Missing the Signs? Abdominal Cancer Detection in GP Practice

Nearly half of patients with undiagnosed cancer initially present with nonspecific symptoms, which can make early detection challenging. When a patient consults a general practitioner (GP), part of the evaluation involves determining the likelihood of a serious underlying condition, including cancer.
A helpful tool for assessing this probability is the positive predictive value (PPV) — the proportion of patients presenting with certain symptoms, signs, or test results who ultimately receive a cancer diagnosis. While common blood tests are frequently used in primary care to investigate these nonspecific symptoms, their accuracy in predicting cancer in this context is not widely understood.
An English cohort study evaluated the PPV, negative predictive value, sensitivity, and specificity of 19 abnormal blood test results in detecting cancer in patients aged 30 years or older who consulted primary care between January 2007 and October 2016. These patients presented with nonspecific abdominal symptoms, such as pain and bloating lasting less than 3 months.
Of 425,549 patients with abdominal pain, 9427 (2.2%) received a cancer diagnosis within 12 months. Similarly, of 52,321 patients with abdominal bloating, 1148 (2.2%) received a cancer diagnosis within the same period. Among men, the most common cancer sites were the colon, prostate, and pancreas, while for women, they were the colon, breast, ovaries, and pancreas.
In both men and women over age 60 years, the PPV for cancer exceeded the 3% risk threshold recommended by the United Kingdom’s National Institute for Health and Care Excellence (NICE) for referral to a specialist.
Blood tests were conducted simultaneously in large proportions of patients: 64% of those with abdominal pain and 70% with bloating.
In patients aged 30-59 years, several abnormal blood tests raised the risk for cancer above the 3% threshold. For example, in women aged 50-59 years with abdominal bloating, the initial cancer risk of 1.6% increased to 10% with elevated ferritin levels, 9% with low albumin, 8% with elevated platelet counts, 6% with elevated inflammatory markers, and 4% with anemia.
When compared with risk assessments based solely on symptoms, age, and sex, those that include blood test results for every 1000 patients with abdominal bloating would result in 63 additional referrals and the identification of three more cancer patients — a 16% relative increase in cancer diagnosis.
Cancer should be considered in patients with elevated inflammatory markers, but their low sensitivity makes them unreliable for ruling out malignancies. Studies across the general population, including cohort studies and meta-analyses, have explored the association between elevated C-reactive protein levels and the risk for colorectal, lung, ovarian, and breast cancers. However, these associations are not strong enough to be clinically useful in diagnosing cancer in symptomatic patients. 
In a 2019 study published in the British Journal of Cancer, which analyzed data from 160,000 primary care patients, those with elevated inflammatory markers showed a 1-year cancer incidence of 3.53% (95% CI, 3.37-3.70). This incidence was higher than the 1.50% (95% CI, 1.43-1.58) seen in patients with normal markers and the 0.97% (95% CI, 0.87-1.07) in untested controls. Cancer risk was higher with abnormal inflammatory marker levels, advanced age, and male gender.
The risk increased further if repeated tests showed consistently elevated markers but decreased if the markers returned to normal. Men over age 50 years and women over age 60 years with elevated inflammatory markers had a cancer risk that exceeded the 3% NICE threshold for urgent specialist referral. However, the sensitivity for detecting cancer was still low at 46.1% for C-reactive protein and 43.6% for erythrocyte sedimentation rate. The authors concluded that inflammatory markers should not be relied upon to rule out cancer.
This large-scale study focused on common digestive disorders and examined abnormalities in 19 inflammatory blood tests, including classic cancer indicators like anemia and thrombocytosis, which should alert GPs to refer patients to a specialist within 28 days. However, it does not provide guidance on imaging or gastrointestinal endoscopies for diagnosing digestive or extradigestive cancers, particularly ovarian cancer, where staging details could have been informative.
The study’s limitations include potential biases related to ICD-10 coding and the likelihood that patients prescribed blood tests by their GPs were already at a higher risk for cancer compared with those with isolated symptoms. Additionally, variations in PPVs may arise when applying these findings to different healthcare settings, given the varying rates of blood test use.
It’s also important to consider the rigid protocols in the National Institutes of Health, where access to specialists is typically granted through GP referrals. Current guidelines rely more on a combination of clinical signs or red flags, such as anemia or elevated inflammatory markers, rather than solely on the blood tests assessed in this study.
The role of certain tumor markers measured in the study, like cancer antigen-125 and prostate-specific antigen, remains unclear. Meanwhile, the recently FDA-approved Shield test, which detects early-stage colorectal cancer DNA in 87% of asymptomatic cases (though with a 10% false-positive rate), presents a promising avenue for future studies.
In conclusion, assessing cancer risk in patients with nonspecific abdominal symptoms should go beyond just considering symptoms, age, and sex. Incorporating additional information from common blood test results can significantly enhance early detection efforts.
In primary care, inflammatory markers should be used judiciously. While they can aid in cancer detection, false positives can cause anxiety, and false negatives may offer inappropriate reassurance. These markers do not definitively rule out cancer but serve as a Bayesian tool, where a positive result increases the risk for cancer and a negative result reduces it — without entirely eliminating the risk.
This story was translated from JIM using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.