Can preoperative assessment of cancer in the elderly (PACE) predict 30-days postoperative outcomes?

Abstract

8537 Background: Surgery is the treatment of choice for solid cancers. Frequent functional impairment/comorbidities in the elderly enhances the risk of treatment related complications. Inability to forecast short term outcomes after cancer surgery in elderly affects clinical practice, denying optimal treatment. No validated instrument is available to help make informed decision; a compilation of validated questionnaires (PACE) is established to preoperatively inform on the health condition of elderly cancer pts. This international multicentre study investigates how components of PACE preoperatively assessed are associated with postoperative outcomes. Methods: A prospective series of consenting elderly cancer pts (≥70 yrs) receiving elective surgery (moderate-major+) were recruited from 8 hospitals (UK, Netherlands, Italy, Japan, Belgium) (07/2003–12/2005) and assessed using PACE (Comorbidities, IADL, ADL, GDS, BFI, PS, MMS, ASA). 30day morbidity, hospital stay and mortality were recorded. Results: 448 pts [breast (48%), GI (30%), GU (16%), miscellaneous (6%)] were recruited and followed postoperatively. Observed morbidity was 36% (161 pts), mortality 4% (16 pts) and median hospital stay was 5 days (range 2 -10). All components of PACE but ASA were associated with morbidity (p<0.05) and hospital stay (p<0.05) on univariate analysis. Operative deaths were too few to look for associations. Forward stepwise logistic regression models (multivariate analysis) identified 3 components of PACE as best describing the occurrence of post-surgical morbidity: BFI (OR (upper quartile)= 2.4; 95% CI=1.2–4.9); PS (OR=1.9; 95% CI=1.1–3.4); IADL (OR=1.7; 95% CI=1.0–3.0). The same components were also selected into the model as best describing hospital stay beyond average: BFI (OR (upper quartile)=18.1; 95% CI=7.2–45.3), PS (OR=2.2; 95%=1.2–4.4) and IADL (OR=2.7; 95% CI=1.4–5.1). Of those pts exposed to all 3 factors entered into the regression models 61% experienced complications and >80% had a longer hospital stay. Conclusions: BFI, PS and IADL appear to be the most relevant prognosticators of short term surgical outcomes. A holistic appraisal of elderly pts undergoing surgery is warmly recommended when consenting the patient and during the decision making process. No significant financial relationships to disclose.