Histological findings of thyroid cancer after lenvatinib therapy

Lenvatinib is a multikinase inhibitor used for treating unresectable or metastatic cancers, including thyroid cancer. As total thyroidectomy followed by radioactive iodine therapy is a commonly recommended initial treatment for thyroid cancer, histological findings of the thyroid after lenvatinib therapy remain unclear. Therefore, the aim of this study was to analyse in‐vivo changes in patients who underwent thyroidectomy after lenvatinib therapy.


Introduction
Lenvatinib is a multikinase inhibitor that primarily inhibits vascular endothelial growth factor receptors. 1 In Japan, lenvatinib is approved for the treatment of unresectable or metastatic cancer. [2][3][4][5] With regard to thyroid cancer, total thyroidectomy, followed by radioactive iodine therapy, is typically performed as the initial treatment. Pharmacotherapies, including lenvatinib, are administered as the second-line therapy. Thus, thyroid specimens surgically resected after lenvatinib therapy are rarely obtained for analysis, and histological findings of the effects of lenvatinib therapy on the thyroid remain unknown. The aim of this study was to investigate the histological findings of thyroid cancer in patients who underwent lenvatinib therapy. The Introduction is detailed in the Supporting information.

Materials and methods
We selected patients who underwent thyroidectomy after lenvatinib therapy for thyroid cancer. Clinicopathological analyses and histological assessment of cancerous thyroid tissues were performed referencing the neoadjuvant therapy for lung cancer. 6 Tumour disappearance rate (%) was also calculated. Areas were considered scarred with no cancer cells if the intrinsic structure of the thyroid gland was unclear without cancer cells but irregular fibrosis, granulation tissue or necrosis was detected. The proportion of residual cancer cells was estimated through histological assessment of the lesions. We examined lenvatinib's angiogenesis inhibition 7 referring to ischaemic colitis patterns. Coagulative necrosis and ischaemic changes in cancer cells were semiquantitatively assessed. 8 Furthermore, microvessel density (MVD) was evaluated according to previously described methods. 9,10 Photographs of CD31-stained sections were subjected to binarisation using Fiji. 11 Subsequently, statistical analyses were conducted. Materials and methods are detailed in the Supporting information.

G E N E R A L F I N D I N G S
Overall, 167 patients with unresectable locally advanced or distant metastatic thyroid cancer who received lenvatinib therapy between June 2015 and June 2022 were screened. Of these, 102, 26 and 39 patients had papillary thyroid cancer (PTC), follicular thyroid cancer (FTC) and anaplastic thyroid cancer (ATC), respectively. Among these, six patients were considered for thyroidectomy. Lenvatinib-induced hypothyroidism was not observed in cases 1-3 but was confirmed in cases 4-6; however, all six patients received levothyroxine (Supporting information, Tables S1 and S2).

H I S T O L O G I C A L F I N D I N G S
In the PTC cases (cases 1-3), a variety of patterns were observed. In case 1, a mixture of coagulative necrosis and viable cancer cells were noted (Supporting information, Figures S1A and S2A). In case 2, fibrosis was prominent at the tumour centre. Case 3 displayed more viable cancer cells with partial ischaemic changes (Supporting information, Figure S3A,B). Inflammatory cell infiltration and lymphoid follicles with germinal centres were seen in non-cancerous thyroid tissues in case 1 ( Figure 1A). Mild lymphocyte infiltration was present in case 2 ( Figure 1B) and case 3 had lymphocyte infiltration, histiocyte reactions and thyroid follicle destruction ( Figure 1C). The metastatic lesions showed more viable cancer cells than the primary lesions (Supporting information, Figure S4A). The thyroid follicular structures in the non-cancerous thyroid tissues were patchy or regionally irregular, but the structure of the thyroid gland was largely preserved (Supporting information, Figure S5A,B).
For FTC (case 4), extensive coagulative necrosis (approximately 60%, Supporting information, Figures S1B and S2B) and ischaemic changes were evident (Supporting information, Figure S3C). Mild lymphocytic inflammatory cell infiltration and partial ischaemic changes were observed in the non-cancerous thyroid tissue. The intravascular carcinoma cells were mostly viable (Supporting information, Figure S4B).
Lastly, in the ATC cases (cases 5 and 6), severe fibrosis with sclerosis was observed in case 5, with noncancerous thyroid tissue not assessable due to fibrosis. Case 6 showed few viable cancer cells with irregular nuclei and hyperchromatin and some lymph follicles (Supporting information, Figure S6). Coagulative necrosis was not apparent in both cases; instead, tumour necrosis or necrotic nests were seen in certain areas (Supporting information, Figures S1C and S2C). The lymphatic metastases showed high densities of viable cancer cells (Supporting information, Figure S4C).

A N A L Y S I S O F M V D
Microvessel density (vessels/mm 2 ) was evaluated in non-cancerous and cancerous thyroid tissues from the six patients treated with lenvatinib (lenvatinibtreated group) and five patients with PTC who did not receive lenvatinib (untreated group). Measurement was not possible in case 5 owing to fibrosis. The median MVD of lenvatinib-treated non-cancerous and cancerous thyroid tissues was 209 and 199.8, whereas that in untreated non-cancerous and cancerous thyroid tissues was 281.8 and 478.1, respectively. Kruskal-Wallis test and Bonferroni correction confirmed significant differences between the respective groups (P < 0.001). The exception was within the lenvatinib-treated group, where no significant difference was found between the non-cancerous and cancerous thyroid tissues (P = 0.371, Figure 2). The results are detailed in the Supporting information.

Discussion
Lenvatinib therapy is administered for unresectable or metastatic cancers. Surgical resection following lenvatinib therapy has been reported. [12][13][14] However, thyroid cancer surgery following lenvatinib therapy is limited. 15 In this study, we identified ischaemic changes in PTC and FTC specimens probably due to inhibition of angiogenesis, a key effect of lenvatinib. 7 Coagulative necrosis was observed around the central lesions in cases 1, 2 and 4. This confirms progressive ischaemia and coagulative necrosis pattern as the final stage of ischaemia. However, some viable cancer cells were detected in the primary lesions. Furthermore, lymph node metastases of PTC and intravascular carcinoma cells of FTC contained significantly more viable cells than the primary lesions. Therefore, it may be difficult to eliminate all cancer cells using lenvatinib alone. Angiogenesis inhibitor therapy may enhance the efficacy of immunotherapy; therefore, combination therapy using immune check-point inhibitors (ICIs) may be promising. [16][17][18] Extensive fibrosis was observed in case 5; however, it should be noted that ATC can cause fibrosis. 19 Some areas of necrosis were also observed in case 5; however, they resembled 'usual necrosis', including nuclear debris. 20 Necrosis may be a process of tumour progression rather than a therapeutic response to lenvatinib therapy. 21 Therefore, whether the fibrosis and necrosis observed were caused by lenvatinib therapy remains unclear. In case 6, the discrepancy between the clinical tumour size reduction rate and histological findings was remarkable because several viable cancer cells were observed. Discrepancies between clinical and histological findings were observed in both ATC cases.
Hypothyroidism is a side effect of lenvatinib therapy 1 ; thus, we examined non-cancerous thyroid tissues. Given the lack of histological knowledge of the thyroid after lenvatinib therapy, we sought to compare lenvatinib with ICIs as therapeutic agents that are likely to cause hypothyroidism. 22 Neutrophilic crypt microabscesses, epithelial cell apoptosis, and crypt atrophy/dropout are present in the intestinal tract after ICI therapy. 23 In contrast, the histology of the non-cancerous thyroid tissues after lenvatinib therapy appeared almost intact. Previous studies reported lower frequency of hypothyroidism after lenvatinib therapy than after ICI therapy. [24][25][26][27] In this study, inflammatory infiltration and ischaemic changes in thyroid follicular cells were not pronounced. It may indicate that the incidence of lenvatinib-associated hypothyroidism is lower than that of ICI-associated hypothyroidism. Tyrosine kinase inhibitors can induce thyroid dysfunction through mechanisms such as thyroid epithelial apoptosis, thyroid blood flow inhibition and unidentified autoimmune processes. [28][29][30] In our cases, apoptosis was not evident and ischaemic changes were sparse. These results suggest that direct epithelial damage may not be the primary cause of hypothyroidism. Therefore, we evaluated MVD in both non-cancerous and cancerous thyroid tissues. MVD decreased significantly in lenvatinib-treated non-cancerous and cancerous thyroid tissues. However, the untreated tissues showed a higher median MVD, suggesting that lenvatinib reduces MVD in cancerous and non-cancerous thyroid tissues. Furthermore, in lenvatinib-untreated tissues, both non-cancerous and cancerous thyroid tissues exhibited high MVD, with the latter showing even higher levels. However, lenvatinib-treated thyroid tissues showed similar MVD regardless of whether they were non-cancerous or cancerous. This implies that although lenvatinib inhibits angiogenesis in both cancerous and non-cancerous thyroid tissues, the effects may be more prominent in cancerous tissues. Decreased MVD in non-cancerous thyroid tissues suggests lenvatinib-associated microvascular damage, potentially leading to hypothyroidism. However, given the small sample size, attributing these outcomes solely to MVD remains uncertain.
Inflammatory cell infiltration in non-cancerous thyroid tissues was a characteristic finding. While not extensive, mild lymphocytic infiltration, pronounced formation of lymphoid follicles with germinal centres and histiocyte reactions were observed in some cases, with significant variance between cases. Case 3, for whom the therapy duration was the shortest, showed the most severe inflammation. The extent of infiltration post-lenvatinib is not solely attributable to dose or duration, suggesting that other factors contribute to the adverse effects. Inflammatory cell infiltration  ). B, Representative images of untreated thyroid papillary cancer (CD31stained specimen, scale bar: 100 lm). C, Comparison of microvessel density in lenvatinib-treated and untreated thyroid tissues. The median MVD in lenvatinib-treated non-cancerous and cancerous thyroid tissues were 209 and 199.8, whereas those in the untreated noncancerous and cancerous thyroid tissues were 281.8 and 478.1, respectively. Kruskal-Wallis test and Bonferroni correction confirmed significant differences between the respective groups (P < 0.001). The exception was within the lenvatinib-treated group, where no significant difference was found between the non-cancerous and cancerous thyroid tissues (P = 0.371). ***P < 0.001; Kruskal-Wallis test and Bonferroni correction. and decreased microvascularity are significant postlenvatinib therapy findings, and their interaction and other factors may contribute to hypothyroidism; however, further investigation is needed to confirm this.

Limitations
Thyroidectomy was performed for patients showing clinical tumour shrinkage, without severe side effects and suitable for surgery; patients at high bleeding risk were excluded. All six patients received levothyroxine before thyroidectomy; their detailed genetic profiles were not investigated. BRAF inhibitors, unapproved for thyroid cancer in Japan, were not used. We evaluated the histological tumour disappearance rate, but this may be related to cancer progression and not lenvatinib therapy.

Conclusion
Lenvatinib probably induces specific ischaemic changes in thyroid cancer cells, with extent depending on the case and histological type. Decreased MVD and variable inflammatory cell infiltration in non-cancerous thyroid tissue may associate to hypothyroidism.

Supporting Information
Additional Supporting Information may be found in the online version of this article: Figure S1. A high-power field of microscopic images showing representative necrosis in thyroid cancer. A, Coagulative necrosis in a papillary thyroid cancer specimen (Case 1). In the eosinophilic necrotic area, the cell outline is preserved, but the nucleus is absent. B, Coagulative necrosis in a follicular thyroid cancer specimen (Case 4). Only the cell outlines and colloids are preserved; the nucleus is not clear. C, Necrosis in an anaplastic thyroid cancer specimen (Case 5). The cell outline is not clear, and nuclear debris is observed. All specimens are stained with H&E. Scale bars: 100 lm. Figure S2. A low-power field of microscopic images showing representative necrosis in thyroid cancer. A, A papillary thyroid cancer specimen (Case 1) showing an irregular mixture of necrotic and viable cancer cell areas. B, A follicular thyroid cancer specimen (Case 4) showing widespread coagulative necrosis. C, An anaplastic thyroid cancer specimen (Case 5) showing partially irregular necrosis. All specimens are stained with H&E. Scale bar: 1000 lm. Figure S3. Microscopic images showing the characteristic cancer cells in papillary and follicular thyroid cancers. A, B, Papillary thyroid cancer lesion in Case 3. Most of the papillary structure observed in the low-power field is preserved; however, some have indistinct structures. In the indistinct area seen in the high-power field, the epithelium is detached, and the nucleus is degenerated. (C) Follicular thyroid cancer lesion in Case 4. Widespread necrosis is prominent; collapse of the follicular structure and concentration of nuclear chromatin are visible. All specimens are stained with H&E. Scale bars: 1000 lm for (A) and (C) and 100 lm for (B). Figure S4. Microscopic images of cancer cells in a lymph node metastasis or intravenous lesion. A, Lymph node metastases in Case 3. Central fibrosis is observed, but the cancer cells seem almost viable. B, Intravascular carcinoma cells in Case 4. The cancer cells preserved their structure and are completely viable. C, Lymph node metastases in Case 6. Cancer cells exhibiting solid proliferation are viable. Neither fibrosis nor necrosis is observed in the background. Specimens are stained with H&E, except for that in image B, which is stained with Victoria blue and H&E. Scale bars: 1000 lm. Figure S5. Microscopic images of thyroid follicular epithelium in non-cancerous thyroid tissue. A, The histological findings of Case 1 as a representative example of the histological features of non-cancerous thyroid tissue. The thyroid follicular structures in a portion of the non-cancerous thyroid tissue are patchy or regionally irregular. B, In the high-power field, the thyroid follicular epithelium shows reduced connectivity, hyperchromatism, and degeneration, indicating ischaemic changes. However, these histological changes were observed in a small proportion of the tissue and were not prominent overall. All specimens are stained with H&E. Scale bars: 400 lm for (A) and 100 lm for (B). Figure S6. Microscopic images showing the characteristic cancer cells in anaplastic thyroid cancer. (A, B) Anaplastic thyroid cancer lesion in Case 5. Severe fibrosis is observed in the low-power field. Residual cancer cells are few and barely visible. Granulation tissue, primarily, fibrosis, is prominent in the high-power field. However, large, irregular, and degenerative carcinoma cells are also visible. (C) Anaplastic thyroid cancer lesions in Case 6. The cancer cells are slightly degenerated but mostly viable. All specimens are stained with H&E. Scale bars: 1000 lm for (A) and 100 lm for (B) and (C). Table S1. Clinical information of the six patients who underwent thyroidectomy after lenvatinib therapy. Table S2. Histological information of the six patients who underwent thyroidectomy after lenvatinib therapy.