases were searched, and 6285 articles (1974-2018) were screened per the Preferred Reporting Items for Systematic Reviews and Meta-analysis guidelines. Fifty-nine randomized trials in untreated LACC without surgical intervention were shortlisted. These used 13 different interventions: RT alone and/or neoadjuvant CT (NACT), adjuvant CT (ACT), concurrent chemoradiation therapy (CTRT) (weekly cisplatin [CDDP]/3-weekly CDDP/combination CT with CDDP/non-CDDP-based CT), hypoxic cell sensitizers, Imm, or HT. Odds ratios (ORs) using random effects network meta-analysis were estimated. Interventions for each endpoint were ranked according to their corresponding surface under cumulative ranking curve values. Of the 9894 patients evaluated, the total events reported for LRC, OS, AM, and LM were 5431 of 8197, 4482 of 7958, 1710 of 7183, and 441 of 6333, respectively. ORs and 95% credible intervals (CrIs) for the 2 best strategies were HT + RT versus CTRT + ACT (OR, 1.23; 95% CrI, 0.49-3.19) for LRC, CTRT (3-weekly CDDP) versus HTCTRT (OR, 1.14; 95% CrI, 0.35-3.65) for OS, RT + ACT versus RT (OR, 0.01; 95% CrI, 0.00-1.04) for AM, and NACT + RT + ACT versus RT + Imm (OR, 0.42; 95% CrI, 0.02-7.39) for LM. The 3 interventions with the highest cumulative surface under cumulative ranking curve values for all 4 endpoints were HTRT, HTCTRT, and CTRT (3-weekly CDDP). Articles with low risk of bias and those published during 2004 to 2018 also retained these interventions as the best. Two-step cluster analysis grouped these 3 modalities in a single distinctive cluster. HTRT, HTCTRT, and CTRT with 3-weekly CDDP were identified as therapeutic modalities with the best comprehensive impact on key clinical endpoints in LACC. This warrants a phase 3 randomized trial among these strategies for a head-to-head comparison and additional validation.

Publication types

  • Meta-Analysis
  • Research Support, Non-U.S. Gov't
  • Systematic Review

MeSH terms

  • Adult
  • Aged
  • Chemoradiotherapy / methods
  • Cisplatin / administration & dosage
  • Cluster Analysis
  • Combined Modality Therapy
  • Female
  • Humans
  • Hyperthermia, Induced
  • Hypoxia
  • Middle Aged
  • Neoadjuvant Therapy
  • Network Meta-Analysis
  • Odds Ratio
  • Patient Safety*
  • Prospective Studies
  • Radiotherapy / methods*
  • Randomized Controlled Trials as Topic
  • Treatment Outcome
  • Uterine Cervical Neoplasms / therapy*


  • Cisplatin