Familial Aggregation and Heritability of Nonmedullary Thyroid Cancer in an Asian Population: A Nationwide Cohort Study

Abstract

Purpose

The purpose of this work is to assess the extent of familial aggregation of nonmedullary thyroid cancer (NMTC) and the relative risks (RRs) of chronic thyroid diseases and common malignancies in first-degree relatives of NMTC patients.

Methods

In the National Health Insurance Research database of Taiwan, all eligible individuals in 2016 were analyzed (n = 23 696 659) and the family structures of 38 686 patients diagnosed with NMTC between 1997 and 2016 were identified. The prevalence and RRs of NMTC, chronic thyroid diseases, and common malignancies in individuals with first-degree relatives with NMTC were examined. The accountability of heritability and environmental factors to NMTC susceptibility was estimated using the polygenic liability model.

Results

The prevalence of NMTC was 0.16% in the general population and 0.64% in individuals with first-degree relatives with NMTC. Regarding affected relatives, the RR (95% CI) for NMTC was 20.12 (4.86-83.29) for twins, 6.43 (4.80-8.62) for siblings, 5.24 (4.55-6.03) for offspring, 5.07 (4.41-5.81) for parents, and 2.07 (1.53-2.81) for spouses. The estimated genetic, common environmental, and nonshared environmental contributions to NMTC were 28.0%, 14.3%, and 57.7%, respectively. A family history of NMTC was associated with higher risks of thyroid nodules (RR, 2.26; 95% CI, 2.18-2.35), Hashimoto thyroiditis (2.11; 1.89-2.36), Graves disease (1.49; 1.42-1.57), lung cancer (1.56; 1.32-1.85), and leukemia and lymphoma (1.24; 1.03-1.50).

Conclusion

Our findings demonstrate the importance of genetic and environmental contributions to NMTC susceptibility and highlight the coaggregation of chronic thyroid diseases and multiple malignancies with NMTC.

Thyroid cancer, the most common neoplasm of the endocrine system, accounted for 3.1% of all new cancer cases worldwide in 2018 and its incidence ranks in ninth place among all cancer sites (1). The latest global cancer statistics (GLOBOCAN 2018) reported that the global age-standardized incidence of thyroid cancer was 6.7 per 100 000 (male, 3.1; female, 10.2), accompanied by a very low mortality rate (worldwide age-standardized rate of 0.4) (1). In Taiwan, the incidence of thyroid cancer has surged rapidly in recent years (standardized incidence rate 12.0 per 100 000 in 2016; male, 5.9; female, 18.1) and ranked fourth among all new cancer sites in women in 2016 (2). This increase might be partially explained by the overdiagnosis of a reservoir of indolent tumors, but clinicopathological analyses suggest there are carcinogenetic factors behind this noticeable worldwide epidemic of thyroid cancer (3). In addition to a family history of thyroid cancer, suspicious carcinogens for thyroid cancer include ionizing radiation exposure, dietary habits, obesity, hormonal exposures, and environmental pollutants; however, ethnic and geographic variations exist (1).

Medullary thyroid carcinoma with parafollicular C-cell origin accounts for 1% to 2% of all thyroid cancer, and it has known high heritability due to autosomal dominant inheritance of MEN2. On the contrary, nonmedullary thyroid cancer (NMTC) originating from follicular cell accounts for more than 90% of all thyroid cancer, but its extent of genetic predisposition is ambiguous. To complicate the analysis of NMTC susceptibility, around 3% to 10% of NMTC cases are classified as familial NMTC (FNMTC), which involves the presence of NMTC in 2 or more first-degree family relatives and is proposed to pass in an autosomal dominant pattern with variable expressivity or in a polygenic pattern (4). Family and twin studies of European populations suggested that thyroid cancer is one of the most inheritable cancers displaying nontypical Mendelian inheritance, with reported risks 5- to 10-fold higher for first-degree relatives of thyroid cancer patients compared to the general population (5, 6). Calculation of the data of a quantitative Nordic twin cohort database revealed that genetic factors accounted for 71% and 67% of the individual differences in thyroid volume and thyroid nodularity, respectively (7, 8). Regarding thyroid cancer, a family cohort of the Norwegian Cancer Registry database between 1960 and 1995 estimated that the familial risk of NMTC was higher in affected first-degree relatives with a standardized incidence ratio of 5.2 for men and 4.9 for women (9). Another Swedish Family-Cancer Database cohort suggested that thyroid cancer had the highest inheritance rate (53%) among 15 cancer sites, with common and unique environmental contributions of 11% and 36%, respectively (10). The latest large twin cohort estimating familial cancer risk in Nordic countries between 1943 and 2010 found the overall heritability of cancer was 33% (95% CI, 30-37%), but the familial risk of thyroid cancer was not evaluated because of the limited number of cases (11). Additionally, no investigation on the heritability of NMTC in Asian populations has been published in the literature to date.

Although many susceptibility loci of NMTC have been found through genome-wide association studies (12, 13), the overall heritable risk conferred by these pathogenic variants is difficult to determine with population-level genotyping. Therefore, the familial inheritance and accountability of environmental factors to NMTC susceptibility in the current epidemic of thyroid cancer remains largely unknown. In addition, family members of NMTC cases are at higher risk of chronic thyroid diseases such as thyroid goiter and autoimmune thyroiditis (14), and they might have a higher risk of other common malignancies contributed by these shared pathogenic mutations or similar carcinogens (15, 16). Therefore, we conducted this nationwide population-based cohort study based on the National Health Insurance (NHI) database of Taiwan in 2016 to examine the familial aggregation of NMTC and assess the accountability of inheritance and environmental factors to NMTC susceptibility. We also investigated the coaggregation of NMTC with chronic thyroid diseases and common malignancies in first-degree relatives of NMTC patients.

Materials and Methods

This population-based cohort study was constructed on the NHI research database of Taiwan. The NHI program in Taiwan, founded in 1995, is a single-payer government-run system. The enrollment of every resident in Taiwan is required by law to provide comprehensive health care, and the NHI program covered more than 99.6% of the residents of Taiwan by the end of 2015 (17). The NHI research database, provided by the Administration of National Health Insurance, Ministry of Health and Welfare, contains the comprehensive medical claims of every beneficiary, and all released data are encrypted to protect patient privacy. In addition, conducting database research and data extraction are restricted to the specialized data science center operated by the administration. All recorded disease diagnosis data during our study period (1997-2016) were encoded with the International Classification of Diseases, Ninth Revision (ICD-9) codes for data coded in 1997 to 2015, or Tenth Revision (ICD-10) for data coded in 2016.

The current study was reviewed and permitted by the institutional review board of Chang Gung Memorial Hospital (approval number 201900677B0). Because all data used in this study were anonymized and untraceable, informed consent was not required.

Study population

All individuals (n = 23 696 659) with a valid insurance status under the NHI in 2016 were included as the target population. In Taiwan, patients with a cancer diagnosis are entitled to a medical copayment waiver (ie, registry of catastrophic illness); expert panels commissioned by the NHI administration critically review their clinical information and confirm their diagnosis before issuing the benefit. We used the registry of catastrophic illness, a subdata set of the NHI research database, as the primary data source to identify patients with a diagnosis of thyroid cancer (ICD-9 code 193.x and ICD-10 code C73.*) during 1997 to 2016; the Taiwan Cancer registry was used to identify thyroid cancer subtypes to derive the NMTC cases for this study. The registry of catastrophic illness in the NHI database and Taiwan Cancer registry is highly correlated, and its clinical consistency was validated in previous studies (18).

Identification of family relatives of nonmedullary thyroid cancer patients

The birth registry of each beneficiary in the NHI database comprised information on the family kinships between the insured individual and his or her first-degree relatives (parent, offspring, sibling) and spouse. Using a unique personal encrypted identification number for each beneficiary in Taiwan for reliable linkages among different subsets of the NHI database, we established family genealogy for every individual in this family study (19). Among these family relationships, parent, offspring, and spousal relationships were identified directly from the NHI database, and full siblings were verified as individuals with the same parents. Full siblings with the same date of birth (±1 day) were identified as twins, but their zygosity could not be distinguished in this database. Finally, the family genealogy was reconstructed after determination of the family relationships of every individual within the family (19). During the statistical analysis, twins were excluded from the sibling analysis and all half-siblings were excluded from the whole analysis.

Ascertainment of thyroid diseases and common malignancies

Previous epidemiologic studies suggested that NMTC patients are at higher association with thyroid diseases and subsequent malignancies, so we intended to estimate the potential coaggregation of NMTC with chronic thyroid diseases and common malignancies in the first-degree relatives of NMTC patients. Each individual was identified as having a specific disease in the NHI database either by registry in the catastrophic illness or by a disease diagnosis made by specialists in at least 1 hospitalization or 2 outpatient visits with those specific diagnostic codes. The analyzed diseases with accompanying ICD codes used in our study were thyroid nodule (ICD-9: 240.x–242.x; ICD-10: E01.2, E04.0–E05.2), Hashimoto thyroiditis (245.2; E06.3), Graves disease (242.x; E05.0), breast cancer (174.x, 175.x; C50.*), colorectal cancer (153.x, 154.x; C18.*-C21.*), liver cancer (155.x; C22.*), and leukemia and lymphoma (200.x–208.x; C81.*-C96.*). These malignancies were chosen because they have relatively high incidences in Taiwan.

Statistical analysis

The prevalence of NMTC for the general population was derived from the number of individuals who were alive in 2016 and were diagnosed with NMTC between 1997 and 2016 then divided by the number of the midyear Taiwanese population in 2016. We used relative risk (RR) and 95% CI for RR to represent the prevalence ratio between the affected family relatives and the general population (20). The prevalence ratio was adjusted for age, sex, place of residence, income level, occupation, and family size to estimate the relative contribution of heritability and familial transmission.

To determine the prevalence ratio in this cross-sectional study, the Breslow-Cox proportional hazard model was used by applying the same follow-up time to all individuals (21). While estimating the relative contributions of heritability and environmental factors to NMTC susceptibility, we applied the polygenic threshold liability model in this study (22). Heritability was defined as the accountability of genetic proportions to phenotypic variance, and familial transmission was the sum of the heritability and shared environmental proportions. Shared environmental factors are the same parental factors or common exposure through childhood to adult life, such as household environment or passive smoking (11). Contributing environmental factors other than shared factors are defined as nonshared environmental factors, such as occupational exposure or viral infection (11). In addition, we assumed a mean of 2 siblings per family and restricted the family history within first-degree relatives to simplify our analysis. Based on the polygenic liability model, spouses are assumed to share the family environment without genetic similarity with other families, and siblings are assumed to have both common genetic and shared environmental factors (19). Therefore, the sibling RR is compared with the population prevalence to calculate familial transmission, whereas the spouse RR is used as a control to derive heritability. Next, under the assumption of polygenic inheritance in NMTC, we applied the polygenic liability model to calculate the probability of a patient with NMTC without a family history (ie, the sporadic proportion of NMTC) (22).

Finally, we further estimated the coaggregation of NMTC with chronic thyroid diseases and other common malignancies in the first-degree relatives of NMTC patients using a Cox proportional hazard regression model by applying the same follow-up time for all participants. The adjusted RR of coaggregation of NMTC with other diseases was estimated using the adjusted prevalence risk ratio of a specific disease between the first-degree relatives of NMTC patients and the general population. Confounders such as age, sex, place of residence, occupation, income level, and family size were adjusted to estimate the familial aggregation of NMTC and coaggregation with other diseases. All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc), and statistical significance was defined as 2-sided P values less than .05.

Results

Nonmedullary thyroid cancer (NMTC) prevalence in general population vs individuals with NMTC-affected relatives

We identified 39 063 patients with a diagnosis of thyroid cancer among the 23 696 659 individuals in the general population of Taiwan in 2016. There were 38 686 patients with a diagnosis of NMTC, equivalent to a crude prevalence of 0.16%. Table 1 compares the demographic characteristics of the first-degree relatives of NMTC patients and the general population. Individuals with NMTC-affected first-degree relatives generally had a higher socioeconomic status (urban residency, income level, and professional occupation) than the general population. Table 2 showed the age-specific prevalence of NMTC in individuals with NMTC-affected first-degree relatives vs the general population of Taiwan. The NMTC prevalence was significantly higher in the first-degree relatives of NMTC patients than in the general population, and women had a significantly higher prevalence of NMTC than men. Among patients with an NMTC-affected first-degree relative, the prevalence of NMTC was 0.64%. Further analysis revealed that 73 754 (0.31%) individuals had at least one NMTC-affected first-degree relative in the general population of Taiwan in 2016: of them, 51 419 had an affected parent, 12 269 had an affected offspring, 10 202 had an affected sibling, and 105 had an affected twin.

Relative risks for nonmedullary thyroid cancer (NMTC) in individuals with NMTC-affected first-degree relatives

Table 3 shows the prevalence and adjusted RRs with 95% CIs for NMTC in individuals with affected first-degree relatives of NMTC by affected relative and sex. Overall, the affected first-degree relatives of NMTC patients had an adjusted RR of 5.27 (95% CI, 4.65-5.97) for NMTC vs the general population, and men had higher RRs (5.84; 4.78-7.13) than women (5.14; 4.49–5.89). The adjusted RR for NMTC among first-degree relatives of NMTC patients was the highest (20.12; 95% CI, 4.86-83.29) for twins (the highest genetic similarity), followed by 6.43 (95% CI, 4.80-8.62) for siblings, 5.24 (95% CI, 4.55-6.03) for offspring, 5.07 (95% CI, 4.41-5.81) for parents, and 2.07 (95% CI, 1.53-2.81) for spouses without genetic similarity. Based on the polygenic liability model, the estimated accountability for the phenotypic variance of NMTC in the Taiwanese population was 28.0%, 14.3%, and 57.7% for heritability, shared environmental factors, and nonshared environmental factors, respectively. Given these parameters including disease prevalence, occurrence risk of family relatives, and heritability on the liability scale, the estimated probability that a patient with NMTC would be sporadic was 83.7%.

Coaggregation of nonmedullary thyroid cancer with chronic thyroid diseases and other common malignancies

Tables 4 and 5 present the prevalence and adjusted RRs (95% CIs) for chronic thyroid diseases and other common malignancies, respectively, in individuals with NMTC-affected first-degree relatives vs the general population of Taiwan. Among all analyzed thyroid diseases with statistical significance, the adjusted RRs of individuals with NMTC-affected first-degree relatives was the highest for thyroid nodule with an RR (95% CI) of 2.26 (2.18-2.35), followed by 2.11 (1.89-2.36) for Hashimoto thyroiditis, and 1.49 (1.42-1.57) for Graves disease. Among analyzed other common malignancies with statistical significance, we found that the adjusted RR was the highest for lung cancer with an RR (95% CI) of 1.56 (1.32-1.85) in affected first-degree relatives of NMTC patients compared to the general population, followed by 1.24 (1.03-1.50) for leukemia and lymphoma, 1.17 (1.03-1.32) for colorectal cancer, and 1.15 (1.03-1.29) for breast cancer.

Discussion

In this nationwide, population-based cohort study, we investigated the risks of NMTC in individuals with NMTC-affected first-degree relatives and the relative contribution of heritability and environmental factors for NMTC in the Taiwanese population in 2016. The observed magnitude of RR was associated with genetic distance, and the overall adjusted RR of NMTC was 5.27-fold higher in the affected first-degree relatives of NMTC patients than that in the general population. The estimated accountabilities of genetic inheritance and environmental factors to NMTC susceptibility in the Taiwanese population were 28.0% and 72.0%, respectively. In addition, individuals with a family history of NMTC were at higher association with chronic thyroid diseases and other malignancies, particularly thyroid nodule, autoimmune thyroiditis, lung cancer, colorectal cancer, breast cancer, and leukemia and lymphoma, suggesting that there might be a common genetic association or carcinogenic roots. Considering the surging incidence of NMTC, 42.3% of familial transmission, and possible association with other diseases, clinical surveillance for NMTC should be reserved for NMTC-affected family members instead of general screening. Our results provide important information regarding the family aggregation of NMTC and coaggregation with other diseases, which is valuable for the counseling and management of patients with a family history of NMTC.

This is the first population-based nationwide cohort study to estimate the genetic and environmental contribution to NMTC susceptibility in an Asian population. After the exclusion of medullary subtype, our estimated genetic contribution was much lower compared to previous estimates on thyroid cancer susceptibility of European populations (10), and nonshared environmental factors had the highest contribution to NMTC susceptibility in our results. Our finding of an association of NMTC between spouses without genetic similarity (RR, 2.07; 95% CI, 1.53-2.81) might possibly originate from shared environmental contribution such as over-screening, radiation exposure, and iodine supply (23). In addition, a large Nordic twin study of the causation of cancer also suggested that environmental factors have a principal role in causing sporadic cancer, among which 58% to 82% of cancer susceptibility was assumed to be attributed to nonshared environmental factors (24). Our result implies that somatic mutation resulted from environmental factors contributes to a higher proportion of NMTC susceptibility compared to heritable germline mutation, but further genomic association studies are required to clarify its clinical implication. However, inheritance might still play a much more important role in NMTC susceptibility in comparison with other common malignancies. Our estimated familial risk of 5.27-fold for NMTC is much higher than familial risks of other common malignancies estimated by pooled analyses of case-controlled studies, such as breast cancer, colorectal cancer, and lung cancer with RR (95% CI) for affected first-degree relatives of 1.86 (1.69-2.06), 1.76 (1.57-1.97), and 1.51 (1.39-1.63), respectively (25-27). This genetic predisposition to NMTC might originate from identified genetic variants of FNMTC such as FOXE1, HABP2, SRRM2, and MYO1F, but the exact involved mutations require further examination (5, 12, 13, 28-32). Another possible explanation for familial aggregation of NMTC is that family members of NMTC patients might be anxious about familial inheritance and request additional thyroid evaluation. Additionally, individuals with NMTC-affected relatives often have higher socioeconomic status (as indicated in Table 1), so they might have better access to thyroid screening and higher detection of thyroid cancer. Further investigations of the causative factors of NMTC and elucidation of its underlying pathogenic mechanisms contributing to the familial aggregation of NMTC in Taiwanese population are required.

Despite apparent familial inheritance in the susceptibility of NMTC, our result estimated that 83.7% of NMTC patients in Taiwan are expected to be sporadic rather than familial under the assumption of polygenic inheritance. This finding is consistent with previous estimations that FNMTC comprised only 3% to 10% of all NMTC cases, and epidemiologic data also support this notion that most NMTC cases are sporadic (4). This phenomenon might be counterintuitive but appears to be the norm for many polygenic diseases such as cancer, psychiatric disorders, and rheumatoid diseases (22).

Aside from higher familial risk of NMTC in affected families, our results also found coaggregation of NMTC with chronic thyroid diseases and other common malignancies. Family members of thyroid cancer patients might ask for additional thyroid evaluation such as ultrasonography and thyroid function tests, which result in a higher incidence of thyroid cancer and thyroid diseases in this population. Among these diseases analyzed in our study, individuals with a family history of NMTC were at the highest risk of thyroid nodules (RR: 2.264). The overdiagnosis of small, indolent forms of papillary thyroid cancer during routine screening for thyroid nodules in NMTC-affected relatives using ultrasonography could potentially explain this coaggregation of thyroid nodules with NMTC (3).

Graves disease is the major cause of 50% to 80% of hyperthyroidism cases (33). Epidemiologic studies have suggested an increased risk of thyroid malignancy in patients with Graves disease, especially in those who require surgical management because of the increased detection of papillary microcarcinoma (34). A recent meta‐analysis found that the prevalence of thyroid cancer in 2582 surgically treated patients with Graves disease was 11.5%, in which the presence of thyroid nodules further increased the risk of detecting thyroid cancer (odds ratio, 5.3; 95% CI, 2.4-11.6) (35).

Hashimoto thyroiditis has long been found to coexist with thyroid cancer in surgical pathology (36). The proposed hypothesis is the pathognomic feature of lymphocytic infiltration in Hashimoto thyroiditis might contribute to early oncogenesis, thus potentiating the development of papillary thyroid carcinoma (37). A large nationwide cohort also observed that patients with Hashimoto thyroiditis had a higher risk (hazard ratio, 11.8) of thyroid cancer than the control group after the adjustment for sex, age, and comorbidities (38).

Our results also found an association between NMTC and other common malignancies, including lung cancer, colorectal cancer, breast cancer, and leukemia and lymphoma. Genetic association studies have identified that genetic alterations such as BRAF and RAS mutations and RET/PTC rearrangement underlying thyroid cancer carcinogenesis might contribute to other subsequent malignancies (39, 40); these mutations might also lead to familial clustering of these malignancies in NMTC-affected families. For example, BRAF mutations are often observed in non–small cell lung cancer and colorectal cancer; RAS mutations have been reported in colorectal cancer and lung cancer; and rearrangements of the RET proto-oncogene have strong correlations with breast cancer, lung cancer, and leukemia (40). These malignancies arising from different tissues might share common cross-cancer heritability or similar somatic mutation due to environmental carcinogen exposure. Smoking is less likely to be the carcinogen responsible for coaggregation of NMTC with these malignancies because current smokers have an inverse association with thyroid cancer risk in previous prospective cohort and case-control studies (23). However, because no comparable observation has been reported in the literature, the clinical interpretation of our findings and its generalizability require further validation.

The present family cohort study has several strengths and important implications. First, the large number of cancer cases and 2–decade-long follow-up in this population-based, nationwide cohort allowed us to provide more precise estimates of the familial risk of NMTC. The accurate linkage between the NHI database of Taiwan and the reliable cancer registration allowed for complete coverage and follow-up of the study population. Second, the analysis method of the threshold liability model was effective for examining familial aggregation and was validated in previous applications (19). We also adjusted for many possible confounders in the statistical analyses such as age, sex, socioeconomic status, region of residence, and family size. Third, the reconstructed familial genealogy could be confirmed for most Taiwanese populations by solid cross-linking with unique personal identification numbers. Fourth, we applied strict case ascertainment for NMTC and other diseases in that eligible cases required registry of catastrophic illness or specialist diagnosis with at least 1 hospitalization or 2 outpatient visits, thereby minimizing selection and recall bias. Taken together, our study was the first to provide quantitative data to estimate the accountability of NMTC susceptibility and coaggregation with other diseases, which is helpful for cancer-risk counseling of individuals in NMTC-affected families.

There are some limitations to the present study. First, our disease identification process was based on inclusion of the ICD-9 and ICD-10 diagnostic codes in the NHI database from 1997 to 2016. Detailed clinical information of twin zygosity, laboratory data, genotyping, lifestyle, dietary habits, smoking, and history of carcinogen exposure was not recorded in this national database. In addition, coding errors and misclassification might be present in this data set despite routine examination. However, the accuracy of the diagnostic codes in the NHI database and its clinical consistency with cancer registry have been validated (18). Second, our threshold liability model includes the basic assumption that spouses share the same percentage of environmental contributions as those shared by siblings, but individual variations exist. This model also assumes no interaction between genetic and environmental components, thus the possibility of an interaction and its accountability to NMTC need to be clarified elsewhere. Third, because twin zygosity was not available in the NHI database, it is impossible to validate our population-based estimation of familial aggregation using a classic twin-study design. Finally, our results are restricted to the population of Taiwan, and further investigations are required to clarify its clinical implication and validate its generalizability to other populations.

In conclusion, this population-based, nationwide cohort study was the first to estimate the genetic and environmental contribution to NMTC susceptibility in an Asian population. Our results showed that, compared to the general population, the first-degree relatives of NMTC patients had a 5.27-fold higher risk of NMTC and a higher risk of thyroid nodule, autoimmune thyroiditis, lung cancer, colorectal cancer, breast cancer, and leukemia and lymphoma. The estimated contributions to NMTC susceptibility in the Taiwanese population were 28.0% for genetic inheritance and 72.0% for environmental factors. Our results provide valuable information for the counseling of family members of NMTC patients in Taiwan.

Abbreviations

Abbreviations
 
• FNMTC

  familial nonmedullary thyroid cancer

 
• ICD

  International Classification of Diseases

 
• NHI

  National Health Insurance

 
• ICD

  International Classification of Diseases

 
• NHI

  National Health Insurance

 
• NMTC

  nonmedullary thyroid cancer

 
• RR

  relative risk

Acknowledgments

This study used data from the National Health Insurance database provided by the Administration of National Health Insurance, Ministry of Health and Welfare, and managed by the National Health Research Institutes. The interpretation and conclusions contained herein do not represent the views of the Administration of National Health Insurance or the National Health Research Institutes. We would like to thank the Center for Big Data Analytics and Statistics in Chang Gung Memorial Hospital for methodology and infrastructure support.

Financial support: This study was supported by Chang Gung Memorial Hospital (CMRPG3H0811-3, CMRPG3I0111-3, and CMRPG3H1191-3) and the Ministry of Science and Technology of Taiwan (MOST 108-2314-B-182A-058-MY2 and MOST 108-2314-B-182A-059-MY2).

Additional Information

Disclosure Summary: The authors have nothing to disclose.

Data Availability: The data used to support the findings of this study are available from the corresponding author on request.

References

Author notes