Thyroidectomy Practice After Implementation of the 2015 American Thyroid Association Guidelines on Surgical Options for Patients With Well-Differentiated Thyroid Carcinoma

Key Points

Question

What are the clinical practice changes following the implementation of the updated 2015 American Thyroid Association guidelines on the extent of thyroidectomy procedures?

Findings

In this cohort study of 169 patients with pathologically proved, well-differentiated thyroid carcinoma, patients who underwent surgery between 2013 and 2014 were compared with patients who underwent surgery during 2016 following implementation of the new 2015 American Thyroid Association guidelines. Rates of up-front total thyroidectomy significantly decreased from 61% to 31%, and the rate of completion thyroidectomy significantly decreased from 74% to 20% following the implementation of the new guidelines.

Meaning

The extent of thyroidectomy was reduced considerably following the 2015 American Thyroid Association guidelines, and only 1 of 5 patients who undergo thyroid lobectomy will require a completion procedure according to the new criteria.

Abstract

Importance

The recommended extent of surgery for well-differentiated thyroid carcinoma has been modified considerably in the updated 2015 American Thyroid Association guidelines published in January 2016. To date, the changes in clinical practice after publication of these new guidelines have not been demonstrated.

Objective

The aim of this study was to evaluate clinical practice changes associated with implementation of the updated guidelines on the surgical procedure rates of total thyroidectomy, thyroid lobectomy, and completion thyroidectomy at a single tertiary medical center.

Design, Setting, and Participants

This is a retrospective cohort study of 169 patients at the Hadassah–Hebrew University Medical Center, Jerusalem, Israel. Patients with pathologically proved, well-differentiated thyroid carcinoma who underwent surgery between January 1, 2013, and December 31, 2014, were compared with patients who underwent surgery from January 1 to December 31, 2016. A total of 434 thyroidectomy procedures were performed during the study period, and 251 had pathologically proved, well-differentiated thyroid carcinoma. Patients with tumors larger than 4 cm, involved lymph nodes, or bilateral nodules were excluded.

Main Outcomes and Measures

Primary outcomes were the rate of up-front total thyroidectomy vs lobectomy and the rates of completion thyroidectomy before and after the implementation of the new guidelines.

Results

Of the 169 patients in the final analysis, 118 (69.8%) were included from 2013 to 2014 and 51 (30.2%) in 2016. The mean (SD) age for the entire cohort was 44 (13.8) years, and 129 (76.3%) were women. Up-front total thyroidectomy was performed in 72 of 118 patients (61.0%) prior to the 2015 American Thyroid Association guidelines and 16 of 51 (31.4%) following their implementation (odds ratio, 0.29; 95% CI, 0.14-0.59). The rate of completion thyroidectomy also significantly decreased between these periods (73.9% vs 20.0%; odds ratio, 0.09; 95% CI, 0.04-0.19).

Conclusions and Relevance

The updated 2015 American Thyroid Association guidelines implementation was associated with a significant decrease in the rates of both up-front total thyroidectomy and completion thyroidectomy. According to these findings, only 1 of 5 patients who undergoes thyroid lobectomy will require a completion procedure.

This cohort study assesses surgical practice changes regarding lobectomy, up-front thyroidectomy, and completion thyroidectomy in a tertiary medical center before and after the implementation of the 2015 American Thyroid Association guidelines for well-differentiated thyroid carcinoma.

Introduction

Well-differentiated thyroid carcinomas (WDTCs) account for most thyroid cancers. Papillary carcinomas comprise 85% and follicular carcinomas comprise 12% of these cancers.¹ The recommended extent of surgery for WDTC has been modified considerably during the past year following the publication of updated guidelines by the American Thyroid Association (ATA) and other international associations.^2,3 The recommended extent of surgery, as well as the need for total thyroidectomy with or without adjuvant radioactive iodine treatment, is influenced by the recurrence risk category. Based on the new guidelines, low-risk patients may be treated with thyroid lobectomy alone, whereas high-risk patients should be treated with up-front total thyroidectomy.² Some of the high-risk characteristics are determined only on final pathology results; therefore, some of the patients who undergo initial thyroid lobectomy may require completion thyroidectomy as a second procedure.

Tumor size necessitating up-front total thyroidectomy is one of the most important changes in the 2015 ATA guidelines. The 2009 ATA guidelines recommended total thyroidectomy for any WDTC larger than microcarcinomas (>1 cm).⁴ The recommendation relied mainly on a study by Bilimoria et al⁵ that demonstrated improved outcomes following total thyroidectomy for tumors ranging from 1 to 4 cm. These results were challenged by several later studies that identified similar outcomes when lobectomy alone was performed for tumors up to 4 cm.^6,7 Consequently, recommendation 35 in the new 2015 ATA guidelines determined that low-risk patients with a WDTC smaller than 4 cm may be treated with thyroid lobectomy alone. This recommendation is limited to low-risk patients without extrathyroidal extension or clinical lymph node metastases.

While the new size limit initiated an expectation for fewer total thyroidectomies, Kluijfhout et al⁸ suggested that as many as 43% of the patients undergoing thyroid lobectomy will require completion thyroidectomy when the final pathology result is available. This study was a theoretical analysis of a cohort of patients who underwent surgery from 2000 to 2010 before the publication of the 2015 ATA guidelines. The authors concluded that surgeons, endocrinologists, and patients need to consider the benefits, risks, and costs of initial surgery extent vs the possible need for completion surgery. Vis-à-vis the theoretical nature of the study by Kluijfhout et al,⁸ no study to date has demonstrated how the new ATA guidelines have changed clinical practice with regard to the rate of thyroid lobectomies or the need for completion thyroidectomies in these patients. The aim of this study was to compare the surgical practice changes in a tertiary medical center before and after the implementation of the new ATA guidelines.

Methods

A retrospective analysis and medical record review were performed to collect demographic data, disease measures, and treatment description for patients with pathologically proved WDTC who underwent surgery at a single tertiary medical center, the Hadassah–Hebrew University Medical Center, Jerusalem, Israel. Two periods were compared: January 1, 2013, to December 31, 2014, and January 1, 2016, to December 31, 2016. Although the formal publication date of the new ATA guidelines was in January 2016, their recommendations became known in 2015, and many departments, including our own, had already discussed and understood the nature of the changes. Therefore, 2015 was considered a transition year for implementation of the guidelines, and that year was not included in this study’s analysis. The study was conducted following the approval of the institutional review board of Hadassah Medical Center according to the World Medical Association Declaration of Helsinki 2013.⁹ Participant informed consent was waived by Hadassah Medical Center.

Three experienced surgeons (N.H., J.M.W., and H.M.) were responsible for performing all of the thyroidectomies during both periods. A multidisciplinary forum including the 3 surgeons, an experienced endocrinologist (B.G.), and an oncologist (B.U.) was established to ensure uniformity, proper implementation, and adherence to the alterations following the establishment of the new ATA guidelines.

Patients with confirmed pathology results of WDTC were included in the study. Patients with benign pathology findings or clear preoperative indication for total thyroidectomy (eg, tumors >4 cm, bilateral nodules, and evidence of lymph node involvement) were excluded. Patients were also excluded if essential data, including pathology and operative reports, were unavailable.

The thyroid surgical procedure was categorized as either thyroid lobectomy, up-front total thyroidectomy, or completion thyroidectomy. The indications for each surgical procedure are described in recommendation 35 in the new 2015 ATA guidelines.²

For each case, the investigators carefully analyzed the pathology report, noting multiple variables, including size, histologic subtype, multifocality of the tumor, presence of extrathyroidal extension, lymphovascular invasion, lymph node status, and surgical margin status. Incidental thyroid carcinoma was defined whenever more than 1 nodule was demonstrated on the final histopathologic specimen and the differentiated carcinoma was of a nodule other than the one that was aspirated prior to surgery.

Statistical Analysis

Primary end points were the rate of lobectomy vs up-front total thyroidectomy and the rates of completion thyroidectomy before and after implementation of the new guidelines. All statistical analyses were performed using IBM SPSS statistics software, version 24.0 (IBM Corporation). All statistical analysis results and their interpretation were independently reviewed by a statistician. The χ² or Fisher exact test was used for comparison of qualitative measures, and the odds ratio (OR) with its 95% CI was presented for each comparison. The unpaired t test and Mann-Whitney test were used for quantitative measures.

Univariate and multivariable Cox proportional hazards regression were used to investigate the association between different potential predictive measures and the need for the completion thyroidectomy procedure. The measures investigated were analyzed for both cohorts together and separately to allow for changes in indications as a consequence of the new guidelines. In those cases (eg, tumor size), the analysis was done separately in 2016, excluding patients who underwent surgery between 2013 and 2014.

Results

Patient Characteristics

A total of 434 thyroidectomy procedures were performed during the study period, of whom 251 had pathologically proved WDTC. After exclusion of patients with benign disease or with clear preoperative indication for up-front total thyroidectomy, a cohort of 169 patients was included in the final analysis: 118 patients (69.8%) in the 2013 to 2014 period and 51 (30.2%) in the 2016 period. The mean (SD) age for the entire cohort was 44 (13.8) years, and 76% were women.

The comparison between the groups is outlined in Table 1. No demographic differences (age, sex, family history, or radiation exposure) were noted, nor was there a difference in any preoperative variables. Multifocal disease and positive surgical margins were more common in the 2013 to 2014 patient cohort.

Table 1. Demographic and Epidemiologic Data With Disease Description of Patients Treated Surgically in 2013-2014 vs 2016.

Variable	No. (%) of Patients		OR (95% CI)
	2013-2014 (n = 118)	2016 (n = 51)
Age, mean (SD), y	45 (14.0)	44 (13.5)	NA
Females	91 (78.0)	38 (74.0)	0.73 (0.34-1.57)
Radiotherapy history	3 (2.5)	1 (1.9)	0.77 (0.08-7.55)
Positive family history for differentiated thyroid carcinoma	9 (7.6)	5 (9.8)	1.26 (0.40-3.98)
Thyroid function
Hypothyroid	9 (7.6)	1 (1.9)	0.24 (0.30-1.95)
Hyperthyroid	8 (6.7)	3 (5.8)	0.85 (0.21-3.41)
Euthyroid	78 (66.0)	37 (72.5)	2.01 (0.63-6.41)
FNA biopsy results, Bethesda classa
III and IV	34 (28.8)	13 (25.4)	0.84 (0.39-1.82)
V and VI	61 (51.7)	32 (62.7)	1.19 (0.54-2.58)
Incidental carcinoma	28 (23.7)	8 (15.6)	0.85 (0.33-2.20)
Multifocal	61 (51.7)	17 (33.0)	0.43 (0.21-0.86)
Extrathyroidal extension	19 (16.1)	4 (7.8)	0.46 (0.14-1.43)
Lymphovascular invasion	17 (14.4)	7 (13.8)	0.83 (0.33-2.08)
Positive surgical margins	27 (22.8)	4 (7.8)	0.26 (0.08-0.80)
Tumor size, mean (SD), cm	1.8 (1.62)	2.1 (1.54)	NA

Abbreviations: FNA, fine-needle aspiration; NA, not applicable; OR, odds ratio.

^aBethesda class: III, atypia of undetermined significance or follicular lesion of undetermined significance; IV, follicular neoplasm or suspicious for a follicular neoplasm; V, suspicious for malignancy; VI, malignant.

Surgical Treatment

Up-front total thyroidectomy was performed in 72 patients (61.0%) prior to the 2015 ATA guidelines and in 16 (31.4%) following the ATA guidelines implementation (OR, 0.29; 95% CI, 0.14-0.59). Of the patients who underwent initial thyroid lobectomy, a significant decrease in the rate of completion thyroidectomy was demonstrated (73.9% vs 20.0%; OR, 0.09; 95% CI, 0.04-0.19). Indications for up-front total thyroidectomy in 2016 included compression symptoms, family or radiotherapy history, patient comorbidities, and patient preference.

Predictive Factors of Completion Thyroidectomy

Univariate and multivariable analysis of the potential predictive factors for completion thyroidectomy are described in Table 2. Women underwent completion thyroidectomy more frequently, whereas age was not associated with completion surgery. Incidental carcinoma (OR, 3.13; 95% CI, 1.16-8.40) as well as lymphovascular invasion identified on final pathology results (OR, 2.32; 95% CI, 1.72-3.11) were associated with higher rates of completion thyroidectomy. Only 2 patients (4.9%) who underwent completion thyroidectomy had extrathyroidal extension. All of those patients with extrathyroidal extension underwent completion thyroidectomy. Patients with disease that involved surgical margins or intermediate Bethesda classification¹⁰ had higher rates of completion thyroidectomy, but the estimates were imprecise and no definitive conclusions can be drawn. The presence of multifocal papillary carcinoma was associated with an 11.0% increase in the rate of completion thyroidectomy, and the data (95% CI, 0.1% to 30.0%) are compatible with a difference as high as 30.0%. The presence of thyrotropin values greater than 4 mIU/L were associated with a 3.6% increase in the rate of completion thyroidectomy, which could be as high as 11.0%.

Table 2. Univariate and Multivariable Analyses of Predictive Measures for Completion Thyroidectomy.

Measure	Thyroidectomy Procedure, No. (%) of Patients		OR (95% CI)
	No Completion (n = 40)	Completion (n = 41)	Univariate	Multivariable
Age, mean (SD), y	42.7 (13.8)	40.3 (12.8)	NA	NA
Females	30 (75.0)	38 (95.0)	6.50 (1.32-31.90)	19.80 (1.20-335.10)
Radiotherapy history	2 (5.0)	0	0.19 (0.01-4.19)	NA
Preoperative thyrotropin levels, mIU/L
<1	6 (23.0)	6 (22.2)	0.95 (0.26-2.89)	NA
1 to ≤4	19 (73.0)	19 (59.2)	0.87 (0.26-2.89)	NA
>4	1 (3.8)	2 (7.4)	2.00 (0.10-23.50)	NA
FNA biopsy, Bethesda classa
III and IV	9 (22.3)	29 (71.9)	8.05 (2.50-26.00)	5.50 (2.90-14.40)
V and VI	27 (67.7)	8 (18.1)	0.12 (0.03-0.40)	0.18 (0.07-0.34)
Surgery period, y
2013-2014	12 (30.0)	34 (82.9)	11.33 (3.93-32.64)	36.40 ‎(5.20-255.20)‎
2016	28 (70.0)	7 (17.1)	0.09 (0.30-0.26)	0.02 (0.003-0.19)
Incidental carcinoma	8 (20.0)	18 (43.9)	3.13 (1.16-8.40)	1.80 (0.76-2.24)
Multifocal	11 (28.0)	16 (39.0)	1.63 (0.63-4.16)	NA
Extrathyroidal extension	0	2 (5.0)	2.08 (1.64-2.64)	NA
Lymphovascular invasion	0	6 (18.0)	2.32 (1.72-3.11)	NA
Positive surgical margins	1 (2.6)	6 (14.6)	6.51 (0.75-56.84)	NA

Abbreviations: OR, odds ratio; FNA, fine-needle aspiration; NA, not applicable.

^aBethesda class: III, atypia of undetermined significance or follicular lesion of undetermined significance; IV, follicular neoplasm or suspicious for a follicular neoplasm; V, suspicious for malignancy; VI, malignant.

The indication for completion thyroidectomy in the 34 patients for the 2013 to 2014 cohort was tumors larger than 1 cm on final pathology results (not confirmed preoperatively). Interestingly, 9 tumors were classified as minimally invasive follicular thyroid carcinoma (MIFTC). The principal causes for the 7 patients who undergo completion thyroidectomy in the 2016 cohort were tumors larger than 4 cm on final pathology results (4 patients [57.1%]) or other specific histologic indications, such as lymphovascular invasion (1 [14.3%]) or extrathyroidal extension (2 [28.6%]). The adherence rate of our team to the new guidelines was as high as 96.5% (83 of 86 cases), considering the extent of primary surgery (51 cases) and the need for a completion thyroidectomy (35 cases). An absolute 23.0% decrease in radioactive iodine treatment was noticed between the 2 periods (95% CI, 7.0%-38.0%).

Discussion

The clinical effect of the updated version of the ATA treatment guidelines for WDTC is not clear. One could assume that the increase in size limit allowance of the tumor in the lobectomy procedure alone would result in a significant decrease in the number of patients undergoing up-front total thyroidectomy. Similarly, the number of completion thyroidectomies would decrease as, again, the tumor size limit increased to 4 cm. Our study is the first to investigate the clinical implications of the new ATA guidelines implementation in a tertiary medical center. When compared with the period before the implementation of the new ATA guidelines, we have demonstrated a 30% decrease in the rate of up-front total thyroidectomy, and the data are compatible with a difference as large as 43%. A similarly large and clinically meaningful decrease (54%) was identified in the number of completion lobectomies required in patients who underwent initial thyroid lobectomy, and the data are compatible with a difference as high as 65%.

Patients included in this cohort were all eligible for lobectomy according to the new ATA guidelines, whereas patients with a clear indication for up-front total thyroidectomy (tumor size >4 cm, lymph node involvement, and contralateral nodules) were excluded from this study. Thus, the analysis demonstrates the net effect of the new ATA guidelines implementation. Since there were no significant differences in the preoperative measures between the periods, it is reasonable to conclude that this dramatic decrease in up-front total thyroidectomy rate is attributed to the implementation of the new guidelines (Table 1).

Another striking difference was the significant decrease in completion thyroidectomy rates between the periods (74% to 20%). Analysis of the indications for completion thyroidectomy in both periods identified tumor size as the most common indication for completion thyroidectomy, a finding that also highlights the clinical effect of recommendation 35 in the new ATA guidelines. This practice change was theoretically analyzed by Kluijfhout et al,⁸ who examined all patients eligible for lobectomy between 2000 and 2010. They suggested that 43% of the patients had postoperative high-risk characteristics (according to the new ATA guidelines) indicating completion thyroidectomy even among small carcinomas (1-2 cm). Their relatively high fraction of expected completion thyroidectomies compared with our findings (20%) may be explained by the theoretical nature of their study.

Two preoperative variables were associated with higher completion thyroidectomy rates—female sex and Bethesda class III and IV results. We assume that the tendency of the referring physician to recommend up-front total thyroidectomy in both male patients and Bethesda class V and VI results may explain these data.

Multifocal disease was more common in the 2013 to 2014 cohort, which raises 2 considerations. Multifocal disease was considered an indication for completion thyroidectomy in the 2013 to 2014 period and may have contributed to the higher completion rates in this group. On the other hand, higher total thyroidectomy rates performed in the 2013 to 2014 cohort resulted in a higher contralateral multifocal disease identified on examination of the contralateral lobe (35% vs 17%).

The decision process for completion thyroidectomy may also be influenced by the recent recognition of low malignancy potential of specific histopathologic subtypes, such as noninvasive encapsulated follicular variant papillary thyroid carcinoma (NIFTP) and MIFTC.^11,12 In this specific study, only 2 patients were diagnosed with NIFTP in the 2016 cohort, and neither patient underwent completion thyroidectomy. In contrast, MIFTC did contribute to the differences between the periods. In the 2013 to 2014 cohort, 9 tumors were categorized as MIFTC and all patients underwent completion thyroidectomy. However, only 5 patients were diagnosed with MIFTC in the 2016 cohort; 2 underwent up-front total thyroidectomy prior to the knowledge of the final pathology results and 3 were treated with lobectomy alone. Thus, these new histologic subtypes affected the management of 5 patients who did not undergo completion thyroidectomy.

Another expected result of the new ATA guidelines is the decreased rates of adjuvant radioactive iodine treatment. In a study from the University of California, San Francisco, only 101 of 394 patients (25.6%) who underwent thyroid lobectomy between 2000 and 2010 should have had adjuvant radioactive iodine treatment according to the new ATA guidelines, whereas 59.6% had radioactive iodine treatment corresponding to the revised 2009 ATA guidelines.¹³ A recent study by Moore et al¹⁴ identified that multidisciplinary conferences with ATA guidelines implementation resulted in a decrease in radioactive iodine treatment rates without an increase in recurrence rates. Similarly, we found a large and clinically meaningful decrease in the use of radioactive iodine treatment in the 2016 cohort compared with the 2013 to 2014 cohort, and the data are compatible with a decrease as large as 38%.

Strengths and Limitations

The meticulous documentation describing the decision process, incorporating the detailed discussions with the patients regarding alternatives, and tumor board opinions and strict adherence to the ATA guidelines strengthen our findings. The major limitation of our study relies on its retrospective nature and potential selection bias associated with more complex cases treated at a tertiary medical center.

Conclusions

Our study analyzing 2 patient populations treated by the same physicians in the same tertiary medical center before and after the implementation of the new 2015 ATA guidelines clearly demonstrates a significant drop of up-front total thyroidectomy and completion thyroidectomy rates. According to our data, only 1 in 5 patients who undergoes thyroid lobectomy according to the new criteria will require a completion procedure. These data may affect the decision process of patients, endocrinologists, and surgeons alike.