Tumor Markers in Primary Care: When to Order, When to Step Away, and Why Screening Causes More Harm Than Good

 

Tumor Markers in Primary Care: When to Order, When to Step Away, and Why Screening Causes More Harm Than Good

PSA, CA-125, CEA, AFP—the tests that cause the most unnecessary panic in your practice.

Tumor markers are among the most misused tests in primary care. They get ordered for vague abdominal pain, pelvic discomfort, fatigue, or "just to be thorough"—and then a mildly elevated result triggers imaging, referrals, biopsies, and patient anxiety for a condition that doesn't exist. The core principle: tumor markers are for monitoring known cancers, not for screening asymptomatic patients (with limited exceptions).

The Markers: What They Do and Don't Tell You

Marker	Primary Cancer Association	Non-Cancer Causes of Elevation	Role in Primary Care
PSA	Prostate cancer	BPH, prostatitis, recent ejaculation, bike riding, UTI, age	Screening (shared decision-making for men 55–69); monitoring after treatment
CA-125	Ovarian cancer	Endometriosis, fibroids, PID, pregnancy, menstruation, cirrhosis, CHF, peritonitis, any peritoneal/pleural inflammation	Monitoring known ovarian cancer. NOT for screening pelvic pain.
CEA	Colorectal cancer	Smoking (most common), IBD, pancreatitis, cirrhosis, peptic ulcer, hypothyroidism	Monitoring colorectal cancer post-treatment. NOT for screening.
AFP	Hepatocellular carcinoma, germ cell tumors	Pregnancy, hepatitis, cirrhosis (regenerating nodules)	HCC surveillance in cirrhosis patients (with ultrasound every 6 months). Germ cell tumor monitoring.
CA 19-9	Pancreatic cancer	Biliary obstruction, cholangitis, pancreatitis, cirrhosis. Not expressed in Lewis antigen-negative patients (~5–10%).	Monitoring known pancreatic cancer. NOT for screening abdominal pain.
HCG	Gestational trophoblastic disease, germ cell tumors	Pregnancy (obviously), marijuana use (rarely), pituitary HCG production in postmenopausal women	Monitoring trophoblastic disease and germ cell tumors post-treatment.

The Rules That Prevent Harm

1. CA-125 Is NOT a Screening Test for Ovarian Cancer

This is the most commonly violated rule. CA-125 is elevated in roughly 80% of advanced ovarian cancer but only 50% of early-stage disease. Meanwhile, it's elevated in dozens of benign conditions: endometriosis, fibroids, PID, menstruation, pregnancy, and any cause of peritoneal or pleural inflammation. The USPSTF recommends against screening for ovarian cancer (including CA-125 and transvaginal ultrasound) in average-risk women because screening does not reduce mortality and leads to unnecessary surgeries, some with serious complications.

The Harm

Ordering CA-125 for "pelvic pain" in a premenopausal woman will frequently return a mildly elevated result from endometriosis, a corpus luteum cyst, or menstruation—triggering imaging, surgical consultation, and sometimes unnecessary oophorectomy. Do not order CA-125 as a screening test in average-risk women. It is appropriate for monitoring response in known ovarian cancer or evaluating a suspicious adnexal mass in a postmenopausal woman in conjunction with imaging.

2. PSA: Shared Decision-Making, Not Reflexive Ordering

Per USPSTF (2018, reaffirmed): for men 55–69, PSA screening should be an individual decision made after shared decision-making about benefits and harms. Not recommended for men ≥70 or men with <10-year life expectancy. Key points:

• PSA >4 ng/mL is traditionally considered the threshold, but many cancers occur below 4 and many benign conditions elevate PSA above 4.
• PSA velocity (rate of change) and PSA density (PSA relative to prostate volume) improve specificity.
• For screening, discuss overdiagnosis and overtreatment. Many screen-detected prostate cancers are low-grade and would never cause symptoms or death.

3. CEA Doesn't Screen for Colorectal Cancer

CEA is only useful for monitoring known CRC post-resection (rising levels may indicate recurrence). Smokers commonly have elevated CEA without cancer. It should never be ordered as a CRC screening test—colonoscopy and stool-based tests (FIT, Cologuard) are the appropriate screens.

4. AFP Has a Defined Surveillance Role

AFP plus abdominal ultrasound every 6 months is the recommended HCC surveillance strategy in patients with cirrhosis (any etiology) and chronic hepatitis B without cirrhosis who meet risk criteria. Outside of this specific surveillance context, AFP should not be ordered for vague abdominal symptoms.

The Pitfalls

• Ordering tumor markers for nonspecific symptoms: This creates false-positive results, unnecessary imaging, invasive follow-up procedures, and patient anxiety—without improving outcomes.
• Using tumor markers to diagnose cancer: Tumor markers support diagnosis in the right clinical context but are NEVER diagnostic on their own. Tissue biopsy is required.
• Not knowing the non-cancer causes: Every tumor marker has a long list of benign elevations. If you don't know the list, you'll over-investigate every positive.
• Ordering panels: "Tumor marker panels" that include CA-125, CEA, CA 19-9, AFP, and CA 15-3 together are almost never indicated in primary care. They generate a statistically inevitable false positive and do not improve cancer detection.
• Repeating tumor markers as "follow-up" without oncology guidance: Post-treatment tumor marker surveillance should be directed by the oncology team, not primary care guesswork.

Bottom Line

Tumor markers are monitoring tools, not screening tools (with the narrow exceptions of PSA with shared decision-making and AFP in cirrhosis surveillance). Ordering them for nonspecific symptoms is one of the most reliably harmful testing decisions in primary care. Know what each marker does, know its benign causes, and resist the urge to order a "cancer panel" for reassurance. The reassurance it provides is false, and the anxiety it generates is real.

Stay sharp out there.