A systematic review of radiotherapy toxicity reporting in randomized controlled trials of rectal cancer: A comparison of patient-reported outcomes and clinician toxicity reporting
A partir d'une revue systématique de la littérature publiée entre janvier 1995 et juillet 2013 (21 essais), cette étude évalue la qualité des informations auto-rapportées sur les effets indésirables à long terme des traitements comportant une radiothérapie ou une chimioradiothérapie pré-opératoire chez les patients atteints d'un cancer rectal
The use of multimodal treatments for rectal cancer has improved cancer-related outcomes but makes monitoring toxicity challenging. Optimizing future radiotherapy regimens requires collection and publication of detailed toxicity data. This review evaluated the quality of toxicity information provided in randomized controlled trials (RCTs) of radiotherapy in rectal cancer and focused on the difference between clinician-reported and patient-reported toxicity.
Medline, EMBASE and the Cochrane Library were searched (January 1995-July 2013) for RCTs reporting late toxicity in patients treated with regimens including preoperative (chemo)radiotherapy. Data on toxicity measures and information on toxicity reported was extracted using QUANTEC (Quantitative Analyses of Normal Tissue Effects in the Clinic) recommendations. International Society for Quality of Life Research (ISOQOL) standards on patient-reported outcomes (PROs) were used to evaluate the quality of patient-reported toxicity.
21 RCT publications met inclusion criteria out of 4144 articles screened. All PRO papers reported higher rates of toxicity symptoms than clinician-reported papers and reported on a wider range and milder symptoms. No clinician-reported paper published data on sexual dysfunction. 55% of clinician-reported papers grouped toxicity data related to an organ system together (e.g. ‘Bowel’) and 45% presented data only on more severe (≥grade 3) toxicity. In comparison, all toxicity grades were reported in 79% of PRO publications and all studies (100%) presented individual symptom toxicity data (e.g. bowel urgency). However, PRO reporting quality was variable. Only 43% of PRO studies presented baseline data, 28% did not use any psychometrically validated instruments and only 29% of studies described statistical methods for managing missing data.
Analysis of these trials highlights the lack of reporting standards for adverse events and reveals the differences between clinician and patient-reporting of toxicity. Recommendations for improving the quality of adverse event data collection are provided with the aim of improving critical appraisal of outcomes for future studies.
International Journal of Radiation Oncology • Biology • Physics , résumé, 2014