CLT is the most common autoimmune disease, and its prevalence has also increased in recent years. In the past decade, the increasing incidence of thyroid cancer together with the simultaneously increasing incidence of CLT leads to the question whether CTL is associated with thyroid cancer. The effect of CLT in the prognosis of PTC has long been a matter of debate. Several previous studies have demonstrated that coexistent CLT was significantly associated with better clinicopathologic characteristics and an improved prognosis among patients with PTC6-9, as demonstrated by reduced incidence of extrathyroidal extension, LN metastasis, distant metastasis and an increased recurrence-free survival duration relative to those in patient without CLT.
It is well known that LN metastasis is an important risk factor for prognosis11, and the relationship between LN metastasis and CLT has been explored. several studies reported that coexisting CLT in patients with PTC was identified as a protective independent predictive factor for LN metastasis6,12 However, other studies did not reveal a protective effect of coexistent CLT on LN metastasis9,13. A large retrospective study of 9210 patients also showed that CLT was not associated with LN metastasis in logistic regression14. Factor that could explain the contradictory results from different studies was differences in the clinicopathological characteristics baseline between coexistent CLT group and PTC alone group in the retrospective study.
In the present study, we observed that the prevalence of LN metastasis was 34.3% in coexistent CLT group and 44.8% in PTC alone group. This result is consistent with previous studies that suggest that CLT is a protective factor for LN metastasis in PTC. However, the clinicopathological characteristics baseline were imbalanced as the previous studies. There was no significant difference in age, primary tumor size, intraglandular dissemination, and bilaterality between the 2 groups in this study, but female patients and the patients with multifocal tumors accounted for a large proportion in coexistent CLT group. There was also a trend toward more extrathyroidal extension in coexistent HT group. These clinicopathological features have been shown to be associated with LN metastasis15. It is difficult to avoid the imbalance of these related factors in the retrospective study, therefor the interference of the related factors will lead to the unreliability of the results.
We performed PSM to match the samples between the groups and balance other related factors affecting LN metastasis. The mean standardized difference analysis and kernel density estimation were used to verify the covariate balance in the matched groups indicated a well-balanced result. The PSM analysis revealed that the prevalence of LN metastasis in patients with CLT (35.0%) was significantly lower than that of patients without CLT(44.7%). This result is consistent with that before PSM. In addition, multivariate logistic regression analysis also showed CLT could decreased risk of LN metastasis in patients with PTC before PSM(OR, 0.63; 95% CI: 0.44 to 0.91; P=0.014) and after PSM(OR, 0.62; 95% CI: 0.40 to 0.96; P=0.032). Both multivariate analysis and PSM analysis showed that CLT had a protective effect on LNM in PTC. And it's this consistency that makes our results even more convincing.
This phenomenon may be attributed to the autoimmune thyroiditis in patients with thyroid cancer. The cancer cells are believed to be destroyed by autoimmunity or thyroid-specific antigens. Fas and the Fas ligand are expressed simultaneously in the follicular cells of CLT, and the Fas-mediated apoptotic pathway plays an important role in thyroid tissue destruction16. The cytotoxic T cell is directly involved in the destruction of carcinoma cells. Moreover, cytokines such as interleukin-1 suppress carcinoma cell growth indirectly.17,18 The autoimmunization process in CLT is associated with antigen-producing B lymphocytes, cytotoxic T cells, and macrophages infiltrating into and accumulating in the thyroid tissue; the autoimmunization extends from thyroid to lymphoid tissue.19 These activities may reflect the mechanisms associated with the protective effects against lymphatic metastasis of thyroid cancer, as observed in this study.
In addition,a significantly more dissected LN and a significantly lower metastatic LN ratio were observed in coexistent CLT group. These results were consistent with the findings of previous studies as well13,20. These findings reflect the phenomenon that, patients with CLT have more noticeably enlarged or hyperplastic LNs caused by long-term inflammation. Grossly enlarged LNs in patients with CLT might lead to complete removal of metastatic LNs which can help us accurately assess tumor staging and develop accurate monitoring and treatment plans. And metastatic LN ratio was proved as an independent predictor for locoregional recurrence in patients with pN1a PTC21.
Despite PSM was performed, the disadvantage of retrospective analysis cannot be entirely avoided, and there were other biases that have not been eliminated, such as the BRAF gene, which was proven as a marker of more aggressive behavior in PTC and a risk factor of LN metastasis22,23. The other limitation of our study is the lack of long-term follow-up data. Recurrence which is a better predictor of prognosis than lymph node metastasis need to be further investigated in future studies.
In conclusion, our findings suggest that CLT may protect against LN metastasis in patients with PTC. Patients with PTC with coexistence CLT had fewer lymph node metastasis, more lymph node dissection, and a smaller metastatic LN ratio was demonstrated in both PSM analysis and multivariate logistic regression analysis.