Blood Tests for Cancer

One of the questions oncologists, internists and family physicians encounter most frequently is: "Can a Blood Test Detect Cancer?". Even many doctors lack sufficient knowledge of this subject and try to test certain cancer-related biological markers in blood as part of a "general examination" or "check-up." Inappropriate use of these tests causes unnecessary anxiety and places a substantial burden on health systems. Indiscriminate testing also causes genuinely necessary cancer screening to be missed.
What you need to know
- Hemoglobin, platelets, serum calcium, liver function tests and markers such as C-reactive protein (CRP) and erythrocyte sedimentation rate, which are inflammatory markers, may be used in primary care (family medicine) as triage blood tests and provide clues to cancer in patients with nonspecific findings.
- Triage tests do not have the performance characteristics of "exclusion / rule-out" tests.
- Scientific evidence supports primary-care use of a small number of cancer markers such as CA125 and PSA, these markers (biomarkers), in specific situations of suspicion rather than as a cancer screening test that everyone should undergo at particular ages.
An example case: A man with a month of back pain, aged 61, visits his general practitioner. He has hypertension, has never smoked and reports several months of fatigue . The pain keeps him awake at night. He has not lost weight. Clinical examination is normal. Pancreatic cancer, multiple myeloma, prostate cancer or any metastatic cancer are among a broad range of potential differential diagnoses.
Identifying cancer can be difficult because most common symptoms are nonspecific and low-risk, and even the best-known cancer warning signshave relatively low positive predictive values for a possible underlying cancer.
Positive predictive value
Positive predictive value (PÖD) is a frequently used term in medical informatics. It expresses, among test results, how many positive results are truly positive . In other words, it is the probability that a positive test result represents actual disease in that population. Its formula is the ratio of true positives to true plus false positives.
For example, weight loss has a positive predictive value (PÖD) of only 0 - 3.3% for underlying cancer, whereas rectal bleeding has a PÖD of 2.2 - 15.8%. Cancer markers / biomarkers used in hospitals have low positive predictive values and high false-positive rates when applied in primary care: they poorly predict the actual condition and frequently mislead. Identifying patients whose vague symptoms may be due to cancer rather than benign disease is therefore a challenge for primary-care physicians.
Although further investigation and diagnosis usually occur in secondary and tertiary centers, initial suspicion often arises in primary care. Patients whose symptoms imply a cancer risk of approximately ≥3% are generally referred urgently. Those with estimated risk <3% may remain in primary care for risk investigation and triage testing. Triage tests may provide clues identifying patients for referral and point to an underlying cancer site. This is particularly useful when vague complaints suggest several cancer types and may guide decisions about further investigation.
This article discusses blood tests used to identify or stratify possible cancer risk in primary care and presents evidence for use in symptomatic patients. First, we address tests not specific to any cancer but potentially helpful to primary-care providers in identifying risk. We then discuss markers specific to certain cancers. Routine cancer screening tests everyone should undergo at particular ages for early diagnosis are outside this article’s scope.
Nonspecific Blood Tests or Clues for Cancer
Some commonly used nonspecific primary-care tests may provide “clues” to possible cancer. Tests with a PÖD >1% for cancer, such as hemoglobin, platelet count, serum calcium, liver function tests and inflammatory markers including C-reactive protein (CRP) and erythrocyte sedimentation rate (ESH) , are summarized in the following imageas a summary. PÖDs of <1% for some underlying cancers are reported in the same infographic. These blood tests should not be measured routinely, but considered in low-risk patients who should not have high-risk symptoms such as unexplained weight loss or persistent fatigue. Rarely, a complete blood count may identify a hematological cancer (leukemia, lymphoma), but a "positive test" makes cancer a more likely explanation.

Although noncancer diseases also frequently cause abnormal results, further investigation or referral may be necessary to exclude cancer. A "negative" result does not establish that cancer is absent: even with normal results, cancer may still be present. None of these tests has sufficient sensitivity to act as an exclusion / rule-out test, except sedimentation rate, which may serve as a simple rule for myeloma diagnosis. If sedimentation rate or blood plasma viscosity is normal, we rule out myeloma.
With low-risk symptoms, negative tests provide some reassurance. However, persistent or changing symptoms require further investigation. Ideally, the rationale and consequences of a negative or positive result should be discussed before testing so patients can share in decisions.

Specific Cancer Markers
The number of cancer biomarkers used in secondary and tertiary care is growing, but no new blood marker for primary care has emerged since Sturgeon and colleagues’ review in 2009 (existing ones were summarized in the TABLE). We emphasize again that these tests are not cancer screening and should be performed in patients with particular symptoms. Even well-known markers such as prostate-specific antigen (PSA) and cancer antigen 125 (CA 125) have a limited primary-care evidence base. For example, regarding PSA, among men aged 70-79, one in every 14 has prostate cancer, and in some of these people it will not become a serious clinical problem. The positive predictive value of a positive PSA result therefore does not necessarily translate into clinical benefit.
Conclusion
Continuation of the example case: The family physician was concerned about night pain and fatigue, which together with age increased the possibility of underlying cancer. The patient was referred for initial blood tests including a complete blood count, liver function tests, serum calcium and erythrocyte sedimentation rate (ESH). Results were notable for a mildly raised platelet count (495 thousand / cubic millimeter) and moderately elevated ESH (34 mm / s). Further specific blood tests, including serum electrophoresis and Bence Jones protein, were performed. Monoclonal immunoglobulins suggesting multiple myeloma were detected, and the patient was urgently referred to a cancer clinic.
The Future of Cancer Markers
For cancers such as pancreatic and ovarian cancer, considered particularly difficult to detect early , many cancer biomarkers are being investigated. However, few studies test candidate biomarkers clinically. Future research evaluating a potential diagnostic cancer marker should aim to determine optimal use, improve physician and patient acceptability, reduce false positives and measure economic implications.
More promising than blood markers for early cancer diagnosis are methods screening blood for tumor cells or tumor cell fragments (such as DNA), called liquid biopsies.
*
Sources
Blood markers for cancer.
BMJ 2019. 14 October 2019.
Was this article helpful?
Your feedback helps us improve our content.





Reader comments
Share your comment
Your name and comment are published only after administrator approval. Your email address is never shown to visitors.
Be the first to comment.