Optimal thyrotropin level for low-risk papillary thyroid carcinoma after ultrasound-guided radiofrequency ablation

Abstract

Purpose

Maintaining an optimal thyroid stimulating hormone (TSH) level is important in the postoperative management of papillary thyroid carcinoma (PTC). However, there is little evidence for TSH target levels in patients undergoing radiofrequency ablation (RFA). This study aimed to determine the optimal TSH level for management in low-risk patients who underwent RFA.

Methods

This retrospective propensity score-matched cohort study included patients with low-risk PTC who underwent RFA from January 2014 to December 2018. The patients were categorized into two groups based on the range of TSH levels: low (≤2 mU/L) and high (>2 mU/L) TSH levels. Local tumor progression and disease-free survival (DFS) were compared between the low TSH and high TSH groups, using propensity score analyses based on patient- and tumor-level characteristics. Univariate analyses were performed to select risk factors for tumor progression.

Results

Overall, our study included 516 patients with low-risk PTC who underwent RFA with a long-term follow-up of 5-years. During follow-up, the overall incidence rate of local tumor progression was 4.8% (25/516), with no significant difference between the matched groups (7/106 [6.6%] vs. 5/53 [9.4%], p = 0.524). DFS did not differ between the two groups (p = 0.5). Moreover, TSH level was not regarded as a significant predictor of tumor progression after Cox analysis; primary tumor size was the only relevant risk factor.

Conclusion

This large propensity-matched study revealed no association between TSH levels and tumor progression. Thus, for patients with low-risk PTC who underwent RFA, the optimalTSH level is recommended at the euthyroid range.

Keywords:

• Papillary thyroid carcinoma
• radiofrequency ablation
• thyrotropin
• recurrence
• ultrasound

Introduction

The incidence of thyroid cancer has been increasing substantially in recent decades [1]. More than 80% of this increase is attributable to the growing detection of papillary thyroid carcinomas (PTCs) [2]. Most PTCs have an indolent and excellent long-term prognosis [3]. Recent guidelines recommend that management of low-risk PTC shift from total thyroidectomy to less aggressive treatment, such as lobectomy or active surveillance (AS) in properly selected patients [4,5].

Thyroid hormone therapy with levothyroxine is commonly used in the postoperative management of patients with differentiated thyroid cancer (DTC) to compensate for the lack of production by their thyroid glands [6]. Furthermore, thyroid stimulating hormone (TSH) has a growth effect on DTC, mediated by TSH receptors on the cell membrane [7]. TSH suppression using supraphysiological doses of levothyroxine could prevent progression or recurrence of thyroid cancers [4,8]. Such TSH suppression may improve outcomes in high-risk patients, but there is little evidence for this benefit in low-risk patients [9]. For patients with low-risk PTC undergoing lobectomy, the American Thyroid Association (ATA) guidelines recommend that TSH may be maintained in the mid to lower reference range (0.5–2mU/L) [4]. Consensus statements have suggested maintaining a low-normal TSH range during AS of low-risk PTC [10].

Ultrasound (US)-guided thermal ablation has also been recommended as a minimally invasive technique by guidelines for low-risk PTC patients who refuse surgery or AS [11,12]. With the growing interest in the use of thermal ablation to treat PTC, it is critical to determine the optimal TSH for individualized postoperative management. However, the appropriate TSH level in patients with low-risk PTC who have undergone RFA have not been reported to date.

Therefore, we conducted a retrospective observational study on a large cohort of patients who underwent US-guided RFA for treatment of PTC, with a follow-up of 5 years, to compare the clinical outcomes of patients with low or high TSH levels by propensity score matching. This study aimed to determine the optimal TSH level that would help guide management in low-risk patients after US-guided RFA.

Materials and methods

Study population

This retrospective cohort study screened consecutive patients with PTC who underwent US-guided RFA at the Chinese People’s Liberation Army General Hospital from 1 January 2014, to 31 December 2018. This study was approved by the Ethics Committee of the Chinese People’s Liberation Army General Hospital (S2019-211-02), which waived the requirement for informed consent.

The inclusion criteria were (a) unifocal T1N0M0 PTC confirmed by pathology with intrathyroid tumor ≤ 2 cm in the greatest dimension; (b) no imaging evidence of capsular infiltration, extrathyroidal extension, lymph node metastases (LNM) and distant metastasis; (c) no history of previous head and neck radiation, or familial thyroid carcinoma; and (d) at least two follow-up TSH data points within a time interval of no more than 2 years. The exclusion criteria were (a) less than 2 years of follow-up; and (b) insufficient TSH data during the follow-up (Figure 1).

Figure 1. Flowchart of patient selection. PTC: papillary thyroid carcinoma; RFA: radiofrequency ablation; TSH: thyroid stimulating hormone.

Pretreatment evaluation

All patients underwent US, CT and laboratory tests prior to ablation. US was performed to evaluate the size, location, US characteristics, and to exclude imaging evidence of capsular infiltration, extrathyroidal extension, LNM or distant metastasis combined with CT. Laboratory tests included complete blood count, coagulation tests, and thyroid function.

Treatment and follow-up

Contrast-enhanced US (CEUS; SonoVue, Bracco) was used to evaluate the enhancement pattern of tumor. All the patients underwent US-guided RFA that was physician-performed by personnel with >20 years of experience in thyroid US and tumor ablation. The equipment comprised (i) a Siemens Acuson Sequoia 512 (Siemens; Mountain View, CA) scanner with a 6.0 MHz linear array transducer for guidance; (ii) a bipolar RFA generator (CelonLabPOWER; Olympus Surgical Technologies, Europe); and (iii) an 18-gauge bipolar radiofrequency applicator with 0.9-cm active tip (CelonProSurge micro100-T09; Olympus Surgical Technologies, Europe). Ablation procedures involved the trans-isthmic, moving-shot, and hydrodissection techniques [13]. The tumor was completely ablated with a 2-mm safety margin to prevent marginal residual tumor and recurrence [14]. The ablation area was evaluated using CEUS. If a residual enhancement was observed, complementary ablation was performed.

All patients were advised regarding the need to maintain TSH levels in the mid to low reference range (0.5–2 mU/L) [4]. For patients with TSH above 2 before thermal ablation, we recommend a starting dose of 25 μg/day of thyroxine. Thyroid function tests (including free thyroxine, free triiodothyronine and TSH, anti-thyroid peroxidase and anti-thyroglobulin antibodies) were performed every 2 months within the first year and every 6–12 months thereafter. Patients were evaluated using US and CEUS at 1, 3, 6, and 12 months and every 6–12 months thereafter. US and CEUS were used to assess the ablation area and neck recurrence, to identify and classify of any suspicious nodules and lymph nodes, and to guide eventual fine-needle aspiration biopsy [15]. Chest CT was applied for the regular detection of lung metastases.

Serum TSH levels during follow-up

Serum TSH levels were measured multiple times during the follow-up period. Every patient had at least two follow-up TSH data points, and the average number of points per patient was 4.27. Serum TSH concentrations were determined using an automated electro chemiluminescence immunoassay (Cobas e 601, Roche Diagnostics, Mannheim, Germany) with a functional sensitivity of at least 0.01 mU/L and a reference range of 0.5–4.0 mU/L. As serum TSH levels may occasionally fluctuate, the average of all serum TSH levels was used to express overall TSH level in each case. Based on the range of TSH levels, patients were categorized into two groups: low (≤2 mU/L) and high (>2 mU/L) TSH levels.

Clinical outcomes after PTC treatment

The primary endpoints of this study were local tumor progression, distant metastasis, and disease-free survival (DFS). Local tumor progression, always established by biopsy under US guidance, was defined by [16,17]: (a) the occurrence of a newly found PTC separated from the treated tumor, (b) the appearance of neck LNM, and (c) the persistence of tumor at the ablation site. Distant metastasis was detected using CT if suspicious signs or symptoms presented from the bone, lung or other soft tissue [18]. DFS was calculated from the time of RFA for PTC to either disease recurrence or last follow-up date.

The secondary endpoints were complications. Complications during follow-up were assessed using the standardization of terminology and reporting criteria for image-guided thyroid ablation [19].

Statistical analyses

Statistical analyses were performed using the SPSS statistical software version 26.0 and R version 4.1.0 (R Project for Statistical Computing). Patient demographic and baseline characteristics were assessed using the chi-squared test or Fisher’s exact test for categorical data and the Mann-Whitney U test for continuous data. Categorical data are presented as numbers and percentages. Continuous data are described as means and standard deviations (SDs) or medians with interquartile ranges.

Propensity score matching was applied to adjust for different baseline characteristics and balance the distribution of biases and confounders between both groups in this retrospective observational study [20]. Patients in the high-TSH group were matched 1:2 to those in the low-TSH group based on age, sex, tumor size, largest diameter, volume, BRAF^V600E, follow-up period, and presence of chronic lymphocytic thyroiditis, which was confirmed by thyroid function test. Propensity score analysis was performed using a greedy, nearest neighbor matching algorithm, with a caliper width of 0.05 SD of the logit of the propensity scores. After propensity score matching, the baseline characteristics were compared between the two groups, with standardized mean differences.

DFS curves were calculated using Kaplan-Meier survival analyses with the log-rank statistic to compare the time of disease recurrence between groups. Cox proportional hazards regression models were used to compute hazard ratios (HRs) and 95% confidence intervals (CIs) for the associations of the identified variables and disease recurrence.

All p values were two-sided, and p < 0.05 was considered statistically significant.

Results

Patient characteristics

A total of 516 patients (mean [SD] age, 43.3 [10.3] years; 404 women [78%]) were included in this study. The mean follow-up period was 59 months. The mean tumor size was 7.0 mm; 89% (459/516) of the patients had a tumor ≤ 1 cm. Overall, the mean TSH of the cohort was 1.21 ± 1.53 mU/L. 463 (90%) and 53(10%) patients had low (≤2 mU/L) and high (>2 mU/L) TSH levels, respectively. Patient demographic and baseline characteristics are shown in Table 1.

Table 1. Baseline characteristics of low-risk PTC patients who underwent radiofrequency ablation with TSH suppressive therapy.

Characteristic	Total (n = 516)	Before propensity score matching			After propensity score matching
		Low TSH (n = 463)	High TSH (n = 53)	p Value	Low TSH (n = 106)	High TSH (n = 53)	p Value
Age, years (%)				0.726			0.377
≤55	460(89)	412(89)	48(91)		100(94)	48(91)
>55	56(11)	51(11)	5(9)		6(6)	5(9)
Sex, No. (%)				0.008*			1.000
Male	112(22)	108(23)	4(8)		8(8)	4(8)
Female	404(78)	355(77)	49(92)		98(92)	49(92)
Tumor size, No. (%)				0.693			0.248
≤1 cm	459(89)	411(89)	48(91)		101(95)	48(91)
1–≤2 cm	57(11)	52(11)	5(9)		5(5)	5(9)
Location				0.308			0.288
Left lobe	216(42)	190(41)	26(49)		44(42)	26(49)
Isthmus	12(2)	12(3)	0(0)		3(3)	0(0)
Right lobe	288(56)	261(56)	27(51)		59(55)	27(51)
Thyroid function				0.448			0.531
CLT	106(21)	93(20)	13(25)		31(29)	13(25)
Normal	410(79)	370(80)	40(75)		75(71)	40(75)
BRAF^V600E mutation				0.214			0.564
Positive	337(65)	306(66)	31(58)		67(63)	31(58)
Negative	179(35)	157(34)	22(42)		39(37)	22(42)
Greatest diameter, mm	7.0(3.1)	7.1(3.1)	6.4(2.8)	0.080	5.9(2.4)	6.4(2.8)	0.269
Volume, mm^3	90.2(147.3)	92.1(156.8)	73.6(89.7)	0.196	42.1(122.0)	73.6(89.7)	0.123
Follow-up time, months	59.3(15.5)	59.8(15.4)	54.6(15.4)	0.012*	56.2(15.6)	54.6(15.4)	0.418

Categorical variables were described by numbers with percentages and continuous variables were described by mean±SD or median (IQR).

CLT: chronic lymphocytic thyroiditis.

*p < 0.05 was considered statistically significant.

Before propensity score matching, patients in the low-TSH group had fewer women (no. [%], 355 [77] vs. 49 [92], p < 0.008) and longer follow-up period (mean [SD] months, 59.8 [15.4] vs. 54.6 [15.4], p = 0.012) than those in the high-TSH group. There were no significant differences in other characteristics between the two groups. After 1:2 propensity score matching, no significant differences were found in age, sex, tumor size, location, thyroid function, BRAF^V600E, maximum diameter, volume and follow-up period between groups (p > 0.05 for all comparisons; Table 1).

Clinical outcomes

With a mean follow-up of 59 (range, 37–97) months, the overall incidence rate of local tumor progression was 4.8% (25/516), including 0.8% LNM (4/516), 3.1% newly found PTC (16/516), and 0.9% persistent lesion (5/516). During the follow-up period, none of the patients had distant metastasis or died of PTC. Treatment decisions about local tumor progression are made by the experience of the physician and by the patient in consultation with the physician based on their preferences. Among these progressive patients, 2 selected surgery, and 23 underwent additional RFA. All patients had no disease progression after re-treatment during the follow-up period. Before propensity score matching, no significant differences were found in local tumor progression between the low- and high-TSH groups (20/463 [4.3%] vs. 5/53 [9.4%], p = 0.100). After propensity score matching, there was no significant difference between the two groups (7/106 [6.6%] vs. 5/53 [9.4%], p = 0.524). Comparisons of local tumor progression are summarized in Table 2. None of the patients had major or minor complications after RFA. Only 21 patients experienced side effects in the form of local pain and discomfort which were spontaneously relieved within 7 days.

Table 2. The comparison of local tumor progression between the two groups.

Characteristic	Total (n = 516)	Before propensity score matching		After propensity score matching
		Low TSH (n = 463)	High TSH (n = 53)	Low TSH (n = 106)	High TSH (n = 53)
Total incidence	25(4.8)	20(4.3)	5(9.4)	7(6.6)	5(9.4)
LNM	4(0.8)	3(0.6)	1(1.9)	0	1(1.9)
Developed time, months	18.0(7.0)	17.3(8.4)	73.6	–	73.6
Greatest diameter, mm	8.5(1.3)	9.0(1.0)	7	–	7
Developed volume, mm^3	99.9(30.2)	103.4(8.0)	20	–	20
Location, lateral: cervical		1:3	0:1	0:0	0:1
Newly found PTC	16(3.1)	12(2.6)	4(7.5)	6(5.7)	4(7.5)
Developed time, months	28.8(14.9)	29.3(16.1)	27.0(12.7)	26.5(19.5)	27.0(12.7)
Greatest diameter, mm	5.1(2.3)	5.6(3.0)	3.8(1.5)	6.5(3.8)	3.8(1.5)
Developed volume, mm^3	32.2(79.3)	46.0(75.7)	9.5(52.1)	67.1(224.1)	9.5(52.1)
Location, psilateral: contralateral		9:5	2:2	5:2	2:2
Persistent lesions	5(0.9)	5(1.1)	0	1(0.9)	0

Categorical variables were described by numbers with percentages and continuous variables were described by mean±SD or median (IQR).

LNM: lymph node metastases; PTC: papillary thyroid carcinomas.

DFS between the low and high TSH groups

The DFS rates at 1, 3, and 5 years were 98.1%, 96.8%, and 95.7%, and 100%, 92.5%, and 90.6% in the low- and high-TSH group, respectively. Before propensity score matching, there were no significant differences between both groups (p = 0.088, Figure 2(A)). After propensity score matching, the DFS rates at 1, 3, and 5 years were 97.2%, 95.3%, and 93.3%, and 100%, 92.5%, and 90.6% in the low- and high-TSH group, respectively. There were still no significant differences between the groups (p = 0.5, Figure 2(B)).

Figure 2. Disease-free survival curves for low vs. high TSH levels in low-risk PTC after RFA. TSH: thyroid stimulating hormone; PTC: papillary thyroid carcinoma; RFA: radiofrequency ablation.

Univariate analysis evaluating the risk factors for local tumor progression

According to Cox proportional hazards regression models, local tumor progression was not significantly associated with TSH levels before (HR [CI] = 0.44[0.16–1.17], p = 0.098) and after propensity score matching (HR [CI] = 0.67[0.21–2.13], p = 0.504) (Table 3). Patients with tumor size > 1 cm were significantly more likely to result in local tumor progression than those with tumor size ≤ 1 cm before (HR [CI] = 0.24 [0.10–0.56], p = 0.010) and after propensity score matching (HR [CI] = 0.11[0.03–0.38], p < 0.001). Cox regression analysis revealed that other factors were not significantly associated with local tumor progression before and after propensity score matching (Table 3).

Table 3. Analysis of hazard ratios of local tumor progression in cox proportional hazards regression models.

Characteristic	Before propensity score matching		After propensity score matching
	Univariate analysis		Univariate analysis
Variable	HR (95% CI)	p Value	HR (95% CI)	p Value
Age, years (%)
≤55	2.99(0.40,22.07)	0.284	0.77(0.10,5.97)	0.802
>55
Sex, No.(%)
Male	0.49(0.15,1.62)	0.240	0.04(0.00,519.17)	0.985
Female
Tumor size, No.(%)
≤1 cm	0.24(0.10,0.56)	0.010*	0.11(0.03,0.38)	0.000*
1–≤2 cm
Thyroid function
CLT	0.38(0.17,0.85)	0.019	0.82(0.25,2.72)	0.74
Normal
BRAF^V600E mutation
Positive	0.85(0.37,1.97)	0.705	1.09(0.35,3.44)	0.881
Negative
TSH
Low (≤ 2 mIU/L)	0.44(0.16,1.17)	0.098	0.67(0.21,2.13)	0.504
High (> 2 mIU/L)

CLT: chronic lymphocytic thyroiditis; HR: hazard ratio; CI: confidence intervals.

*p < 0.05 was considered statistically significant.

Discussion

Although the safety and efficacy of RFA for PTC have been demonstrated by many studies, there is no evidence to guide the appropriate TSH level in the management of patients who underwent RFA for low-risk PTC. This study involved a propensity-matched analysis of 516 patients with low-risk PTC who underwent RFA with a long-term follow-up of 5-years. The results showed no significant differences in local tumor progression between the low and high TSH groups (9.4% vs. 4.3%). The association remained insignificant after propensity score matching. DFS did not differ according to serum TSH levels, in patients with low-risk PTC after RFA. In the Cox analysis, TSH levels was not regarded as a significant predictor of tumor progression; primary tumor size was the only relevant risk factor.

TSH suppression therapy using levothyroxine supplementation is proposed in the postoperative management of patients with PTC [4]. Decreased disease recurrence and cancer-related mortality rates for PTC have been demonstrated in high-risk patients treated with TSH suppression [21–23]. However, for patients with low-risk PTC who have undergone thyroid lobectomy or AS, the evidence for the effect of TSH suppression therapy is insufficient [9,10,24]. Active TSH suppression may cause several non-negligible adverse effects, including osteoporosis, fracture, and cardiovascular disease, and it can also affect the quality of life with side effects like anxiety and insomnia [7,8]. Therefore, optimal TSH levels should balancethe benefit of TSH suppression with the harm of adverse effects [25].

RFA differs from thyroid lobectomy in that it causes focal hyperthermic injury to ablated tumors and spares healthy thyroid parenchyma [17]. Patients may not require levothyroxine replacement because the remaining thyroid tissue maintains thyroid hormone in the euthyroid range [11,26]. In this study, patients were advised to keep serum TSH levels in the range of 0.5–2 mU/L with reference to ATA guidelines for patients of low-risk papillary carcinoma involving lobectomy [4]. Fifty-three patients with high TSH levels did not meet this target. After propensity score matching, high TSH levels did not differ in terms of tumor progression rates (5/53 [9.4%] vs. 7/106 [6.6%], p = 0.524) or DFS (p = 0.5), when compared with low TSH levels subsequent to treatment of patients with low-risk PTC who underwent RFA. Moreover, the association between the two groups remained nonsignificant after Cox analysis (p = 0.504). Although elevated TSH level is theoretically followed by an increased risk of recurrence of PTC, the association between TSH and recurrence in post-RFA patients may be complicated by the effects of residual thyroid. Several studies have investigated the appropriate TSH level in low-risk patients who underwent lobectomy. Xu et al. [27] found no association between the mean TSH level and tumor recurrence in post-lobectomy PTC patients after a 70-month follow-up period. Park et al. [28] found no clinical benefits of TSH suppression therapy in patients with low-risk DTC who underwent lobectomy. Park et al. [29] found that TSH suppression therapy is not necessary for patients with low-risk DTC who underwent lobectomy, considering the excellent prognosis of low-risk DTC and the limitation of TSH suppression therapy. The results of the present study are in line with the evidence obtained in patients treated by lobectomy. More than 80% of PTC cases are classified as low risk, which has an excellent prognosis with recurrence rates of only 1–6% [4,24]. Low-risk patients may derive little or no benefit from suppressing their serum TSH levels. Therefore, patients treated with RFA can maintain normal thyroid function and do not require additional thyroid hormone therapy to control TSH levels.

Moreover, our study provided evidence for the efficacy of RFA in the treatment of low-risk PTC. A large cohort of 516 patients with unifocal T1N0M0 PTC were treated with RFA. During a follow-up period of nearly 5 years, the overall incidence rate of local tumor progression was 4.8%. Other studies have reported that the tumor progression rate in a similar series was 0.4–6.6% [30–34]. This relative variation in recurrence may be explained by differences in sample sizes, selection bias, follow-up period, and intrinsic heterogeneity of the tumor. In addition, recent observational studies based on RFA vs. surgery, controlling for confounding variables, found that the lack of difference in recurrence between the two interventions is supported for those patients with low-risk PTC [16,17]. In summary, these results show that RFA is a promising technique for the treatment of low-risk PTC. However, the association between tumor variables and recurrence after RFA has been relatively under-examined. Our study revealed that primary tumor size was a significant risk factor associated with recurrence in univariate analysis. For PTCs < 2 cm, papillary microcarcinomas have a good prognosis after RFA. This finding is similar to that in patients after surgery, and the prognosis for DTC is incrementally poorer as the tumor size increases [35,36].

Our study has some limitations. First, although propensity score matching was applied to minimize the effects of confounders, the study was based on retrospective data. Randomized controlled clinical trials are required to improve this evidence. Second, the levothyroxine supplementation dose and potential adverse events were not recorded during the follow-up period. The study cannot determine how to balance the adverse events and benefits of TSH suppression. Third, the tumor progression rate was low owing to the limited follow-up period, and this was a single institution study. Thus, the findings may not be generalizable. Larger cohorts and longer-term follow-up studies are required to overcome these limitations.

Conclusions

Our findings suggest no association between TSH levels and tumor progression in patients with low-risk PTC who underwent RFA. For patients with low-risk PTC who have undergone RFA, the optimal TSH level is recommended at the euthyroid range. Patients treated with RFA can maintain normal thyroid function and do not require additional thyroid hormone therapy to control TSH levels. Rigorous prospective randomized studies are required to determine the optimal TSH level and to optimize patient selection for TSH suppression.

Disclosure statement

No potential conflict of interest was reported by the author(s).