Patients with recent weight loss are at higher risk of having cancer, specifically in correlation with a separate clinical feature that may also indicate a cancer diagnosis, according to a study published in the British Journal of General Practice.
Researchers from the University of Oxford used electronic databases (EMBASE, the Science Citation Index & Conference Proceedings Citation Index, and Medline) to identify studies that included patients aged ≥18 years with a cancer diagnosis who presented to a primary care clinician. Studies were included if 2×2 diagnostic accuracy data could be extracted. No restrictions were placed on the definition of weight loss; cancer diagnosis was included if confirmed by a cancer registry, histology, the general practice electronic record, or another reliable source. The QUADAS-2 tool (Quality Assessment of Diagnostic Accuracy Studies) was used to assess methodologic quality.
After applying exclusion criteria, 25 studies published between 1994 and 2015 were included; 23 of these (92%) used primary care records. In all, 13 were case-control studies and 12 were cohort studies. Control patients were matched for age, sex, primary care practice, and consultation year. Cohort entry was defined as the first incidence of recorded weight loss for the exposed group or study entry, or 12 months after registering with the practice for the unexposed group. The sample size of patients ranged from 398 to 2,140,194.
All studies reported weight loss linked to 10 distinctive cancer types: colorectal, pancreatic, gastroesophageal, ovarian, lung, renal tract, myeloma, non-Hodgkin lymphoma, biliary tree, and prostate. Sensitivity ranged from 2% to 47% and specificity from 92% to 99% across all cancer types; only colorectal and pancreatic cancer were included in sufficient studies to warrant separate analyses. Pooled sensitivity for colorectal cancer was 14% and pooled specificity was 97%. For pancreatic cancer, pooled sensitivity was 13% and pooled specificity was 99%. Hazard ratios ranged from 1.6 for ovarian cancer to 12.5 for pancreatic cancer. Adjustment for other symptoms of cancer did not abolish the association between weight loss and a diagnosis of cancer.
Sensitivity for any cancer in men was 3% and specificity was 99%; for women, sensitivity was 5% and specificity was 99%. No significant difference was found for hazard ratios between men and women after adjustments for age and body mass index. Positive predictive values (PPVs) for weight loss ranged from 0.0% for biliary tract cancer to 3.3% for prostate cancer. One study reported higher PPVs for colorectal cancer for people aged ≥70 years (1.7%) vs those aged <70 years (0.4%). A total of 6 studies reported that the PPV for weight loss paired with another clinical feature ranged from 0.1% for chest pain associated with myeloma to 31.6% for jaundice in pancreatic cancer. Weight loss associated with another clinical feature (compared with weight loss alone, clinical feature alone, or by adding together the individual PPVs of weight loss or clinical feature) yielded the strongest PPV.
“Patients presenting to primary care with weight loss are at higher risk of having cancer than patients without recorded weight loss,” the authors concluded. “This review suggests that patients aged ≥60 years presenting to primary care settings with weight loss that prompts a clinical record entry warrant rapid investigation for possible cancer, potentially across multiple sites.”
Nicholson BD, Hamilton W, O’Sullivan J, Aveyard P, Hobbs FDR. Weight loss as a predictor of cancer in primary care: a systematic review and meta-analysis. Br J Gen Pract. 2018;68(670):e311-e322.