The Influence of Incidental Detection of Thyroid Nodule on Thyroid Cancer Risk and Prognosis - a Systematic Review

Clinically unapparent thyroid nodules discovered serendipitously on imaging for non-thyroid indications are termed ‘thyroid incidentalomas’. The increase in the detection of these incidentalomas (which are known to be very common) has been attributed to the widespread use of diagnostic imaging and the increase in sensitivity and resolution of these modalities. It is unclear whether these incidentalomas have a lower prevalence of thyroid cancer or slower tumour progression compared to symptomatic thyroid nodules. This systematic review aimed to determine the risk of malignancy in incidentally detected thyroid nodules and its impact on prognosis in patients with thyroid cancer. PubMed and MEDLINE® on Web of Science databases were searched from inception to March 2020 for English language articles reporting on human studies of thyroid cancer risk and/or prognosis in incidental and non-incidental nodules. Quality of included studies was assessed using the Newcastle-Ottawa Scale (NOS). Seventeen observational studies published between 1998 and 2018 were eligible for analysis; 4 studies reported on risk, 8 studies on prognosis and 5 studies on both risk and prognosis. In the risk review, the odds ratios calculated from the six case-control studies (3246 patients) ranged from 0.64 to 2.86 whilst the relative risks calculated from the three cohort studies (489 patients) ranged from 0.13 to 6.27. NOS score for included risk studies (n=9) ranged from 22.2% to 66.7%. A meta-analysis of the eligible case-control studies (n=3) showed a non-significant summated odds ratio of 1.04 (95% CI=0.63-1.70, p=0.88). In the prognosis review of thirteen studies, three direct and thirteen indirect markers of prognosis were compared between the incidental (1923 patients) and non-incidental (2639 patients) groups. NOS score for included prognosis studies ranged from 66.7% to 100%. Incidentally detected thyroid nodules were significantly more likely to be smaller, have lower rates of extra-thyroidal and extra-nodal extension and lymph node metastasis, and interestingly more likely to have advanced disease. Other indirect prognostic markers were not shown to be significantly different between the two groups. A meta-analysis was not possible but incidentally detected thyroid cancer had better progression-free and overall survival; this finding may be affected by ‘lead time’ bias. Current evidence suggests that the investigation and management of thyroid nodules should not be influenced by the mode of detection.