Prevalence and outcome of metastases to the thyroid gland: a nationwide database study Authors: Monique S Dombrée, Yvette Sloot, Niek Hugen, Quirinus J Voorham, Ilse van Engen-van Grunsven, Iris D Nagtegaal, Romana T Netea-Maier Published: 2026 ABSTRACT ObjectiveMetastases to the thyroid are seen relatively often in autopsy studies. In clinical studies, they are described as a rare occurrence. This study aimed to provide a more extensive representation of the origin, timing, and survival of patients diagnosed with thyroid metastases in the Netherlands.MethodsWe performed a retrospective database study using two Dutch nationwide resources. In the Dutch Nationwide Pathology Databank (PALGA) database, we identified 575 patients between 1989 and 2014 with proven thyroid metastases. These patient records were linked with data from the Netherlands Cancer Registry (NCR) to assess whether the presence of these metastases had any impact on patients’ survival. Complete survival data were available for 463 patients.ResultsThe most frequently diagnosed primary tumors were lung (29.4%), kidney (16.5%), and breast carcinoma (10.1%). However, thyroid metastases can originate from any primary tumor. The median time to diagnosis of thyroid metastasis was 17.2 months after the primary diagnosis. In 36 patients, the latency period was over 10 years. In 77% of patients, thyroid metastasis was part of pathologically proven widespread disease. Survival after diagnosis of thyroid metastasis was poor, with a median survival of 12.3 months.ConclusionClinically relevant metastases to the thyroid are relatively rare and occur in a large variety of tumor types. Most of them occur in the context of advanced metastatic disease, and therefore, prognosis and indication for surgical treatment depend on the extent of the primary tumor, general health of the patient, and presence of local symptoms. Introduction Diagnostic evaluation of thyroid nodules is common in daily practice. However, despite the large number of fine needle aspirations for cytological examination (FNAC) routinely performed for thyroid nodules, metastases to the thyroid gland are an uncommon finding. One study retrospectively reviewed almost 20,000 thyroid aspirates and identified only 24 patients with metastases over a period of 20 years (1). Another study by Battistella et al. only identified nine patients with thyroid metastases among 5,000 thyroidectomies performed between 2007 and 2018 (2). However, a much higher prevalence of metastatic lesions in the thyroid has been reported in autopsy studies, ranging between 1.9 and 26.4% of cases (3, 4). In recent literature, thyroid metastases have been described more frequently in clinical studies (5, 6, 7, 8). However, robust data on the prevalence and clinical relevance of metastatic lesions in the thyroid remain limited. Most previous literature consists of case reports and small case series from specific primary tumor sites or the studies are limited to one or a few medical centers. In 2017, Nixon et al. performed a literature review of all previously reported metastases to the thyroid (4). They showed that these metastases can originate from almost every primary tumor location. In line with previous reviews, they demonstrated that, in autopsy studies, lung carcinoma is the most prevalent primary tumor, followed by breast carcinoma. In clinical series, renal cell carcinoma (RCC) is the most frequent primary tumor. This is likely explained by the relatively indolent course of renal cell carcinoma, making further investigation and treatment of new metastases during the clinical course of the disease more relevant. Because of the increasing treatment options, more aggressive treatment of oligometastases is also becoming more relevant in other primary tumors. While thyroid metastases most often occur in the context of extensive metastatic disease, a different management strategy may be appropriate for oligometastatic disease. This study aimed to provide insights into the origin and timing of thyroid metastases and the survival of patients diagnosed with these metastases in the Netherlands using nationwide databases that provide autopsy, cytology, histology, and survival data. Methods Database This was a retrospective database study using two Dutch nationwide databases. First, we queried the Dutch nationwide pathology databank (PALGA), which contains all cytology, histology, and autopsy reports from the 46 pathology laboratories in the Netherlands (9). PALGA was founded in 1971, was technically connected to all hospitals nationwide in 1989, and has had nationwide coverage of data collection since 1991. We then extracted all records containing the words ‘thyroid’ and ‘metastasis’ between 1989 and 2014. In total, 916 records were extracted. Patients with incorrect, incomplete, or equivocal data were manually excluded. Reasons for exclusion were recorded. The remaining 575 patient records were included in the primary database (Fig. 1). This database was used for the descriptive analysis. The patient records were subsequently linked with data from the Netherlands Cancer Registry (NCR). This registry is maintained by the Netherlands Comprehensive Cancer Organization (IKNL). Since 1989, it has collected information from hospital records about primary tumor type, TNM classification (using the applicable criteria at the time of diagnosis), and primary treatment. Recurrence, progression, follow-up treatment, and cause of death were not registered. Follow-up of vital status is obtained through the NCR database and through linkage of the PALGA database with the Municipal Personal Records Database, which is updated annually and was updated at the time of data analysis until December 31, 2024. Pseudonymized selected records from the PALGA database were linked to records in the NCR database. The data were then manually incorporated into the combined final database. In 112 records, it was not possible to include survival data for the reasons outlined in Fig. 1. The remaining 463 records were included in the survival analysis. Case selection. Data revision Both databases represent diagnostic assessments performed in routine practice and contain real-world data. Discrepancies between the different databases and equivocal results were evaluated on a case-by-case basis. Nine patients were excluded from the follow-up data because the data were conflicting. When age at diagnosis differed between databases, the age reported in the NCR database was used because this database contains complete patient records. However, when there was a discrepancy in the registration of the primary tumor diagnosed on the same date, pathology records in the PALGA database were used as the primary source. In nine cases, PALGA did not register the primary tumor, whereas the NCR database did. If the primary site registered in the NCR database matched the pathology, this diagnosis was used in our database. Survival analysis The time between the pathological diagnosis of the primary tumor and the pathological diagnosis of thyroid metastasis was recorded in months. For the survival analysis, the NCR database was used as the primary source; if no data were available in the NCR database, available data from the Municipal Personal Records Database were used. Overall survival (OS) data were calculated as the number of days between the registered date of diagnosis and the date of death or until last follow-up (December 31, 2024). These values were converted to months. For the survival analysis, we excluded cases in which thyroid metastasis was diagnosed at autopsy. All statistical analyses were conducted using SPSS for Windows version 29.0.1 (SPSS Inc., USA). Categorical variables were reported as numbers and percentages. Continuous variables were reported as median and range. OS data were generated using the Kaplan–Meier method. A P-value ≤ 0.05 was considered to indicate statistical significance. No ethical approval was required for this study because it used available fully anonymized databases. Results Primary database The primary database included 575 patients, of whom 271 (47.1%) were males and 304 (52.9%) were females. Thyroid metastasis was diagnosed primarily by cytology in 253 (44%) patients and by histological biopsy in 51 (8.9%) patients. In 25 (4.3%) patients, histological biopsy was required after cytology to confirm the diagnosis. An additional 88 (15.3%) patients were diagnosed after tumor excision and 154 (26.8%) during autopsy. In four patients, it was unclear whether cytology or needle biopsy was used. In patients with lung cancer, thyroid metastases were diagnosed at autopsy in 42.6% of cases. In contrast, in patients with RCC, diagnosis of metastasis was often made after thyroid resection (41.1%). Figure 2 shows the different primary tumor sites of the patients included in the study. Lung carcinoma was the most frequently diagnosed primary tumor. Other frequent primary sites included the kidney, breast, colon, head and neck, and upper gastrointestinal tract. Patients included in the ‘other’ category had a wide variety of primary tumor locations, including prostate, gallbladder, ovary, pancreas, salivary gland, testis, bladder, adrenal, thymus, non-Hodgkin lymphoma, synoviocarcinomas, and sarcomas. In 40 (7.0%) patients, histology and IKNL data could not identify the primary tumor (cancer of unknown primary). For these patients, histopathology showed adenocarcinoma in 37.5% and squamous cell carcinoma in 22.5%. Pathology reports showed both a metastasis from a distant organ and a primary thyroid carcinoma in seven patients and concomitant benign thyroid disease in 63 patients. This included thyroiditis, benign solitary nodules, and multinodular goiter. Proportion of patients with metastases to the thyroid according to primary tumor site and type of thyroid surgery. In 176 (30.6%) patients, thyroid metastasis was detected synchronously with the primary tumor. In 288 patients (50.1%), it was diagnosed metachronously, and in 111 (19.3%) patients, the interval between primary tumor and thyroid metastases could not be confirmed because the specific date of diagnosis of the primary tumor was unavailable. In 30 patients, thyroid metastasis was diagnosed before identification of the primary tumor. Overall, 453 patients (78.8%) also had other lymph node and/or solid organ metastases proven by histopathology or cytopathology during follow-up. In 235 patients (40.9%), these metastases were found synchronously with thyroid metastases. However, there was also a sizable group of patients (180 patients, 31.3%) in whom the timing between thyroid metastases and other metastases could not be determined from the available data. The median time until diagnosis of thyroid metastasis was 17.2 months after the diagnosis of primary tumor. This ranged from 8 months prior to the pathological diagnosis of the primary tumor in a patient with lung adenocarcinoma to 26 years after diagnosis of the primary tumor in a patient with RCC. Overall, the latency period was shortest for patients with lung carcinoma and longest for patients with RCC. In 36 patients, the latency period was over 10 years after diagnosis of the primary tumor, and the majority of these patients (n = 22) had RCC. Figure 3 (and Supplementary Table 2) shows the time until diagnosis of thyroid metastasis for each primary tumor site. Time until diagnosis of thyroid metastasis for each primary tumor site. In the total cohort, 150 (26.1%) patients underwent some form of thyroid surgery. Of these, 27 patients (18%) underwent total thyroidectomy, 81 patients (54%) underwent lobectomy, 2 patients (1.3%) underwent isthmus resection, and in 40 patients (26.7%), the extent of surgery was unclear from the pathology report. Even among patients without pathologically proven metastases at other sites, only 32% underwent surgery. Figure 2 and Supplementary Table 2 (see the section on Supplementary materials at the end of the article) show the proportion of patients who underwent surgery for each primary tumor location. Survival analysis Complete and matching follow-up data were available for 463 patients. In 121 of these patients, thyroid metastasis was diagnosed at autopsy, and these patients were excluded from the survival analysis. Only 19 (5.6%) patients were still alive at the time of linkage with the Municipal Personal Records Database. The median survival for all patients was 37.3 months (IQR: 84.1). In 255 patients, the time between diagnosis of thyroid metastasis and the end of follow-up could be determined. In these patients, the median survival after diagnosis of thyroid metastasis was only 13.6 months (IQR: 33.8). In 74 patients without other pathologically proven metastases, the median survival after diagnosis of thyroid metastasis was 11.9 months (IQR: 19.3). Table 1 and Supplementary Table 2 show the survival data for each primary tumor site from the diagnosis of thyroid metastasis. Table 1Median survival after diagnosis of thyroid metastasis.Location primaryMedian months of survival from diagnosis of thyroid metastasis (IQR)Lung9.8 (14.5)Kidney48.9 (67.1)Breast15.0 (29.4)Colon16.2 (35.1)Head and neck5.01 (14.6)Upper GI7.9 (13.5)Skin5.8 (52.2)Not otherwise specified * Other140.6 (134.3)*Out of 40 cases in this category, 38 did not have a date of diagnosis; therefore, survival time could not be calculated. Median survival after diagnosis of thyroid metastasis. Out of 40 cases in this category, 38 did not have a date of diagnosis; therefore, survival time could not be calculated. There was a significant difference (P =< 0.001) between patients who did and did not undergo thyroid surgery (Fig. 4) (and Supplementary Table 2). Median survival after diagnosis of thyroid metastasis was 8.8 months in patients who did not undergo surgery (IQR: 15.2) and 34.5 months in those who underwent surgery (IQR: 71.2). There were no significant differences between the different types of surgery (lobectomy, total thyroidectomy, and other procedures) with respect to OS. Among the different primary tumors, patients with RCC most frequently underwent thyroid surgery for removal of metastases (Fig. 2). Within this subgroup, patients who underwent thyroid surgery also had significantly better survival compared with those who did not undergo thyroid surgery (P = 0.002), with a median survival of 66.3 months (IQR: 69.7) in those who underwent surgery and 13.6 months (IQR: 25.2) in those who did not. Even after excluding the patients with RCC from the analysis, survival remained significantly better in patients who underwent thyroid surgery (P = 0.002), although the differences were less pronounced. In patients with breast carcinoma, thyroid surgery was also associated with better survival (P =< 0.001). Median survival in patients without surgery was 11.2 months (12.3) and in patients with surgery was 51.4 months (IQR: 65.8). In patients with colon carcinoma (P = 0.67), surgery was not associated with significantly better survival. A multivariate survival analysis using Cox Regression with primary tumor location, synchronous vs metachronous metastases, and type of operation showed similar results, with better survival associated with asynchronous metastasis, total thyroidectomy or hemithyroidectomy, and a primary tumor originating from the kidney (Supplementary Table 3). Survival with and without thyroid surgery. Discussion Metastases to the thyroid are uncommon, and knowledge about their origin and outcomes remain limited. The present study assessed the prevalence and characteristics of these metastases and the clinical outcomes of patients in whom they occur. This is the first comprehensive study based on pathology and cancer registries with nationwide coverage, over a period of 25 years addressing this issue. Our data support that pathologically confirmed metastases to the thyroid are an uncommon finding and can occur in many types of malignancies. The vast majority of these metastases are diagnosed in the context of advanced disease and are associated with poor prognosis. Although only a quarter of patients in this retrospective series underwent thyroid surgery to remove the metastasis, this was associated with prolonged survival, which likely reflects the more indolent course of the primary tumors in these patients. The present study is based on records of pathological examinations. Among all PALGA nationwide pathology results between 1989 and 2014, only 575 records matched a definitive diagnosis of thyroid metastasis. This is likely an underestimation of the total number of cases with thyroid metastases because many thyroid metastases are overlooked in the absence of symptoms (10, 11). Moreover, in cases with more advanced metastatic disease, pathological examination of a thyroid nodule might not be considered relevant, even if symptoms occur. Finally, only a minority of patients undergo autopsy in the Netherlands. Our results show that lung, kidney, breast, colon, and head and neck tumors are the most common primary tumors, which is consistent with earlier studies (2, 4, 5, 6, 8, 11, 12). A history of one of these tumors should prompt clinicians to consider the possibility of thyroid metastasis when a patient presents with a thyroid nodule. In half of the cases in this study, thyroid metastases were diagnosed more than 6 months after diagnosis of the primary tumor. This was particularly observed for specific indolent primary tumors, such as RCC, breast cancer, and skin cancer, which is in line with previous reports (2, 4, 5, 6, 11, 12). Clinicians should be aware of the possibility of metastatic lesions when evaluating thyroid nodules in patients with a history of such malignancies, even years after diagnosis of the primary tumor. However, as demonstrated by the diversity of primary tumors identified in this study, a previous history of any tumor could indicate thyroid metastasis in a patient with a new thyroid nodule. Although rare, metastases to the thyroid should also be considered in the differential diagnosis of a thyroid nodule in patients without a previous malignancy. In 30 patients in our database, the pathological diagnosis of thyroid metastasis was made before the pathological diagnosis of the primary tumor. A thyroid nodule could therefore represent the first manifestation of malignancy from different types of primary tumors (6, 11, 12). One previous study showed that 13 of 36 RCC patients presented with a thyroid nodule before the primary tumor was identified (8). Previous literature has suggested that thyroid metastases occur more frequently in patients with underlying thyroid pathology, such as thyroiditis, benign solitary nodules, or multinodular goiter (4, 11, 12). Willis hypothesized that pre-existing pathological changes in the thyroid may favor the arrest of tumor cells and provide a favorable environment for tumor growth; however, this has not been robustly investigated (13). A study of 36 patients showed that 42% of patients with thyroid metastases had previous thyroid disease (8). In our cohort, concomitant thyroid disease was reported in only 12.2% of patients. Potential explanations for the lower percentage of thyroid disease in our cohort include that it was either not reported because it was not considered relevant at the time, was not recognizable in the cytology or biopsy, or was not reported in the report available for review from the PALGA database. The diagnostic workup of thyroid nodules is not different when a thyroid metastasis is suspected. However, in the presence of other metastatic lesions, clinicians must consider whether additional investigations are relevant for further treatment and prognosis. Imaging modalities increase the detection of thyroid nodules; however, they cannot definitively differentiate between benign nodules, primary thyroid carcinoma, and thyroid metastases (14). FNAC can, in the majority of cases, provide an accurate diagnosis of thyroid metastasis (1, 9, 11, 15). In our cohort, after excluding autopsy findings, thyroid metastasis was diagnosed primarily by cytology in 60% of cases. When cytology is inconclusive, histological biopsy or surgical resection can be considered (16). The increasing use of molecular tumor markers may improve the accuracy of diagnoses made using cytology in patients with an unknown primary tumor and in cases with equivocal morphological features (17). Survival after diagnosis of thyroid metastasis is poor. This is likely attributable to thyroid metastasis most often being a marker of widespread metastatic disease with an associated high mortality rate. We excluded patients with a diagnosis of thyroid metastasis at autopsy because these cases could have skewed the survival outcomes. In our series, surgical removal of the thyroid metastases was associated with improved survival. However, because of retrospective nature of the study, a causal relationship cannot be assessed because several confounding factors, such as performance status, clinical extent and progression, and radiological findings, were not reported in the databases used in this study. Several previous small studies have reported the effect of thyroidectomy on survival. Russel et al. summarized all reported cases from the English literature between 1994 and 2013 and added their own 17-patient series to review 818 cases, of which the surgical treatment status could be determined for 580 patients (12). Of these cases, 33.5% did not undergo surgery, 34.2% underwent subtotal thyroidectomy, and 34% underwent total thyroidectomy. They showed improved OS from diagnosis of the primary tumor in patients who underwent thyroid metastasis resection in RCC, gastrointestinal cancer, and lung cancer. The effect of thyroid surgery on survival could not be assessed in patients with breast cancer because most cases in their database underwent surgery. Consistent with our results, the extent of surgery had no significant impact on survival. However, a significant probability of bias should be considered, both in this retrospective series and in our study, because clinicians likely selected surgical removal of thyroid metastases for patients with a more indolent course of the primary tumor. Moreover, published case reports may overrepresent cases in which physicians opted for surgical treatment. This is less likely in our series, which may also explain the lower percentage of patients undergoing surgery in our cohort (26.3%). In patients with only a proven thyroid metastasis, survival was better, although they had only a slightly higher percentage of thyroid surgeries. Most likely, the longer median survival in these patients is related to their lower disease burden. Primary tumors with a relatively indolent course, such as RCC and ductal breast carcinoma, represent specific entities in which thyroid surgery may provide additional survival benefit. Both our results and the literature review by Russel et al. demonstrate an increased survival after thyroid surgery in these groups. However, Iesalnieks et al. found in a retrospective analysis of 45 patients with RCC and thyroid metastasis that survival was mainly related to overall health status rather than the thyroid surgery (18). Therefore, the longer survival observed in these patients is largely determined by the slow progression of the disease rather than by removal of the thyroid metastasis. In the present study, the databases used do not provide information on the clinical considerations for or against surgery in the included patients. Based on the expected biological behavior of these tumors, it seems likely that the apparent survival benefit observed in patients who underwent removal of thyroid metastasis reflects the indolent course of the primary tumor rather than the effect of tumor removal. Therefore, the indication for surgical management should be evaluated on a case-by-case basis, preferably within a multidisciplinary team, considering the course and prognosis of the primary tumor, the general condition of the patient, the possibility of curative resection, local complaints, and potential complications of surgery. This study represents the largest nationwide cohort of patients with thyroid metastases to date. However, the use of registries for data collection has inherent limitations. First, there is a possibility of misclassification because these data represent real-world clinical practice. Misclassification could occur both in the primary databases used and during transfer of these data to our database. This is illustrated by several cases in our database in which characteristics did not match between the two databases. To reduce this risk, patients with insufficient, discrepant, or equivocal reports were excluded from the analysis. Second, because keywords were used to select relevant cases, there could be errors in the automatic selection of cases from the different databases. For this reason, all records were manually reviewed for accuracy. We cannot exclude the possibility that a small number of patients who did not match the selected keywords were missed. The significant proportion of patients with missing data represents another limitation. Although this is inherent to any database research, missing data could skew both patient and tumor characteristics and analyses of progression-free and overall survival. Finally, the databases used in the present study do not provide information on the clinical context of the patients. Therefore, we cannot assess the clinical reasoning behind the diagnosis and treatment of metastases, nor can we provide information on patients who may not have undergone histological or cytological examination. Conclusion Clinically relevant metastases to the thyroid are rare and can occur in a wide variety of tumor types. The spectrum and course of different malignant tumors have changed significantly in recent years, with more patients surviving longer and undergoing more imaging during follow-up. Therefore, thyroid metastases may be encountered more frequently in the future. Most thyroid metastases occur in the context of advanced metastatic disease, and their removal is largely performed in a palliative setting. Although removal of thyroid metastases appears to be associated with improved overall survival, a causal relationship has not been established. Therefore, prognosis and the indication for surgery depend primarily on the extent of the primary tumor, the general health of the patient, and the presence or local symptoms. Supplementary materials Declaration of interest The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported. Funding This research did not receive any specific grant from any funding agency in the public, commercial, or not-for-profit sector. Author contribution statement RNM, IDN, and ACHEG conceived the study. RNM and YS contributed to the formation of the study design. QV assembled the PALGA database, and AE and IDN provided the data and insights for the IKNL database. YS and MD created the merged PALGA and IKNL database. MD analyzed the data and led the writing of the manuscript under the supervision of RNM. All authors provided critical feedback and contributed to shape the research, analysis, and the final manuscript.