Management of Medullary Thyroid Cancer: Patterns of Recurrence and Outcomes of Reoperative Surgery

Abstract Background There remains uncertainty regarding the optimal extent of initial surgery and management of recurrent disease in medullary thyroid cancer (MTC). We aim to describe the patterns of disease recurrence and outcomes of the reoperative surgery in a cohort of consecutively treated patients at a specialized tertiary referral center. Patients and Methods A retrospective cohort study of 235 surgically treated patients with MTC at a tertiary referral center was performed using prospectively collected data. Results In the study period 1986-2022, 235 patients underwent surgery for MTC. Of these, 45 (19%) patients had reoperative surgery for cervical nodal recurrence at a median (range) 2.1 (0.3-16) years following the index procedure. After a median follow-up of 4 years, 38 (84%) patients remain free of structural cervical recurrence, although 15 (33%) underwent 2 or more reoperative procedures. No long-term complications occurred after reoperative surgery. Local cervical recurrence was independently predicted by pathologically involved nodal status (OR 5.10, P = .01) and failure to achieve biochemical cure (OR 5.0, P = .009). Local recurrence did not adversely affect overall survival and was not associated with distant recurrence (HR 0.93, P = .83). Overall survival was independently predicted by high pathological grade (HR 10.0, P = .002) and the presence of metastatic disease at presentation (HR 8.27, P = 0018). Conclusion Loco-regional recurrence in MTC does not impact overall survival, or the development of metastatic disease, demonstrating the safety of the staged approach to the clinically node-negative lateral neck. When recurrent disease is technically resectable, reoperative surgery can be undertaken with minimal morbidity in a specialized center and facilitates structural disease control.


Introduction
Medullary thyroid cancer (MTC) is a rare neuroendocrine tumor of parafollicular (C cells) and accounts for 2% of thyroid malignancies; however, it is responsible for 8% of thyroid cancer-related deaths. 1 Most medullary thyroid carcinomas are sporadic and typically present between the fourth and sixth decades of life, often with nodal involvement or distant metastatic disease. 2,3Approximately, 15%-20% of MTC cases are familial and associated with a germline mutation in the rearranged during transfection (RET) protooncogene (multiple endocrine neoplasia type 2, MEN2). 4Because of its propensity for vascular invasion, lymph node involvement 5,6 and distant metastasis, MTC is notoriously difficult to cure.However, it is often associated with an indolent clinical course and may take years or decades to progress, with an overall survival of approximately 70%-80% at 10 years. 7,8herefore, clinicians must strike an effective balance between achieving adequate oncological management and minimizing surgical morbidity.
In the treatment of MTC with curative intent, guidelines recommend total thyroidectomy and bilateral central compartment dissection, with additional lateral compartmental dissection if there is evidence of lateral nodal involvement. 94][15] Historically, due to the lack of effective adjuvant therapies for MTC and the inaccuracy associated with staging ultrasound (US) 16 and intraoperative nodal assessment, 5 aggressive surgical management of the lymph node basins was often favored. 7,8However, with the increasing diagnostic sensitivity of high-resolution US and CT, 17,18 as well as the emergence of effective-targeted therapies for patients with metastatic disease, [19][20][21] there is a trend to de-escalating surgical treatment of the clinically negative lateral neck compartments. 22Aggressive lymphadenectomy has not been demonstrated to result in a long-term survival benefit 14,23,24 despite resulting in lower postoperative calcitonin levels. 25atients with a normal (<10 pg/mL) postoperative calcitonin level achieve "biochemical cure" and have a 97% 10-year survival, 26 with a 3% recurrence risk. 27However, elevated serum calcitonin after seemingly adequate surgical excision is a common clinical problem, and the approach to further investigation and surgical management is variable.In particular, the decision to consider further nodal dissection must weigh the often indolent course of the disease, the risk of distant failure, and potential treatment-related morbidity. 28,291][32][33][34] Recently, a pathological grading system has been described and validated as an independent predictor of survival in MTC. 35e aim to describe the patterns of recurrent disease in patients with MTC and explore predictive clinical and pathological factors.In addition, we aim to describe the safety and outcomes of reoperative surgery for the management of locoregional recurrence.

Study Design and Study Population
A retrospective cohort study was undertaken.The study population included patients treated surgically for MTC in the period 1986 and 2022 at a single tertiary referral endocrine surgery unit.A comprehensive thyroid cancer database, which contains prospectively collected preoperative, operative, and follow-up data, was utilized to identify cases.Approval to access these data and perform the study was obtained from the Northern Sydney Local Health District Ethics Committee (reference 2020/ETH02787).

Initial Surgical Management
Patients with clinically apparent tumors and a clear preoperative diagnosis of MTC were treated with a total thyroidectomy and bilateral central compartment lymph node dissection (CLND).Patients with atypical preoperative cytology underwent a staged completion thyroidectomy and consideration of bilateral CLND once histology was confirmed.The decision to perform lateral compartmental dissection was individualized and made by the treating surgeon based on results of serum calcitonin and imaging.The extent of lateral nodal dissection was defined according to accepted definitions 36 and described according to the levels dissected.Our routine approach to the lateral neck in MTC includes dissection of levels II, III, IV, and Vb.

Histopathological Analysis
A specialist endocrine pathologist confirmed the diagnosis of MTC and provided a structured report on all tumor specimens using SDs. 37,38MTC grade was classified according to The International Medullary Thyroid Cancer Grading System, with high-grade defined as a tumor demonstrating one of: tumor necrosis, mitotic rate ≥5 per 2 mm 3 , or Ki67 proliferation index ≥5%.Other pathological features included tumor size, vascular invasion, extrathyroidal extension (ETE), nodal status, and the presence of extranodal extension.Tumors were staged according to the American Joint Committee on Cancer (AJCC) eighth edition. 39

Definitions of Recurrence
Local cervical recurrence was defined as radiological evidence of recurrent disease, visualized on CT, US, or DOTATATE PET scan.Local recurrence was confirmed on cytological analysis of fine-needle aspiration biopsy (FNAB) or on histopathological analysis of reoperative surgical specimens.Distant metastases were defined as structural abnormalities on cross-sectional imaging and were confirmed on histopathological analysis of a core biopsy specimen if required.

Statistical Analysis and Outcomes
Data were analyzed using Jamovi (The jamovi project [2022].jamovi (Version 2.3) [Computer Software].Retrieved from https://www.jamovi.org).Data were expressed as mean ± SD if normally distributed, as median (range) for nonparametric data, or n (%) for descriptive statistics.The Student's t test was used to compare continuous variables.Pearson's chisquare test or logistic regression analysis was used for categorical variables.Kaplan-Meier survival analysis was used to estimate the overall recurrence and survival rate.The Cox proportional hazard regression model was used to identify prognostic factors for overall survival and disease-specific survival.The log-rank test was used to compare differences in survival outcomes.A P-value of <.05 was considered statistically significant.

Results
A total of 235 consecutive patients with previously untreated MTC underwent surgical intervention in the period 1986 and 2022.Twenty-one (9%) patients were lost to follow up.The median age at presentation was 53 years (range 10-96 years), there were 133 females (57%), and the mean diameter of the primary tumor was 22 mm (SD = 18 mm).Thirty patients (13%) had germline mutations of the RET gene.The median serum calcitonin level at presentation was 1200 pg/mL (range 5-65 000).The demographic, clinical, and pathological data are summarized in Table 1.Twenty-one patients (9%) had distant metastatic disease at presentation, and 126 (54%) had advanced (stage III or IV) disease.Vascular invasion was seen in 91 patients (39%), and 123 (52%) had node-positive disease.Lateral CLND was performed in 100 (43%) patients, including 17 (7%) who underwent bilateral lateral compartment dissections.Twelve (5%) patients underwent prophylactic lateral neck dissection.

Patterns of Local Disease Recurrence
The median follow-up time was 48 months (range 1-370).Cervical nodal recurrence occurred in 55 (23%) patients at a median time of 25 months (range 3-192) after the index operation, and 45 (80%) patients were treated with repeat surgery.
Overall, 10 (4%) patients remain under observation for asymptomatic, low-volume recurrent cervical nodal disease.Local irresectable cervical failure occurred in 2 (0.9%) patients, and irresectable mediastinal nodal disease in 29 (12%) patients, including 5 (2%) patients who presented with advanced mediastinal nodal disease.Independent prognostic factors for local recurrence included an increased risk with pathologically involved nodal status (OR 5.10, P = .01),and a decreased risk associated with biochemical cure after the index operation (OR 0.20, P = .009;Table 2).

Reoperative Surgery
A total of 65 reoperative procedures were performed in 45 (19%) patients during the study period.Two patients (4%) remain biochemically cured.Four (9%) remain under observation for low-volume nodal recurrence in the setting of other medical comorbidities, and 39 (87%) patients remain free of structural disease in the neck at the last follow-up.However, 15 (33%) have required 2 or more reoperative procedures.The median time to second nodal recurrence, and third local recurrence after initial surgery were 38 months (range 10-120) and 12 months (range 7-36), respectively.The median (range) calcitonin levels were significantly reduced by reoperative surgery (500 [14-34 600] pg/mL, and 138 [1-7560] pg/mL, respectively, P = .04).
The rate of temporary hypoparathyroidism after revision central compartment dissection was 16%, and there were no cases of permanent hypoparathyroidism.There were no temporary or permanent recurrent laryngeal nerve palsies, and there were no major complications of reoperative surgery in the lateral compartment.
Eighteen patients (8%) died from MTC during the study period, and 30 (13%) died from other causes.Among patients with metastatic disease, overall survival from the time of diagnosis of metastatic disease was shorter in patients with visceral metastases (estimated medial survival 22 months vs. 38 months, P = .050).One patient was palliated with locally advanced irresectable disease, before the advent of TKI therapy.

Prediction of Survival
The 10-year overall survival (OS) and disease-specific survival (DSS) were 71% and 87%, respectively (Fig. 2).The predictors of OS and DSS are summarized in Tables 4 and 5. On univariate analysis for OS, nodal positivity, vascular invasion, extrathyroidal extension, high pathological grade, the presence of metastatic disease, and age >55 years predicted shorter survival.Local recurrence did not adversely impact overall survival.Similarly, failure to achieve biochemical cure was not a significant adverse prognostic factor for survival.On multivariate analysis, only high pathological grade and the presence of metastatic disease on presentation remained independently predictive.The multivariate analysis results were similar for DSS.

Discussion
This study examines the patterns of failure and the results of reoperative surgical management in MTC.Failure to achieve biochemical cure after initial therapy and pathological N1  nodal status are independent predictors of local recurrence, which occurs in approximately 23% of patients at extended follow-up.Reoperative surgery can be safely performed in a specialist center and is effective in achieving local structural disease control, however, up to a third may require additional reoperative procedures.Similarly, in patients with resectable mediastinal nodal disease, or oligometastatic disease of the liver or lung, metastectomy should be considered, as it often provides extended structural disease control.Local recurrence does not increase the risk of metastatic disease or death.In predicting survival, high pathological tumor grade and the presence of metastatic disease on presentation are the key independent prognostic factors.
In MTC, local recurrence or persisting disease is a common problem, occurring in up to 79% of patients at extended follow-up. 40In the setting of non-metastatic, locally recurrent disease, the data guiding reoperative management are limited.A correlation between survival and radiological remission has been previously demonstrated, indicating that reoperative surgery to excise recurrent disease appears to be beneficial. 41ur data indicate that reoperative surgery can be performed with minimal morbidity and effectively provides structural disease control, even if biochemical cure is not achieved.This includes resection of nodal disease within the mediastinum, and oligometastatic disease of the liver or lung.As systemic treatments improve and allow prolonged survival even in the setting of visceral metastatic disease, 19 the importance of local control and the role of reoperative surgery may increase.
In predicting survival, adverse prognostic factors on univariate analysis were similar to previous reports 30,34,40,[42][43][44] and included nodal positivity, vascular invasion, extrathyroidal extension, failure to achieve excellent response to initial therapy and the presence of metastatic disease.In addition, the pathological tumor grade was highly predictive of both overall and disease-specific survival.When tumor grade is included in the multivariate analysis, the pathological features of vascular invasion and extrathyroidal extension lose significance, and only tumor grade and the presence of distant metastases at presentation remain independently predictive.Interestingly, excellent response to initial therapy was not independently predictive of survival.This suggests that the group of patients with persistently elevated calcitonin, but a low pathological grade is likely to follow an indolent disease course.
The role of prophylactic lateral compartmental dissection in clinically N0 patients is controversial.With the quality of contemporary high-resolution US, the benefit of prophylactic versus staged lateral compartment dissection in cN0 patients is debatable, especially when considering the increased risk of complications and the lack of proven long-term survival benefit. 14,23,24,45International guidelines lack consensus on this issue, with current American Thyroid Association (ATA) guidelines commenting that lateral lymph node dissection should be considered based on serum calcitonin levels, 10 informed predominantly by the work of Machens et al who noted that preoperative calcitonin levels of >200 pg/mL were predictive of lateral lymph node metastases. 46More recently, Pena et al reported the safety of observation of the cN0 lateral neck, with 5% of patients developing ipsilateral lateral nodal recurrence amenable to salvage surgery at a median follow-up of 2.3 years. 22Similarly, Spanheimer et al noted that no difference in the cumulative incidence of distant recurrence, disease-specific survival, or overall survival between cN0 patients who were electively dissected, compared to those who were observed. 45Our data reinforce these findings.All cervical nodal recurrences that occurred in patients with cN0 disease who did not undergo lateral compartment dissection were amenable to salvage surgery.Furthermore, local recurrence did not influence the pace of disease with respect to distant recurrence, DSS or OS.This demonstrates the safety of a staged approach to the lateral neck in the cN0 patient.Our data support our current approach to the lateral neck in MTC: all patients undergo dedicated high-resolution ultrasonographic assessment of bilateral lateral compartments, performed by a sonographer and reviewed by a specialist radiologist.A FNAB is performed if there is suspicion of nodal involvement.If there is no evidence of disease, the lateral compartments are not dissected, irrespective of the preoperative calcitonin level.Postoperatively, calcitonin levels are closely monitored, and the lateral compartments carefully observed with serial US.In addition, we frequently utilize DOTATATE PET scan in the postoperative period if calcitonin levels suggest persistent or recurrent disease.
The main strength of this study is the large cohort of consecutively treated patients at a tertiary referral center for  MTC, by highly specialized thyroid surgeons, endocrinologists, radiologists, pathologists, and oncologists, with a meticulously maintained prospective data collection process and a low proportion of patients lost to follow up.The large sample size enabled analysis of the predictors for local recurrence and survival outcomes.The main limitation of this study is its retrospective design, which impacted our ability to capture all data, especially with respect to preoperative and postoperative calcitonin levels.Furthermore, the cohort of patients treated at our center may be subject to selection bias toward more advanced tumors.

Conclusion
Local recurrence in MTC does not predict distant failure or adversely impact survival, demonstrating the safety of the staged approach to the clinically N0 lateral neck.When recurrent disease is technically resectable, reoperative surgery can be undertaken with minimal morbidity in a specialist center and facilitates structural disease control.

Figure 1 .
Figure 1.Development of distant metastatic disease stratified by local cervical recurrence.Abbreviation: LR: local recurrence.

Table 1 .
Demographic, clinical, and pathological features of patients with medullary thyroid cancer.
Data are presented as mean ± SD or median and range unless otherwise indicated.a Data available for 233 patients.b Data available for 84 patients.c Data available for 87 patients.Abbreviation: RET: rearranged during transfection gene.

Table 2 .
Medullary thyroid cancer characteristics as prognostic factors for local recurrence.

Table 3 .
Predictors of development of distant metastases.