EPIC Cohort Study: High Thyroglobulin and Low TSH Precede Differentiated Thyroid Cancer, but Neither Can Be Used in Screening
In a study of the association of thyroid-stimulating hormone (TSH), thyroglobulin, and thyroid hormones with risk of differentiated thyroid cancer reported in Journal of the National Cancer Institute, Rinaldi et al found that high thyroglobulin levels can precede thyroid cancer by many years and that low TSH may be associated with thyroid cancer risk. Neither marker had sufficient accuracy to be suitable for use in screening.
Study Details
In the study, 300 women and 57 men with incident thyroid cancer were identified from the EPIC cohort and matched for study center, age, and date, time, and fasting status at prediagnostic blood collection with two (for cases in women) or three (for cases in men) control subjects. Levels of total and free thyroxine (T4) and triiodo-thyronine (T3), TSH, thyroglobulin, and antithyroglobulin antibody were measured by commercially available immunoassays.
TSH and Thyroglobulin Associated With Risk
Mean age at blood collection (51.5 years) and cancer diagnosis (58 years) and mean number of years between blood collection and diagnosis (6.4 years) were similar in women and men. Thyroid cancer cases in both men and women were associated with lower TSH levels (P = .01, P = .05) and higher thyroglobulin levels (P < .001 for both) than in controls.
Median free T3 levels were similar in cases and controls, but the lower levels at lower percentiles in control subjects resulted in a significant difference (P = .04). No significant differences were found between cases and controls for free T3 in men or for T3, free T4, or T4 in men or women.
Comparison of highest vs lowest quartiles showed that lower TSH level (odds ratio [OR] = 0.56, P = .001 for trend) and higher thyroglobulin level (OR = 9.15, P < .001 for trend) were significantly associated with cancer risk. Risk was also higher in the second and fourth quartiles vs the lowest quartile, but there was no statistically significant linear trend (P = .10 for trend).
No significant associations with risk were found for free T3, T3, free T4, or T4. Subjects who were thyroglobulin antibody-positive were at significantly increased risk of cancer (OR = 1.50, P = .03). Odds ratios did not differ substantially after adjustment for weight and height. No significant interaction between TSH and thyroglobulin or thyroglobulin antibody was observed.
Subgroup Differences
No significant heterogeneity in the association of TSH and thyroglobulin with thyroid cancer risk was observed according to sex, age at blood collection, country, education level, smoking, or TNM stage.
Odds ratios for cancer did not differ according to lower or higher body mass index (BMI) for TSH, but the odds ratio for thyroglobulin was higher (P = .05) among subjects with lower BMI (< 26 kg/m2; OR = 2.70, 95% confidence interval [CI] = 2.11–3.45) vs higher BMI (≥ 26 kg/m2; OR = 1.93, 95% CI = 1.53–2.43). The association of thymoglobulin with cancer risk was stronger (P = 0.03) for follicular (OR = 4.45, 95% CI = 2.31–8.56) vs papillary thyroid cancer (OR = 2.10, 95% CI = 1.74–2.54).
Odds ratios for TSH did not vary over time between blood collection and thyroid cancer diagnosis. The odds ratio for thyroglobulin was highest at < 4 years before diagnosis, but the trend for time was not significant (P = .12 for trend).
Time Between Measurement and Diagnosis
Among cancer cases, TSH levels were stable over the years before diagnosis (P = .57 for trend). Thyroglobulin levels were highest with closer proximity of blood collection to diagnosis (P = .002 for trend), with adjustment for age having no substantial effect on this finding.
Area under the receiver operating characteristic curve values for distinguishing subjects who did vs did not develop thyroid cancer were 57% for TSH and 74% for thyroglobulin. At 90% specificity, the sensitivity of thyroglobulin level for detection of cancer was 43%. At 80% sensitivity, specificity was reduced to 42%.
The investigators concluded, “High [thyroglobulin] levels precede by up to 8 years the detection of [thyroid cancer], pointing to a long sojourn time of the disease. Low TSH levels may predispose to [thyroid cancer] onset. Neither marker has sufficient accuracy to be a screening test.”
The study was supported by the World Cancer Research Fund and others.
Silvia Franceschi, MD, of the International Agency for Research on Cancer, Lyon, is the corresponding author for the Journal of the National Cancer Institute article.
The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.
