Disease-specific survival for patients in groups 1, 2, and 3.
Overall survival for patients in groups 1, 2, and 3.
Palme CE, Waseem Z, Raza SN, Eski S, Walfish P, Freeman JL. Management and Outcome of Recurrent Well-Differentiated Thyroid Carcinoma. Arch Otolaryngol Head Neck Surg. 2004;130(7):819-824. doi:10.1001/archotol.130.7.819
The AMES (age, distant metastasis, tumor extent, and size), AGES (age, tumor size, histologic grade, tumor extent, distant metastasis), and MACIS (distant metastasis, age, completeness of primary tumor resection, local invasion, and tumor size) prognostic systems for well-differentiated thyroid carcinoma (WDTC) are well known. The development of disease recurrence is associated with a poor outcome; however, the prognostic importance of multiple treatment failures has not been clearly reported.
To identify patient, tumor, and treatment factors that may be associated with the development of multiple recurrences in WDTC.
Design and Setting
All patients treated for residual or recurrent WDTC were retrospectively identified from the thyroid cancer database at the Department of Otolaryngology–Head and Neck Surgery, Mount Sinai Hospital, Toronto, Ontario (1963-2000). Data on relevant patient, tumor, and treatment factors were collected.
Main Outcome Measures
Patient, tumor, and treatment factors predicting the development of multiple treatment failures, disease-specific survival, and overall survival.
A total of 574 patients (115 male, 459 female; median age, 42 years [range, 9-92 years]) were identified, whose final histopathologic diagnosis was papillary carcinoma in 468, follicular carcinoma in 76, and mixed in 30 cases. TNM staging was as follows: 409 (71%) stage I, 66 (12%) stage II, 68 (12%) stage III, and 31 (5%) stage IV. Initial management included total thyroidectomy for 217 patients (38%), subtotal thyroidectomy for 357 (62%), and adjuvant iodine 131 therapy for 492 (86%). Seventy-three patients (13%) developed recurrent WDTC (21 male, 52 female; median age, 44 years [range, 18-84 years]). Patients were divided into 3 groups: group 1 (no recurrence, n = 501), group 2 (1 recurrence only, n = 42), and group 3 (multiple recurrences, n = 31). Group 2 data were as follows: site of recurrence (locoregional, 25; distant, 7; unspecified, 10) and treatment (surgery, 12; iodine 131, 42) and for group 3: site of first recurrence (locoregional, 16; distant, 11; unspecified, 4) and treatment (surgery, 14, iodine 131, 22; palliation, 1). Actuarial disease-specific survival at 20 years was 100%, 94%, and 60%, respectively, for the 3 groups (median follow-up, 7 years; range, 1-34 years). Male sex, advanced stage, extrathyroidal spread, and primary treatment with total thyroidectomy were predictive factors for multiple recurrences on multivariate regression (all P<.05).
Male sex, advanced initial stage, and presence of extrathyroidal spread within the primary tumor are the most significant independent predictors of developing multiple recurrences in patients with WDTC. These patients have a poor prognosis with a significant reduction in tumor-free survival.
Well-differentiated thyroid carcinoma (WDTC) accounts for 2% of all cancers and is responsible for fewer than 0.5% of all cancer deaths.1 Combination therapy including surgery and adjuvant iodine 131 therapy is the treatment of choice. It results in 10-year survival rates of greater than 90%. Factors such as age, sex, size of the tumor, stage of disease, presence of extrathyroidal spread (ETS), and completeness of resection have been found to significantly influence prognosis.1 Despite best practice, recurrence rates are reported in the vicinity of 8% to 23%.2,3 Both thyroglobulin estimation and iodine 131 scanning have a high yield in detecting recurrent disease with a combined sensitivity of over 90% and are therefore used routinely in the posttreatment monitoring of these patients.1 In addition, both clinical examination and imaging with modalities such as computed tomographic (CT) scanning are used in the follow-up but are less sensitive and specific in the detection of recurrent WDTC.4 Most patients with recurrent WDTC undergo salvage treatment with further surgery and/or iodine 131 therapy. However, a small number of patients will die as a result of uncontrolled locoregional or distant disease. The mortality of patients with a recurrence has been reported as high as 38% to 69%.5,6 One would assume that those who have a recurrence once are at an increased risk of multiple recurrences. The theory behind this is possible dedifferentiation of the tumor, lack of iodine 131 uptake, and a significant risk of leaving behind microscopic disease after salvage surgery. However, long-term survival is possible in these patients with persistent and recurring disease.7 The purpose of our study was to evaluate outcome in patients with residual or recurrent WDTC and identify patient, tumor, treatment, and first-recurrence factors that may predict for the development of multiple treatment failures.
A retrospective chart review of the thyroid carcinoma database at Mount Sinai Hospital (1963-2000) was carried out to identify all patients treated for WDTC, including those who developed a recurrence after primary therapy. Recurrence was defined as any evidence of disease requiring further therapy after initial curative treatment. Patient, tumor, treatment, and first-recurrence data were collected and included age, sex, original tumor data (pathologic diagnosis, size, multifocal/solitary, stage, ETS, extent of primary surgery, adjuvant iodine 131 therapy) and information about the site (locoregional, distant, and unspecified), mode of detection (clinical, imaging, thyroglobulin estimation), and treatment (surgery, iodine 131 therapy) of any disease recurrence. Extent of primary disease was retrospectively staged according to the current American Joint Committee on Cancer (AJCC) staging system for WDTC.8 Time to recurrence was defined as the time interval between completion of treatment of the primary lesion and detection of residual or recurrent WDTC. The location of the recurrence was labeled as unspecified in patients with an elevated thyroglobulin level and no positive localization of disease based on clinical examination and standard imaging modalities (ie, ultrasonography, iodine 131 scanning, and CT) available at the time of presentation. Patients were then separated into 3 groups. Group 1 were patients who developed no recurrence after primary therapy. Group 2 consisted of those who only experienced a single recurrence. Group 3 included patients who had multiple recurrences. Management and outcome data pertaining to the first recurrence were collected in group 3. Treatment success was determined on the basis of undetectable thyroglobulin estimation and/or normalization of imaging modalities including iodine 131 scanning, ultrasonography, and CT scanning.
Outcome was documented as alive, no disease; alive with disease; dead, no disease; and dead of disease. Follow-up was counted from completion of treatment of the last known recurrence.
Disease-specific outcome data were calculated using the Kaplan-Meier method and curves were compared when using the log-rank test. P<.05 was deemed statistically significant. Patient, tumor, and treatment factors associated with having multiple recurrences were statistically analyzed using the method of bivariant correlation. Multiple regression was carried out using the Cox proportional hazards model and Backward Wald Test (entry P<.05; removal P>.1). All statistics were carried out using SPSS software version 11.0 (SPSS Inc, Chicago, Ill).
Our analysis identified 574 consecutive patients who were managed for previously untreated WDTC at the Department of Otolaryngology–Head and Neck Surgery, Mount Sinai Hospital, Toronto (3 surgeons and 1 senior endocrinologist). Patient, tumor, and treatment data are summarized in Table 1. Seventy-three patients (13%) with either residual or recurrent disease required further active intervention. On the basis of treatment outcome, patients were divided into group 1 (no recurrence), group 2 (1 recurrence only), and group 3 (>1 recurrence) (Table 2).
In group 1 (n = 501), there were 407 female patients and 94 male patients (median age, 45 years; range 9-92 years). The final histopathological diagnosis was papillary carcinoma in 407, follicular carcinoma in 64, and mixed in 30 cases. The median size of the dominant nodule was 2.3 cm (range, 1-8 cm). The disease was multifocal in 167 cases. Extrathyroidal spread was identified in 27 tumors (5%). The AJCC staging was stage I in 394, stage II in 36, stage III in 49, and stage IV in 22 patients. The majority of patients (300) were treated with a subtotal thyroidectomy while 201 received a total thyroidectomy. Adjuvant radioactive iodine was used in 422 cases (84%). The median follow-up was 7 years (range, <1-34 years). The final disease outcome was alive with no disease in 479 and 22 patients died of causes unrelated to their WDTC. The disease-specific and overall survival as actuarially predicted by the Kaplan-Meier method at 20 years was 100% and 89%, respectively.
There were 31 female and 11 male patients in group 2 (n = 42) with a median age of 45 years (range, 18-82 years). Primary histologic diagnosis included papillary carcinoma in 35 patients and follicular carcinoma in 7. The tumor was staged as stage I in 25, stage II in 3, stage III in 11, and stage IV in 3 patients. Median tumor size was 2 cm (range, 1-4 cm). Multifocality of WDTC was present in 21 patients. Extrathyroidal spread was present in 6 patients (14%). Primary therapy included total thyroidectomy in 9 patients, subtotal thyroidectomy in 33, and adjuvant iodine 131 therapy in 42.
The median age of group 3 patients (n = 31) was 49 years (range, 24-79 years). There were 21 female and 10 male patients. Tumor histologic grade included papillary carcinoma in 26 and follicular carcinoma in 5 patients. The primary disease was stage I in 12, stage II in 3, stage III in 3, and stage IV in 13 patients. The median tumor size was 3 cm (range, 1-6 cm). The tumor was multifocal in 21 patients. Extrathyroidal spread was present in 15 patients (48%). Seven patients were treated initially with total thyroidectomy and 24 received a subtotal thyroidectomy. Adjuvant iodine 131 therapy was given to 28 patients (90%).
Median time to recurrence in group 2 patients was 7 months (range, 1-194 months). The location of the recurrence was identified as locoregional in 25, at distant sites in 7, and at an unspecified location in the remaining 10 patients. The recurrence was detected in group 2 patients based primarily on clinical examination in 6, on thyroglobulin estimation in 19, or by imaging (CT and or ultrasonography) in 17 patients. These patients were all treated with iodine 131. Salvage surgery was possible in 12 cases (29%). This included resection of locoregional disease in 11 and a solitary bone metastasis in 1 patient. External beam radiotherapy was used in 2 patients with extensive locoregional disease. At the last point of contact 37 patients are cured, 2 are alive with disease, 1 died of disease, and 2 patients died of other causes. The actuarial disease-specific and overall survival at 20 years was 94% and 83%, respectively. The median follow-up in this group was 10 years (range, 1-26 years).
Time to the first recurrence in group 3 patients was 7 months (range, 1-189 months). The location of the first recurrence was identified as locoregional in 21 and at distant sites in 6 patients. The site was unspecified in 4 patients. Primary detection of the recurrence was by clinical palpation in 11, via thyroglobulin estimation in 11, or by imaging (CT and/or ultrasonography) in 9 patients. Surgery was possible in 12 patients with locoregional disease and in 2 with isolated bone metastases (45%). Iodine 131 was given as primary therapy in 16 and in an adjuvant setting in 6 patients. One patient received palliation only. The actuarial disease-specific and overall survival at 20 years was 60% and 58%, respectively. Six patients are alive with no disease, 16 are alive with disease, 8 died of their disease, and 1 patient died of an unrelated cause. Follow-up time was calculated from completion of treatment of the last recurrence and is 5 years (range, 1- 20 years).
Univariate analysis confirmed the importance of advanced initial stage, size, and the presence of ETS, in addition to both clinical detection and surgical management of the first recurrence, as significant for the development of multiple treatment failure (P<.01, <.01, <.01, .03, and .02, respectively). Male sex, advancing AJCC stage, the presence of ETS, and treatment with total thyroidectomy were independent predictors of multiple treatment failures on multivariate regression (P = .04, .01, .02, .02, respectively).
We carried out a separate evaluation of patient, tumor, and treatment factors considering only those patients who developed a recurrence. Univariate analysis confirmed the importance of advanced initial stage and the presence of ETS, in addition to both clinical detection (P = .03) and surgical management (P = .02) of the first recurrence, as significant for the development of multiple treatment failure (P<.01, <.01, .03, .02, respectively) (Table 3). Extrathyroidal spread was the only independent factor significant on multivariate regression (P = .03).
Factors that were not found to be significant for the development of multiple treatment failures included age, histopathologic grade, multifocality, and time to the first recurrence. In addition, location of the first recurrence (ie, locoregional, distant, or unspecified) and treatment with iodine 131 did not predict for those that occurred on multiple occasions.
Actuarially predicted disease-specific survival at 20 years for patients who were cured, who had only 1 recurrence, and those who developed multiple treatment failures was 100%, 94%, and 60%, respectively. This was statistically significant (P<.001) (Figure 1). Similarly, actuarially predicted overall survival at 20 years was 89% for group 1, 83% for group 2, and 58% for group 3 patients, also statistically significant (P<.001) (Figure 2).
Well-differentiated thyroid carcinoma is associated with good disease-specific outcome in the majority of patients. The most important prognostic factors are age and sex.9 In addition, other tumor and treatment factors form the basis of the AMES (age, distant metastasis, tumor extent, and size), AGES (age, tumor size, histologic grade, tumor extent, distant metastasis), and MACIS (distant metastasis, age, completeness of primary tumor resection, local invasion, and tumor size) prognostic scoring systems.10 This has resulted in the recognition of low- and high-risk patient categories and allowed meaningful comparison of a variety of treatment approaches. In general, young female patients have the best disease-specific outcome and have allowed a more conservative surgical approach. However, male patients and those older than 45 years have warranted more aggressive combination therapy, which includes total thyroidectomy and adjuvant radioactive iodine. Despite this, recurrence rates are significant and portend a worse prognosis in some patients. However, treatment failure does not universally lead to death, as the majority of patients have successful salvage with further therapy, which includes surgery, radioactive iodine, and possibly external beam radiotherapy. To our knowledge, little has been reported on patients who develop multiply recurrent and persistent WDTC. It was the aim of the present study to evaluate patient, tumor, treatment, and first-recurrence factors that may predict for the development of multiple recurrences among patients for whom primary therapy for WDTC fails. In theory, this may allow a more targeted approach with close follow-up and prompt therapy to achieve better cure rates in this group of patients. In addition, it may provide extra information that is useful when adequately counseling patients presenting both with primary WDTC or recurrent disease.
The most significant factors on univariate analysis for the development of multiple recurrences in our study included male sex, advanced stage of the primary tumor (ie, III and IV), size of the tumor, presence of ETS, primary total thyroidectomy, concomitant neck dissection, and adjuvant iodine 131 therapy. In addition, the detection of the recurrence by physical examination and subsequent surgical management predicted for the development of multiple treatment failures and living with persistent disease. Among these, male sex, advancing stage, ETS, and primary treatment with total thyroidectomy were the most important independent predictors on multivariate regression. This clearly demonstrates that male patients with recurrent WDTC and advanced primary lesions will most likely develop further treatment failures despite initial successful salvage therapy. The prognostic importance of sex and advanced stage are widely recognized.1,3,4 McConahey et al11 in a review of the experience by the Mayo Clinic identified a clear association between ETS and the development of both locoregional and distant recurrence. Similarly Andersen et al12 demonstrated that patients with ETS were more likely to develop treatment failure with a significant reduction in disease-specific survival. The aim of treatment should focus on complete surgical removal and adjuvant radioactive iodine in those in whom this is not possible. A study by Mazafferi and Jhiang2 found that rates of recurrence in patients receiving surgery, thyroid hormone suppression, and iodine 131 therapy were 16% compared with 38% for patients treated with surgery and thyroid hormone suppression alone. This is a controversial issue and there is currently no randomized trial confirming the efficacy of iodine 131 in reducing the incidence of recurrence in patients with WDTC. Nevertheless, it is standard therapy in our institution to give iodine 131 in patients with adverse prognostic findings such as ETS or large tumor size. In the present study 86% of all patients received a postoperative dose of iodine 131. Interestingly, statistical analysis demonstrated that the use of adjuvant iodine 131 therapy was associated with developing recurrent disease (P<.01). There are a number of potential explanations. Our results may suggest that iodine 131 therapy is not as efficient in controlling any residual locoregional or distant disease. This is certainly contrary to reports in the literature; however, clear evidence regarding the usefulness of iodine 131 therapy via prospective randomized controlled trial is still lacking. However, we believe that the most probable reason based on our data is the fact that a higher number of subtotal thyroidectomies were performed in the past, rendering adjuvant iodine 131 therapy less effective in controlling any residual locoregional or distant disease. This lends support to our surgical approach, which is to perform a total thyroidectomy on every patient presenting with WDTC in order to completely remove the disease, including that which is potentially present in the contralateral lobe, and allow optimal conditions for the potential beneficial effects of iodine 131 therapy. This is despite the fact that both total thyroidectomy and concomitant neck dissection were associated with a higher incidence of recurrence in our study. We explain this result by the fact that a more aggressive surgical approach acts as a surrogate marker for disease extent. That is, patients with more advanced disease, as perceived by the physician at the time, were treated with a total thyroidectomy and/or a neck dissection rather than with a more conservative procedure. The practice of subtotal thyroidectomy may cause patients to have to undergo completion surgery in the event that the final pathologic diagnosis demonstrates an aggressive variant, significant nodal involvement, or ETS. This, in our view, unnecessarily exposes the patient to all the inherent problems associated with completion surgery.13 In addition, follow-up of these patients is in our opinion less than optimal because we rely on serum thyroglobulin estimation and iodine 131 scanning to detect recurrent disease.
Primary detection of recurrent WDTC by physical examination is of prognostic importance. Coburn et al14 reviewed the results of salvage therapy in patients with recurrent WDTC detected by iodine 131 scanning vs physical examination. This study clearly demonstrates that the probability of dying from recurrent WDTC is significantly less in patients in whom the disease is detected via iodine 131 scanning compared with physical examination. In addition, treatment with iodine 131 alone was superior compared with those who also received surgical salvage. The findings in our study are in agreement with these results, as detection of the first recurrence by clinical examination and surgical salvage were statistically significantly associated with the development of multiple recurrences (P = .03 and .02, respectively). This indicates that a palpable recurrence represents a more advanced and aggressive form of disease then one detected by imaging or thyroglobulin alone and therefore is less amenable to curative treatment. These results support our approach, which is to follow up patients with WDTC with both thyroglobulin estimation and iodine 131 scanning in an attempt to diagnose any disease recurrence before it becomes clinically detectable. Interestingly, we were unable to localize the recurrence using standard imaging modalities (ie, ultrasonography, CT, or iodine 131 scanning) in 14 patients (group 2, 10/42; group 3, 4/31) with elevated levels of thyroglobulin. These patients were all treated with therapeutic doses of iodine 131. The majority of patients (group 2, 10/10; group 3, 2/4) remain alive and free of disease. Results by Pineda et al,15 who reported on 17 patients treated for WDTC with iodine 131, all of whom had elevated thyroglobulin estimation and negative imaging results, support these findings. Contemporary imaging modalities useful in the evaluation of recurrent WDTC, in addition to iodine 131 scanning and ultrasonography, include CT, magnetic resonance imaging, or positron emission tomography.16 Clearly these may have been beneficial in demonstrating the location of recurrence in our 14 patients in whom no focus could be demonstrated. However, limited availability and varying treatment philosophy during the management phase of these patients precluded the routine use of these advanced imaging modalities.
Our study has demonstrated that the development of multiple treatment failures is associated with a significant reduction in disease-specific survival (Figure 1). However, prolonged survival is possible, with an actuarially predicted 10-year disease-specific survival rate of 81% in patients who developed multiple recurrences. Interestingly, patients with only 1 recurrence have a disease-specific survival equal to those who are cured with primary therapy. This knowledge is important in the counseling of every patient with primary or recurrent WDTC regarding the extent of therapy in addition to stressing the importance of careful and regular lifelong follow-up.
Male patients presenting with recurrent WDTC detected primarily on physical examination and an advanced primary tumor (ie, presence of ETS, stage IV), have a very high likelihood of experiencing ongoing treatment failures. Both a high degree of suspicion in follow-up and prompt treatment of subclinical recurrent disease appear beneficial in avoiding the poor outcome in this group of patients. The future may well lie in the identification of biological markers that help identify patients at risk of multiple recurrences and thereby provide another tool in the management of WDTC.
Submitted for publication April 15, 2003; final revision received January 23, 2004; accepted January 29, 2004.
We would like to acknowledge the support given by Temmy Latner-Dynacare to the Department of Otolaryngology–Head and Neck Surgery at Mount Sinai Hospital, Toronto.
Correspondence: Jeremy L. Freeman, MD, FRCSC, Department of Otolaryngology, University of Toronto, Temmy Latner/Dynacare Chair in Head and Neck Oncology, Mount Sinai Hospital, Toronto, Suite 401, 600 University Ave, Toronto, Ontario, Canada M5G 1X5 (firstname.lastname@example.org).