Predictors of poor prognosis in long-term survivors of differentiated thyroid cancer with psychiatric disorders.

Abstract

Background: Thyroid malignancies, while accounting for a small proportion of cancer diagnoses globally, have demonstrated a consistent upward trend in occurrence across diverse populations, with gender-specific analyses revealing a disproportionate burden among women. Despite the characteristically indolent nature of most thyroid carcinomas and their associated high survival rates, emerging evidence points to significant unmet needs regarding psychosocial adaptation and neuropsychiatric sequelae in this growing survivor population. The spectrum of mental health conditions, ranging from affective and anxiety disorders to cognitive impairments, presents substantial barriers to functional recovery and may potentially influence disease trajectories through complex psychoneuroimmunological pathways. Clinical observations consistently report elevated rates of mood disturbances and executive function deficits persisting throughout the cancer continuum, from active treatment into extended follow-up periods. These findings highlight a critical knowledge gap in understanding the dynamic interplay between thyroid cancer biology, its therapeutic interventions, and the development of treatment-resistant psychiatric manifestations that complicate long-term patient care.

Aim: To analyse the factors influencing the poor prognosis of patients surviving long-term differentiated thyroid cancer with psychiatric disorders and to construct a prediction model.

Methods: Forty-eight patients with mental disorders combined with differentiated thyroid cancer who were treated in our hospital during the period of March 2018 to March 2023 were retrospectively selected as the study subjects (thyroid cancer group), and 30 cases each of patients with mental disorders combined with benign thyroid nodules (benign nodules group) and patients with mental disorders alone (mental disorders group), who were treated during the same time period, were selected as controls. The patients with differentiated thyroid cancer were further divided into a poor prognosis group (10 cases) and a good prognosis group (38 cases). The study outcome was poor prognosis as shown by whole body bone imaging within 2 years after thyroid cancer surgery. Factors influencing poor prognosis in survivors of differentiated thyroid cancer were analyzed by univariate and multivariate logistic regression analyses, receiver operating characteristic (ROC) curve analysis was used to assess the predictive efficacy of these factors for poor prognosis, and the DeLong test was used to determine whether there was a statistically significant difference in the area under the curve (AUC) of the model.

Results: One-way logistic regression analysis showed that tumour diameter [odds ratio (OR) = 19.190, P = 0.002], T-stage (OR = 7.692, P = 0.018), extra-glandular infiltration (OR = 37.000, P = 0.003), degree of differentiation (OR = 24.667, P = 0.008), serum free T3 (OR = 22.348, P = 0.025), serum free T4 (FT4) (OR = 1.158, P = 0.002), total bilirubin (TBil) (OR = 1.792, P = 0.004), albumin (OR = 0.675, P = 0.003), cortisol (OR = 1.180, P = 0.003), norepinephrine (OR = 1.047, P = 0.002), angiotensin II (OR = 1.975, P = 0.002), and superoxide dismutase (OR = 0.515, P = 0.005) all increased the risk of poor prognosis in patients with psychiatric disorders and long-term differentiated thyroid cancer. Multifactorial logistic regression analysis showed that tumour diameter (OR = 16.570, P = 0.021), extra-glandular infiltration (OR = 53.145, P = 0.010), FT4 (OR = 1.186, P = 0.007), and TBil (OR = 2.823, P = 0.048) were independent risk factors for poor prognosis of patients with psychiatric disorders with long-term differentiated thyroid cancer, and the regression equation was: Y = 2.808 × tumour diameter + 3.973 × extra-glandular infiltration + 0.171 × FT4 + 1.038 × TBil - 88.138. ROC analysis showed that the predictive power of the overall model (AUC = 0.992, P = 0.000) was significantly higher than that of independent risk factors (DeLong test P < 0.05).

Conclusion: Tumour diameter, extra-glandular infiltration, FT4, and TBil are independent risk factors for poor prognosis in patients with psychiatric disorders with long-term differentiated thyroid cancer, and the combination of these factors is of higher value in predicting the prognosis of patients. These risk factors can be used as a basis to develop a reasonable prognostic management plan in clinical practice for patients with long-term differentiated thyroid cancer with mental disorders, so as to improve the prognosis and quality of life of patients.