The Prognostic Impact of Radioiodine Therapy in Patients with Papillary Thyroid Cancer

Abstract

Radioiodine (RAI) therapy after surgery, is an important component for the treatment of patients with papillary thyroid cancer (PTC), the most common thyroid cancer. In this study we sought to evaluate the cancer-specific survival (CSS) impact of RAI in specific thyroid cancer subgroups. The Surveillance, Epidemiology, and End Results (SEER) database were used to identify patients with PTC who underwent surgery between 2000 and 2019. Patients not treated with RAI were compared to those treated with RAI using propensity score matching (PSM) on the basis of identical inclusion criteria. A total of 106 333 patients were identified from the SEER database. RAI therapy was associated with improved CSS in the matched cohort (HR: 0.83; 95% CI: 0.72–0.96, p=0.01) but not in the unmatched data set (HR: 1.46; 95% CI: 1.30–1.64, p<0.001) among all PTC patients regardless of disease stage. Detailed analyses, however, showed that only patients with high-risk disease (pT3N1, pT4N1) experienced the greatest benefit in CSS. In the lower disease stages, no significant differences were recognized in the group of PTC patients with or without RAI therapy. One exception: in the group of PTC patients in stage pT1bN0, a significant difference was seen towards RAI. This is, however, most likely due to the large number of patients investigated. In summary, RAI therapy should not be used in low-risk PTC patients and might be used to some extent in intermediate-risk PTC patients. The histological suptype of the tumor needs to be considered in this context.

 

Scroll to top

Keywords

papillary thyroid cancer - radioiodine therapy - cancer-specific mortality - SEER database

Introduction

Thyroid cancer is the most common new malignancy diagnosed in Europe and the United States. It also carries an expanding global disease burden, in part due to the increased detection of tumors that may otherwise remain asymptomatic and not impact mortality [1]. The vast majority (more than 90%) of these cases are well-differentiated thyroid carcinomas (DTC), consisting of papillary and follicular histology, which carry an overall excellent prognosis, particularly in early-stage disease. With appropriate treatment, 5-year survival exceeds 95% [2]. Surgery is often the first step in successful therapy for thyroid cancer. In the past, nearly all patients received adjuvant radioiodine (RAI) therapy. However, concerns about a possible carcinogenic effect have also been raised since the introduction of RAI therapy. An increased incidence of secondary primary malignancies (SPMs) in patients with DTC, possibly related to disease-specific therapy or genetic predisposition, was found in a meta-analysis of 13 studies [3]. In this regard, we have recently found that thyroid cancer appears to be positively associated with the risk of developing a subsequent lymphoma [4]. In addition, the occurrence of SPM in DTC may be associated with the use of RAI, as suggested by the results of a previous meta-analysis including two multicenter studies [5]. Accordingly, guidelines have recently shifted toward a more individual and stratified approach. Today, only selected cases of DTC warrant RAI therapy, typically indicated for thyroid remnant ablation, adjuvant therapy, or treatment of known persistent disease. In the 2015 American Thyroid Association guidelines for the management of thyroid cancer, RAI adjuvant therapy is routinely recommended after total thyroidectomy for patients with high-risk disease and should be considered for intermediate-risk disease [6]. It is not recommended for patients with low-risk disease [6]. The evidence supporting RAI therapy in high-risk patient subgroups is relatively undisputed [7]. In the intermediate-risk DTC subgroups, there is controversy surrounding appropriate use. A growing body of literature indicates that the risk/benefit ratio is in favor of avoiding RAI therapy in low-risk populations. Most recently, the European Thyroid Association (ETA) published a consensus statement regarding post-surgical radioiodine therapy in DTC [8]. The recommendations are mostly in line with the recommendations of the 2015 American Thyroid Association guidelines. The ETA recommendations were, however, critically questioned by other authors [9]. A new German guideline for the treatment of DTC patients is now in preparation and will be published in the near future.

Thus, the aim of the present propensity score matched case-control study was to evaluate the impact of RAI therapy on cancer-specific mortality in a large cohort of patients with surgically resected papillary thyroid cancer (PTC) using real-world data from the Surveillance, Epidemiology, and End Results (SEER) database.

 

Scroll to top

Materials and Methods

Study cohort

The study was conducted retrospectively as case-control study by retrieving data on PTC from the SEER-17 registries of the SEER program of the National Institutes of Health (NIH) released in November 2021 using the SEER*stat software, version 8.4.0.1 [10]. The SEER database is a publicly available, federally funded cancer reporting system that is a collaboration between the US Centers for Disease Control and Prevention, the National Cancer Institute, and regional and state cancer registries [11]. The SEER-17 registries, which include cancer patients with diagnoses between 2000 and 2019, were utilized to identify patients with a first diagnosis of thyroid cancer (C73.9). PTC cases were identified using ICD-O-3 codes 8050/3, 8260/3, 8340–8344/3, 8350, and 8450–8460/3. The following states are included in the SEER-17 registries: Alaska Native Tumor Registry, Connecticut, Atlanta, Greater Georgia, Rural Georgia, San Francisco-Oakland, San Jose-Monterey, Greater California, Hawaii, Iowa, Kentucky, Los Angeles, Louisiana, New Mexico, New Jersey, Seattle-Puget Sound, and Utah. Data collection was conducted on May 16, 2022. Patients were excluded if they were younger than 18 years of age at the time of initial diagnosis, had no histopathologic confirmation of diagnosis or incomplete TNM classification, were diagnosed before the introduction of the 6th edition of the UICC/AJCC TNM classification in 2004, had unknown cause of death (COD), had received any form of radiotherapy other than RAI therapy, and had unclear or unknown surgical therapy, or no total thyroidectomy ([Fig. 1]).

Fig. 1 Patient selection. COD: Cause of death, NOS: Not otherwise specified, RAI: radioiodine.

 

Scroll to top

Primary outcome and statistical analysis

The present study was aimed to investigate the impact of radioiodine therapy on cancer-specific survival (CSS) in patients with PTC. Missing values were imputed using the multivariate imputation by chained equations (MICE) algorithm [12]. It should be noted that only variables in which less than 20% of the data were missing were included in the study. With the exception of the variable “focality”, which exhibited a 14.15% rate of missing data, all other variables were complete. Consequently, the replacement of the missing variables was limited to the variable “focality”. Patients with PTC who did not receive RAI therapy were compared with patients who underwent RAI therapy based on identical inclusion criteria using propensity score matching (PSM). The matching parameters included age, gender, race, first malignant primary disease, histologic variant, focality, extent of the primary tumor (T), lymph node metastasis (N), and distant metastasis (M). In this approach, a 1:1 nearest neighbor propensity score matching without replacement was performed. Propensity scores were estimated using logistic regression of treatment on covariates. After matching, all standardized mean differences for the covariates were found to be less than 0.1, confirming adequate balance ([Fig. 2]).

Fig. 2 Love plot displays the covariate balance before and after 1:1 propensity score matching. It depicts the standardized mean differences before and after 1:1 matching. Purple circles indicate the standardized mean differences before matching, cyan circles indicate the differences after matching. The dotted line represents a standardized mean difference of 0.1.

CSS, with and without RAI, were analyzed using Kaplan–Meier curves. A comparison of these curves was analyzed using the log-rank test. Additionally, a univariate Cox regression analysis was performed to identify associations between RAI and CSS. The results of the Cox regression model were expressed as hazard ratio (HR) with 95% confidence interval (CI). Separate Kaplan–Meier curves and Cox regression analyses were performed for focality, histologic variants, T category, and metastatic burden. A p-value of less than 0.05 was considered statistically significant. Lifetime tables were used to determine the 5- and 10-year CSS rates. The statistical analyses were performed using the software R (version 1.4.1106), with the packages readxl, mice, matchit, cobalt, survival, and survminer [13] [14] [15] [16] [17] [18] [19].

 

Scroll to top

 

Scroll to top

Results

RAI usage

In total, 51.08% (n=54 320) of patients with PTC underwent RAI therapy ([Table 1]). As illustrated in [Fig. 2] using a balance plot and [Table 1], patients in the unmatched cohort who received RAI were younger (average 48.11±14.70 years) compared to patients who did not receive RAI treatment (50.51±14.75 years). Furthermore, we observed a higher prevalence of multifocal and lymphonodular metastatic PTCs as well as T2 and T3 tumors in the group of patients who received RAI prior to PSM. In contrast, only 20.68% of patients with microcarcinomas (T1aN0M0) underwent RAI therapy.

Table 1 Patient characteristics stratified by RAI.

	Before PSM					After PSM (1:1)
Variable	No RAI		RAI		SMD	No RAI		RAI		SMD
	n=52.013	(%) 48.92	n=54.320	(%) 51.08		n=33.232	(%) 50.00	n=33.232	(%) 50.00
Age (years)
Mean (SD)	50.51	14.75	48.11	14.70	− 0.1629	49.24	15.04	48.75	14.81	− 0.0335
Sex
Female	41.768	39.28	40.408	38.00	− 0.1355	25.845	38.89	24.435	36.76	− 0.0972
Male	10.245	9.63	13.912	13.08	0.1355	7.387	11.11	8.797	13.24	0.0972
Race
Asian or Pacific Islander	5.187	4.88	6.743	6.34	0.0740	3.730	5.61	3.626	5.46	− 0.0095
Black	3.442	3.24	2.569	2.42	− 0.0890	1.797	2.70	2.078	3.13	0.0400
White	42.266	39.75	44.157	41.53	0.0008	27.102	40.78	26.808	40.33	− 0.0227
Others	1.118	1.05	851	0.80	− 0.0469	604	0.91	720	1.08	0.0281
First malignant primary
No	6.385	6.00	5.295	4.98	− 0.0852	3.935	5.92	4.254	6.40	0.0324
Yes	45.628	42.91	49.025	46.11	0.0852	29.297	44.08	28.978	43.60	− 0.0324
Primary tumor
T1a	26.743	25.15	9.807	9.22	− 0.8674	8.091	12.17	9.105	13.70	0.0793
T1b	12.418	11.68	14.794	13.91	0.0755	12.290	18.49	10.404	15.65	− 0.1275
T2	6.386	6.01	11.358	10.68	0.2123	6.385	9.61	6.637	9.99	0.0186
T3	5.654	5.32	15.840	14.90	0.4024	5.654	8.51	5.814	8.75	0.0106
T4	812	0.76	2.521	2.37	0.1464	812	1.22	1.272	1.91	0.0658
Regional lymph node
N0	45.063	42.38	32.876	30.92	− 0.5343	26.290	39.56	24.988	37.60	− 0.0802
N1	6.950	6.54	21.444	20.17	0.5343	6.942	10.44	8.244	12.40	0.0802
Distant metastasis
M0	51.805	48.72	53.664	50.47	− 0.0740	33.030	49.70	32.850	49.43	− 0.0496
M1	208	0.20	656	0.62	0.0740	202	0.30	382	0.57	0.0496
Focality
Solitary	32.135	30.22	25.377	23.87	− 0.3020	17.297	26.02	17.072	25.69	− 0.0136
Multifocal	19.878	18.69	28.943	27.22	0.3020	15.935	23.98	16.160	24.31	0.0136
Histological Subtype
Papillary	35.879	33.74	36.799	34.61	− 0.0264	22.243	33.47	21.420	32.23	− 0.0530
Follicular	14.884	14.00	15.664	14.73	0.0049	10.040	15.11	10.429	15.69	0.0258
Encapsulated	584	0.55	536	0.50	− 0.0138	415	0.62	442	0.67	0.0082
Columnar	459	0.43	1.003	0.94	0.0716	385	0.58	711	1.07	0.0729
Oxyphilic	115	0.11	122	0.11	0.0007	87	0.13	104	0.16	0.0108
Sclerosing	92	0.09	196	0.18	0.0307	62	0.09	126	0.19	0.0321

Abbreviation: RAI, radioiodine; NOS, not otherwise specified.

 

Scroll to top

Cancer-specific mortality

The overall mortality rate in the SEER cohort observed over a ten-year period was 8.9%, while the tumor-specific mortality rate during this period was only 1.7%. Increasing thyroid cancer-specific mortality after 10 years of follow-up was seen with more advanced T category (T1a 0.6%, T1b 0.6%, T2 1.1%, T3 3.2%, and T4 16.4%, p<0.001) ([Fig. 3a]). A similar trend was seen with N and M category (N0 1.1%, N1a 2.1%, N1b 6.1%, p<0.001; and M0 1.4%, M1 37.9%, p<0.001) ([Fig. 3b, c]). Patients with a biologically aggressive histological type (e. g., columnar cell variant and diffuse sclerosing variant) had a worse prognosis after 10 years of observation (mortality: 6.6%) compared to other phenotypes (mortality: 1.6%, p<0.001) ([Fig. 3d]).

Fig. 3 Cancer Specific Survival (CSS) in PTC patients depending on pathological variables. Kaplan–Meier curves for CSS of the unmatched PTC cohort according to (a) the extend of the primary tumor (T), (b) lymph node metastasis (N), (c) distant metastasis (M) and histological type (HT). Histological type was classified according to the biological behavior into aggressive variants of PTC (i. e., diffuse sclerosing variant and columnar cell variant) and other variants of PTC.

 

Scroll to top

RAI and cancer-specific mortality

The primary objective of the present study was to determine whether RAI is associated with a prognostic advantage in patients with PTC. However, due to an unbalanced distribution of the factors age, tumor focality, and extent of the primary tumor as well as lymph node metastasis between the group of patients who received RAI and those who did not, a PSM was performed to achieve an adequate balance between both groups ([Fig. 2] and [Table 1]).

In the unmatched cohort, the Cox regression analysis (p<0.001) and the log-rank test (p<0.001) demonstrated that RAI was associated with a less favorable CSS ([Table 2] and [Fig. 4a]). The difference in survival curves between the RAI-treated and untreated cohorts became apparent after approximately 8 years, with survival after 10 years being only 0.8% worse in the RAI group. However, after matching, this survival difference was reversed. Patients receiving RAI therapy had a significant (p=0.01) yet relatively modest survival benefit (10-year CSS rate: no RAI: 98.3% vs. RAI: 98.4%) in the matched cohort ([Table 2] and [Fig. 4b]). After PSM, there was no longer any difference in the RAI-dependent survival curves in patients without lymph node metastases or with solitary tumors ([Table 2], [Fig. 6a, e]), which was seen before PSM for some subgroups ([Fig. 5a–e and g]). In contrast to the unmatched cohort, RAI showed a significant but very small prognostic benefit in the group of PTCs with less aggressive biological behavior ([Table 2], [Fig. 5g] and [Fig. 6g]). In the aggressive PTC variants, no influence of RAI on CSS could be demonstrated either before or after matching ([Table 2], [Fig. 5h] and [Fig. 6h]). While RAI demonstrated no survival benefit in multifocal PTCs in the unmatched cohort ([Table 2] and [Fig. 5f]), this was significant after PSM in both the Cox regression analysis and the log-rank test ([Table 2] and [Fig. 6f]). However, it is important to note that the minimal difference in survival rates of 0.6% in the first 5 years, which subsequently decreases, should also be considered.

Fig. 4 Kaplan–Meier curves for Cancer Specific Survival (CSS) depending on radioiodine (RAI) therapy. Survival curves for CSS in patients who did not receive RAI (None) or who were treated with RAI in the (a) unmatched and (b) matched PTC cohort.

Fig. 6 Prognostic influence of RAI therapy in the matched cohort depending on pathological tumor characteristics. Cancer-specific survival (CSS) for patients with PTC in relation to lymph node metastasis (a) N0 or (b) N1, distant metastasis (c) M0 or (d) M1, focality (e) solitary or (f) multifocal, histological type (g) good or (h) poor biological behavior and whether a RAI therapy was performed (Yes) or not performed (None).

Fig. 5 Prognostic influence of RAI therapy in the unmatched cohort depending on pathological tumor characteristics. Cancer-specific survival (CSS) for patients with PTC in relation to lymph node metastasis (a) N0 or (b) N1, distant metastasis (c) M0 or (d) M1, focality (e) solitary or (f) multifocal, histological type (g) good or (h) poor biological behavior and whether a RAI therapy was performed (Yes) or not performed (None).

Table 2 Cox regression analysis.

	Before PSM			After PSM (1:1 matching)
	Cancer-Specific-Survival			Cancer-Specific-Survival
Group/Subgroup	HR	95% CI	p-Value	HR	95% CI	p-Value
RAI (Ref. No RAI)
All	1.46	1.30–1.64	<0.001	0.83	0.72–0.96	0.01
Distant metastasis (M)
M0	1.40	1.23–1.59	<0.001	0.81	0.69–0.95	0.008
M1	0.59	0.44–0.80	<0.001	0.49	0.35–0.69	<0.001
Lymph node metastasis (N)
N0	1.52	1.29–1.79	<0.001	1.00	0.82–1.22	0.990
N1	0.64	0.54–0.76	<0.001	0.62	0.50–0.76	<0.001
Focality
Solitary	1.92	1.63–2.27	<0.001	0.99	0.81–1.21	0.950
Multifocal	1.04	0.88–1.22	0.660	0.69	0.56–0.84	<0.001
Histologic subtype (biological behavior)
Good	1.44	1.28–1.63	<0.001	0.80	0.69–0.93	0.003
Poor	1.03	0.63–1.70	0.896	0.82	0.49–1.39	0.467
Tumor stages without distant metastasis (M0)
T1aN0	1.15	0.75–1.75	0.525	1.19	0.67–2.13	0.549
T1bN0	0.46	0.28–0.73	0.001	0.50	0.30–0.82	0.006
T2N0	1.17	0.71–1.94	0.530	1.26	0.75–2.13	0.388
T3N0	0.91	0.66–1.25	0.542	0.96	0.66–1.38	0.806
T4N0	0.78	0.50–1.22	0.277	0.74	0.46–1.19	0.216
T1aN1	1.39	0.56–3.43	0.473	1.54	0.62–3.87	0.355
T1bN1	0.69	0.30–1.60	0.384	1.03	0.40–2.67	0.950
T2N1	0.79	0.40–1.58	0.512	0.89	0.40–1.98	0.772
T3N1	0.57	0.43–0.77	<0.001	0.61	0.41–0.91	0.015
T4N1	0.49	0.36–0.66	<0.001	0.44	0.30–0.64	<0.001

A prognostic advantage of RAI was shown in both the cohort before matching and after matching in the presence of lymph node or distant metastases ([Table 2], [Fig. 5b and d], [Fig. 6b and d]). The 5-year and 10-year survival rates of patients with lymph node metastases were higher with RAI by 1.4% and 0.9% in the unmatched cohort and by 1.3% and 1.3% in the PSM cohort. In patients with distant metastases, tumor-specific 5-year survival was 82.8% with RAI therapy and 70.6% without RAI therapy. After matching, the 5-year survival rates were 87% and 69.9%, respectively.

 

Scroll to top

RAI, cancer-specific mortality, and TNM stage

Given the substantial survival benefits that could be demonstrated for patients with distant metastases, the prognostic value of RAI therapy in patients without distant metastases was subsequently investigated. Thus, we performed a detailed analysis according to the extent of the primary tumor and lymph node metastasis of all patients ([Table 2], [Fig. 7a–j] and [Fig. 8a–j]). With the exception of patients with T1bN0, T3N1, and T4N1 tumors, there were no significant differences in CSS between the RAI and non-RAI groups ([Table 2], [Fig. 7b, i and j]; [Fig. 8b, i, and j]). However, a comprehensive examination of Kaplan–Meier curves and survival tables of the T1bN0 patients revealed that the survival advantage was marginal, amounting to 0.3% (before PSM: No RAI: 99.6% CSS vs. RAI: 99.9% CSS; after PSM: No RAI: 99.6% CSS vs. RAI: 99.9% CSS) and 0.4% (before PSM: No RAI: 99.3% CSS vs. RAI: 99.7% CSS; after PSM: No RAI: 99.3% CSS vs. RAI: 99.7% CSS) for the survival rate after 5 and 10 years, respectively, before and after PSM. However, we assume that the minimal, clinically unimportant differences in the T1bN0-PTC group are statistical effects mainly due to the large number of patients in this subgroup (before PSM: n=20.505 and after PSM: n=19.223). Although Kaplan–Meier curves indicated an improvement of up to 4.9% in survival in the RAI-treated group of patients with T4N0 PTC in the first eight years after PSM, Cox regression analysis and the log-rank test demonstrated no significant difference. Importantly, however, improvement of CSS was especially seen in PTC patients with lymphonodular metastasized tumor stages, for example, T3N1 and T4N1, respectively, which were statistically significant ([Fig. 7i, j] and [Fig. 8i, j], [Table 2]).

Fig. 7 Prognostic influence of RAI therapy in the unmatched cohort depending on the tumor extend and lymph node metastasis. Cancer-specific survival (CSS) depending on the administration of a RAI in patients (a–e) without lymph node metastasis (N0) and (a) T1a, (b) T1b, (c) T2, (d) T3 or (e) T4 PTC and in patients (f–j) with lymph node metastasis (N1) and (f) T1a, (g) T1b, (h) T2, (i) T3 or (j) T4 PTC.

Fig. 8 Prognostic influence of RAI therapy in the matched cohort depending on the tumor extend and lymph node metastasis. Cancer-specific survival (CSS) depending on the administration of a RAI in patients (a–e) without lymph node metastasis (N0) and (a) T1a, (b) T1b, (c) T2, (d) T3 or (e) T4 PTC and in patients (f–j) with lymph node metastasis (N1) and (f) T1a, (g) T1b, (h) T2, (i) T3 or (j) T4 PTC.

 

Scroll to top

 

Scroll to top

Discussion

RAI therapy is a key component in the treatment of patients with DTC. Still, there is a major debate within the scientific community on the use in patients with different disease stages. This has been further fostered by the observation that the risk of developing SPM is increased in patients treated with RAI for DTC [5]. In this context, a recent nationwide population study from South Korea demonstrated a strong association with the risk of developing leukemia for RAI doses above 100 mCi, but not for lower doses of RAI [20]. Although beneficial effects of RAI are undisputed in high-risk patients, controversy remains in intermediated-risk and some low-risk patients. The aim of the present study was to evaluate the impact of a RAI therapy in patients with PTC on the cancer-specific mortality in these patients. Therefore, we conducted a large, retrospective, propensity score-matched case-control study of thyroidectomized PTC patients recruited from the SEER database. We were able to show that only patients with high-risk and lymph node metastatic disease (T3N1, and T4N1) experienced a benefit in cancer-specific mortality. In the lower disease stages, no significant differences were seen in the group of PTC patients with or without RAI therapy (with the exception of T1bN0 PTC patients).

It is important to note that, due to the retrospective nature of our study, the findings should be considered as preliminary evidence and not an establishment of quantification of treatment effect. National observational databases like SEER are useful in evaluating low-probability events (like mortality in thyroid cancer) because they provide large sample sizes. Unfortunately, they lack key information regarding the extent of surgery, RAI dose, histologic subtype, genetic mutations, and other tumor or nodal factors that may have influenced the decision to give RAI. Furthermore, controlling for multiple covariates does not guarantee an absence of confounders and collinear variables with complex interactions. Disease recurrence is another outcome not attainable from these large datasets that is particularly relevant in thyroid cancer [21].

In summary, our data clearly show an advantage for CSS in PTC patients with advanced disease stages, for example, pT3N1 and pT4N1. Based on our study results, PTC patients with lower tumor stages are unlikely to benefit from RAI therapy in terms of CSS. However, there was one exception: in the group of pT1bN0 PTC patients, there was a significant difference in favor of RAI, but this is most likely due to the large number of patients studied.

In conclusion, RAI should not be used in low-risk PTC patients and could be used to some extent in intermediate-risk PTC patients. The histologic subtype must be considered in this context.

 

Scroll to top

 

Scroll to top

Conflict of Interest

The authors declare that they have no conflict of interest.

• References

• 1 Vaccarella S, Franceschi S, Bray F. et al. Worldwide thyroid-cancer epidemic? The increasing impact of overdiagnosis. N Engl J Med 2016; 375: 614-617
• 2 Sherman SI. Thyroid carcinoma. Lancet 2003; 361: 501-511
• 3 Subramanian S, Goldstein DP, Parlea L. et al. Second primary malignancy risk in thyroid cancer survivors: a systematic review and meta-analysis. Thyroid 2007; 17: 1277-1288
• 4 Krieg A, Krieg S, Al Natour OMS. et al. Risk of lymphoma and leukemia in thyroid cancer patients: a retrospective cohort study in Germany. Horm Metab Res 2024; 56: 559-565
• 5 Sawka AM, Thabane L, Parlea L. et al. Second primary malignancy risk after radioactive iodine treatment for thyroid cancer: a systematic review and meta-analysis. Thyroid 2009; 19: 451-457
• 6 Haugen BR, Alexander EK, Bible KC. et al. 2015 American thyroid association management guidelines for adult patients with thyroid nodules and differentiated thyroid cancer: the American thyroid association guidelines task force on thyroid nodules and differentiated thyroid cancer. Thyroid 2016; 26: 1-133
• 7 Jonklaas J, Sarlis NJ, Litofsky D. et al. Outcomes of patients with differentiated thyroid carcinoma following initial therapy. Thyroid 2006; 16: 1229-1242
• 8 Pacini F, Fuhrer D, Elisei R. et al. 2022 ETA consensus statement: what are the indications for post-surgical radioiodine therapy in differentiated thyroid cancer?. Eur Thyroid J 2022; 11: e210046
• 9 Schmidt M, Bartenstein P, Bucerius J. et al. Individualized treatment of differentiated thyroid cancer: the value of surgery in combination with radioiodine imaging and therapy – a German position paper from surgery and nuclear medicine. Nuklearmedizin 2022; 61: 87-96
• 10 Surveillance, Epidemiology, and End Results (SEER) Program www.seer.cancer.gov SEER*Stat Database: Incidence – SEER Research Plus Data, 17 Registries, Nov 2021 Sub (2000–2019) – Linked to County Attributes – Total U.S., 1969–2019 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, released April 2021, based on November 2020 submission
• 11 Doll KM, Rademaker A, Sosa JA. Practical guide to surgical data sets: surveillance, epidemiology, and end results (SEER) database. JAMA Surg 2018; 153: 588-589
• 12 Zhang Z. Multiple imputation with multivariate imputation by chained equation (MICE) package. Ann Transl Med 2016; 4: 30
• 13 Wickham H, Bryy J. Read Excel Files, R Package Version 1.4.2. 2022. [(accessed on 14 November 2022)]. Available online https://CRAN.R-project.org/package=readxl 2022
• 14 Van Buuren S, Groothuis-Oudshoorn K. Mice: multivariate imputation by chained equations in R. J Stat. Softw. 2011 45. 1-67
• 15 Ho D, Imai K, King G. et al. MatchIt: nonparametric preprocessing for parametric causal iunference. J Statist Software 2011; 42: 1-28
• 16 Greifer N. Cobalt: covariate balance tables and plots. R package version 4.5.5.9000 https://github.com/ngreifer/cobalt 2024
• 17 Survival analysis. R package version 3.3-1 https://cran.r-project.org/web/packages/survival/index.html
• 18 Drawing Survival Curves using 'ggplot2'. R package version 0.4.9 https://cran.r-project.org/web/packages/survminer/survminer.pdf
• 19 Therneau TM, Grambsch PM. Modeling survival data: extending the Cox model. New York: Springer; 2000
• 20 Seo GH, Cho YY, Chung JH. et al. Increased risk of leukemia after radioactive iodine therapy in patients with thyroid cancer: a nationwide, population-based study in Korea. Thyroid 2015; 25: 927-934
• 21 Orosco RK, Hussain T, Noel JE. et al. Radioactive iodine in differentiated thyroid cancer: a national database perspective. Endocr Relat Cancer 2019; 26: 795-802

Publikationsverlauf

Eingereicht: 20. August 2024

Angenommen nach Revision: 22. September 2024

Artikel online veröffentlicht:
07. November 2024