Lower Collagen Area and Contralateral Lymph Node Metastasis Type Associated with Lower Disease-Free Survival in Thyroid Papillary Carcinoma

Objectives: To introducing a novel prognostic risk scheme which based on collagen area and lymph node metastasis type in thyroid papillary carcinoma (PTC). Method: Tumor collagen area and lymph node matastasis type as well as several other histomorphological factors with disease-free survival (DFS) were investigated in a corhort of 101 PTC patients. Results: Median follow-up time of DFS was 76 months(inter quartile range: 71–83 months). Low collagen area and contralateral lymph node matastasis type were associated with dismal patient disease-free survival (DFS) signicantly. While tumors with high collagen area showed a mean DFS of 77.37 months, gradually decreased to 76.53 months for tumors with moderate collagen area and to 66.67 months for tumors with low collagen area (p =0.028). Negtive, central and ipsilateral lymph node metastasis showed mean DFS of 72.72, 74.8 and 72.81 months respectively, decreased to 62.17 months with bilateral lymph node metastasis and 59.43 months with contralateral lymph node metastasis (p=0.029). In the cohort, uinivariate statistical analysis of the novel risk scheme revealed that the hazard ratio (HR) for DFS was 5.18 for R2 and 15.53 for R3 tumors compared to R1 PTC (p=0.003). Disease free survival dropped from 77.74 months for R1 tumors to 68.7 months for R2 and 56.38 months for R3. Multivariate statistical analysis of the novel risk scheme revealed that the HR for DFS was 3.259 (95% condent interval [CI] 1.122–9.469, p=0.03). Conclusion: Our novel risk algorithm incorporating tumor collagen area and lymph node matastasis type allows strongly prognostic stratication of PTC. We suggest this risk algorithm as a morphology-based parameter for the routine diagnostic assessment of this tumor entity.

However, these potential prognostic markers did not include accurate prognostic patient strati cation according to morphological factors.
Guidelines of the WHO introduced a risk strati cation system according to clinical outcomes following total thyroidectomy and radioiodine remnant ablation or adjuvant therapy 5 . Also, WHO-based grading of PTC does not allow prognostic patient strati cation according to morphological factors.
To clarify this question, we evaluated three common morphological factors: tumor collogen area, lymphatic follicles around tumor and calci cation, as well as other clinicopathological factors (age, gender, operative method, lymph node matastasis type and T stage). All factors regarding disease-free survival(DFS) were assessed.

Patients
The records of 222 patients treated surgically for PTC between 2012 and 2015 were identi ed from database of the rst a liated hospital of southern china. Of these, one hundred and three patients had pathologic con rmation and available follow-up information. Two patients died and were excluded from analysis. Finally, 101 patients were included in this study.

Histological evaluation
Full block haematoxylin and eosin stained PTC slides of every case were evaluated by two pathologists (CCJ, TF)who were blinded to clinicopathological data and follow-up. Furthermore, 10 randomly selected cases from the cohort were evaluated by ZQ consecutively in order to investigate interobserver reproducibility. According to tumor-node-metastasis staging system of the American Joint Committee on Cancer 4 , regional lymph node metastasis is classi ed as central lymph node metastasis (N1a) and lateral lymph node metastasis (N1b). As was shown in table S1, we subdivided lymph node metastasis type into ve subgroups (Table S1). Collagen types were divided into three categories according to the area of collagen and tumor area( Figure S1). Calci cation condition were divided into two groups: present and absent of calci cation. Lymphatic follicles around tumor were also evaluated: area with more than 5 lymphatic follicles /HPF around tumor were in one group; the rest were in the other group.

Statistics
Statistical analysis was performed using SPSS 20. Desease free survival probabilities were plotted with the Kaplan-Meier method, a log-rank test was used to probe for the signi cance of differences in survival   Figure S2A). Lymph node metastasis type also showed prognostic for DFS: negtive, central and ipsilateral lymph node metastasis showed mean DFS of 72.72, 74.8 and 72.81 months respectively, decreased to 62.17 months with bilateral lymph node metastasis and 59.43 months with contralateral lymph node metastasis (p = 0.029, Table S1 and gure S2B). Other clinicopathological factors showed no association with DFS in the test (Table S1).

Composition of the grading system
We established a novel risk scheme based on collagen area and lymphnode metastasis type (Table 1), attributing a score to both collagen type (1-2 points) and lymphnode metastasis type(1-2 points).
Tumors with moderate and high collagen area received a score of 1 and tumors with low collagen area had a score of 2.

Discussion
Age and gender were reported as signi cant parameters in DFS for thyroid cancer in some studies 8-10 . In our study, age and gender did not signi cantly associated with DFS. The critical reason may be the heterogeneous conditions of other studies; including all the differentiated thyroid cancers in the study other than papillary thyroid carcinoma.
American thyroid association (ATA) introduced a risk strati cation system according to clinical outcomes following total thyroidectomy and radioiodine remnant ablation or adjuvant therapy. Intrathyroidal PTCs of all sizes were included in the ATA low-risk category, the risk of structural disease recurrence decreased from 8-10% in intrathyroidal PTC > 4 cm 11 , to 5-6% in 2-4 cm intrathyroidal PTC 11 , to 4-6% in multifocal papillary microcarcinomas and 1-2% in unifocal papillary microcarcinomas 12,13 . In our study, T stage classi cation not only included intrathyroidal PTCs, but also other parameters. Therefore, the signi cance of T stage in DFS may not be displayed, and this was in accordance with Yuksel UM et.al's study 14 .
A highly prognostic grading approache for uterine cervix 15 , pulmonary 16 , and several squamous cell carcinomas was based on tumor budding and cell nest size. We also assessed these histopathological parameters in PTC. However, interobserver reproducibility between pathologists were not high, and probably due to glands features. A prognostic morphological factor that associated with DFS was collagen area in PTCs. Moderate and high collagen area associated with higher DFS than low's (p = 0.028). Zeng R et al. reported that there is a positive assosiation between hashimoto's thyroiditis and PTC in children and adolescents. A prominent characteristic of hashimoto's thyroiditis is the presence of lymphatic follicles 17 . However, we did not found signi cant association between lymphatic follicles around tumor and DFS. Calci cation could be found in PTC occassionally. Neither, this morphology feature showed association with survival in the cohort.
The prognostic signi cance of nodal metastases from papillary thyroid carcinoma can be strati ed based on the size and number of metastatic lymph nodes, as well as the presence of extranodal extension 18 . With ipsilateral lymph node metastases, contralateral lymph node metastases are more prevalent 19 . However, it is not impossible that papillary thyroid carcinoma located in one lobe showing contralateral but not ipsilateral N1b. Based on N1a and N1b of American Joint Committee on Cancer, lympho node metastasis type were subdivided into 5 types in our study. Bilateral and contralateral lymph node metastasis type were signi cantly associated with shortened DFS (p = 0.029). This was inaccordance with Yuksel, U. M.'s study 14 . Two operative methods were included in this study: total thyroidectomy with lymph node dissection and subtotal thyroidectomy with lymph node dissection. This operative factor showed no association with DFS.
Knezević-Obad A et al. reported that positive cytologic examination played an important role in determing total thyroidectomy for patients 20 . To the best of our knowledge, no risk scheme could indicate a higher or lower DFS with collagen area morphological factor. Our risk scheme contained collagen area and metastasis type and showed as an independent prognostic factor for DFS in PTCs.
Our retrospective study is limited by the fact that our analyses contains a small sample size. The initial corhort included two patients who were died due to PTC recurrence. However, the sample was not enough for stastistical analysis.
Taken together, the novel risk scheme contains collagen area and metastasis type and is easy to implement. Our data underlines the assumption that tumor collagen area and lymph node metastases type have the potential to constitute the pillars of a highly prognostic DFS risk scheme in PTC.

Declarations
Ethics approval and consent to participate Ethical approval was approved by the EthicsCommittee of the First A liated Hospital of Universityof Southern China and informed consent wasobtained from all patients involved in this study

Availability of data and material
The data that support the ndings of this study are available on request from the corresponding author.