Abstract
Aim: Programmed death-ligand 1 (PD-L1) is a component of the tumor microenvironment, which is closely involved in the occurrence and development of tumors. We investigated the correlation between the expression of PD-L1 and ultrasound characteristics in papillary thyroid carcinoma (PTC) and its effect on recurrence. Patients and Methods: Fifty-two patients at the Department of General Surgery of Qingdao Municipal Hospital underwent thyroid ultrasonic examination before surgery and their clinicopathological variables were collected. Then, immunohistochemistry staining was conducted to evaluate the PD-L1 expression in tumors and adjacent normal tissues. The correlations of PD-L1 expression with clinicopathological and ultrasound characteristics were analyzed. Results: The expression of PD-L1 was positive in 59.7% (40/67) of PTC tumor tissues. In clinicopathological analyses, PD-L1 positivity was related to multifocality of tumors (p=0.031). In analyses of ultrasound characteristics, the expression of PD-L1 was positively correlated with halo sign (p=0.035), capsular invasion (p=0.003), microcalcification (p=0.02), and recurrence (p=0.009). In multivariate logistic analysis of ultrasonic characteristics and recurrence of thyroid carcinoma, microcalcification [odds ratio=13.349, 95% confidence interval (CI)=2.052-86.832, p=0.007] and the halo sign (odds ratio=15.273, 95% CI=1.451-160.747, p=0.023) were factors associated with recurrence of PTC. In the multivariate Cox regression analysis, positive PD-L1 staining [hazard ratio (HR)=5.031, 95% CI=1.092-23.172, p=0.038] and a halo sign (HR=4.998, 95% CI=1.084-23.051, p=0.039) were independent predictors for poorer recurrence-free survival. Positive expression of PD-L1 predicted worse recurrence-free survival in the subgroup of patients with a halo sign (HR=6.537, 95% CI=1.863-22.94, p=0.037). Conclusion: Positive expression of PD-L1 in PTC affects the clinical and ultrasonic characteristics of the tumor and may negatively affect the prognosis of patients with PTC.
Thyroid cancer is a common malignant tumor of the endocrine system. The incidence of thyroid cancer in the world, especially in the female population, continues to grow (1, 2). According to the pathological type, most patients with thyroid cancer have differentiated thyroid cancer (3), among which papillary thyroid carcinoma (PTC) accounts for 85-90% of all thyroid cancer. As most thyroid cancers progress slowly and have a good prognosis, with an overall median survival time of 35-40 years, they are often regarded as inert tumors. However, lymph node metastasis can occur in about 30-40% of patients (4, 5) and 5% develop distant metastases (6). In addition, disease in about 2% of patients progresses to anaplastic thyroid cancer, which also leads to 50% of deaths related to thyroid cancer every year (7). At present, treatment for these patients is limited, and the median survival time is reduced to only 3 to 5 months (8).
Programmed death-ligand 1 (PD-L1) has been identified as an integral part of the tumor microenvironment and is closely associated with the occurrence and development of cancer (9-11). PD-L1 is abnormally expressed in tumor cells in response to the ongoing stimulus of an oncogenic inert tumor-signaling pathway (12). In addition, PD-L1 is also expressed in tumor-infiltrating lymphocytes, including intertumoral and peritumoral lymphocytes. Tumors that overexpress PD-L1 are more likely to resist innate autoimmunity than those that do not, which makes it easier for the tumor to progress (13-15).
Ultrasound is the preferred mode for examination of thyroid disease and reflects the biological characteristics of the tumor. As far as we are aware, there is no study exploring the relationship between PD-L1 expression in PTC and tumor imaging and its predictive value in prognosis. This study aimed to explore the effect of PD-L1 on the biological characteristics of PTC, its correlation with ultrasonic characteristics, and their combined effect on the prognosis of thyroid cancer, in the hope of monitoring and evaluating its long-term prognosis.
Patients and Methods
Patients. Fifty-two patients who underwent thyroidectomy at the Department of General Surgery of Qingdao Municipal Hospital from 2018 to 2019, with their postoperative pathology being identified as PTC, were included. Inclusion criteria: Those without previous history of thyroid surgery, but with complete medical history, clear pathological diagnosis of PTC, and complete preoperative ultrasound examination. Patients without PTC, patients without preoperative ultrasound images, or with nonstandard and incomplete ultrasound images were excluded. A total of 67 cancer tissues were collected after surgery and all tissue samples were fixed in formalin and embedded in paraffin. According to postoperative microscopic pathology, the tumor tissue was re sliced into thin sections, and normal thyroid tissues about 5 mm away from the tumor were obtained. All of the tissue sections were re-examined and classified according to the current standards of the World Health Organization (16) by pathologists blind to the outcome of disease.
Qingdao Municipal Hospital Ethics Committee approved the protocol of this study (2020-010). Since it was a retrospective analysis, the requirement for informed consent was waived. Clinical details of all surgical patients were obtained from the case database.
Clinical variables. The clinical variables studied were age at diagnosis, sex, maximal tumor diameter, tumor multifocality (defined as two or more tumor sites within the thyroid; either unilateral or bilateral), lateral lymph node metastasis and central lymph node metastasis. Patients were divided into those 40 years old or under and those over 40 years old. The prognosis of tumor was analyzed as a binary variable to indicate whether any local recurrence or distant metastasis had occurred.
Ultrasonic examination. Before surgery, all patients routinely underwent ultrasonic thyroid examination, and the ultrasound images were assessed by two doctors (WH. H. and XL. C.) with more than 10 years’ work experience with the Philips IU22 (Philips, Best, the Netherlands) and Mindray Resona7S (Shenzhen Mindray Bio-Medical Electronics Co., Ltd., Shenzhen, PR China) color Doppler ultrasound systems. When there was any difference in image interpretation between the two, a third doctor was brought in to reach an agreement. The lesion was evaluated from the following aspects: (i) Location: The left or right lobe and isthmus of thyroid gland; (ii) orientation: tendency for the long axis of the nodule and the skin line to be vertical (height exceeded width) or horizontal (width exceeded height); (iii) edge: regular, irregular, or ill-defined according to clarity and regularity; (iv) boundary: divided into well-defined and poorly defined according to clarity; (v) halo: thin or thick halo by a demarcation of 2 mm (thick halo, ≥2 mm; thin halo, <2 mm in thickness) (17), further divided into complete and incomplete halo according to uniformity, and, based on the homogeneity of the halo thickness, divided into halo of even or uneven thickness (vi) echogenicity: compared with the echogenicity of the thyroid parenchyma and the strap muscles of the neck, the echogenicity of the tumor was divided into hyperechoic, isoechoic, hypoechoic, markedly hypoechoic, and mixed (cystic with solid components); (vii) echotexture: homogeneous or heterogeneous according to the diversity and consistency of the tumor echoes; (viii) microcalcification: punctate echogenic foci of less than about 1 mm (18); (ix) size (cm); (x) color Doppler flow imaging: by Alder classification (19), divided into class 0, I, II, or III.
Immunohistochemistry. The tissue paraffin blocks were cut into 4-μm sections for analysis. Primary antibody to PD-L1 (rabbit monoclonal antibody, Ab228462, 1:50 dilution; Abcam, Cambridge, UK) was applied to test for the expression of PD-L1 in thyroid nodules and adjacent normal tissues. Paraffin-embedded sections were deparaffinized with xylene and rehydrated with descending series of ethanol. Heat-induced antigen retrieval was carried out by boiling sections in EDTA buffer (pH 9.0, diluted 1:50) at 100°C for 3 min and sections were then incubated with 10% H2O2 for 10 min to quench the endogenous peroxidase activity. After that, sections were incubated with anti-PD-L1 at 37″C for 1 h, and then the secondary antibody combined with horseradish peroxidase was applied at 37°C for 15 min. The immunoreactivity was demonstrated with diaminobenzidine solution for 1-2 min. The section was counterstained with hematoxylin for 30 s-3 min, sealed, and then fixed.
The tissue sections stained by immunohistochemistry were interpreted by two experienced pathologists blind to the clinical parameters of each case. When there was any discrepancy in assessment, the final result was obtained by consensus. Partial or total staining of the cell membrane was considered to be PD-L1-positive (Figure 1), while PD-L1 staining of the cytoplasm of cancer cells was ignored.
Expression of programmed death-ligand 1 (PD-L1) in papillary thyroid cancer tissues by immunohistochemistry. Partial or total staining of the cell membrane was considered PD-L1-positive. A: Positive expression of PD-L1 (100×). Yellow and black arrows indicate positive PD-L1 expression, with partial or complete staining of the cell membrane showing a brownish yellow color. B: Higher magnification (400×) of the area indicated by the yellow arrow in part A. The black arrow indicates clear staining of the cell membrane.
Statistical analysis. According to the results, quantitative data that had an abnormal distribution are expressed by the median; for those with a normal distribution, data are presented as the mean±standard deviation. Chi-squared test was carried out for analysis of basic clinical information and ultrasonic characteristics in grouped cases, such as age, sex, and echogenicity of lesions. Mann-Whitney U-test was conducted on ordered data, for example, blood flow signal grading, for statistical analysis. Univariate logistic regression analysis was adopted to screen meaningful independent variables that influenced the recurrence of thyroid cancer and then multivariate logistic regression was adopted to analyze factors affecting the recurrence of thyroid carcinoma comprehensively. The effect of a single factor between groups on recurrence-free survival (RFS) was analyzed according to the Kaplan-Meier survival curves with log-rank test, and the effect of multiple factors on survival time was analyzed by Cox regression. RFS was defined as the time between the date of the first operation and the first recurrence, or death. Associations and differences were considered statistically significant when p<0.05.
Results
Clinical characteristics of patients. In this study, 52 patients with PTC were recruited, aged from 22 to 72 years (with a median age of 48 years). All patients with primary thyroid tumors underwent thyroid surgery. Among them, 17 (32.69%) cases underwent lateral thyroidectomy and ipsilateral central lymph node dissection, 35 (67.31%) cases underwent total thyroidectomy, and each PTC patient generally underwent central lymph node dissection. Lymph node metastasis only involved the central region (N1a) in 16 (30.77%) cases, the central region and the lateral region (N1b) concurrently in seven (13.46%) cases, and only the lateral region in one (1.92%) case. The patients had no preoperative history of head and neck radiotherapy, or distant metastasis during preoperative diagnosis.
Correlation between the PD-L1 expression and clinicopathological characteristics. The partial or total staining of the membrane of cancer cells was considered to represent positive PD-L1 expression. PD-L1 was detected by immunohistochemical staining in thyroid nodules and adjacent normal tissues. In this study, PD-L1 expression was positive in 59.7% (40/67) of tumor tissues, but none was observed in peripheral normal tissues (p<0.001). In terms of clinical features, there were 20 (38.46%) multifocal PTC cases, of which 13 (65%) had PD-L1 positive expression. PD-L1 expression was positively associated with multifocality (p=0.031) (Table I). But PD-L1 positivity was not significantly associated with age, sex, lymph node metastasis, or B-Raf proto-oncogene serine/threonine kinase (BRAF) V600E (p>0.05) (Table I).
Correlation of programmed death-ligand 1 (PD-L1) expression with clinical parameters of papillary thyroid carcinoma.
Correlation between PD-L1 expression and ultrasonic signs in cancer tissues. The correlation between characteristics of tumor ultrasonic imaging and PD-L1 expression was determined. The median diameter of thyroid nodules was 1.4 cm (ranging from 0.3 to 5.8 cm). According to the PD-L1 expression in the tumors, patients were divided into PD-L1-positive and PD-L1-negative groups. Next, the correlation between PD-L1 expression in PTC nodules and ultrasound features and outcomes was determined (Table II). The expression of PD-L1 was positively correlated with halo sign (p=0.035), capsular invasion (p=0.003), microcalcification (p=0.02), and recurrence (p=0.009), but was not correlated with tumor size, composition, echogenicity, boundary, location, CDFI, orientation or margin (p>0.05). In further analysis of the halo sign, most of the patients with positive PD-L1 expression presented an incomplete halo (p=0.003) (Table II).
Correlation of programmed death-ligand 1 (PD-L1) expression with ultrasound features of papillary thyroid carcinoma nodules.
Correlation between ultrasonic signs of tumor and recurrence. The follow-up was performed until March 2021, resulting in a median follow-up of 23 months (ranging from 7 to 34 months). A total of seven (13.46%) patients had focal or distant metastasis and thyroid nodules that had recurrence after thyroid cancer surgery were analyzed. The prognosis of patients (whether there was focal or distant metastasis) was converted into a binary variable and univariate logistics regression was employed to explore the correlation between sonographic signs of nodules and the occurrence of local or distant recurrence. The results showed that an obscure boundary, microcalcification, capsular involvement and the presence of a halo were the main ultrasonic characteristics associated with lesion recurrence (p<0.05) (Table III).
Univariate analysis of association of ultrasound characteristics with positive programmed death-ligand 1 expression in papillary thyroid carcinoma.
In clinical practice, the ultrasonic signs of thyroid lesions cannot be used as direct signs to diagnose the disease independently, so our study not only included meaningful independent variables in the above univariate analysis, but also incorporated the aspect of nodules, as a matter of experience. After multivariate logistic analysis, microcalcification [odds ratio (OR)=13.349, 95% confidence interval (CI)=2.052-86.832, p=0.007] and the halo sign (OR=15.273, 95% CI=1.451-160.747, p=0.023) remained factors significant for the recurrence of PTC and it was inferred that these two factors were independent predictors for the recurrence of PTC.
Impact of PD-L1 expression and ultrasonic characteristics in cancer on RFS. The effects of clinical variables of cases, PD-L1 expression, and ultrasonic manifestations on time to recurrence were analyzed by Kaplan-Meier survival curves with the log-rank test. This indicated that positive tumor PD-L1 expression [hazard ratio (HR)=5.890, 95% CI=1.893-18.33, p=0.008] (Figure 2A), and ultrasound features of poorly-defined boundary (HR=4.049, 95% CI=1.289-12.72, p=0.020), capsular involvement (HR=4.209, 95% CI=1.336-13.27, p=0.016), microcalcification (HR=7.526, 95% CI=2.378-23.82, p=0.020) and peripheral halo sign (HR=5.844, 95% CI=1.878-18.18, p=0.008) were associated with worse RFS (Figure 2B-E, respectively), while the remaining ultrasonic characteristics had no statistical significance for RFS. Furthermore, multivariate Cox regression was used to analyze the correlation between risk factors and RFS in patients with PTC. Limited to recurrent cases, only variables significant in univariate analysis were included. In the multivariate analysis, positive staining of PD-L1 (HR=5.031, 95% CI=1.092-23.172, p=0.038) and ultrasonic halo sign (HR=4.998, 95% CI=1.084-23.051, p=0.039) were independent predictors for local or distant recurrence after adjusting for ultrasound and pathological features: halo, orientation, capsule invasion, microcalcification and boundary. According to the ultrasonic halo sign, patients were divided into halo-positive and -negative groups. Kaplan-Meier survival curves with log-rank test were used to judge the difference in the impact of PD-L1 staining in tumor tissue on survival between the two groups. It was found that positive expression of PD-L1 was also related to worse RFS in patients with an ultrasonic halo sign (HR=6.537, 95% CI=1.863-22.94, p=0.037) (Figure 2F and G).
Kaplan-Meier recurrence-free survival curves based on expression of programmed death-ligand 1 (PD-L1) in papillary thyroid cancer by immunohistochemistry (A) and ultrasonic characteristics of papillary thyroid carcinoma, namely boundary (B), capsular invasion (C), microcalcification (D) and halo (E). Kaplan-Meier survival curves according to the expression of PD-L1 by immunohistochemistry in papillary thyroid cancer tissues highlight the negative influence of a halo finding in ultrasonography (F and G).
Discussion
The prognosis of PTC is generally good, but some patients still experience local or distant metastasis. Chowdhury et al. have shown that immune system disorder, which silences immunological monitoring, is an important factor that enhances the invasion of tumor cells and causes metastasis. PD-L1 has been identified as an integral molecule of the tumor microenvironment and is closely associated with the development of cancer. Expression of PD-L1 on the tumor cell membrane inhibits the cytotoxic T-cell immune response and leads to immune tolerance of tumor cells (20). In previous studies of PD-L1 expression by immunohistochemistry, the frequency of PD-L1 positivity was different, ranging from 6.1-82.5% in patients with PTC (21). In our study, PD-L1 positivity was observed in tumor tissue of PTC, but not in normal tissue, and was related to multifocality of tumor, which is consistent with previous findings (22, 23). This confirmed the adverse effects of PD-L1 positivity and leading to it being considered a risk factor for long-term recurrence (24).
Previous research on the effect of PD-L1 in PTC mainly focused on clinical and histopathological manifestations (25-27), while the relationship between PD-L1 and ultrasound and study on the joint effect of the two on thyroid cancer prognosis were seldom reported. Our work examined whether PD-L1 was associated with ultrasonic characteristics.
Ultrasonic examination is convenient and is the preferred examination for diagnosing thyroid disease and has high diagnostic value (28). In this study, the routine ultrasonic characteristics of thyroid cancer were fully summarized and analyzed; PD-L1-positive thyroid nodules often showed a halo sign, microcalcification and capsular involvement on ultrasonography and patients had distant recurrence. In previous work, views on the value of the halo sign around lesions differed (29). In our work, 53.7% of the nodules showed a halo sign, of which 69.7% were PD-L1 positive, and the halo sign was incomplete in most cases. In addition, it was found that PD-L1 was mostly positive in nodules containing microcalcifications, which may be related to the microenvironment of the tumor. PD-L1 expressed on the surface of tumor cells binds to programmed cell death protein 1 (PD1), thereby inducing an immune response in tumor. The complex interaction between lesion tissues, peripheral T-cells, macrophages, dendritic cells, and other immune cells enriched in the tumor microenvironment dramatically affects biological characteristics and facilitates invasion by a tumor into surrounding and even distant tissues (30-32). Studies have even shown that the high expression of PD-L1 in lung and thyroid tumors was associated with CD68+ macrophages, and that osteopontin secreted by CD68+ macrophages can stimulate the formation of microcalcifications in PTC (33, 34). In the multivariate analysis of ultrasound signs for thyroid cancer recurrence, we showed that a halo sign and microcalcification were predictors for thyroid cancer recurrence, and their presence heightened the risk of recurrence.
In further analysis of the RFS of patients with PTC, it was verified that PD-L1 positivity of PTC cells had a negative impact on time to local/distant recurrence of PTC and was an independent negative prognostic marker. This was consistent with the results of Chowdhury et al. (7). In our study, after multivariate adjustment for clinicopathological factors, the effect of PD-L1-positive expression in tumor tissues, on recurrence was still notable, indicating that the negative effect of positive PD-L1 expression is an independent risk factor for long-term recurrence. Moreover, the halo sign in ultrasonography also had a negative impact on the recurrence of PTC, which indirectly reflects the role of PD-L1 expression in the occurrence and development of tumor biological characteristics.
In this work, the clinicopathological characteristics and ultrasonic signs of PD-L1 and PTC were analyzed systematically. Our findings may help to improve clinicians’ prediction of thyroid cancer prognosis, to allow early intervention and prevent recurrence.
Limitations
Our work investigated the expression of PD-L1 in PTC and analyzed its correlation with the clinical and ultrasonic characteristics of PTC. However, all study patients were from the same large hospital in the region. Secondly, this study analyzed the relationship between PD-L1 expression and patient prognosis but did not include evaluation of endocrine factors. Thirdly, the sample size of the study was relatively small, so that it only reveals possible correlation, and only represents one race. In the next step, we intend to expand the sample source to cross regional joint hospitals at the same level to eliminate the bias between different regions.
Acknowledgements
We thank the two ultrasound doctors WH. H. and XL. C. for the ultrasound image assessment in our study.
Footnotes
↵* Senior Authors.
Authors’ Contributions
R. L. wrote the article. Q. L. designed the study. C. J. and C. P. performed the histopathology reporting of all the patients’ PTC tissues. M. L., B. S. and H. L. contributed with analytical tools. R. L., S. Z. and M. Q. did the statistical analysis. The article was drafted by R. L. and submitted to all the Authors for comments. All Authors approved the final version of the article.
Funding
This research was supported by the Capital Medical University Research and Cultivation Fund (No. PYZ22123).
Conflicts of Interest
The Authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
- Received June 20, 2023.
- Revision received July 29, 2023.
- Accepted August 7, 2023.
- Copyright © 2023, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.