1586P - Potential drug interactions in older patients with cancer: updated data from the ELCAPA cohort survey.

Date 10 September 2017
Event ESMO 2017 Congress
Session Poster display session
Topics Supportive measures
Geriatric Oncology
Supportive and Palliative Care
Presenter Guillaume Beinse
Citation Annals of Oncology (2017) 28 (suppl_5): v543-v567. 10.1093/annonc/mdx388
Authors G. Beinse1, D. Reitter2, L. Segaux3, M. Carvalho-Verlinde2, C. Tournigand4, B. Rousseau4, T. Cudennec5, E. Paillaud6, F. Canouï-Poitrine3, P. Caillet1
  • 1Geriatric Oncology, Centre hospitalier universitaire Henri-Mondor, 94010 - Créteil/FR
  • 2Department Of Pharmacy, Centre hospitalier universitaire Henri-Mondor, Créteil/FR
  • 3Public Health, Centre hospitalier universitaire Henri-Mondor, 94010 - Créteil/FR
  • 4Medical Oncology, Centre hospitalier universitaire Henri-Mondor, 94000 - Creteil/FR
  • 5Geriatric Oncology, Centre hospitalier universitaire Ambroise Paré, Boulogne-Billancourt/FR
  • 6Geriatrics, Centre hospitalier universitaire Henri-Mondor, 94010 - Créteil/FR



Because of polypharmacy, older cancer patients are at risk of adverse events related to potential drug interactions (PDI). We aim to identify PDI in daily medications, between daily medications and chemotherapy (CT), and related potential clinical outcomes (PCO).


All cancer patients aged ≥70 years, referred for geriatric assessment at Henri Mondor Hospital (Paris’ area, Créteil, France), included in the prospective ELCAPA cohort survey (2007–2014), and who received CT were included. PDI were identified using Lexicomp® (LexiComp, Hudson, USA) software and the Theriaque® website. PDI were classified as: A, no interactions; B, no action needed; C, monitor therapy; D, consider therapy modification; X, avoid combination. Factors associated with grade C or D/X PDI were analyzed using ordered multivariate logistic regression.


We analyzed 442 patients (median age: 78 years; 49% women). Main tumor sites were upper digestive tract (23%), colorectal (21%), urological tract (19%), lymphoid malignancies (15%), and breast (12%); 23% had metastasis. Median number of drugs/patients/day was 3 (Q1-Q3 [1-6]). We identified 1742 PDI: 87% in daily medications (183 patients had grade C PDI (41%), 128 grade D/X (29%)), and 13% between daily medications and CT (66 patients had grade C PDI (15%), 56 grade D/X PDI (13%)). Main PCO involving daily medications were hypotension risk (33%), psychotropic effects (17%), glycemic (12%) and hemostasis (9%) dysregulations. Main PCO related to PDI involving CT were risk of CT over-exposure (34%), hypotension risk (20%), and hemostasis dysregulation (11%). In multivariable analysis, adjusted for number of drugs, factors associated with grade D/X PDI, both with or without CT were: ≥2 metastatic sites (p = 0.01) and lymphoid malignancies (p = 0.01). Patients living alone had less grade D/X PDI in daily medications (p = 0.003), while breast cancer (p = 0.04) was associated with more grade D/X PDI in daily medications. Higher body mass index was associated with grade D/X PDI involving CT (p = 0.03).


The high prevalence of PDI in older cancer patients highlights the need to assess precisely the iatrogenic risk before anti-cancer treatment.

Clinical trial identification

Legal entity responsible for the study

Assistance Publique - Hopitaux de Paris




All authors have declared no conflicts of interest.