The role of Hashimoto thyroiditis in predicting radioiodine ablation efficacy and prognosis of low to intermediate risk differentiated thyroid cancer

Differentiated thyroid cancer (DTC) is the most common endocrine malignancy, accounting for 1% of all cancers diagnosed each year [1, 2] and it is considered a slowly growing cancer with an overall good outcome, except of cases with distant metastases or radioiodine-refractory disease [3, 4]. The baseline treatment of DTC consists of thyroidectomy followed by postoperative risk-adapted radioiodine therapy (RAIT) when indicated [2, 5].

The choice of most appropriate RAI activities to administer with the aim to reach an efficient remnant ablation and reduce the risk of recurrence is yet an open issue. This point is even more debated in low-risk and intermediate-risk DTC, where it seems crucial to find the best balance between the effective cure and the risk of side effects (reduced radiation exposure, cost, time of hospitalization) [6‐9]. Consensus guidelines recommend a dose of 1.1–3.7 GBq for patients with low to intermediate-risk DTC and higher activities for high-risk DTC, residual disease or aggressive histological subtypes [2]. Several papers [10, 11] demonstrated the equivalence of successful ablation between low-activity (1.1 GBq) and high-activity (3.7 Gbq) and between thyroid hormone withdrawal and recombinant human thyroid-stimulating hormone injections, without significant difference in outcome survival.

To better understand which activity inject in these patients successfully while minimizing the side effects, it seems fundamental to know the factors that influence the success of ablation and to predict the outcome.

The most shared known factors include the preablation stimulated thyroglobulin (sTg) level, the completeness of surgery, the preablation radioiodine scan findings and the histotype [12‐14]. The potential influence of Hashimoto Thyroiditis (HT) in the effectiveness of RAIT is not well-investigated. Only Kwon et al. [15] showed that coexisting HT together with sTg were negative predictive factors for successful low-dose RAIT (1.1 GBq).

Moreover, also in predicting treatment response and prognosis the role of HT is not well-investigated. The association between DTC and HT has been long suggested with several studies in favor [16‐18]. Probably, HT is frequently associate with DTC, especially papillary thyroid carcinoma, and may be considered a risk factor for developing thyroid cancer but the reason of this link is not yet understood [19, 20]. Moreover, DTC patients with concomitant HT seems to have a better prognosis that DTC patients without [21, 22].

The main aim of this study was to retrospectively investigate in a consecutive cohort of patients with low- to intermediate-risk DTC treated with total thyroidectomy and RAIT, the role of HT in predicting 1-year and 5-year treatment response and disease status. The second aim was to analyze the potential prognostic role of HT in this population.

We retrospectively screened all patients who performed thyroidectomy from January 2010 until January 2018 in our Surgery Department. Inclusion criteria were (1) a total thyroidectomy followed by RAIT, (2) age more or equal to 18 years at time of surgery, (3) a histological confirmation of DTC, (4) a classification as low risk or intermediate risk according to 2015 ATA guidelines [2], (5) at least 12 months of clinical and imaging follow-up after RAIT. Exclusion criteria were (1) age less than 18-years old, (2) hypothyroidism or hyperthyroidism condition before surgery, (3) a previous history of other neoplastic disease. Finally, 314 patients were included: 174 low-risk class stage after surgery and RAIT and 140 intermediate-risk class. There was a prevalence of female (F:M = 2.8:1) and mean age at diagnosis was 50-years old (Table 1). The first administered activity of RAI ablation ranged from 1.1 to 4.4 GBq (average 2.7 GBq) and it was established according to the risk class based on the TNM staging of the AJCC currently in use. Patients treated before the introduction of new American Joint Committee on Cancer/International Union against Cancer (AJCC) 8th edition were reclassified according to the last guideline [23]; this explains why some patients were treated with high-activities due to the downstage from intermediate-risk class to low-risk one.

All patients underwent total thyroidectomy and had a histopathological diagnosis of DTC: 146 classic variant of papillary carcinoma, 105 follicular variant of papillary carcinoma, 25 follicular carcinoma (17 minimally invasive and 8 widely invasive), 26 aggressive papillary variants (18 tall cells variant of papillary carcinoma and 8 sclerosing diffuse variant of papillary carcinoma) and 12 Hürthle cell carcinoma. One hundred and fifty patients underwent also a central lymphadenectomy and also lateral in fifty-four cases. HT was reported at the histological sample in 120 patients (38%), while in the remaining 194 cases (62%) no foci of HT were described. All patients received RAIT 2–3 months after thyroidectomy: 137 patients discontinued levothyroxine treatment 4 weeks before radioactive-iodine remnant ablation and replaced by triiodothyronine for 2 weeks, while 177 patients received recombinant human thyrotropin (rhTSH-Thyrogen, Genzyme Corporation) intramuscularly with a dose of 0.9 mg over 2 consecutive days during treatment with levothyroxine and RAI was administered the day after the second injection. Serum thyrotropin (TSH), fT3, fT4, Tg, antithyroglobulin antibodies (TgAb), and ioduria were measured before RAIT. All patients followed a low-iodine diet for at least 2 weeks and mean TSH level was higher than 30 UI/L in all cases. Three days after RAIT, the patients underwent a whole body scintigraphy (WBS) followed by single photon emission computed tomography/computed tomography (SPECT/CT), by hybrid dual-detector SPECT/CT (Infinia Hawkeye II, GE Healthcare, Haifa Israel), equipped with 1 inch StarBrite^TMCrystal and a high-energy collimator. The StarBrite™ Crystal has a sensitivity more than twofolds higher than that provided by a 3/8 inch crystal without a significant loss in resolution, therefore contributing to improve the reliability of SPECT with RAI. The WBS was performed in continuous mode with high-energy general purpose (HEGP) parallel holes collimator, 364-keV photopeak with ± 10% energy windows setting and scatter correction. The infrared-based real-time automatic body contouring system was activated for simultaneous anterior and posterior view with a matrix of 256 × 1024. Hybrid SPECT/CT scans from skull base to the lung bases were routinely acquired in all cases and additional SPECT/CT scans of other areas were done according to the WBS findings. SPECT images were acquired with HEGP collimator, matrix size of 128 × 128, 364-keV photopeak with ± 10% energy and scatter windows, dual-detector 180° acquisition, angular step of 3°, 15″ time per step/view. The CT parameters were 140 kV, 2.5 mA, 30″ rotation speed, 10 mm slice thickness, 256 × 256 matrix. CT acquisition was performed with a 2-slices elicoidal acquisition. An ordered subset expectation maximization iterative reconstruction with CT-based attenuation correction and scatter correction was performed. Moreover, about one-year after RAIT, a WBS after a diagnostic activity of 185 MBq or after a therapeutic activity in case of persistent disease was executed. Also sTg, TSH and TgAb were done at the same time and the combination of biochemical and radiological findings were used to classify the ablation success rate and the treatment response categories according to ATA groups (excellent response, biochemical incomplete response, structural response and indeterminate response) [2]. Disease status 5 year after RAIT was evaluated considering a combination of clinical, biochemical (Tg, AbTg) and imaging (ultrasound, CT, WBS and/or PET/CT) data.

HT is the most common autoimmune inflammatory thyroid disease with a worldwide annual incidence of 0.3–1.5 cases for 1000 persons and it is considered the first cause of hypothyroidism in the iodine-sufficient areas of the world [24]. The etiopathogenesis is probably a combination between environmental factors and genetic condition which causes the loss of immunological tolerance, with a subsequent autoimmune attack to the thyroid tissue. HT brings a chronic inflammatory condition that awakens an immune response leading to a continuous damage of surrounding stromal cells; overall this phenomenon causes potential genetic alterations, an inappropriate cell proliferation, and increases the risk of neoplastic transformation. Historically, HT was demonstrated significantly associate with an increased risk of developing thyroid lymphoma [25, 26]. But also the association between HT and DTC has been investigated with positive evidences [19, 20]. First in 1955 Dailey et al. [27] reported an increased association between HT and DTC, mainly papillary TC. However, the reason of this link remains yet unclear. The incidence of HT with concomitant PTC were reported as 26.8–38.1% in Italy, and 14.7–30.1% in the United States [28]. There are present in literature several hypotheses: someone suggested that HT could be a “cancer-impending effect” meaning an attempt of autoimmune system to fight against the tumor with the aim to cure the oncological disease and it could explain the favorable outcome reported in DTC with concomitant HT [29]; others speculated that the elevated TSH levels, often secondary to HT, might stimulate follicular epithelial cells proliferation and gained also the proliferation of cancer cells [30], but no shared consensus is available.

In our sample, the two groups (HT vs. no HT) had similar epidemiological and histological features, despite the histotype variants distribution and the age were near to statistical significance (0.090 for histotype and 0.058 for age). Patients with HT had lower age at diagnosis and less cases of follicular DTC in comparison with no HT patients.

Beside the linkage between HT and DTC, it seems crucial to clarify if conventional therapy of DTC may be affected by the presence or not of HT. Kwon et al. [15] conducted a retrospective analysis on 691 DTC patients who underwent thyroidectomy plus 1.1 GBq of RAI. Successful RAI ablation was registered in 62.3% of patients, and stimulated Tg at ablation and coexisting HT were the only two factors significantly associated with the success rate. The efficacy of low-activity of RAI was negatively influenced by sTg and HT. Moreover, patients with HT underwent more RAI therapies and higher total RAI activity. These evidences were confirmed also by our analysis: in DTC patients with coexisting HT the rate of 1-year excellent response was lower than DTC patients without HT (54% vs 69%; p < 0.001) and also the 5-year excellent response was lower than DTC patients without HT (79% vs 94%; p < 0.001). One possible explanation for RAI failure is the reduction of lower sodium–iodide symporter (NIS) expression in patients with coexisting HT, which it is strictly related to the decreased radioactive iodine uptake into the remnant thyroid tissue [31]. Kollecher et al. [32] demonstrated that thyroid tissue with HT showed significantly lower levels of NIS membrane staining than normal thyroid tissue. Moreover, the inflammatory background typical of HT, may reduce the therapeutic efficacy of RAI. Also for high-activity (3.7 GBq) HT seems to have a significant role in predicting ablation success after RAI [33]: in patients with HT the success rate was 33.4% in comparison with DTC without HT with a rate of 66.6%. In our study, we included both low and intermediate risk class DTC, thus low–intermediate RAI activity was injected. The presence of HT was related with higher cumulative RAI activities administrated and higher number of radiometabolic therapies; also in this case, the reason could be related with the less efficacy of radioiodine to enter in the thyroid cells due to the inflammatory micro-environment and the loss NIS expression.

In the last 2015 ATA guidelines [2], for the first time a new category response was introduce called indeterminate response which consist of patients without clear biochemical or structural disease, but without a complete response. In this group it is included also the case of DTC patient with TgAb stable or declining, but positive, without definitive structural evidence of disease. Two studies have demonstrated that only 13–20% of patients with an indeterminate response to therapy are reclassified as persistent/recurrent disease over approximately 10 years of follow-up [34, 35]

The rate of patients with indeterminate response after RAIT was significantly higher in HT patients compared to not HT both 1 year and 5 years after RAIT. This is directly associate with the presence of TgAb in patients with HT compared in patients without. The high rate of indeterminate response could partially explain the different excellent response rate between the two groups, even though the prognosis is not affected.

On the other hand, HT does not show a prognostic role in predicting PFS and OS. Other parameters, like age at diagnosis, sTg, and disease status after first-line therapy were significantly correlated with the risk of recurrence. Stimulated Tg confirmed to be an independent prognostic variable also for OS. Several papers underlined a possible association between HT and a good outcome, describing that DTC with concomitant HT had a lower rate of lymph node and distant metastases [22, 36, 37]. In our analysis, no significant impact of HT in prognosis was showed also due to the selection of low- to intermediate-risk DTC, which for definition have a good prognosis. In fact, the number of events (progression/relapse or death) was relatively low. As previously demonstrated [2], preablation sTg level has been considered a poor prognostic factor for successful ablation and prognosis. In this manuscript, a sTg threshold of 3.8 ng/mL was derived as the best compromise between sensitivity and specificity; other cutoff values are suggested in literature [38, 39].

Our study had several limitations, including the retrospective nature of the study design, the heterogeneity of patients’ features and the relatively not long follow-up time (88 months in this study) for prognostic evaluation considering the outcome of low to intermediate risk DTC.

In conclusion, we have demonstrated that HT may affect the efficacy of RAIT in low- to intermediate-risk DTC, particularly reducing the successful rate of excellent response after RAIT. Instead, HT does not seem to have a prognostic role considering PFS and OS.