- Email: [email protected]
Correct extent of thyroidectomy is poorly predicted preoperatively by the guidelines of the American Thyroid Association for low and intermediate risk thyroid cancers
Surgery 163 (2018) 81–87
Contents lists available at ScienceDirect
Surgery j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y m s y
Thyroid
Correct extent of thyroidectomy is poorly predicted preoperatively by the guidelines of the American Thyroid Association for low and intermediate risk thyroid cancers Mashaal Dhir, MD a, Kelly L. McCoy, MD a, N. Paul Ohori, MD b, Cameron D. Adkisson, MD a, Shane O. LeBeau, MD c, Sally E. Carty, MD a, and Linwah Yip, MD a,* a
Division of Endocrine Surgery, Department of Surgery, University of Pittsburgh, PA Department of Pathology, University of Pittsburgh, PA c Division of Diabetes, Endocrinology, and Metabolism, Department of Medicine, University of Pittsburgh, PA b
A R T I C L E
I N F O
Article history: Accepted 5 April 2017 Available online 8 November 2017
Background. Recent guidelines from the American Thyroid Association recommend thyroid lobectomy for intrathyroidal differentiated thyroid cancers <4 cm. Our aim was to examine histology from patients with cytologic results that were positive or suspicious for malignancy to assess the extent of initial thyroidectomy based on criteria from the 2015 American Thyroid Association guidelines. Methods. We studied consecutive patients who had either a positive or suspicious for malignancy cytologic diagnosis and under prior American Thyroid Association guidelines underwent initial total thyroidectomy ± lymphadenectomy. Results. Among 447 patients, high-risk features necessitating total thyroidectomy were present in 19% (72/380) of positive and 15% (10/67) of suspicious for malignancy patients (P = .5). Intermediate-risk features on histology were identified postoperatively in 46% (175/380) with positive and 15% (18/67) with suspicious for malignancy fine-needle aspiration results. In multivariable analysis, preoperative factors associated with intermediate-risk disease included age ≥45 years, women, larger tumor size, positive fineneedle aspiration cytology, and BRAF V600E or RET/PTC positivity. Conclusion. When patients are considered for lobectomy under the 2015 American Thyroid Association guidelines, ~ 60% with positive and 30% with suspicious for malignancy cytology would need completion thyroidectomy based on intermediate-risk disease. The cost and risk implications of the new American Thyroid Association strategy were substantial and better tools are needed to improve preoperative risk stratification. © 2017 Elsevier Inc. All rights reserved.
Thyroid cancer is increasing in incidence, and an estimated 64,300 new cases were diagnosed in 2016.1 The increasing incidence is likely multifactorial, although improved detection is a substantial contributor.2 Several studies have demonstrated that the diagnosis of small, subcentimeter carcinomas is increasing at the greatest rate.3-5 Differentiated thyroid cancer, however, has excellent 10-year survival rates, and the recent trend is toward judicious management to avoid overtreatment.6 The 2015 guidelines of the American Thyroid Association (ATA) now advise a risk-directed approach to the management of thyroid cancer. For example, thyroid lobectomy may be sufficient treatment for intrathyroidal, differentiated thyroid cancers <4 cm (ATA recommendation 35).6 This observation was based on
Presented as an oral presentation at the 38th annual meeting of The American Association of Endocrine Surgeons, April 2–4, 2017, Orlando, FL. * Reprint requests: Linwah Yip, MD, Division of Endocrine Surgery, University of Pittsburgh, 3471 Fifth Ave, Kauffman Bldg, Ste 101, Pittsburgh, PA 15213. E-mail: [email protected]. https://doi.org/10.1016/j.surg.2017.04.029 0039-6060/© 2017 Elsevier Inc. All rights reserved.
findings from studies demonstrating that 1) lobectomy does not affect disease-specific survival,7-9 2) salvage therapy for recurrences is efficacious,10 and 3) experts now advocate a more selective use of radioactive iodine (RAI)6,11; however, if lobectomy is performed and more aggressive features are seen on histology, such as microscopic extrathyroidal extension or lymphovascular invasion, completion thyroidectomy may be needed to facilitate RAI ablation and/or subsequent surveillance. Thus, recommendations for treatment are now more dependent on the results of the final pathology than previously, and the appropriate extent of thyroidectomy in particular can be difficult or impossible to assess preoperatively. Results of fine-needle aspiration (FNA) biopsy are used typically to diagnose thyroid cancer and determine if operative intervention is needed. Studies have demonstrated that low-risk differentiated thyroid cancers are often associated with cytologically indeterminate results on FNA.12,13 It is unknown if and how frequently initial lobectomy is adequate treatment when preoperative FNA is positive (POS) or suspicious for malignancy (SUSP). The aim of the present study was to examine the histologic features of patients who had
82
Dhiret al / Surgery 163 (2018) 81–87
a POS or SUSP cytologic result and to assess the recommended extent of initial thyroidectomy using criteria from the 2015 ATA Guidelines.
Table I Study population. N = 447 (%)
Materials and Methods Study cohort The present study was a retrospective review of consecutive patients from May 2007 to December 2012 with preoperative cytology either POS (Bethesda Category 6) or SUSP (Bethesda Category 514) who underwent initial total thyroidectomy with or without lymphadenectomy according to the 2006 and 2009 ATA guidelines15,16; this was our routine surgical approach during the study period. Molecular analysis was performed routinely for BRAF, KRAS, HRAS, NRAS mutations, and RET/PTC1, RET/PTC3, and PAX8/PPARG rearrangements, as reported previously.17 Some of these patients have been described previously in other studies. Based on a previous study in which phenotypically similar thyroid cancers were observed in 2 groups (BRAFV600E + RET/PTC and RAS + PAX8/PPARG),17 we analyzed cancer genotype using these same 2 groupings. After approval by the Quality assessment and quality improvement (QA/QI)-institutional review board, clinicopathologic data were collected. Follow-up including postoperative RAI ablation, suppression with thyroid stimulating hormone, and radiologic surveillance, including neck ultrasonography and whole body iodine-131 imaging, was performed by the treating endocrinologist and/or surgeon based on the extant ATA guidelines.15,16 Information on last follow-up interval and date of histologically confirmed recurrent disease was also collected. Mean follow-up for the study population was 52.4 months (0.1–112.6). ATA risk stratification Based on the clinicopathologic features, patients were classified retrospectively into ATA high, intermediate, or low risk categories to determine the recommended extent of thyroidectomy per the 2015 guidelines.6 Briefly, high-risk tumors were defined here by clinical N1, clinical T4, tumor size ≥4 cm, or synchronous distant metastases. Intermediate-risk tumors were defined by extrathyroidal extension, >5 positive lymph nodes, vascular invasion, and/or aggressive histologic variants, including tall cell, hobnail variant, or columnar cell carcinoma. The remainder were classified as low risk. Statistics Continuous variables were summarized as means (±standard deviation) or medians (interquartile range). Categorical data were summarized as frequencies and percentages. χ2 test was used for comparison of categorical variables, and Mann-Whitney U test was used to compare continuous variables. Univariable analysis was performed using logistic regression. Data were reported as odds ratios and 95% confidence intervals. Variables were introduced into the multivariable model to predict intermediate versus low risk (referent) based on statistical significance and clinical relevance. Multivariable model was adjusted for the following covariates: age (<45 or ≥45 years), sex, tumor size (continuous variable), FNA category (POS or SUSP), and mutational status (negative, BRAF + RET/ PTC, or RAS + PAX8PPARG). All analyses were performed using STATA 14.1 (Statacorp, College Station, TX). Results Patient characteristics A total of 447 patients were reviewed, including 380 patients with POS and 67 patients with SUSP FNA results. Table I provides
Age (y) Women Histologic subtype Classic PTC Follicular variant PTC Tall cell variant PTC Other Size (cm) Size ≥4 Extrathyroidal extension Multifocal Bilateral Angiolymphatic invasion Positive margins Clinical N1 Central compartment LN dissection Pathologic LN positive M1 Thyroiditis Mutation data Mutation positive BRAF positive Recurrence
47.7 (36.3–58.6) 345 (77.2) 317 (71.2) 48 (10.8) 66 (14.8) 14 (3.2) 1.5 (1–2) 25 (5.6) 229 (51.5) 260 (58.3) 170 (38.2) 272 (61.1) 113 (25.4) 61 (13.6) 406 (90.8) 231 (51.7) 7 (8.54) 154 (34.5) 431 359 (83.3) 326 (75.6) 46 (10.3)
Continuous variables are summarized as medians with interquartile range (25th percentile to 75th percentile). Categorical variables are summarized as frequencies and percentages.
the details of the demographic and clinicopathologic features. One POS patient had false positive cytology (0.03%). The overall median age was 47.7 years, and women comprised 77.2% of the study population. Classic papillary thyroid cancer (PTC) was the most common histologic subtype (71.2%), and the median tumor size was 1.5 cm. The majority of PTC had extrathyroidal extension (51.5%), were multifocal (58.3%), demonstrated angiolymphatic invasion (61.1%), and had lymph node metastases (51.7%). Although 38.2% (170/447) had bilateral cancers, only 3% of patients had contralateral cancers >1 cm. Clinicopathologic characteristics POS versus SUSP FNA results. We compared the clinicopathologic features of PTC associated with POS FNA to those of PTC with SUSP FNA (Table II). PTC with POS cytology were more often seen in younger patients (median age 46.6 vs 53.7, P = .021) and were more likely to have extrathyroidal extension (55% vs 31.3%, P < .001), angiolymphatic invasion (65.1% vs 38.8%, P < .001), clinical N1 disease (15% vs 5.9%, P = .047), central and lateral LN positive disease (54.5% vs 35.8%, P = .005), and histologic background thyroiditis (36.8% vs 20.9%, P = .011). Eight patients had distant metastases, and the rate did not differ by preoperative FNA cytologic category (POS 6/380, 1.6% vs SUSP 2/67, 2.9%; P = .4). Recurrent/ persistent disease requiring reoperation was also equally likely in the 2 groups (P = .6). Distribution in ATA risk categories Overall, 82 (18%), 193 (43%), and 171 (38%) were classified as ATA high-, intermediate-, and low-risk tumors (Table III). Based on the current ATA recommendations for extent of the initial extent of resection for high-risk disease, total thyroidectomy would therefore have been definitively indicated by preoperative staging for only 18%. The proportion of patients with high-risk disease did not differ between patients with POS or SUSP cytology (19% vs 15%, P = .5). In the 374 patients who were not high risk, classification into the intermediate-risk category due solely to >5 positive lymph nodes
Dhiret al / Surgery 163 (2018) 81–87
Table II Comparison of clinicopathologic features of patients with positive versus suspicious FNA.
Age (y) Woman Histologic subtype Classic PTC Follicular variant PTC Tall cell variant PTC Other Size (cm) Size ≥4 Extrathyroidal extension Multifocal Bilateral Angiolymphatic invasion Positive margins Clinical N1 Central compartment LN dissection Pathologic LN positive M1 Thyroiditis Mutation data Mutation positive BRAF positive Recurrence/Persistent disease
Positive FNA N = 380
Suspicious FNA N = 67
P value
46.6 (36.1–57.6) 296 (77.9)
53.7 (36.5–64.9) 49 (73.1)
.021 .392 <.001
275 (72.7) 30 (7.9) 61 (16.1) 12 (3.3) 1.5 (1–2) 19 (5) 208 (55) 221 (58.3) 147 (38.9) 246 (65.1) 101 (26.7) 57 (15) 349 (91.8)
42 (62.7) 18 (26.9) 5 (7.5) 2 (2.9) 1.5 (0.9–2) 6 (8.9) 21 (31.3) 39 (58.2) 23 (34.3) 26 (38.8) 12 (17.9) 4 (5.9) 57 (85)
207 (54.5) 6 (1.6) 140 (36.8) 369 314 (85.1) 287 (77.8) 38 (10)
24 (35.8) 2 (2.9) 14 (20.9) 62 47 (75.8) 39 (62.9) 8 (11.9)
.874 .194 <.001 .987 .479 <.001 .127 .047 .077 .005 .423 .011 .067 .012 .635
Continuous variables are summarized as medians with interquartile range (25th percentile to 75th percentile). Categorical variables are summarized as frequencies and percentages.
in the central compartment comprised 21 patients (5.6%). Among patients with POS FNA results, the incidence of ATA intermediateand low-risk disease was 46% and 35%, respectively. For those with SUSP FNA results, the incidence of ATA intermediate and low-risk disease was 27% and 58%, respectively (Fig 1, P < .001). Among these remaining 364 patients who could have had an initial lobectomy by the 2015 ATA guidelines, final histopathologic findings would have identified intermediate-risk tumors in 175/307 (59%) POS FNA and 18/57 (32%) SUSP FNA patients.
83
Stratification of low versus intermediate risk by preoperative factors We further evaluated if any preoperative features in addition to FNA Bethesda category can help differentiate low from intermediaterisk tumors preoperatively to inform surgical decision-making. On univariable analysis, patients with ATA low-risk disease were more likely to be younger (age <45y), have smaller tumors, and have SUSP FNA results (Table IV). Data on somatic tumor mutations also were available for 431/447 study patients (96%). Because the tumor genotype is another variable that can be readily obtained preoperatively, we also evaluated if the genotype could be a factor in preoperative risk stratification. ATA intermediate-risk disease was present in 50.6% of BRAFV600E or RET/PTC positive PTC, but in only 10.5% of RAS or PAX8/PPARG positive PTC (P = .002, Fig 2). In multivariable analysis and compared with low-risk disease, intermediaterisk disease was associated with age ≥45 years (OR 1.96, P = .005), female sex (OR 1.92, P = .03), larger tumor size (OR 1.73, P = .001), POS FNA results (OR 2.48, P = .008), and BRAF or RET/PTC positive PTC (OR 3.86, P < .001). Discussion We set out to address the problematic reasoning in the 2015 ATA guidelines, which advise paradoxically that decisions about the operative extent of the thyroidectomy should be based on risk factors available only after the operation. The 2015 ATA guidelines now suggest that thyroid lobectomy alone may be sufficient initial treatment for low-risk papillary and follicular carcinoma (Recommendation 35),6 but total thyroidectomy may be needed for disease that is at greater risk of recurrence or persistent and can facilitate RAI ablation therapy and aid in surveillance. Patients with ATA intermediate-risk tumors have features, such as aggressive histology, microscopic extrathyroidal extension, and lymph node metastases, which are not known until the final pathologic analysis. Most large population-level studies of the outcomes of treatment of PTC have stratified patients by their histologic diagnosis, which makes extrapolation to clinical care difficult, because patients present typically by cytologic diagnosis. Thus, the design of this study was to identify what proportion of patients with SUSP or POS FNA
Table III Comparison of clinicopathologic features of patients in the ATA risk categories. ATA 2015
Age (y) Woman Histologic subtype Classic PTC Follicular variant PTC Tall cell variant PTC Other Size (cm) Size ≥4 cm Extrathyroidal extension Multifocal Bilateral Angiolymphatic invasion Positive margins Clinical N1 Central compartment LN dissection Pathologic LN positive M1 Thyroiditis Recurrence/persistent disease
High risk N = 82 (%)
Intermediate risk N = 193 (%)
Low risk N = 171 (%)
P value all groups
P value intermediate versus low
48.4 (37.3–59.7) 48 (58.5)
49.3 (37.4–60.5) 163 (84.5)
44 (35.5–57.4) 134 (78.4)
.114 <.001 <.001
.041 .134 <.001
130 (67.4) 9 (4.7) 52 (26.9) 2 (1) 1.5 (1–2) 0 172 (89.1) 111 (57.5) 75 (38.9) 136 (70.5) 82 (42.5) 0 173 (89.6) 109 (56.5) 0 66 (34.2) 10 (5.2)
131 (76.6) 31 (18.1) 0 9 (5.3) 1.3 (0.9–1.9) 0 0 95 (55.6) 55 (32.2) 63 (36.8) 4 (2.3) 0 152 (88.9) 51 (29.8) 1 (0.6) 64 (37.4) 8 (4.7)
<.001 <.001 <.001 .321 .031 <.001 <.001 <.001 .043 <.001 <.001 .339 <.001
.002 — <.001 .707 .183 <.001 <.001 — .818 <.001 .287 .521 .825
56 (69.1) 8 (9.9) 14 (17.3) 3 (3.7) 2.2 (1.1–4) 25 (30.5) 57 (70.4) 53 (65.4) 40 (49.4) 73 (90.1) 27 (33.3) 61 (74.4) 80 (97.6) 71 (86.6) 7 (8.5) 23 (28) 28 (34.1)
Continuous variables are summarized as medians with interquartile range (25th percentile to 75th percentile). Categorical variables are summarized as frequencies and percentage.
84
Dhiret al / Surgery 163 (2018) 81–87
Fig 1. Frequencies of high-, intermediate-, and low-risk tumors stratified by the preoperative FNA results. White bars depict patients with preoperative positive FNA, and black bars depict patients with preoperative suspicious FNA. High-risk tumors were defined by clinical N1, tumor size ≥4 cm, or synchronous distant metastases. Intermediaterisk tumors were defined by extrathyroidal extension, >5 positive lymph nodes, and/or aggressive histologic variants. The remainder of the tumors were classified as low risk. P value was computed using the χ2 test.
biopsies fall into each of the ATA risk categories as well as to discover any preoperative variables that can be used to help determine the extent of initial thyroidectomy. Our findings demonstrate that preoperative ultrasonography was able to identify ATA high-risk disease in 19% and 15% of patients with POS and SUSP FNA results, respectively. In the remaining patients, the FNA category can help further risk stratify. Many patients with a POS FNA result have ATA intermediate-risk tumors (46%), while the majority of patients with a SUSP FNA result have ATA lowrisk tumors (58%). We also confirmed that POS FNA are more likely than SUSP FNA results to be associated with intermediate-risk tumors on multivariable analysis. Thus, preoperative FNA results and ultrasonography are important, risk-stratification tools and can be helpful in guiding the extent of initial thyroidectomy. Interestingly, lymph node dissection of the central compartment was performed in ~98% of patients; however, having >5 positive lymph nodes was infrequently (5.6%) the only cause for intermediaterisk classification. Differentiating between low- and intermediate-risk tumors using cytology alone, however, is suboptimal. Excluding the patients with
high-risk disease, if lobectomy were performed as the initial operation for all remaining patients, completion thyroidectomy to treat ATA intermediate-risk disease would potentially be required for 53% (193/365) of patients, including the 57% with POS FNA and the 32% with SUSP FNA. Although the complication rate of completion thyroidectomy is comparable with that of initial total thyroidectomy,18 additional risk is associated with a second general anesthesia as well as added costs of care.19,20 In contrast, initial total thyroidectomy for all patients with a SUSP or POS FNA biopsy potentially over treats 47% of patients, including the 43% with POS FNA and the 68% with SUSP FNA results. Cost-efficacy studies to evaluate the impact of lobectomy versus total thyroidectomy for thyroid cancer patients in the long-term including the use of preoperative mutation testing on FNA specimens to guide the extent of thyroidectomy are beyond the scope of the present study but are most certainly needed. Our results are similar to those of Lang et al21 who analyzed 600 patients with PTC 1–4 cm in size without any clinical high-risk features, such as distant metastases, clinical nodal disease, size >4 cm, gross extrathyroidal extension, incomplete tumor resection, macroscopic bilateral disease, and family history or a history of radiation.
Table IV Univariable and multivariable analysis of preoperative factors associated with ATA intermediate-risk versus low risk-tumors. Univariable OR Age Sex Size* FNA Mutation
<45 y ≥45 y Female Male cm Suspicious Positive Negative BRAF or RET/PTC RAS or PAX8/PPARG
* Size was assessed as a continuous variable.
1.822 0.666 1.615 2.872 4.052 0.4
Multivariable 95% CI referent 1.198–2.769 referent 0.391–1.136 1.209–2.156 referent 1.572–5.247 referent 2.134–7.696 0.080–1.987
P value
OR
95% CI
P value
.005
1.956
1.227–3.118
.005
.136 .001
0.519 1.726
0.287–0.939 1.244–2.392
.03 .001
.001
2.475
1.262–4.855
.008
<.001 .263
3.86 0.381
1.973–7.551 0.073–1.972
<.001 .25
Dhiret al / Surgery 163 (2018) 81–87
85
Fig 2. ATA risk categories by 7-gene mutation testing results. P value was computed using the χ2 test.
They reported that 257/600 (42.8%) of the patients had ≥1 unrecognized histologic, high-risk feature (aggressive histology, lymphovascular invasion, microscopic extrathyroidal extent, positive resection margin, and pN1 >0.5 cm), making it likely that these patients would need completion thyroidectomy. Similarly, Kluijfhout et al22 reported that 43% (122/287) of 1- to 4-cm well differentiated thyroid cancers eligible for lobectomy under the current guidelines may need completion thyroidectomy based on histologic results. Based on the present study and previous studies, ~ 40% to 50% (i.e., 1 in 2) patients would require completion thyroidectomy if lobectomy was the initial operative procedure as recommended by the current 2015 ATA Guidelines. Better preoperative risk stratification is needed. We further considered what, if any, preoperative factors could differentiate patients with ATA low-risk disease from those with intermediate-risk disease. ATA intermediate-risk was associated on univariable analysis with age ≥45 years, larger tumor size, and positive FNA. On multivariable analysis, female sex was associated additionally with intermediate-risk disease. We also analyzed available results with mutation testing, because it is another reasonable preoperative variable. We demonstrated previously that thyroid cancers with somatic BRAF V600E mutations or RET/PTC rearrangements have a high incidence of lymph node metastasis, distant metastasis, and recurrence.17 We report here that BRAF V600E or RET/PTC-positivity also was associated with a 4-fold increased risk of ATA intermediate-risk disease. Furthermore, the status of any molecular mutation remained a significant and independent risk factor for ATA intermediate-risk disease on multivariable analysis. Mutation testing can be obtained readily on preoperative FNA biopsy specimens and has been demonstrated to contribute effectively to decision making for indeterminate cytology results.23,24 In the context of the 2015 ATA Guidelines, preoperative mutation testing even on biopsies that are positive for malignancy also can contribute productively to the decision for initial lobectomy or total thyroidectomy. Prospective studies that use preoperative mutation testing to guide the extent of the initial thyroidectomy are ongoing. The limitations of our study include the potential selection bias that may be present in any retrospective single institutional series. We did not re-review ultrasonographic features, because the ability
of ultrasonography to predict the subtle features of the disease accurately that differentiate low and intermediate-risk disease is variable.25 Furthermore, we do not utilize intraoperative frozen analysis routinely to assess lymph node or margin status, which may also contribute incrementally to risk stratification. Because all patients had an initial total thyroidectomy, we do not know how many patients had concerning intraoperative findings that may have changed the extent of thyroidectomy if a lesser resection was planned. We did not include in this analysis any other indications for initial total thyroidectomy, and it is possible that identification of contralateral nodules, underlying thyroiditis, or other clinical considerations may lead surgeons to favor total thyroidectomy instead of lobectomy. Thus, the number of patients who would be true candidates for initial lobectomy may be overestimated here. In summary, total thyroidectomy for high-risk disease was indicated by the 2015 ATA guidelines preoperatively in 18% of patients with POS or SUSP FNA biopsies. If an initial lobectomy was performed for the remaining patients, 53% would have intermediaterisk disease on final histopathology and thus require completion thyroidectomy. Although the risk category based on the FNA biopsy can help risk stratify ATA low- and intermediate-risk disease, additional preoperative variables including age, sex, tumor size, and the presence of BRAF V600E or RET/PTC rearrangements are also important and useful considerations. The implications in terms of cost and risk of the new ATA strategy are substantial, and adjunct tools can improve preoperative risk stratification. References 1. Miller KD, Siegel RL, Lin CC, et al. Cancer treatment and survivorship statistics, 2016. CA Cancer J Clin 2016;66:271-89. 2. Aschebrook-Kilfoy B, Kaplan EL, Chiu BC, Angelos P, Grogan RH. The acceleration in papillary thyroid cancer incidence rates is similar among racial and ethnic groups in the United States. Ann Surg Oncol 2013;20:2746-53. 3. Ahn HS, Welch HG. South Korea’s thyroid-cancer “epidemic”—turning the tide. N Engl J Med 2015;373:2389-90. 4. Davies L, Ouellette M, Hunter M, Welch HG. The increasing incidence of small thyroid cancers: where are the cases coming from? Laryngoscope 2010;120:2446-51. 5. Brito JP, Davies L. Is there really an increased incidence of thyroid cancer? Curr Opin Endocrinol Diabetes Obes 2014;21:405-8.
86
Dhiret al / Surgery 163 (2018) 81–87
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. Adam MA, Pura J, Gu L, et al. Extent of surgery for papillary thyroid cancer is not associated with survival: an analysis of 61,775 patients. Ann Surg 2014;260:601-5, discussion 5-7. 8. Matsuzu K, Sugino K, Masudo K, et al. Thyroid lobectomy for papillary thyroid cancer: long-term follow-up study of 1,088 cases. World J Surg 2014;38:68-79. 9. Haigh PI, Urbach DR, Rotstein LE. Extent of thyroidectomy is not a major determinant of survival in low- or high-risk papillary thyroid cancer. Ann Surg Oncol 2005;12:81-9. 10. Vaisman F, Shaha A, Fish S, Michael Tuttle R. Initial therapy with either thyroid lobectomy or total thyroidectomy without radioactive iodine remnant ablation is associated with very low rates of structural disease recurrence in properly selected patients with differentiated thyroid cancer. Clin Endocrinol (Oxf) 2011;75:112-9. 11. Jonklaas J, Cooper DS, Ain KB, et al. Radioiodine therapy in patients with stage I differentiated thyroid cancer. Thyroid 2010;20:1423-4. 12. Gweon HM, Koo HR, Son EJ, et al. Prognostic role of the Bethesda System for conventional papillary thyroid carcinoma. Head Neck 2016;38:1509-14. 13. Ohori NP, Wolfe J, Hodak SP, et al. Colloid-rich” follicular neoplasm/suspicious for follicular neoplasm thyroid fine-needle aspiration specimens: cytologic, histologic, and molecular basis for considering an alternate view. Cancer Cytopathol 2013;121:718-28. 14. Cibas ES, Ali SZ. The Bethesda system for reporting thyroid cytopathology. Am J Clin Pathol 2009;132:658-65. 15. American Thyroid Association (ATA) Guidelines Taskforce on Thyroid Nodules and Differentiated Thyroid Cancer, Cooper DS, Doherty GM, et al. Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid 2009;19:1167-214.
16. Cooper DS, Doherty GM, Haugen BR, et al. Management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid 2006;16:10942. 17. Yip L, Nikiforova MN, Yoo JY, et al. Tumor genotype determines phenotype and disease-related outcomes in thyroid cancer: a study of 1,510 patients. Ann Surg 2015;262:519-25, discussion 24-5. 18. Li YJ, Wang YZ, Yi ZB, Chen LL, Zhou XD. Comparison of completion thyroidectomy and primary total surgery for differentiated thyroid cancer: a meta-analysis. Oncol Res Treat 2015;38:528-31. 19. Lang BH, Wong CK. Lobectomy is a more cost-effective option than total thyroidectomy for 1 to 4 cm papillary thyroid carcinoma that do not possess clinically recognizable high-risk features. Ann Surg Oncol 2016;23:3641-52. 20. Yip L, Farris C, Kabaker AS, et al. Cost impact of molecular testing for indeterminate thyroid nodule fine-needle aspiration biopsies. J Clin Endocrinol Metab 2012;97:1905-12. 21. Lang BH, Shek TW, Wan KY. The significance of unrecognized histological high-risk features on response to therapy in papillary thyroid carcinoma measuring 1–4 cm: implications for completion thyroidectomy following lobectomy. Clin Endocrinol (Oxf) 2017;86:236-42. 22. Kluijfhout WP, Pasternak JD, Lim J, et al. Frequency of high-risk characteristics requiring total thyroidectomy for 1–4 cm well-differentiated thyroid cancer. Thyroid 2016;26:820-4. 23. Yip L, Wharry LI, Armstrong MJ, et al. A clinical algorithm for fine-needle aspiration molecular testing effectively guides the appropriate extent of initial thyroidectomy. Ann Surg 2014;260:163-8. 24. Nikiforov YE, Steward DL, Robinson-Smith TM, et al. Molecular testing for mutations in improving the fine-needle aspiration diagnosis of thyroid nodules. J Clin Endocrinol Metab 2009;94:2092-8. 25. Lee CY, Kim SJ, Ko KR, Chung KW, Lee JH. Predictive factors for extrathyroidal extension of papillary thyroid carcinoma based on preoperative sonography. J Ultrasound Med 2014;33:231-8.
Discussion Dr Askok R. Shaha (New York, NY): Very nice study. Very well analyzed. There are 2 points I want to bring up. When we talk about minor extrathyroidal extension, as you know very well, the staging system is going to change. The AJCC staging manual is already out, but it will be implemented next year. It does not increase the T stage. In the past, it used to be certainly T3. Now it is not going to be T3 anymore. And there are very nice papers from Sloan Kettering and Mayo Clinic showing that minor extrathyroidal extension does not have any major prognostic implication. The second, when you change the staging from low risk to intermediate risk, there is an emotional concern, but I am not convinced that all those patients will require a completion and will benefit from RAI. The entire premise of doing a completion is to benefit the patient with RAI. What is more important is to define who needs RAI, and those are the only patients who should go for completion thyroidectomy. The remaining intermediate-risk group can be easily monitored. Thank you. Dr Linwah Yip: Thank you. You are absolutely right. As we all know, staging and the risk categories are a moving target, and so for the purposes of the study we used what was described in the most recent set of ATA Guidelines. And you are correct, radioactive iodine may not always be necessary for an intermediate risk lesion. In fact, our endocrinologists do not give radioactive iodine frequently either, but they often still would like to follow those thyroglobulin levels and sometimes do request to have a completion for that reason. Thank you. Dr Julie Ann Sosa (Durham, NC): So, Lin, fabulous study from, really, the leading edge of innovation in molecular testing. But I think that sort of begs my first question, which is, many of us live in a resource limited world where we do not have access yet to molecular testing to allow us to preoperatively profile tumors to try to
better risk-stratify them. Instead, we use ultrasound, and, obviously, this may have been a limitation of your study, but I think many of us believe ultrasound, at least combined with molecular testing, or used on its own, is a very powerful tool to help guide extent of resection. So my first question is, if you had been able to retrospectively look at those ultrasounds, do you believe that might have changed your results? And, going forward, incorporating other institutions into this study who maybe rely less on molecular profiling and more on ultrasound, might that add to the wealth of the data you accumulate, so a multi-institutional approach? My second question builds on Dr. Shaha’s, and that is the evolving staging system. Extrathyroidal extension is one, but the second is debate about micro versus macro metastatic disease. And are those really equal or are micrometastasis more similar to the absence of metastasis? I wondered if you had done a sensitivity analysis taking those out of a higher risk category, micrometastatic disease, and putting them into a lower risk category. Thank you, again. Terrific study. Dr Linwah Yip: So to answer your second question, we used histologic variables and correlation with published ATA risk categories as our study outcome. We did not use disease related outcomes such as recurrence and mortality, specifically, so I cannot answer if reclassifying extrathyroidal extension or extent of lymph node involvement would make a difference in preoperative risk stratification. But I think that is a very interesting question and certainly begs for more than our 447-patient cohort to be able to probably reach those answers. Multi-institutional studies are a great idea and we would love to do that. Our ultrasound data was a bit limited, because this was retrospective. For example, the degree of extrathyroidal extension is not routinely reported in every single one of our ultrasound reports, and is unfortunately not assessed with every patient es-
Dhiret al / Surgery 163 (2018) 81–87
pecially if cancer is not known during the initial diagnostic study. It would be something to look for prospectively, as well. Thank you. Dr Douglas Evan (Milwaukee, WI): Lin, that was fantastic. Maybe you could spend 2 minutes on your clinical trial and the sample size that you need and what your primary endpoint is. And also since your bias is reflected in your data that you presented, historically you performed total thyroidectomy, and I suspect probably because you agreed with doing that, not just that you read it in the ATA guidelines, how successful have you been in enrolling patients, and how do you answer the question when the patient says, “Well, Dr Yip, what would do you if this were you or your mom?” Dr Linwah Yip: So our endpoint for the trial is dependent on histologic variables, and we are specifically looking at how accurately preoperative molecular testing helps us predict the need for lobectomy when the biopsy is positive or suspicious. When using that specific endpoint, and the results of this study, ~ 40% were being overtreated. Our calculated sample size is 90 patients. And so we just opened the trial in February and have already accrued 5 patients.
87
It has not been difficult to enroll and the conversation that we all currently have with patients is complicated. For example, if someone comes in with a 2-cm nodule that is positive for PTC and their ultrasound does not show any lymph node disease, then we have this very long extensive discussion e.g., you can have a lobe, you can have a total, you may need to have a completion, etc. As we all know, that is a very long discussion that can be confusing to patients. With the trial, I now say, we can potentially use molecular testing to help us sort out if we need to do lobe or a total. We explain that we do not know if it is going to help, but that is an extra piece of information that we can provide if you enter this trial. And I have actually had patients say, well, that makes a lot of sense, why do we not do that for everybody. So it has not been difficult to talk to patients about it at all. Dr Douglas Evans (Milwaukee, WI): So you are 5 for 5 now? No one has refused? Dr Linwah Yip: Five of 5, correct. No one refused.
Contents lists available at ScienceDirect
Surgery j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y m s y
Thyroid
Correct extent of thyroidectomy is poorly predicted preoperatively by the guidelines of the American Thyroid Association for low and intermediate risk thyroid cancers Mashaal Dhir, MD a, Kelly L. McCoy, MD a, N. Paul Ohori, MD b, Cameron D. Adkisson, MD a, Shane O. LeBeau, MD c, Sally E. Carty, MD a, and Linwah Yip, MD a,* a
Division of Endocrine Surgery, Department of Surgery, University of Pittsburgh, PA Department of Pathology, University of Pittsburgh, PA c Division of Diabetes, Endocrinology, and Metabolism, Department of Medicine, University of Pittsburgh, PA b
A R T I C L E
I N F O
Article history: Accepted 5 April 2017 Available online 8 November 2017
Background. Recent guidelines from the American Thyroid Association recommend thyroid lobectomy for intrathyroidal differentiated thyroid cancers <4 cm. Our aim was to examine histology from patients with cytologic results that were positive or suspicious for malignancy to assess the extent of initial thyroidectomy based on criteria from the 2015 American Thyroid Association guidelines. Methods. We studied consecutive patients who had either a positive or suspicious for malignancy cytologic diagnosis and under prior American Thyroid Association guidelines underwent initial total thyroidectomy ± lymphadenectomy. Results. Among 447 patients, high-risk features necessitating total thyroidectomy were present in 19% (72/380) of positive and 15% (10/67) of suspicious for malignancy patients (P = .5). Intermediate-risk features on histology were identified postoperatively in 46% (175/380) with positive and 15% (18/67) with suspicious for malignancy fine-needle aspiration results. In multivariable analysis, preoperative factors associated with intermediate-risk disease included age ≥45 years, women, larger tumor size, positive fineneedle aspiration cytology, and BRAF V600E or RET/PTC positivity. Conclusion. When patients are considered for lobectomy under the 2015 American Thyroid Association guidelines, ~ 60% with positive and 30% with suspicious for malignancy cytology would need completion thyroidectomy based on intermediate-risk disease. The cost and risk implications of the new American Thyroid Association strategy were substantial and better tools are needed to improve preoperative risk stratification. © 2017 Elsevier Inc. All rights reserved.
Thyroid cancer is increasing in incidence, and an estimated 64,300 new cases were diagnosed in 2016.1 The increasing incidence is likely multifactorial, although improved detection is a substantial contributor.2 Several studies have demonstrated that the diagnosis of small, subcentimeter carcinomas is increasing at the greatest rate.3-5 Differentiated thyroid cancer, however, has excellent 10-year survival rates, and the recent trend is toward judicious management to avoid overtreatment.6 The 2015 guidelines of the American Thyroid Association (ATA) now advise a risk-directed approach to the management of thyroid cancer. For example, thyroid lobectomy may be sufficient treatment for intrathyroidal, differentiated thyroid cancers <4 cm (ATA recommendation 35).6 This observation was based on
Presented as an oral presentation at the 38th annual meeting of The American Association of Endocrine Surgeons, April 2–4, 2017, Orlando, FL. * Reprint requests: Linwah Yip, MD, Division of Endocrine Surgery, University of Pittsburgh, 3471 Fifth Ave, Kauffman Bldg, Ste 101, Pittsburgh, PA 15213. E-mail: [email protected]. https://doi.org/10.1016/j.surg.2017.04.029 0039-6060/© 2017 Elsevier Inc. All rights reserved.
findings from studies demonstrating that 1) lobectomy does not affect disease-specific survival,7-9 2) salvage therapy for recurrences is efficacious,10 and 3) experts now advocate a more selective use of radioactive iodine (RAI)6,11; however, if lobectomy is performed and more aggressive features are seen on histology, such as microscopic extrathyroidal extension or lymphovascular invasion, completion thyroidectomy may be needed to facilitate RAI ablation and/or subsequent surveillance. Thus, recommendations for treatment are now more dependent on the results of the final pathology than previously, and the appropriate extent of thyroidectomy in particular can be difficult or impossible to assess preoperatively. Results of fine-needle aspiration (FNA) biopsy are used typically to diagnose thyroid cancer and determine if operative intervention is needed. Studies have demonstrated that low-risk differentiated thyroid cancers are often associated with cytologically indeterminate results on FNA.12,13 It is unknown if and how frequently initial lobectomy is adequate treatment when preoperative FNA is positive (POS) or suspicious for malignancy (SUSP). The aim of the present study was to examine the histologic features of patients who had
82
Dhiret al / Surgery 163 (2018) 81–87
a POS or SUSP cytologic result and to assess the recommended extent of initial thyroidectomy using criteria from the 2015 ATA Guidelines.
Table I Study population. N = 447 (%)
Materials and Methods Study cohort The present study was a retrospective review of consecutive patients from May 2007 to December 2012 with preoperative cytology either POS (Bethesda Category 6) or SUSP (Bethesda Category 514) who underwent initial total thyroidectomy with or without lymphadenectomy according to the 2006 and 2009 ATA guidelines15,16; this was our routine surgical approach during the study period. Molecular analysis was performed routinely for BRAF, KRAS, HRAS, NRAS mutations, and RET/PTC1, RET/PTC3, and PAX8/PPARG rearrangements, as reported previously.17 Some of these patients have been described previously in other studies. Based on a previous study in which phenotypically similar thyroid cancers were observed in 2 groups (BRAFV600E + RET/PTC and RAS + PAX8/PPARG),17 we analyzed cancer genotype using these same 2 groupings. After approval by the Quality assessment and quality improvement (QA/QI)-institutional review board, clinicopathologic data were collected. Follow-up including postoperative RAI ablation, suppression with thyroid stimulating hormone, and radiologic surveillance, including neck ultrasonography and whole body iodine-131 imaging, was performed by the treating endocrinologist and/or surgeon based on the extant ATA guidelines.15,16 Information on last follow-up interval and date of histologically confirmed recurrent disease was also collected. Mean follow-up for the study population was 52.4 months (0.1–112.6). ATA risk stratification Based on the clinicopathologic features, patients were classified retrospectively into ATA high, intermediate, or low risk categories to determine the recommended extent of thyroidectomy per the 2015 guidelines.6 Briefly, high-risk tumors were defined here by clinical N1, clinical T4, tumor size ≥4 cm, or synchronous distant metastases. Intermediate-risk tumors were defined by extrathyroidal extension, >5 positive lymph nodes, vascular invasion, and/or aggressive histologic variants, including tall cell, hobnail variant, or columnar cell carcinoma. The remainder were classified as low risk. Statistics Continuous variables were summarized as means (±standard deviation) or medians (interquartile range). Categorical data were summarized as frequencies and percentages. χ2 test was used for comparison of categorical variables, and Mann-Whitney U test was used to compare continuous variables. Univariable analysis was performed using logistic regression. Data were reported as odds ratios and 95% confidence intervals. Variables were introduced into the multivariable model to predict intermediate versus low risk (referent) based on statistical significance and clinical relevance. Multivariable model was adjusted for the following covariates: age (<45 or ≥45 years), sex, tumor size (continuous variable), FNA category (POS or SUSP), and mutational status (negative, BRAF + RET/ PTC, or RAS + PAX8PPARG). All analyses were performed using STATA 14.1 (Statacorp, College Station, TX). Results Patient characteristics A total of 447 patients were reviewed, including 380 patients with POS and 67 patients with SUSP FNA results. Table I provides
Age (y) Women Histologic subtype Classic PTC Follicular variant PTC Tall cell variant PTC Other Size (cm) Size ≥4 Extrathyroidal extension Multifocal Bilateral Angiolymphatic invasion Positive margins Clinical N1 Central compartment LN dissection Pathologic LN positive M1 Thyroiditis Mutation data Mutation positive BRAF positive Recurrence
47.7 (36.3–58.6) 345 (77.2) 317 (71.2) 48 (10.8) 66 (14.8) 14 (3.2) 1.5 (1–2) 25 (5.6) 229 (51.5) 260 (58.3) 170 (38.2) 272 (61.1) 113 (25.4) 61 (13.6) 406 (90.8) 231 (51.7) 7 (8.54) 154 (34.5) 431 359 (83.3) 326 (75.6) 46 (10.3)
Continuous variables are summarized as medians with interquartile range (25th percentile to 75th percentile). Categorical variables are summarized as frequencies and percentages.
the details of the demographic and clinicopathologic features. One POS patient had false positive cytology (0.03%). The overall median age was 47.7 years, and women comprised 77.2% of the study population. Classic papillary thyroid cancer (PTC) was the most common histologic subtype (71.2%), and the median tumor size was 1.5 cm. The majority of PTC had extrathyroidal extension (51.5%), were multifocal (58.3%), demonstrated angiolymphatic invasion (61.1%), and had lymph node metastases (51.7%). Although 38.2% (170/447) had bilateral cancers, only 3% of patients had contralateral cancers >1 cm. Clinicopathologic characteristics POS versus SUSP FNA results. We compared the clinicopathologic features of PTC associated with POS FNA to those of PTC with SUSP FNA (Table II). PTC with POS cytology were more often seen in younger patients (median age 46.6 vs 53.7, P = .021) and were more likely to have extrathyroidal extension (55% vs 31.3%, P < .001), angiolymphatic invasion (65.1% vs 38.8%, P < .001), clinical N1 disease (15% vs 5.9%, P = .047), central and lateral LN positive disease (54.5% vs 35.8%, P = .005), and histologic background thyroiditis (36.8% vs 20.9%, P = .011). Eight patients had distant metastases, and the rate did not differ by preoperative FNA cytologic category (POS 6/380, 1.6% vs SUSP 2/67, 2.9%; P = .4). Recurrent/ persistent disease requiring reoperation was also equally likely in the 2 groups (P = .6). Distribution in ATA risk categories Overall, 82 (18%), 193 (43%), and 171 (38%) were classified as ATA high-, intermediate-, and low-risk tumors (Table III). Based on the current ATA recommendations for extent of the initial extent of resection for high-risk disease, total thyroidectomy would therefore have been definitively indicated by preoperative staging for only 18%. The proportion of patients with high-risk disease did not differ between patients with POS or SUSP cytology (19% vs 15%, P = .5). In the 374 patients who were not high risk, classification into the intermediate-risk category due solely to >5 positive lymph nodes
Dhiret al / Surgery 163 (2018) 81–87
Table II Comparison of clinicopathologic features of patients with positive versus suspicious FNA.
Age (y) Woman Histologic subtype Classic PTC Follicular variant PTC Tall cell variant PTC Other Size (cm) Size ≥4 Extrathyroidal extension Multifocal Bilateral Angiolymphatic invasion Positive margins Clinical N1 Central compartment LN dissection Pathologic LN positive M1 Thyroiditis Mutation data Mutation positive BRAF positive Recurrence/Persistent disease
Positive FNA N = 380
Suspicious FNA N = 67
P value
46.6 (36.1–57.6) 296 (77.9)
53.7 (36.5–64.9) 49 (73.1)
.021 .392 <.001
275 (72.7) 30 (7.9) 61 (16.1) 12 (3.3) 1.5 (1–2) 19 (5) 208 (55) 221 (58.3) 147 (38.9) 246 (65.1) 101 (26.7) 57 (15) 349 (91.8)
42 (62.7) 18 (26.9) 5 (7.5) 2 (2.9) 1.5 (0.9–2) 6 (8.9) 21 (31.3) 39 (58.2) 23 (34.3) 26 (38.8) 12 (17.9) 4 (5.9) 57 (85)
207 (54.5) 6 (1.6) 140 (36.8) 369 314 (85.1) 287 (77.8) 38 (10)
24 (35.8) 2 (2.9) 14 (20.9) 62 47 (75.8) 39 (62.9) 8 (11.9)
.874 .194 <.001 .987 .479 <.001 .127 .047 .077 .005 .423 .011 .067 .012 .635
Continuous variables are summarized as medians with interquartile range (25th percentile to 75th percentile). Categorical variables are summarized as frequencies and percentages.
in the central compartment comprised 21 patients (5.6%). Among patients with POS FNA results, the incidence of ATA intermediateand low-risk disease was 46% and 35%, respectively. For those with SUSP FNA results, the incidence of ATA intermediate and low-risk disease was 27% and 58%, respectively (Fig 1, P < .001). Among these remaining 364 patients who could have had an initial lobectomy by the 2015 ATA guidelines, final histopathologic findings would have identified intermediate-risk tumors in 175/307 (59%) POS FNA and 18/57 (32%) SUSP FNA patients.
83
Stratification of low versus intermediate risk by preoperative factors We further evaluated if any preoperative features in addition to FNA Bethesda category can help differentiate low from intermediaterisk tumors preoperatively to inform surgical decision-making. On univariable analysis, patients with ATA low-risk disease were more likely to be younger (age <45y), have smaller tumors, and have SUSP FNA results (Table IV). Data on somatic tumor mutations also were available for 431/447 study patients (96%). Because the tumor genotype is another variable that can be readily obtained preoperatively, we also evaluated if the genotype could be a factor in preoperative risk stratification. ATA intermediate-risk disease was present in 50.6% of BRAFV600E or RET/PTC positive PTC, but in only 10.5% of RAS or PAX8/PPARG positive PTC (P = .002, Fig 2). In multivariable analysis and compared with low-risk disease, intermediaterisk disease was associated with age ≥45 years (OR 1.96, P = .005), female sex (OR 1.92, P = .03), larger tumor size (OR 1.73, P = .001), POS FNA results (OR 2.48, P = .008), and BRAF or RET/PTC positive PTC (OR 3.86, P < .001). Discussion We set out to address the problematic reasoning in the 2015 ATA guidelines, which advise paradoxically that decisions about the operative extent of the thyroidectomy should be based on risk factors available only after the operation. The 2015 ATA guidelines now suggest that thyroid lobectomy alone may be sufficient initial treatment for low-risk papillary and follicular carcinoma (Recommendation 35),6 but total thyroidectomy may be needed for disease that is at greater risk of recurrence or persistent and can facilitate RAI ablation therapy and aid in surveillance. Patients with ATA intermediate-risk tumors have features, such as aggressive histology, microscopic extrathyroidal extension, and lymph node metastases, which are not known until the final pathologic analysis. Most large population-level studies of the outcomes of treatment of PTC have stratified patients by their histologic diagnosis, which makes extrapolation to clinical care difficult, because patients present typically by cytologic diagnosis. Thus, the design of this study was to identify what proportion of patients with SUSP or POS FNA
Table III Comparison of clinicopathologic features of patients in the ATA risk categories. ATA 2015
Age (y) Woman Histologic subtype Classic PTC Follicular variant PTC Tall cell variant PTC Other Size (cm) Size ≥4 cm Extrathyroidal extension Multifocal Bilateral Angiolymphatic invasion Positive margins Clinical N1 Central compartment LN dissection Pathologic LN positive M1 Thyroiditis Recurrence/persistent disease
High risk N = 82 (%)
Intermediate risk N = 193 (%)
Low risk N = 171 (%)
P value all groups
P value intermediate versus low
48.4 (37.3–59.7) 48 (58.5)
49.3 (37.4–60.5) 163 (84.5)
44 (35.5–57.4) 134 (78.4)
.114 <.001 <.001
.041 .134 <.001
130 (67.4) 9 (4.7) 52 (26.9) 2 (1) 1.5 (1–2) 0 172 (89.1) 111 (57.5) 75 (38.9) 136 (70.5) 82 (42.5) 0 173 (89.6) 109 (56.5) 0 66 (34.2) 10 (5.2)
131 (76.6) 31 (18.1) 0 9 (5.3) 1.3 (0.9–1.9) 0 0 95 (55.6) 55 (32.2) 63 (36.8) 4 (2.3) 0 152 (88.9) 51 (29.8) 1 (0.6) 64 (37.4) 8 (4.7)
<.001 <.001 <.001 .321 .031 <.001 <.001 <.001 .043 <.001 <.001 .339 <.001
.002 — <.001 .707 .183 <.001 <.001 — .818 <.001 .287 .521 .825
56 (69.1) 8 (9.9) 14 (17.3) 3 (3.7) 2.2 (1.1–4) 25 (30.5) 57 (70.4) 53 (65.4) 40 (49.4) 73 (90.1) 27 (33.3) 61 (74.4) 80 (97.6) 71 (86.6) 7 (8.5) 23 (28) 28 (34.1)
Continuous variables are summarized as medians with interquartile range (25th percentile to 75th percentile). Categorical variables are summarized as frequencies and percentage.
84
Dhiret al / Surgery 163 (2018) 81–87
Fig 1. Frequencies of high-, intermediate-, and low-risk tumors stratified by the preoperative FNA results. White bars depict patients with preoperative positive FNA, and black bars depict patients with preoperative suspicious FNA. High-risk tumors were defined by clinical N1, tumor size ≥4 cm, or synchronous distant metastases. Intermediaterisk tumors were defined by extrathyroidal extension, >5 positive lymph nodes, and/or aggressive histologic variants. The remainder of the tumors were classified as low risk. P value was computed using the χ2 test.
biopsies fall into each of the ATA risk categories as well as to discover any preoperative variables that can be used to help determine the extent of initial thyroidectomy. Our findings demonstrate that preoperative ultrasonography was able to identify ATA high-risk disease in 19% and 15% of patients with POS and SUSP FNA results, respectively. In the remaining patients, the FNA category can help further risk stratify. Many patients with a POS FNA result have ATA intermediate-risk tumors (46%), while the majority of patients with a SUSP FNA result have ATA lowrisk tumors (58%). We also confirmed that POS FNA are more likely than SUSP FNA results to be associated with intermediate-risk tumors on multivariable analysis. Thus, preoperative FNA results and ultrasonography are important, risk-stratification tools and can be helpful in guiding the extent of initial thyroidectomy. Interestingly, lymph node dissection of the central compartment was performed in ~98% of patients; however, having >5 positive lymph nodes was infrequently (5.6%) the only cause for intermediaterisk classification. Differentiating between low- and intermediate-risk tumors using cytology alone, however, is suboptimal. Excluding the patients with
high-risk disease, if lobectomy were performed as the initial operation for all remaining patients, completion thyroidectomy to treat ATA intermediate-risk disease would potentially be required for 53% (193/365) of patients, including the 57% with POS FNA and the 32% with SUSP FNA. Although the complication rate of completion thyroidectomy is comparable with that of initial total thyroidectomy,18 additional risk is associated with a second general anesthesia as well as added costs of care.19,20 In contrast, initial total thyroidectomy for all patients with a SUSP or POS FNA biopsy potentially over treats 47% of patients, including the 43% with POS FNA and the 68% with SUSP FNA results. Cost-efficacy studies to evaluate the impact of lobectomy versus total thyroidectomy for thyroid cancer patients in the long-term including the use of preoperative mutation testing on FNA specimens to guide the extent of thyroidectomy are beyond the scope of the present study but are most certainly needed. Our results are similar to those of Lang et al21 who analyzed 600 patients with PTC 1–4 cm in size without any clinical high-risk features, such as distant metastases, clinical nodal disease, size >4 cm, gross extrathyroidal extension, incomplete tumor resection, macroscopic bilateral disease, and family history or a history of radiation.
Table IV Univariable and multivariable analysis of preoperative factors associated with ATA intermediate-risk versus low risk-tumors. Univariable OR Age Sex Size* FNA Mutation
<45 y ≥45 y Female Male cm Suspicious Positive Negative BRAF or RET/PTC RAS or PAX8/PPARG
* Size was assessed as a continuous variable.
1.822 0.666 1.615 2.872 4.052 0.4
Multivariable 95% CI referent 1.198–2.769 referent 0.391–1.136 1.209–2.156 referent 1.572–5.247 referent 2.134–7.696 0.080–1.987
P value
OR
95% CI
P value
.005
1.956
1.227–3.118
.005
.136 .001
0.519 1.726
0.287–0.939 1.244–2.392
.03 .001
.001
2.475
1.262–4.855
.008
<.001 .263
3.86 0.381
1.973–7.551 0.073–1.972
<.001 .25
Dhiret al / Surgery 163 (2018) 81–87
85
Fig 2. ATA risk categories by 7-gene mutation testing results. P value was computed using the χ2 test.
They reported that 257/600 (42.8%) of the patients had ≥1 unrecognized histologic, high-risk feature (aggressive histology, lymphovascular invasion, microscopic extrathyroidal extent, positive resection margin, and pN1 >0.5 cm), making it likely that these patients would need completion thyroidectomy. Similarly, Kluijfhout et al22 reported that 43% (122/287) of 1- to 4-cm well differentiated thyroid cancers eligible for lobectomy under the current guidelines may need completion thyroidectomy based on histologic results. Based on the present study and previous studies, ~ 40% to 50% (i.e., 1 in 2) patients would require completion thyroidectomy if lobectomy was the initial operative procedure as recommended by the current 2015 ATA Guidelines. Better preoperative risk stratification is needed. We further considered what, if any, preoperative factors could differentiate patients with ATA low-risk disease from those with intermediate-risk disease. ATA intermediate-risk was associated on univariable analysis with age ≥45 years, larger tumor size, and positive FNA. On multivariable analysis, female sex was associated additionally with intermediate-risk disease. We also analyzed available results with mutation testing, because it is another reasonable preoperative variable. We demonstrated previously that thyroid cancers with somatic BRAF V600E mutations or RET/PTC rearrangements have a high incidence of lymph node metastasis, distant metastasis, and recurrence.17 We report here that BRAF V600E or RET/PTC-positivity also was associated with a 4-fold increased risk of ATA intermediate-risk disease. Furthermore, the status of any molecular mutation remained a significant and independent risk factor for ATA intermediate-risk disease on multivariable analysis. Mutation testing can be obtained readily on preoperative FNA biopsy specimens and has been demonstrated to contribute effectively to decision making for indeterminate cytology results.23,24 In the context of the 2015 ATA Guidelines, preoperative mutation testing even on biopsies that are positive for malignancy also can contribute productively to the decision for initial lobectomy or total thyroidectomy. Prospective studies that use preoperative mutation testing to guide the extent of the initial thyroidectomy are ongoing. The limitations of our study include the potential selection bias that may be present in any retrospective single institutional series. We did not re-review ultrasonographic features, because the ability
of ultrasonography to predict the subtle features of the disease accurately that differentiate low and intermediate-risk disease is variable.25 Furthermore, we do not utilize intraoperative frozen analysis routinely to assess lymph node or margin status, which may also contribute incrementally to risk stratification. Because all patients had an initial total thyroidectomy, we do not know how many patients had concerning intraoperative findings that may have changed the extent of thyroidectomy if a lesser resection was planned. We did not include in this analysis any other indications for initial total thyroidectomy, and it is possible that identification of contralateral nodules, underlying thyroiditis, or other clinical considerations may lead surgeons to favor total thyroidectomy instead of lobectomy. Thus, the number of patients who would be true candidates for initial lobectomy may be overestimated here. In summary, total thyroidectomy for high-risk disease was indicated by the 2015 ATA guidelines preoperatively in 18% of patients with POS or SUSP FNA biopsies. If an initial lobectomy was performed for the remaining patients, 53% would have intermediaterisk disease on final histopathology and thus require completion thyroidectomy. Although the risk category based on the FNA biopsy can help risk stratify ATA low- and intermediate-risk disease, additional preoperative variables including age, sex, tumor size, and the presence of BRAF V600E or RET/PTC rearrangements are also important and useful considerations. The implications in terms of cost and risk of the new ATA strategy are substantial, and adjunct tools can improve preoperative risk stratification. References 1. Miller KD, Siegel RL, Lin CC, et al. Cancer treatment and survivorship statistics, 2016. CA Cancer J Clin 2016;66:271-89. 2. Aschebrook-Kilfoy B, Kaplan EL, Chiu BC, Angelos P, Grogan RH. The acceleration in papillary thyroid cancer incidence rates is similar among racial and ethnic groups in the United States. Ann Surg Oncol 2013;20:2746-53. 3. Ahn HS, Welch HG. South Korea’s thyroid-cancer “epidemic”—turning the tide. N Engl J Med 2015;373:2389-90. 4. Davies L, Ouellette M, Hunter M, Welch HG. The increasing incidence of small thyroid cancers: where are the cases coming from? Laryngoscope 2010;120:2446-51. 5. Brito JP, Davies L. Is there really an increased incidence of thyroid cancer? Curr Opin Endocrinol Diabetes Obes 2014;21:405-8.
86
Dhiret al / Surgery 163 (2018) 81–87
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. Adam MA, Pura J, Gu L, et al. Extent of surgery for papillary thyroid cancer is not associated with survival: an analysis of 61,775 patients. Ann Surg 2014;260:601-5, discussion 5-7. 8. Matsuzu K, Sugino K, Masudo K, et al. Thyroid lobectomy for papillary thyroid cancer: long-term follow-up study of 1,088 cases. World J Surg 2014;38:68-79. 9. Haigh PI, Urbach DR, Rotstein LE. Extent of thyroidectomy is not a major determinant of survival in low- or high-risk papillary thyroid cancer. Ann Surg Oncol 2005;12:81-9. 10. Vaisman F, Shaha A, Fish S, Michael Tuttle R. Initial therapy with either thyroid lobectomy or total thyroidectomy without radioactive iodine remnant ablation is associated with very low rates of structural disease recurrence in properly selected patients with differentiated thyroid cancer. Clin Endocrinol (Oxf) 2011;75:112-9. 11. Jonklaas J, Cooper DS, Ain KB, et al. Radioiodine therapy in patients with stage I differentiated thyroid cancer. Thyroid 2010;20:1423-4. 12. Gweon HM, Koo HR, Son EJ, et al. Prognostic role of the Bethesda System for conventional papillary thyroid carcinoma. Head Neck 2016;38:1509-14. 13. Ohori NP, Wolfe J, Hodak SP, et al. Colloid-rich” follicular neoplasm/suspicious for follicular neoplasm thyroid fine-needle aspiration specimens: cytologic, histologic, and molecular basis for considering an alternate view. Cancer Cytopathol 2013;121:718-28. 14. Cibas ES, Ali SZ. The Bethesda system for reporting thyroid cytopathology. Am J Clin Pathol 2009;132:658-65. 15. American Thyroid Association (ATA) Guidelines Taskforce on Thyroid Nodules and Differentiated Thyroid Cancer, Cooper DS, Doherty GM, et al. Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid 2009;19:1167-214.
16. Cooper DS, Doherty GM, Haugen BR, et al. Management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid 2006;16:10942. 17. Yip L, Nikiforova MN, Yoo JY, et al. Tumor genotype determines phenotype and disease-related outcomes in thyroid cancer: a study of 1,510 patients. Ann Surg 2015;262:519-25, discussion 24-5. 18. Li YJ, Wang YZ, Yi ZB, Chen LL, Zhou XD. Comparison of completion thyroidectomy and primary total surgery for differentiated thyroid cancer: a meta-analysis. Oncol Res Treat 2015;38:528-31. 19. Lang BH, Wong CK. Lobectomy is a more cost-effective option than total thyroidectomy for 1 to 4 cm papillary thyroid carcinoma that do not possess clinically recognizable high-risk features. Ann Surg Oncol 2016;23:3641-52. 20. Yip L, Farris C, Kabaker AS, et al. Cost impact of molecular testing for indeterminate thyroid nodule fine-needle aspiration biopsies. J Clin Endocrinol Metab 2012;97:1905-12. 21. Lang BH, Shek TW, Wan KY. The significance of unrecognized histological high-risk features on response to therapy in papillary thyroid carcinoma measuring 1–4 cm: implications for completion thyroidectomy following lobectomy. Clin Endocrinol (Oxf) 2017;86:236-42. 22. Kluijfhout WP, Pasternak JD, Lim J, et al. Frequency of high-risk characteristics requiring total thyroidectomy for 1–4 cm well-differentiated thyroid cancer. Thyroid 2016;26:820-4. 23. Yip L, Wharry LI, Armstrong MJ, et al. A clinical algorithm for fine-needle aspiration molecular testing effectively guides the appropriate extent of initial thyroidectomy. Ann Surg 2014;260:163-8. 24. Nikiforov YE, Steward DL, Robinson-Smith TM, et al. Molecular testing for mutations in improving the fine-needle aspiration diagnosis of thyroid nodules. J Clin Endocrinol Metab 2009;94:2092-8. 25. Lee CY, Kim SJ, Ko KR, Chung KW, Lee JH. Predictive factors for extrathyroidal extension of papillary thyroid carcinoma based on preoperative sonography. J Ultrasound Med 2014;33:231-8.
Discussion Dr Askok R. Shaha (New York, NY): Very nice study. Very well analyzed. There are 2 points I want to bring up. When we talk about minor extrathyroidal extension, as you know very well, the staging system is going to change. The AJCC staging manual is already out, but it will be implemented next year. It does not increase the T stage. In the past, it used to be certainly T3. Now it is not going to be T3 anymore. And there are very nice papers from Sloan Kettering and Mayo Clinic showing that minor extrathyroidal extension does not have any major prognostic implication. The second, when you change the staging from low risk to intermediate risk, there is an emotional concern, but I am not convinced that all those patients will require a completion and will benefit from RAI. The entire premise of doing a completion is to benefit the patient with RAI. What is more important is to define who needs RAI, and those are the only patients who should go for completion thyroidectomy. The remaining intermediate-risk group can be easily monitored. Thank you. Dr Linwah Yip: Thank you. You are absolutely right. As we all know, staging and the risk categories are a moving target, and so for the purposes of the study we used what was described in the most recent set of ATA Guidelines. And you are correct, radioactive iodine may not always be necessary for an intermediate risk lesion. In fact, our endocrinologists do not give radioactive iodine frequently either, but they often still would like to follow those thyroglobulin levels and sometimes do request to have a completion for that reason. Thank you. Dr Julie Ann Sosa (Durham, NC): So, Lin, fabulous study from, really, the leading edge of innovation in molecular testing. But I think that sort of begs my first question, which is, many of us live in a resource limited world where we do not have access yet to molecular testing to allow us to preoperatively profile tumors to try to
better risk-stratify them. Instead, we use ultrasound, and, obviously, this may have been a limitation of your study, but I think many of us believe ultrasound, at least combined with molecular testing, or used on its own, is a very powerful tool to help guide extent of resection. So my first question is, if you had been able to retrospectively look at those ultrasounds, do you believe that might have changed your results? And, going forward, incorporating other institutions into this study who maybe rely less on molecular profiling and more on ultrasound, might that add to the wealth of the data you accumulate, so a multi-institutional approach? My second question builds on Dr. Shaha’s, and that is the evolving staging system. Extrathyroidal extension is one, but the second is debate about micro versus macro metastatic disease. And are those really equal or are micrometastasis more similar to the absence of metastasis? I wondered if you had done a sensitivity analysis taking those out of a higher risk category, micrometastatic disease, and putting them into a lower risk category. Thank you, again. Terrific study. Dr Linwah Yip: So to answer your second question, we used histologic variables and correlation with published ATA risk categories as our study outcome. We did not use disease related outcomes such as recurrence and mortality, specifically, so I cannot answer if reclassifying extrathyroidal extension or extent of lymph node involvement would make a difference in preoperative risk stratification. But I think that is a very interesting question and certainly begs for more than our 447-patient cohort to be able to probably reach those answers. Multi-institutional studies are a great idea and we would love to do that. Our ultrasound data was a bit limited, because this was retrospective. For example, the degree of extrathyroidal extension is not routinely reported in every single one of our ultrasound reports, and is unfortunately not assessed with every patient es-
Dhiret al / Surgery 163 (2018) 81–87
pecially if cancer is not known during the initial diagnostic study. It would be something to look for prospectively, as well. Thank you. Dr Douglas Evan (Milwaukee, WI): Lin, that was fantastic. Maybe you could spend 2 minutes on your clinical trial and the sample size that you need and what your primary endpoint is. And also since your bias is reflected in your data that you presented, historically you performed total thyroidectomy, and I suspect probably because you agreed with doing that, not just that you read it in the ATA guidelines, how successful have you been in enrolling patients, and how do you answer the question when the patient says, “Well, Dr Yip, what would do you if this were you or your mom?” Dr Linwah Yip: So our endpoint for the trial is dependent on histologic variables, and we are specifically looking at how accurately preoperative molecular testing helps us predict the need for lobectomy when the biopsy is positive or suspicious. When using that specific endpoint, and the results of this study, ~ 40% were being overtreated. Our calculated sample size is 90 patients. And so we just opened the trial in February and have already accrued 5 patients.
87
It has not been difficult to enroll and the conversation that we all currently have with patients is complicated. For example, if someone comes in with a 2-cm nodule that is positive for PTC and their ultrasound does not show any lymph node disease, then we have this very long extensive discussion e.g., you can have a lobe, you can have a total, you may need to have a completion, etc. As we all know, that is a very long discussion that can be confusing to patients. With the trial, I now say, we can potentially use molecular testing to help us sort out if we need to do lobe or a total. We explain that we do not know if it is going to help, but that is an extra piece of information that we can provide if you enter this trial. And I have actually had patients say, well, that makes a lot of sense, why do we not do that for everybody. So it has not been difficult to talk to patients about it at all. Dr Douglas Evans (Milwaukee, WI): So you are 5 for 5 now? No one has refused? Dr Linwah Yip: Five of 5, correct. No one refused.