No consensus exists on the association between papillary thyroid carcinoma (PTC) and Hashimoto's thyroiditis (HT). To resolve this controversy, this study aimed to evaluate the relationship between the two conditions using a meta-analysis.
We searched relevant published studies using citation databases including PubMed, Embase, and ISI Web of Science. The effect sizes of clinicopathologic parameters were calculated by odds ratio (OR), weighted mean difference, or hazard ratio (HR). The effect sizes were combined using a random-effects model.
Thirty-eight eligible studies including 10 648 PTC cases were selected. Histologically proven HT was identified in 2471 (23.2%) PTCs. HT was more frequently observed in PTCs than in benign thyroid diseases and other carcinomas (OR=2.8 and 2.4; P<0.001). PTCs with coexisting HT were significantly related to female patients (OR=2.7; P<0.001), multifocal involvement (OR=1.5; P=0.010), no extrathyroidal extension (OR=1.3; P=0.002), and no lymph node metastasis (OR=1.3; P=0.041). Moreover, PTCs with HT were significantly associated with long recurrence-free survival (HR=0.6; P=0.001).
Our meta-analysis showed that PTC is significantly associated with pathologically confirmed HT. PTC patients with HT have favorable clinicopathologic characteristics compared with PTCs without HT. However, patients with HT need to be carefully monitored for the development of PTC.
Papillary thyroid carcinoma (PTC) is the most prevalent form of thyroid cancers, comprising about 80% of all diagnosed thyroid cancers. Hashimoto's thyroiditis (HT) – chronic lymphocytic thyroiditis or autoimmune thyroiditis – is a well-defined clinicopathologic entity and its incidence has increased over the past 50 years (1). HT is characterized by hypothyroidism, the presence of serum antithyroglobulin and antiperoxidase antibodies, and widespread lymphocytic infiltration with depletion of follicular cells. In addition to the classical HT, recent studies have proposed that IgG4-related thyroiditis may be considered as a variant of HT (2). The association between HT and PTC has been a subject of long and ongoing debate (1, 2).
HT has shown a wide range of occurrence from 5 to 85% in thyroid specimens resected for PTC (3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 33, 34, 35, 36, 37, 38, 39, 40). In addition, the clinicopathologic characteristics of PTCs with concomitant HT have not been definitely proposed (3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 33, 34, 35, 36, 37, 38, 39, 40). Therefore, the present meta-analysis was conducted to clarify the relationship between PTCs and histologically proven conventional HT and to investigate the clinicopathologic features of PTCs with coexistent HT.
Materials and methods
Data collection and eligibility criteria
We searched the following online databases using the keywords ‘thyroiditis’ and ‘cancer’: i) Medline using PubMed (http://www.ncbi.nlm.nih.gov/pubmed), ii) Embase (www.embase.com), and iii) ISI Science Citation Index using the ISI Web of Science search interface (http://apps.isiknowledge.com). We also manually searched the reference lists of the identified articles. Duplicate data or overlapping articles were excluded by examining the authors' names and affiliations. The following types of articles were included: i) original articles demonstrating that the association between PTC and classical HT was assessed only in thyroid specimens by histopathologic examination; ii) articles published before September 2011; iii) when multiple articles were published by the same authors or institutions, the most recent or informative single article was selected. Articles lacking clinicopathologic data for meta-analysis, review articles without original data, conference abstracts, and single case reports were excluded. Study quality was independently scored by two reviewers using the Newcastle–Ottawa Scale (41). The Newcastle–Ottawa Scale is frequently used for nonrandom studies such as case–control and cohort studies. The maximum scores of case–control and cohort studies are 9 and 13 respectively. Quality scores of the 38 studies ranged from 5 to 7 with a mean of 5.9 (Table 1). All were considered adequate for meta-analysis. Neither language nor geographic restriction was defined. The selection process of the articles is shown in Fig. 1.
Characteristics of individual studies included in the meta-analysis.
|References||Year||Country||Study design||Study objective||HT/PTC (%)||Quality score|
|(3)||2011||Korea||Case–control||Association between PTC and HT||307/1028 (29.9)||6|
|(4)||2010||Italy||Case–control||Association between PTC and HT||25/101 (24.8)||6|
|(5)||2010||Greece||Case–control||Association between PTC and HT||12/32 (37.5)||6|
|(6)||2009||Italy||Case–control||Serum thyroid autoantibody in PTC with HT||257/304 (84.5)||6|
|(7)||2007||Turkey||Case–control||Association between PTC and HT||37/199 (18.6)||6|
|(8)||2006||Japan||Case–control||Niban expression in thyroid tumor and HT||6/54 (11.1)||5|
|(9)||2005||Italy||Case–control||Association between PTC and HT||19/71 (26.8)||6|
|(10)||2002||Saudi Arabia||Case–control||Association between PTC and HT||34/59 (57.6)||6|
|(11)||1997||Ireland||Case–control||Thyroid diseases in west Ireland||1/14 (7.1)||6|
|(12)||1995||USA, Japan||Case–control||Association between PTC and HT in three races||210/312 (67.3)||6|
|(13)||1998||USA||Case–control||Association between PTC and HT||30/143 (21.0)||6|
|(14)||2011||Taiwan||Case–control||Association between PTC and HT||85/1788 (4.8)||6|
|(15)||2008||USA||Case–control||Association between PTC and HT||63/292 (21.6)||6|
|(16)||2004||USA||Case–control||HT in pediatric thyroid tumor||3/6 (50.0)||6|
|(17)||2002||Argentina||Cohort||PTC and HT in relation to iodine prophylaxis||31/87 (35.6)||7|
|(18)||1999||USA||Case–control||Association between PTC and HT||125/564 (22.2)||6|
|(19)||1999||USA||Case–control||Association between PTC and HT||57/388 (14.7)||6|
|(20)||1998||Germany||Case– control||Association between PTC and HT||23/92 (25.0)||6|
|(21)||1993||Italy||Case–control||HT in thyroid tumor||4/22 (18.2)||5|
|(22)||1983||Italy||Case–control||Pathologic characteristics in thyroid cancer||14/79 (17.7)||6|
|(23)||1957||USA||Case–control||HT in thyroid lesion||2/16 (12.5)||6|
|(24)||2012||Korea||Case–control||Association between PTC and HT||56/195 (28.7)||6|
|(25)||2010||Italy||Case–control||Association between PTC and HT||128/343 (37.3)||6|
|(26)||2010||Korea||Case–control||Association between PTC and HT||105/323 (32.5)||6|
|(27)||2010||USA||Case–control||FoxP3+ regulatory T cell frequency in PTC||37/100 (37.0)||6|
|(28)||2009||Korea||Case–control||Association between PTC and HT||214/1441 (14.9)||6|
|(29)||2009||Korea||Case–control||BRAF mutation in PTC and HT||37/101 (36.6)||6|
|(30)||2009||Norway||Case–control||PDGFC expression in PTC||7/18 (38.9)||5|
|(31)||2009||Turkey||Case–control||HT and tumor infiltrating lymphocytes in PTC||16/61 (26.2)||6|
|(32)||2007||Japan||Case–control||Ultrasonographic finding in PTC with HT||29/83 (34.9)||6|
|(33)||2001||Austria||Case–control||Latent thyroid cancer in Austria||6/10 (60.0)||6|
|(34)||2001||USA||Case–control||Association between PTC and HT||41/136 (30.1)||6|
|(35)||1998||Japan||Case–control||Association between PTC and HT||281/1533 (18.3)||6|
|(36)||1998||Japan||Case–control||Association between PTC and HT||15/69 (21.7)||6|
|(37)||1997||Spain||Case–control||Association between PTC and HT||6/129 (4.7)||6|
|(38)||1995||Japan||Case–control||Association between PTC and HT||36/95 (37.9)||6|
|(39)||2008||Italy||Case–control||Association between PTC and HT||72/189 (38.1)||6|
|(40)||2010||Turkey||Case–control||Association between PTC and HT||40/171 (23.4)||6|
|Total||2471/10 648 (23.2)|
HT, Hashimoto's thyroiditis; PTC, papillary thyroid cancer.
Data pooling and statistics
Meta-analysis was performed as previously described (42, 43). Briefly, effect sizes for each study were calculated by odds ratio (OR) or weighted mean difference (WMD) with 95% confidence intervals (CIs). For studies without hazard ratios (HRs) for survival, we assessed HRs and CIs using a published approximation method (44). The ORs, WMDs, or HRs were combined using a random-effects model (DerSimonian–Laird method). For identifying and quantifying inter-study heterogeneity, Q statistics were calculated, which is an adaptation of the χ2 goodness-of-fit test. P<0.10 was considered statistically significant. Sensitivity analyses according to study design (case–control vs cohort studies) were performed to examine the influence of each study on the pooled OR, WMD, or HR by serially omitting an individual study and pooling the remaining studies. Publication bias was examined by funnel plots and Egger's tests for the degree of asymmetry. P<0.05 was considered statistically significant. The pooled analysis was conducted using Comprehensive Meta-analysis software version 2.0 (Biostat, Englewood, NJ, USA).
A total of 38 articles satisfied the eligibility criteria (3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 33, 34, 35, 36, 37, 38, 39, 40). The eligible studies consisted of 37 case–control studies and one cohort study, all of which were hospital based. The eligible studies are summarized in Table 1. The number of patients in each study ranged from six to 1788, for a total of 10 648 PTC patients. Among the PTC patients, HT was present in 2471 (23.2%) cases.
PTC vs benign lesions
Eleven studies compared the occurrences of HT in PTCs and in benign thyroid diseases such as nodular hyperplasia and follicular adenoma (3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13). HT was found in 938 (40.5%) of 2317 PTC patients, whereas it was found in 634 (21%) of 3019 benign thyroid diseases. The coexistence of HT was significantly associated with PTCs than benign lesions (OR=2.766; 95% CI 1.947–3.929; P<0.001) (Fig. 2). Significant statistical heterogeneity was found among the studies (Q=39.664, df=10, P<0.001).
PTC vs other carcinomas
Sixteen studies investigated the frequencies of HT in PTCs and in other carcinomas such as follicular carcinoma and medullary carcinoma (3, 5, 8, 10, 11, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23). HT was present in 797 (17.1%) of 4664 PTC patients and in 57 (7.9%) of 725 other carcinoma patients. The coexistence of HT was more related to PTCs than other thyroid carcinomas (OR=2.432; 95% CI 1.614–3.665; P<0.001) (Fig. 3). There was significant statistical heterogeneity among the studies (Q=22.727, df=15, P=0.090).
Clinicopathologic characteristics of PTCs with HT
The incidence of HT in PTCs according to gender was compared in 23 studies (3, 4, 5, 6, 7, 14, 15, 19, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38). HT in PTCs was observed in 1677 of 7346 (23%) female patients and in 180 of 573 (11%) male patients. On the basis of this finding, there was a high association of HT in PTCs with females but not with males (OR=2.678; 95% CI 1.755–4.087; P<0.001). Significant statistical heterogeneity was found among the studies (Q=78.712, df=22, P<0.001).
Fourteen studies addressed the frequency of HT in PTCs according to patients' mean age (3, 4, 5, 14, 24, 28, 29, 30, 32, 33, 36, 37, 38, 39). The mean ages of PTC patients with HT ranged from 39.5 to 69.0 years, whereas the mean ages of PTC patients without HT ranged from 38.2 to 56.3 years. There was no association between the mean age of PTC patients and the incidence of HT in PTC (WMD=−0.081; 95% CI, −0.024 to 0.042; P=0.195). Statistical heterogeneity was detected among the studies (Q=22.788, df=13, P=0.044).
Eleven studies identified the prevalence of HT in PTCs according to average tumor size (3, 4, 14, 24, 28, 32, 33, 36, 37, 38, 39). The mean tumor sizes of PTCs with HT ranged from 0.6 to 4.8 cm, whereas those of PTCs without HT ranged from 0.6 to 3.0 cm. The tumor size was not related to the frequency of HT in PTC (WMD=−0.355; 95% CI, −1.224 to 0.514; P=0.424). There was statistical heterogeneity among the studies (Q=998.329, df=10, P<0.001).
Eleven studies presented 4128 PTCs without extrathyroidal extension and 2897 PTCs with extrathyroidal involvement (3, 14, 19, 24, 27, 28, 29, 35, 38, 39, 40). HT was found in 722 (17.5%) of 4128 PTCs without extrathyroidal extension and in 500 (17.2%) of 2897 PTCs with extrathyroidal extension. The coexistence of HT in PTCs was associated with no extrathyroidal involvement of PTC (OR=1.295; 95% CI 1.098–1.527; P=0.002). No significant statistical heterogeneity was detected among the studies (Q=11.656, df=10, P=0.309).
Lymph node metastasis
Sixteen studies reported 4185 PTC patients without lymph node metastasis and 3462 patients with lymph node metastasis (3, 4, 7, 14, 19, 24, 25, 27, 28, 29, 30, 35, 37, 38, 39, 40). HT was seen in 746 (17.8%) of 4185 PTC cases without lymph node metastasis and in 622 (17.9%) of 3462 cases with lymph node metastasis. PTCs with HT were related to the absence of lymph node metastasis (OR=1.287; 95% CI 1.010–1.639; P=0.041). There was significant statistical heterogeneity among the studies (Q=29.899, df=15, P=0.012).
Twelve studies addressed the frequencies of HT in single and multifocal PTCs (3, 4, 7, 24, 26, 28, 33, 34, 36, 37, 39, 40). The studies included 1378 cases with multifocal PTC and 2549 cases with single PTC. HT was present in 359 (26%) of 1378 multifocal PTCs and in 541 (21%) of 2549 single PTCs. HT was more often observed in multifocal PTCs than in single PTCs (OR=1.467; 95% CI 1.096–1.964; P=0.010) (Fig. 4). Significant statistical heterogeneity was found among the studies (Q=23.514, df=11, P=0.015).
Four studies including 616 patients of PTC with HT and 4241 of PTC without HT presented recurrence-free survival outcomes (14, 28, 35, 38). The estimated unadjusted HRs ranged from 0.547 to 0.781. The presence of HT in PTCs was significantly associated with a long duration of recurrence-free survival (HR=0.576; 95% CI 0.421–0.790; P=0.001) (Fig. 5). There was no significant statistical heterogeneity among the studies (Q=0.303, df=3, P=0.960).
Sensitivity analysis and publication bias
The sensitivity analyses revealed that all studies or case–control studies did not affect the pooled ORs and HR with CIs. However, seven studies influenced the result of lymph node metastasis (4, 7, 14, 29, 30, 37, 40). In the funnel plots and the Egger's regression tests, there was no evidence of publication bias (Fig. 6).
This meta-analysis showed that pathologically confirmed HT is more often found in PTC than in benign thyroid diseases and other carcinomas. Moreover, this analysis revealed that PTCs with coexisting HT are associated with female, multifocal involvement, the absence of extrathyroidal extension, no lymph node metastasis, and high recurrence-free survival rates.
Our pooled analysis indicates that the frequency of HT in PTCs was about 23%, ranging from 5 to 85%. The varying incidence rates of HT in PTC may be due to several factors such as different diagnostic criteria for HT, various surgical procedures, and heterogeneous patient characteristics. Most studies presented the incidence of HT in surgically resected PTC cases. In cytology specimens of HT patients, follicular cells often exhibit nuclear elongation, nuclear grooves, and even intranuclear inclusions, leading to a misdiagnosis of PTC (45, 46). Therefore, the current meta-analysis included only the studies that presented cases of HT confirmed by histopathologic diagnosis.
Our meta-analysis showed that the occurrence rate of HT in PTC patients was 2.8 times higher than HT patients in benign thyroid diseases. In addition, the incidence of HT in patients with PTC was 2.4 times higher than in those patients with other types of thyroid carcinoma. This result was similar to a previous result of another meta-analysis (19). In addition, some studies supported the tight association between HT and PTC, based on the fact that RET/PTC rearrangements were found in about 90% of HT cases (47) and transgenic mice expressing RET/PTC developed HT and PTCs (48).
Interestingly, this meta-analysis revealed that PTC patients with coexisting HT had distinctive clinicopathologic characteristics such as female gender, multifocality, no extrathyroidal extension, no lymph node metastasis, and long recurrence-free survival. Considerable controversy exists concerning the prognostic significance of HT in PTC patients. Loh et al. (18) and Yoon et al. (24) reported that PTC with HT was significantly associated with females and a lower incidence of extrathyroidal invasion and lymph node metastasis. Several studies found that PTC with HT had a tendency for multifocal involvement (3, 4, 26, 34). In contrast, other studies failed to present any significant clinicopathologic characteristics in PTC with HT (5, 25, 40).
This pooled analysis identified a paradoxical role of HT in the development and progression of PTC. The meta-analysis suggests a possible tight link between HT and the development of PTC rather than a chance occurrence of two relatively common diseases. Paradoxically, HT in PTC patients appears to play a role in impeding cancer progression. Therefore, the cross-link of these two conditions may represent a cause and effect relationship or a predisposing factor. It is hypothesized that PTC is induced or facilitated by a pre-existing lymphocytic infiltration. Conversely, lymphocytic infiltration of HT may be due to autoimmune thyroiditis and/or immune reaction to tumor-specific antigens from a pre-existing PTC (1).
Our pooled study indicates a close relationship between HT and PTC. As the incidence of HT is increased in PTC patients, careful clinical monitoring for the patients with HT and meticulous histopathologic examination of surgical specimens from these patients are required. The PTCs with HT are characterized by female predominance, multifocality, no extrathyroidal extension, no lymph node metastasis, and better recurrence-free survival outcomes.
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.
This study was supported by a Korea University Grant.
Author contribution statement
J-H Lee performed statistical analyses and wrote the manuscript draft. Y Kim and J-W Choi performed the literature search and data collection. Y-S Kim designed this study and edited the manuscript.
FioreERagoTScutariMUgoliniCProiettiADi CoscioGProvenzaleMABertiPGrassoLMariottiS. Papillary thyroid cancer, although strongly associated with lymphocytic infiltration on histology, is only weakly predicted by serum thyroid auto-antibodies in patients with nodular thyroid diseases. Journal of Endocrinological Investigation200932344–351.
OkayasuIFujiwaraMHaraYTanakaYRoseNR. Association of chronic lymphocytic thyroiditis and thyroid papillary carcinoma. A study of surgical cases among Japanese, and white and African Americans. Cancer1995762312–2318. (doi:10.1002/1097-0142(19951201)76:11<2312::AID-CNCR2820761120>3.0.CO;2-H).
SchafflerAPalitzschKDSeiffarthCHohneHMRiedhammerFJHofstadterFScholmerichJRuschoffJ. Coexistent thyroiditis is associated with lower tumour stage in thyroid carcinoma. European Journal of Clinical Investigation199828838–844. (doi:10.1046/j.1365-2362.1998.00363.x).
MuzzaMDegl'InnocentiDColomboCPerrinoMRavasiERossiSCirelloVBeck-PeccozPBorrelloMGFugazzolaL. The tight relationship between papillary thyroid cancer, autoimmunity and inflammation: clinical and molecular studies. Clinical Endocrinology201072702–708. (doi:10.1111/j.1365-2265.2009.03699.x).
FrenchJDWeberZJFretwellDLSaidSKlopperJPHaugenBR. Tumor-associated lymphocytes and increased FoxP3+ regulatory T cell frequency correlate with more aggressive papillary thyroid cancer. Journal of Clinical Endocrinology and Metabolism2010952325–2333. (doi:10.1210/jc.2009-2564).
KimEYKimWGKimWBKimTYKimJMRyuJSHongSJGongGShongYK. Coexistence of chronic lymphocytic thyroiditis is associated with lower recurrence rates in patients with papillary thyroid carcinoma. Clinical Endocrinology200971581–586. (doi:10.1111/j.1365-2265.2009.03537.x).
MatsubayashiSKawaiKMatsumotoYMukutaTMoritaTHiraiKMatsuzukaFKakudohKKumaKTamaiH. The correlation between papillary thyroid carcinoma and lymphocytic infiltration in the thyroid gland. Journal of Clinical Endocrinology and Metabolism1995803421–3424. (doi:10.1210/jc.80.12.3421).