Tumour burden, thyroglobulin and Ki-67 as prognostic indicators in papillary thyroid cancer

Each year, more than 700 patients in Sweden are diagnosed with thyroid cancer, with papillary thyroid cancer (PTC) being the most common type, accounting for approximately 80% of cases. The incidence of PTC has increased in high- and middle-income countries, doubling in Sweden over the past two decades. Despite this incidence rise, mortality rates remain low, highlighting concerns about overdiagnosis and overtreatment. While most patients have an excellent prognosis - demonstrating a 93% disease-specific survival rate over ten years - a subset of patients experience recurrence and poorer outcomes. Currently, treatment decisions for PTC rely on anatomical criteria, underscoring the need for additional prognostic biomarkers to identify high-risk patients. In recent years, the role of tumour biology and tumour microenvironment, particularly the tumour stroma, has gained recognition as a crucial factor in cancer progression and treatment resistance. Histopathological biomarkers are increasingly integrated into risk stratification and treatment planning for various cancers - yet their application in PTC remains limited.

Aims: This thesis aimed to identify and evaluate potential histopathological biomarkers in primary PTC tumours and metastatic lymph nodes for improved prognostic risk stratification.

Methods: Studies I, II and IV were retrospective analyses of patients surgically treated for PTC at Karolinska University Hospital (2006-2017). We examined the immunohistochemical expression of the Ki-67 labeling index, thyroglobulin, and the lymph node ratio (LNR) in relation to clinical data and PTC recurrence.

Study III was a feasibility study on AI-assisted digital pathology, assessing thyroglobulin expression H-score and tumour-stroma ratio in primary tumours and their corresponding metastatic lymph nodes.

Results: In study I, 165 patients were included. Lymph node metastases were significantly more common in younger patients, those with multifocal tumours, a Ki-67 index >3%, and extrathyroidal extension. The Ki-67 index in the primary tumour and LNR significantly correlated with tumour recurrence.

In study II an expanded cohort of 327 patients was used. Findings from Study I were confirmed, with LNR ≥21% independently associated with tumour recurrence. The Ki-67 index was higher in the lymph node metastases (mean 4%) than in primary tumours (mean 3%). Thyroglobulin expression was generally higher in the primary tumour than in metastatic lymph nodes. Low thyroglobulin expression (0-25%) in lymph node metastases was associated with shorter recurrence-free survival and a higher risk of recurrence in PTC - likely due to reduced response to adjuvant radioiodine therapy.

In study III, we found that AI-assisted digital image analysis was adequate for assessing tumour-stroma ratio and estimating thyroglobulin expression in primary tumours and lymph node metastases. However, the clinical significance of these biomarkers requires future investigation in a larger patient cohort.

Study IV included a subgroup of patients (n=245) from the Study II cohort with clinically node-negative PTC - NO, Nx and Nla stage. We found that microscopic extrathyroidal extension (mETE) was associated with larger tumour size, a higher Ki-67 index, and a sixfold increased risk of recurrence compared to tumours without mETE. Tumours ≤2 cm had significantly fewer recurrences (2.8%) and longer recurrence-free survival than tumours >2 cm (12%). Tumours with a high Ki-67 index (>3%) had a worse prognosis, though neither the Ki-67 index nor tumour size were independent risk factors for recurrence.

Conclusion: Our findings highlight the potential prognostic value of the Ki-67 index and thyroglobulin expression in primary tumours and lymph node metastasis for predicting PTC recurrence. A high metastatic burden (LNR ≥21 %) significantly impacts recurrence risk in N1b-stage PTC. In clinically node-negative PTC, microscopic ETE might be associated with recurrence. These insights reinforce the need to integrate histopathological biomarkers into risk stratification strategies for PTC, potentially improving individualised treatment decisions.

List of scientific papers

I. Lindfors H, Ihre Lundgren C, Zedenius J, Juhlin CC, Shabo I. The Clinical Significance of Lymph Node Ratio and Ki-67 Expression in Papillary Thyroid Cancer. World Journal of Surgery. 2021 Jul;45(7):2155-2164. https://doi.org/10.1007/s00268-021-06070-y

II. Lindfors H, Karlsen M, Karlton E, Zedenius J, Larsson C, Ihre Lundgren C, Juhlin CC, Shabo I. Thyroglobulin expression, Ki-67 index, and lymph node ratio in the prognostic assessment of papillary thyroid cancer. Scientific Reports. 2023 Jan 19;13(1):1070. https://doi.org/10.1038/s41598-023-27684-3

III. Lindfors H, Hellgren LS, Zedenius J, Ihre Lundgren C, Juhlin CC, Shabo I. Assessment of tumor-stroma ratio and thyroglobulin immunohistochemical staining intensity scoring in papillary thyroid cancer using automated digital image analysis: A feasibility study. [Manuscript]

IV. Lindfors H, Zedenius J, Ihre Lundgren C, Juhlin CC, Shabo I. The roles of tumor size and biological criteria in estimating the treatment options in papillary thyroid cancer. [Manuscript]