Neoadjuvant Chemoradiation Therapy for All Elderly Patients With Locally Advanced Rectal Cancer?
Menée à partir de données 2004-2016 portant sur 3 868 patients âgés atteints d'un cancer du rectum localement avancé (53 % d'hommes ; âge moyen : 83,4 ans), cette étude analyse la survie globale et la mortalité en fonction des traitements reçus (radiochimiothérapie néoadjuvante suivie d'une résection, intervention chirurgicale suivie ou non d'un traitement adjuvant)
Although the safety of rectal cancer surgery in very carefully selected elderly patients has been established, data on the safety and efficacy of chemoradiation therapy in this patient population is limited. This may explain the relative underuse of radiation therapy in elderly individuals, first noted in 2001.Nassoiy et al, using the National Cancer Data Base in a multicenter retrospective design, demonstrate that only 58.8% of patients older than 80 years diagnosed with a locally advanced rectal cancer between 2004 and 2016 received neoadjuvant chemoradiation therapy (NACRT). Their secondary analysis led them to conclude that NACRT is associated independently with a 25% decreased risk of death compared with alternate treatment sequences (surgery + adjuvant therapy or surgery alone). However, their conclusion is based on an analysis of groups that, according to Table 1 in their article, are statistically different. Although inverse probability of treatment weighting methodology was used to balance the groups, the authors do not provide any evidence that the groups were similar after inverse probability of treatment weighting adjustment. A preferred approach may have been to select patients with similar baseline characteristics, group them by treatment received, and then compare outcomes. Did the authors consider using propensity score matching methodology? Of equal concern is their decision to use binary logistic regression to compare 2 arbitrary groups (NACRT vs surgery ± adjuvant therapy) instead of the 3 individual treatments. Therefore, we take issue with their usage of the words “independently associated” because they do not actually compare the 3 treatment options independently; at best, there is an association.
JAMA Surgery , éditorial, 2021