Introduction: As frailty is associated with inflammation, biomarkers of inflammation may represent objective measures that could facilitate the identification of frailty. Glasgow prognostic score (GPS), combines C-reactive protein (CRP) and albumin, and is scored from 0 to 2 points. Higher score indicates a higher degree of inflammation.
Objectives: To investigate whether (1) GPS is associated with frailty, (2) GPS could be used to screen for frailty, (3) IL-6 and TNF-α add to the accuracy of GPS as a screening tool, and (4) GPS adds prognostic information in frail older patients with cancer.
Methods: Prospective, observational study of 255 patients ≥70 years with solid malignant tumours referred for medical cancer treatment. At baseline, frail patients were identified by a modified Geriatric Assessment (mGA), and blood samples were collected.
Results: Mean age was 76.7 years, 49.8% were frail, and 56.1% had distant metastases. The proportion of frail patients increased with higher GPS (GPS zero: 43.2%, GPS one: 52.7%, GPS two: 94.7%). GPS two was significantly associated with frailty (OR 18.5), independent of cancer type, stage, BMI and the use of anti-inflammatory drugs. The specificity of GPS was high (99%), but the sensitivity was low (14%). Frail patients with GPS two had poorer survival than patients with GPS zero-one. TNF-α and IL-6 did not improve the accuracy of GPS when screening for frailty.
Conclusion: Frailty and GPS two are strongly associated, and GPS two is a significant prognostic factor in frail, older patients with cancer. The inflammatory biomarkers investigated are not suitable screening tools for frailty.
Copyright © 2018 Elsevier Inc. All rights reserved.