Radioactive Iodine Therapy Decreases the Recurrence of Intermediate-Risk PTC With Low Thyroglobulin Levels

Abstract

Context

Whether radioactive iodine therapy (RAIT) is necessary for intermediate-risk papillary thyroid cancer (PTC) after total thyroidectomy is still lacking reliable evidence, especially for patients with low postoperative thyroglobulin (Tg) levels.

Objective

This study conducted a propensity score matching (PSM) analysis to investigate whether RAIT is effective in reducing the recurrence of intermediate-risk PTC with low Tg levels.

Methods

In total, 1487 patients with intermediate-risk PTC with unstimulated Tg ≤ 1 ng/mL or stimulated Tg ≤ 10 ng/mL after total thyroidectomy were enrolled retrospectively. The clinicopathological characteristics were compared between the non-RAIT and RAIT groups before and after PSM (1:4 matching). The impact of RAIT on biochemical recurrence and structural recurrence was evaluated.

Results

Overall, 1349 (90.7%) patients underwent RAIT, and 138 (9.3%) did not. After a median follow-up time of 51 months, 30 patients presented with recurrence, including 11 structural and 19 biochemical recurrences. After PSM, the non-RAIT group had a higher rate of structural recurrence (5/138 vs 5/552, P = .046) and biochemical recurrence (6/138 vs 4/552, P = .005) than the RAIT group. Multivariate analysis showed that not receiving RAIT was an independent risk factor for structural recurrence (hazard ratio [HR] 10.572, 95% CI 2.439-45.843, P = .002) and biochemical recurrence (HR 16.568, 95% CI 3.670-74.803, P < .001). Kaplan–Meier analysis showed that the non-RAIT group had more unfavorable recurrence-free survival (structural and biochemical, all P < .05).

Conclusion

RAIT could decrease the recurrence risk of intermediate-risk PTC in patients with unstimulated Tg ≤ 1 ng/mL or stimulated Tg ≤ 10 ng/mL. Further prospective randomized studies are needed to confirm these findings.

papillary thyroid cancer, intermediate-risk, radioactive iodine, thyroglobulin, recurrence

Differentiated thyroid cancer (DTC) is the most common malignant tumor of the endocrine system, accounting for 90% of all thyroid cancers (1-3). Papillary thyroid cancer (PTC) is the most common pathological type of DTC, which generally has a good prognosis (4). The American Thyroid Association (ATA) management guidelines have proposed a clinicopathological risk stratification system to classify patients with DTC as low, intermediate, or high risk for recurrence after thyroidectomy (5). Radioactive iodine therapy (RAIT) is recommended routinely for patients with high-risk DTC after total thyroidectomy, while it is considered selectively for patients with intermediate-risk DTC because of conflicting or inadequate data.

Several retrospective studies based on public data sources have shown that RAIT can improve overall survival or disease-specific death in patients with intermediate-risk PTC (6, 7). Notably, the data from these studies were obtained from public databases without detailed information on the specific causes of death and the progression of PTC, which may reduce the applicability of the research results. Other studies with contrasting perspectives thought RAIT did not decrease the risk of recurrence or death in intermediate risk patients (8-11). However, the retrospective design of these studies was prone to selection bias, and some of the studies had a relatively small sample size. Furthermore, none of the above studies considered the effect of thyroglobulin (Tg) on clinical outcomes.

Tg levels after total thyroidectomy provide an excellent predictive value for persistent and recurrent disease in patients with PTC (12, 13). Our previous study demonstrated that disease persistence/recurrence occurred in only 4.1% of intermediate-risk patients with a stimulated Tg ≤ 10 ng/mL, much less than in patients with Tg > 10 ng/mL (37.9%) (14). In a prospective study enrolling intermediate- to high-risk patients with stimulated Tg > 10 ng/mL, functional or structural disease was observed in 28.4% of the cohort after RAIT treatment, and more than 80% of patients achieved nonstructural incomplete response at final follow-up; therefore, the authors suggested that DTC patients with hyperthyroxinemia could benefit from RAIT (15).

The necessity of RAIT in intermediate-risk patients with low Tg levels lacked evidence from a random, prospective study, with the previous retrospective studies yielding biased results. The propensity score matching method (PSM) is a statistical method used to establish a new control group which attempts to alleviate the interference of confounding bias from observational cohorts, allowing for the proper assessment of the intended variable (16, 17). Thus, our study investigated the efficacy of RAIT in patients with intermediate-risk PTC with unstimulated Tg ≤ 1 ng/mL or stimulated Tg ≤ 10 ng/mL after total thyroidectomy using the PSM method.

Materials and Methods

Patients

This was a single-center, retrospective, observational cohort study. The present study was approved by the Institutional Review Board of Sichuan University, West China Hospital, and the requirement for obtaining written informed consent from the patients was waived. Inclusion criteria were (1) patients who underwent total thyroidectomy or near-total thyroidectomy at our center between August 2009 and June 2020; (2) patients with postoperative pathologically confirmed PTC; (3) patients defined as intermediate risk according to the ATA initial risk stratification system; (4) and patients with postoperative serum thyroglobulin antibody (TgAb) ≤ 40 IU/mL (14, 18) and stimulated Tg level ≤ 10 ng/mL or unstimulated Tg level ≤ 1 ng/mL after total thyroidectomy. Patients with the following conditions were excluded: neck ultrasound, computed tomography, or ¹³¹I whole-body scan indicating the existence of a structural disease within 6 months after surgery, follow-up duration <1 year, or incomplete follow-up information.

Based on clinical and surgical histopathological information, the intermediate-risk group was defined according to the ATA risk criteria (5). Tumor–Node–Metastasis staging was determined by 8th edition of the American Joint Committee on Cancer Staging Manual (19).

Surgical Treatment

All patients underwent total thyroidectomy and lymph node dissection in the central or lateral cervical region at the discretion of experienced surgeons based on the Chinese guidelines for the management of thyroid nodules and differentiated thyroid cancer (20), with a therapeutic or prophylactic central-compartment neck dissection (CND) being performed. A lateral compartment neck dissection was performed in patients with biopsy-proven metastatic neck lymphadenopathy.

The entire surgical specimen of the thyroid and the resected lymph node were sent for pathological examination. The tumor size, extrathyroidal extension (ETE), and the total number of retrieved and lymph node metastases (LNM) were recorded.

Thyrotropin, Tg, TgAb Measurement

All the patients were tested for thyrotropin (TSH), Tg, and TgAb after total thyroidectomy in the same laboratory at the West China Hospital, Sichuan University. A highly sensitive electrochemiluminescence immunoassay was used for measuring TSH (Elecsys TSH Kit; Roche Cat# 11731459, RRID:AB_2756377, Roche Diagnostics; measurement range 0.005-100 µIU/mL), Tg (Elecsys anti-Tg antibody from 2009 to 2015; Roche Cat# 04738578, RRID:AB_2631045, Roche Diagnostics, Mannheim, Germany, measurement range 0.1-1000 ng/mL; Elecsys Tg II assay from 2016; Roche Cat# 06445896, RRID:AB_2894917, Roche Diagnostics, measurement range 0.04-5000 ng/mL), and TgAb (Elecsys Anti-Tg Kit; Roche Cat# 07026919, RRID:AB_28949220, Roche Diagnostics, measurement range 10.0-4000 IU/mL). The stimulated Tg was defined as the Tg measured at TSH > 30 µIU/mL following thyroid hormone withdrawal after total thyroidectomy. Stimulated Tg was measured before the first RAIT treatment, while unstimulated Tg was measured 1 to 2 months after total thyroidectomy. TSH, Tg, and TgAb were tested every 6 months to 1 year during the follow-up period.

Radioactive Iodine Administration

RAIT was initiated 1 to 6 months after surgery. After maintaining a low-iodine diet for 2 weeks and implementing the withdrawal from levothyroxine for 2 to 3 weeks to ensure that TSH levels were greater than 30 μIU/mL in all the patients, we empirically adopted 3.7 GBq of ¹³¹I for all included patients. Planar anterior and posterior ¹³¹I- whole-body scan were obtained 3 to 5 days after the administration of ¹³¹I, and single-photon emission computed tomography/computed tomography was performed, if necessary. The imaging results were interpreted by 2 nuclear medicine physicians.

Primary Endpoint

The primary endpoint was recurrence-free survival (RFS) between the RAI and non-RAI groups after PSM. Structural recurrence was defined as the new biopsy-confirmed PTC metastasis (cytology or histology) occurring after 6 months postoperatively. Biochemical recurrence was defined as unstimulated Tg > 1 ng/mL during the follow-up and without evidence of structural disease. RFS_structural was defined as the time interval between the total thyroidectomy and the occurrence of structural recurrence. RFS_biochemical was defined as the time interval between the total thyroidectomy and the occurrence of biochemical recurrence.

Statistical Analysis

All the statistical analyses were performed using SPSS software (version 22.0 IBM Corp., Armonk, NY, USA). The continuous variables are presented as mean with SD, and the categorical variables are presented as numbers with percentages. The Student’s t-test and Pearson’s chi-square test or Fisher’s exact test were used to comparing the continuous and categorical variables, respectively. Univariate analysis used Pearson’s chi-square test to screen for factors associated with disease recurrence, and parameters that were statistically different in the univariate analysis were included in the Cox regression analysis for multivariate analysis to validate the predictors of recurrence. RFS was compared using the Kaplan–Meier survival analysis with a log-rank test. Two-tailed P < .05 was considered to be statistically significant.

We used PSM to eliminate possible confounding factors and selection bias with SPSS 22.0 (IBM Corp, Armonk, NY, USA) embedded with the PSM plug-in (IBM Corporation, Armonk, NY, USA). PSM analysis was performed with 1:4 matching and a caliper value of 0.02 using the following clinicopathological characteristics: sex, age, multifocality, bilateral tumor, ETE, number of LNM, a ratio of LNM, primary tumor size, N stage, and American Joint Committee on Cancer stage.

Results

Study Cohort

From August 2009 to June 2020, 4263 consecutive patients with PTC with total thyroidectomies were referred to our department. Of these, 2625 patients were classified as intermediate risk according to ATA guidelines. A total of 1138 patients were excluded based on exclusion criteria, and the remaining 1487 PTC constituted our study cohort (Fig. 1).

Figure 1.

Flow chart of patient selection. PTC, papillary thyroid cancer; TgAb, thyroglobulin antibody; sTg, stimulated thyroglobulin; uTg, unstimulated thyroglobulin; SR, structural recurrence; BR, biochemical recurrence.

A total of 1487 intermediate-risk patients formed the cohort of this study, with a median age of 41 years (range 18-75 years). Of these, 979 (65.8%) were female, and 508 (34.2%) were male. Most patients were classified as stage I (93.0%, 1383/1487), while 104 (7.0%) were classified as stage II. The size of the primary tumor ranged from 0.1 to 8.0 cm, with 2.4% (36/1487) of patients having a larger size (≥4 cm). Microscopic ETE was found in 23.8% (354/1487) of patients. The ratio of LNM had a wide range of 0 to 100%, with a median value of 44.0% (Table 1).

Table 1.

Comparison of clinicopathological characteristics before and after propensity score matching

Parameters	Total n=1487	Before PSM			After PSM
		TT + RAI n = 1349	TT alone n = 138	P value	TT + RAI n = 552	TT alone n = 138	P value
Age, years	41 (18-75)			.001			.890
<55	1288 (86.6%)	1181 (87.5%)	107 (77.5%)		431 (78.1%)	107 (77.5%)
≥55	199 (13.4%)	168 (12.5%)	31 (22.5%)		121 (21.9%)	31 (22.5%)
Sex				.727			.518
Male	508 (34.2%)	459 (34.0%)	49 (35.5%)		180 (32.6%)	49 (35.5%)
Female	979 (65.8%)	890 (66.0%)	89 (64.5%)		372 (67.4%)	89 (64.5%)
Number of LNM	3 (0-35)			.010			.611
<5	996 (67.0%)	890 (66.0%)	106 (76.8%)		435 (78.8%)	106 (76.8%)
≥5 or xa	491 (33.0%)	459 (34.0%)	32 (23.2%)		117 (21.2%)	32 (23.2%)
Ratio of LNM(%)	44.1 (0-100)			.037			1
<50	1061 (71.4%)	952 (70.6%)	109 (79.0%)		436 (79.0%)	109 (79.0%)
≥50	426 (28.6%)	397 (29.4%)	29 (21.0%)		116 (21.0%)	29 (21.0%)
ETE				<.001			.885
Negative	1133 (76.2%)	1061 (78.7%)	72 (52.2%)		288 (52.2%)	72 (52.2%)
Positive	354 (23.8%)	288 (21.3%)	66 (47.8%)		264 (47.8%)	66 (47.8%)
Multifocality				.100			1
Negative	995 (66.9%)	894 (66.3%)	101 (73.2%)		404 (73.2%)	101 (73.2%)
Positive	492 (33.1%)	455 (33.7%)	37 (26.8%)		148 (26.8%)	37 (26.8%)
Bilateral tumor				.206			.420
Negative	1108 (74.5%)	999 (74.1%)	109 (79.0%)		418 (75.7%)	109 (79.0%)
Positive	379 (25.5%)	350 (25.9%)	29 (21.0%)		134 (24.3%)	29 (21.0%)
Primary tumor size				.843			1
≤4 cm	1451 (97.6%)	1316 (97.6%)	135 (97.8%)		541 (98.0%)	135 (97.8%)
>4 cm	36 (2.4%)	33 (2.4%)	3 (2.2%)		11 (2.0%)	3 (2.2%)
N stage				.006			.860
N0 + N1a	1169 (78.6%)	1048 (77.7%)	121 (87.7%)		481 (87.1%)	121 (87.7%)
N1b	318 (21.4%)	301 (22.3%)	17 (12.3%)		71 (12.9%)	17 (12.3%)
Stageb				.001			1
I	1383 (93.0%)	1245 (92.3%)	138 (100%)		550 (99.6%)	138 (100%)
II	104 (7.0%)	104 (7.7%)	0 (0%)		2 (0.4%)	0

Abbreviations: ETE, extra-thyroidal extension; LNM, lymph node metastasis; TT, total thyroidectomy.

Indicates that information about that characteristic was not available.

TNM staging was determined by 8th American Joint Cancer Committee TNM stage system.

Comparison of Clinicopathological Characteristics Before and After PSM

Table 1 shows the baseline clinicopathological characteristics of the patients before and after PSM. A total of 1349 (90.7%) underwent RAIT after a total thyroidectomy, while 138 (9.3%) did not. Compared with the RAIT group, patients who did not receive RAIT were characterized by older age (P = .001), less N1b disease (P = .006), lower stage (P = .001), fewer numbers of LNM (P = .010), and lower proportions of LNM (P = .037).

Six hundred and ninety patients were matched successfully by the PSM analysis (1:4 matching), including 138 patients without RAIT and 552 patients with RAIT. There were no significant differences in the clinicopathological characteristics between the matched groups (all P > .05, Table 1).

Clinical Outcome

The median follow-up duration for the enrolled patients with intermediate risk PTC was 51 months (range 12-142 months). Overall recurrence was observed in 30 (2.0%) patients, including 11 structural recurrences and 19 biochemical recurrences, and the median RFS was 52 months, with a range of 10 to 142 months. Before PSM, the overall recurrence rates were 1.4% (19/1349) and 8.0% (11/138) in the RAIT and non-RAIT groups, respectively (χ² = 27.275, P < .001). After PSM, the overall recurrence rates were 1.6% (9/552) and 7.9% (11/138) in the RAIT and non-RAIT groups, respectively (χ² = 15.770, P < .001).

Impact of RAIT on Structural Recurrence in Patients With Intermediate-Risk PTC

Structural recurrence was observed in 11 patients (0.7%), all of which were in the cervical lymph nodes. Table 2 shows the characteristics of the patients with structural recurrence. The structural recurrence rate was 0.4% (6/1349) in the RAIT group compared with 3.6% (5/138) in the non-RAIT groups (χ² = 17.224, P < .001). As shown in Kaplan–Meier curves, the cumulative RFS_structural rate in the non-RAIT group was 53.8% at 5 years, lower than that in the RAIT group (98.1% at 5 years) (P < .001, Fig. 2A). Ten patients presented with structural recurrence in the cohort after calibration using PSM. The structural recurrence rate was 0.9% (5/552) in the RAIT group, and 3.6% (5/138) in the non-RAIT group (χ² = 3.964, P = .046), and the 5-year RFS_structural rates were 98.8% and 53.8% in the RAIT and non-RAIT groups, respectively (P = .007, Fig. 2B).

Figure 2.

Kaplan–Meier curves for RFS_structural of patients with intermediate-risk PTC with unstimulated Tg ≤ 1 ng/mL or stimulated Tg ≤ 10 ng/mL according to RAIT before (A) and after (B) propensity score matching. RFS_structural, structural recurrence-free survival; PTC, papillary thyroid cancer; RAIT, radioactive iodine therapy.

Table 2.

Characteristics of patients with structural recurrence

Patients	RAIT	Age	Sex	Stage	LNM 1st	Tg 1st	Tg 2nd	LNM 2nd	Recurrence site	RFS (month)
1	No	41	Female	I	16.7% (1/6)	0.08	0.30	13.3% (2/15)	CLN	60
2	No	39	Female	I	75.0% (6/8)	0.2	0.65	25.6% (11/43)	CLN	10
3	No	29	Male	I	0	0.2	0.85	2.4% (1/42)	CLN	50
4	No	27	Female	I	11.1% (1/9)	4.2(s-Tg)	0.85	20% (1/5)	CLN	25
5	No	40	Female	I	16.7% (2/12)	0.04	<0.04	13.3 (2/15)	CLN	30
6	Yes	61	Male	II	26.1% (6/23)	2.27(s-Tg)	4.71	3.2% (1/31)	CLN	55
7	Yes	51	Male	I	8.3% (3/36)	3.95(s-Tg)	NA	NA	CLN	81
8	Yes	47	Female	I	50.0% (1/2)	4.2(s-Tg)	<0.04	5% (1/20)	CLN	30
9	Yes	70	Female	II	33.3% (2/6)	7.1(s-Tg)	1.07	NA	CLN	39
10	Yes	25	Female	I	10.5% (2/19)	4.24(s-Tg)	<0.04	NA	CLN	36
11	Yes	18	Female	I	40.9% (9/22)	6.75(s-Tg)	0.20	20% (1/5)	CLN	12

Abbreviations: CLN, cervical lymph node; LNM 1st, lymph node metastasis at the time of total thyroidectomy; LNM 2nd, lymph node metastasis in lymph node dissection after the appearance of recurrence; NA, information about that characteristic was not available; RAIT, radioiodine therapy; RFS, recurrence-free survival; s-Tg, stimulated thyroglobulin; Tg 1st, thyroglobulin after total thyroidectomy; Tg 2nd, thyroglobulin levels at the time of recurrence were found.

Univariate and Cox regression analyses revealed that the absence of RAIT was a risk factor for structural recurrence, both before and after PSM (Table 3). After PSM, univariate analysis revealed that ETE (P = .006), primary tumor size (P < .001), stage (P < .001), and RAIT (P = .046) were associated with structural recurrence. The Cox regression analysis identified that non-RAIT (hazard ratio [HR] 10.572, 95% CI 2.439-45.843, P = .002), primary tumor size > 4 cm (HR 5.787, 95% CI 1.114-30.056, P = .037), and stage II (HR 29.764, 95% CI 3.718-238.290, P = .001) were risk factors for disease recurrence, while ETE was not a risk factor for biochemical recurrence (P = .932).

Table 3.

Univariate and multivariate analyses of risk factors for structural recurrence

Parameters	Before PSM				After PSM
	NED	SR	χ2	Cox regression HR (95% CI, P value)	NED	SR	χ2	Cox regression HR (95% CI, P value)
Age, years			0.639				0.589
<55	1279	9			529	9
≥55	197	2			151	1
Sex			0.629				0.225
Male	505	3			226	3
Female	971	8			454	7
Number of LNM			0.684				0.792
<5	988	8			534	7
x or ≥5	488	3			146	3
Ratio of LNM			0.441				0.394
<50%	1052	9			537	8
≥50%	424	2			143	2
ETE			0.250				0.006
Negative	1123	10			350	10		1
Positive	353	1			330	0		−(P = .932)
Multifocality			0.817				0.819
Negative	988	7			498	7
Positive	488	4			182	3
Bilateral tumor			0.210				0.518
Negative	1098	10			518	9
Positive	378	1			162	1
Tumor size			0.001				<0.001
≤4 cm	1442	9		1	668	8		1
>4 cm	34	2		6.71 (1.42-31.64, P = .016)	12	2		5.79 (1.11-30.06, P = .037)
N stage			0.051				0.227
N0/N1a	1163	6			594	8
N1b	313	5			86	2
Stage			0.784				<0.001
I	1373	10			680	8		1
II	103	1			0	2		29.76(3.72-238.29, P = .001)
Treatment			<0.001				0.046
TT + RAI	1343	6		1	547	5		1
TT alone	133	5		22.18(6.01-81.91, P < .001)	133	5		10.57 (2.44-45.83, P = .002)

Abbreviations: ETE, extrathyroidal extension; HR, hazard ratio; LNM, lymph node metastasis; NED, no evidence of disease; PSM, propensity score matching; RAI, radioactive iodine; SR, structural recurrence; TT, total thyroidectomy.

Impact of RAIT on Biochemical Recurrence in Patients With Intermediate-Risk PTC

Biochemical recurrence was observed in 19 patients (1.4%). At the last follow-up, patients with biochemical recurrence had a median unstimulated Tg of 1.96 ng/mL (range 1.06-22.21 ng/mL). Before PSM, the biochemical recurrence rates were 1.0% (13/1349) and 4.3% (6/138) in the RAIT and non-RAIT groups, respectively (χ² = 8.842, P = .003); the Kaplan–Meier curves showed longer RFS_biochemical in the RAIT group than in the non-RAIT group (5-year RFS_biochemical: 98.3% vs 63.6%, P < .001, Fig. 3A). Ten (1.9%) patients experienced biochemical recurrence in the cohort after calibration with PSM. The recurrence rates were 0.7% (4/552) and 4.3% (6/138) in the RAIT and non-RAIT groups, respectively (χ² = 7.769, P = .005); the Kaplan–Meier curves showed that the 5-year RFS_biochemical was 98.9% and 63.6% in the RAIT and non-RAIT groups, respectively (P < .001, Fig. 3B).

Figure 3.

Kaplan–Meier curves for RFS_biochemical of intermediate-risk PTC patients with unstimulated Tg ≤ 1 ng/mL or stimulated Tg ≤ 10 ng/mL according to RAIT before (A) and after (B) the propensity score matching. RFS_biochemical, biochemical recurrence-free survival; PTC, papillary thyroid cancer; RAIT, radioactive iodine therapy.

Univariate and Cox regression analyses revealed that the absence of RAIT was a risk factor for biochemical recurrence, both before and after PSM (Table 4). After PSM, univariate analysis revealed that the numbers of LNM (P = .010), primary tumor size (P = .016), N stage (P = .002), and RAIT (P = .005) were associated with biochemical recurrence. Cox regression indicated that only RAIT (HR 16.568, 95% CI 3.670-74.803, P <.001) was the independent predictor of biochemical recurrence.

Table 4.

Univariate and multivariate analyses of risk factors for biochemical recurrence

Parameters	Before PSM				After PSM
	NED	BR	χ2	Cox regression HR (95% CI, P value)	NED	BR	χ2	Cox regression HR (95% CI, P value)
Age, years			0.756				0.876
<55	1272	16			530	8
≥55	196	3			150	2
Sex			0.804				0.829
Male	501	7			226	3
Female	967	12			454	7
Number of LNM			0.020				0.010
<5	988	8		1	537	4		1
x or ≥5	480	11		2.85 (0.99-8.17, P = .051)	143	6		4.59(0.87-24.18, P = .073)
Ratio of LNM			0.212				0.755
<50%	1045	16			538	7
≥50%	423	3			142	3
ETE			0.801				0.146
Negative	1118	15			352	8
Positive	350	4			328	2
Multifocality			0.726				0.896
Negative	983	12			497	8
Positive	485	7			183	2
Bilateral tumor			0.477				0.786
Negative	1092	16			519	8
Positive	376	3			161	2
Tumor size			0.010				0.016
≤4 cm	1435	16		1	668	8		1
>4 cm	33	3		3.83 (1.04-14.09, P = .043)	12	2		4.03 (0.67-24.47, P = .130)
N stage			0.002				0.002
N0/N1a	1160	9		1	597	5		1
N1b	308	10		2.29 (0.79-6.68, P = .128)	83	5		2.68 (0.51-14.077, P = .244)
Stage			0.766				0.864
I	1365	18			678	10
II	103	1			2	0
Treatment			0.003				0.005
TT + RAI	1336	13		1	548	4		1
TT alone	132	6		14.31 (4.96-41.29, P < .001)	132	6		16.57 (3.67-74.80, P < .001)

Abbreviation: BR, biochemical recurrence; HR, hazard ratio; PSM, propensity score matching; NED; no evidence of disease; RAI, radioactive iodine; TT, total thyroidectomy.

Discussion

PTC is the most common pathological type of DTC, with a fair prognosis and a 10-year survival rate of >90% (21). However, some patients develop local recurrence or distant metastases, which may lead to patient death (22-24). The recurrence risk of intermediate-risk PTC patients ranges from 5% to 20% (5). In recent years, there have been controversies regarding RAIT decisions for intermediate-risk patients, especially for the relatively low recurrence subgroup (6-11).

Several studies on public databases suggest that RAIT improves survival in intermediate-risk patients (6, 7). Ruel et al (6) analyzed the data on intermediate-risk patients from the National Cancer Database and showed that RAIT decreased mortality risk by 29% and 36% in the subgroup aged <45 years. Thus, they suggested that RAIT can improve the survival of intermediate-risk patients and should be considered in such patients. A recent retrospective study (7) that analyzed the SEER database showed that RAIT significantly decreased disease-specific mortality in patients with intermediate-risk PTC and that male patients aged ≥45 years and with tumor size >20 mm may benefit from RAIT. Although the effect of RAIT on disease recurrence has not been investigated in these studies, we suppose that our results support the above studies to some extent due to death occurring from disease persistence or recurrence. In our study, patients who did not receive RAIT had an approximately 10-fold and 16-fold increased risk of structural and biochemical recurrence, respectively. Structural recurrence was found in 0.7% of patients, and all recurrence lesions were in the cervical lymph nodes, indicating a better prognosis for patients with intermediate-risk PTC with low Tg levels. Even with such a low recurrence rate, the regression analysis after adjusting for confounders confirmed that RAIT significantly decreased the risk of structural recurrence and improved RFS_structural in intermediate-risk patients with PTC. RAIT may help intermediate-risk patients with low Tg levels avoid secondary surgery. However, it is worthwhile considering whether RAIT should be withheld in patients willing to undergo reoperation and concerned about the financial burden and side effects of RAIT because the recurrence rate of this group was relatively low, rarely leading to death or distant metastasis.

When referring to the studies that have indicated RAIT lacks value for intermediate-risk patients, it is noteworthy that Kim et al (8) enrolled more than 8297 patients with intermediate-risk PTC, 7483 of whom underwent RAIT. The results showed that RAIT did not decrease the risk of locoregional recurrence (HR 0.852, P = .413), even in patients with aggressive features such as BRAF^V600E mutation, tumor size >1 cm and ETE. However, this study did not consider Tg levels, and many baseline characteristics differed between groups. In other studies, RAIT also failed to improve recurrence, but the sample size was small, and there was selection bias in selecting patients (9-11). A retrospective study (25) showed that RAIT did not improve RFS in intermediate-risk patients with Tg < 1 ng/mL and that the baseline characteristics differed between the groups. Moreover, the risk stratification system used in that study was the GAMES criteria, which was designed to predict the risk of death rather than the risk of recurrence.

All patients in our cohort underwent total thyroidectomy and therapeutic or prophylactic CND. A recent study with a large sample size confirmed that prophylactic CND does not decrease locoregional recurrence in patients (26). However, prophylactic CND identified 28% to 33% of LNM that could not be detected on preoperative examination, thus altering the patients’ postoperative staging and management (27, 28). Therefore, all patients undergoing total thyroidectomy in our center received prophylactic CND, which is not common practice in the rest of the world. After total thyroidectomy, multiple factors influence whether a patient receives RAIT. Intermediate-risk patients with younger ages (≤55 years), higher numbers, and a higher ratio of LNM were more likely to be advised to receive RAI adjuvant therapy at our center. This result can be explained as clinicians are more likely to recommend RAIT for intermediate-risk patients with such aggressive features. Besides, a patient’s financial situation, level of education, and perception of the disease may also influence a patient's choice of RAIT.

Patients with low Tg levels have been shown to have an excellent prognosis, especially those with stimulated Tg ≤ 10 ng/mL or unstimulated Tg ≤ 1 ng/mL (12, 14, 29). A total of approximately 4000 patients with DTC included in the study by Webb et al (12) indicated that stimulated Tg < 10 ng/mL was a predictor of disease remission, with a negative predictive value of 94.2%. Vaisman et al (13) compared the clinical outcomes in patients with different Tg levels, and they concluded that it is appropriate to categorize stimulated Tg > 10 ng/mL or unstimulated Tg > 1 ng/mL as an incomplete biochemical response in a dynamic risk assessment stratification. A recent study demonstrated that a stimulated Tg level was an independent predictor of recurrence and that the risk of recurrence was 2.48-fold higher in intermediate-risk patients with stimulated Tg > 10 ng/mL than in other patients (30). As mentioned above, intermediate-risk patients with low Tg levels present a low recurrence rate (12, 14, 26, 27); therefore, the need for RAIT in intermediate-risk PTC patients with low Tg levels is more controversial than in those with high Tg levels. Thus, our study focused on determining whether RAIT helps decrease recurrence in intermediate-risk patients with unstimulated Tg ≤ 1 ng/mL or stimulated Tg ≤ 10 ng/mL.

Our study had several limitations. Firstly, this study was retrospective; we could not exclude selection bias resulting from physician and patient preferences, even though we performed adjustments using the PSM method. Secondly, due to the small number of recurrences observed in the study, subgroup analysis could not be performed; which subgroup that may benefit from RAIT still needs to be identified. Thirdly, since the available studies support that patients with elevated Tg after total thyroidectomy have a higher risk of recurrence, most patients with stimulated Tg > 10 ng/mL and unstimulated Tg > 1 ng/mL underwent RAIT at their doctor’s recommendation in our center, which makes it difficult to compare the recurrence rates between the RAIT and the non-RAIT groups in this subgroup. Fourthly, Tg and TgAb were not comparable between the RAIT and non-RAIT groups since these biochemical indicators were still under the influence of residual thyroid in the non-RAIT group. Besides, the cut-off value for classifying elevated TgAb as a biochemical disease remains ambiguous (5); thus, patients with increased TgAb were not considered as biochemical recurrence.

Conclusion

Intermediate-risk PTC with unstimulated Tg ≤ 1 ng/mL or stimulated Tg ≤ 10 ng/mL exhibits low risk of relapse. RAIT decreases the risk of structural and biochemical recurrence in patients with intermediate-risk PTC with low Tg levels. Further prospective, randomized studies are needed to confirm these findings and to identify the subclinical characteristics of patients who should receive RAIT.

Funding

This study was supported by the National Natural Science Foundation of China (no. 81972502).

Conflict of Interest

All the authors declare that they have no conflict of interests.

Data Availability

Some or all datasets generated and/or analyzed during the current study are not publicly available but are available from the corresponding author upon reasonable request. Restrictions apply to the availability of some or all data generated or analyzed during this study to preserve patient confidentiality or because they were used under license. The corresponding author will on request detail the restrictions and any conditions under which access to some data may be provided.

Abbreviations

 
• ATA

  American Thyroid Association

 
• CND

  central-compartment neck dissection

 
• DTC

  differentiated thyroid cancer

 
• ETE

  extrathyroidal extension

 
• HR

  hazard ratio

 
• LNM

  lymph node metastasis

 
• PSM

  propensity score matching

 
• PTC

  papillary thyroid cancer

 
• RAIT

  radioactive iodine therapy

 
• RFS

  recurrence-free survival

 
• Tg

  thyroglobulin

 
• TgAb

  thyroglobulin antibody

 
• TSH

  thyrotropin

Author notes

Tian Tian and Zhibing Qi contributed equally to the present study.