Prognosis of medullary thyroid carcinoma
Demographic, clinical, and pathologic predictors of survival in 1252 cases
Abstract
BACKGROUND.
Medullary thyroid cancer (MTC) is a rare cancer. There is a relative paucity of data over the last decade with regard to the prognosis of these patients. Therefore, the authors used the population-based Surveillance, Epidemiology, and End Results (SEER) registry to update what to their knowledge is one of the largest series of patients with MTC reported to date.
METHODS.
All patients with a diagnosis of MTC with active follow-up in the SEER database from 1973 to 2002 were included. Univariate and multivariate regression analyses were used to assess the associations between demographic, clinical, and pathologic characteristics of patients and survival.
RESULTS.
A total of 1252 patients with MTC were identified over 29 years of follow-up. In all, 87% of patients were white and 60% were female, with a mean age of 50 years. Although many variables were significant on univariate analysis, SEER stage and age at diagnosis were found to be the strongest predictors of survival in the multivariate analysis. Prognosis was poor in patients with advanced disease (hazards ratio [HR], 4.47), or those age >65 years (HR, 6.55). Patients who underwent surgery fared better than those who did not. Overall, 51% of patients had less than the currently recommended treatment guidelines for MTC. Adjuvant radiation therapy was found to be independently associated with a decreased survival (HR, 1.65).
CONCLUSIONS.
Stage of disease and age at diagnosis were found to be the strongest predictors of survival for patients with MTC. To the authors' knowledge there has been no change in stage at diagnosis or asignificant improvement in survival noted over the last 30 years. Many patients underwent surgery that was deemed less than optimal for stage of disease. Cancer 2006. © 2006 American Cancer Society.
Medullary thyroid carcinoma (MTC) is a relatively rare cancer, accounting for up to 10% of all thyroid cancers and up to 14% of all thyroid cancer-related deaths.1 The relative paucity of new cases of MTC has made the study of this disease difficult in institution-based series. Cooperative groups have established guidelines for the diagnosis and treatment of patients with MTC. These guidelines recommend primary complete surgical extirpation of disease with total thyroidectomy and varying level lymphadenectomy.2, 3 For example, in 2006, the National Comprehensive Cancer Network Practice Guidelines in Oncology recommend that bilateral central neck dissection is preferable for patients with pathologically demonstrated MTC as well as for patients with multiple endocrine neoplasia (MEN)-2B. A modified ipsilateral radical neck dissections may be considered for those patients with primary tumors that measure ≥1 cm in greatest dimension, or for those patients with central compartment lymph node metastases.4
Implementation of these guidelines has been inconsistent, and there has been variability in practice patterns as a result. Variation in practice patterns has implications for variations in the quality and cost of care for patients with MTC. Controversy continues regarding the extent of lymph node dissection, as well as its usefulness in reducing the rate of disease recurrence. Although some studies suggest that attempted complete extirpation and reresection of recurrences can improve long-term outcome, other studies suggest that, irrespective of disease stage, complete surgical resection does not preclude recurrence.5, 6 In a recent review of MTC therapy strategies, Kebebew et al.7 found no significant changes in disease stage at diagnosis and treatment modalities over the last 20 years. They concluded that the creation of practice guidelines had little effect on the diagnosis and treatment of MTC.
Although many studies have focused on controversies in MTC therapy, to our knowledge studies evaluating prognosis have been few. Many now are over a decade old. The German Medullary Thyroid Carcinoma Study group evaluated a total of 741 patients in 1993; the Swedish national MTC study examined the clinical characteristics of 249 patients in 1990; a Japanese nationwide survey included 634 patients in 1996; and, more recently, Kebebew et al.7 reviewed the experience of 104 patients with MTC at a single institution.7-11 The largest American study to date examining factors associated with the prognosis of MTC was from the Surveillance, Epidemiology, and End Results (SEER) Program database; it reviewed records from 1973–1991 and included 516 patients with MTC.12
We designed this study using the population-based SEER registry to update what to our knowledge is one of the largest continuous series of patients with MTC and to evaluate the association between demographic, clinical, and pathologic variables and survival of patients with MTC.
MATERIALS AND METHODS
Data Sources and Study Subjects
The SEER database was used to identify people in the U.S. diagnosed with MTC (International Classification of Diseases for Oncology histology codes 8510/3 and 8511/3) in 13 SEER regions during the period 1973–2002.13 The 13 SEER regions are Atlanta, Connecticut, Detroit, Hawaii, Iowa, New Mexico, San Francisco-Oakland, Seattle-Puget Sound, Utah, Los Angeles, San Jose-Monterey, Alaska, and rural Georgia. Collectively, these regions represent 26.2% of the total U.S. population. The SEER population is comparable to the general U.S. population with regard to poverty and education level. However, it is more urban and has a higher proportion of foreign-born persons than the overall U.S. population.13 SEER is the only population-based database in the U.S. that provides information regarding stage of disease at the time of diagnosis and survival rates for each stage.14
In this retrospective cohort study, all patients with histologically confirmed MTC were identified. Only patients who were actively followed by SEER were included. Patients were excluded if they were dead based on autopsy or death certificate or if they were not followed in a continuous fashion. The registries collect data regarding patient demographics, primary tumor site, histology, stage at diagnosis, first course of treatment (surgery and/or radiation therapy), and vital status at follow-up; all of these data were included in our analyses.
Associations between the demographic, clinical, and pathologic characteristics of patients and survival time were assessed. Demographic variables included gender, age at diagnosis, year of diagnosis, race/ethnicity, and marital status at diagnosis. In our analysis of the association between age and survival, we analyzed age as a continuous variable and as a categoric variable, with the following age intervals: age <40 years, ages 41–65 years, and age >65 years. Year of diagnosis was also treated as a categoric variable, with the following time intervals: 1973–1982, 1983–1992, and 1993–2002. Race/ethnicity was classified as white and nonwhite. Marital status was classified as married, single/divorced/widowed, and unknown.
Clinical information included number of primary tumors, surgical therapy, radiation therapy, and survival time. Surgical therapy was defined by the site-specific thyroid surgery codes and the scope of regional lymph node surgery. We classified surgical therapy into: 1) no operation; 2) lobectomy or lobectomy with isthmusectomy; and 3) total/subtotal thyroidectomy. The lymph node dissections were classified as: 1) no lymph node dissection; 2) limited lymph node dissection; and 3) radical/modified radical lymph node dissection. Radiation therapy included external beam radiation, radioisotopes, a combination of the 2, and “other”; it was treated as a binomial variable, based on whether patients had received radiation in any form.
Pathologic information included stage at diagnosis, tumor grade, size, number of regional lymph nodes examined, and the number of metastatic lymph nodes removed. Stage at diagnosis was categorized according to the SEER staging system: localized (tumor confined entirely to the thyroid gland), regional (extension beyond the thyroid directly into surrounding tissues or regional lymph node metastases); distant (metastasis to extracervical lymph nodes or organs), and unstaged (insufficient information to assign a stage).13, 14 Tumors were graded as well differentiated, moderately differentiated, poorly differentiated, and anaplastic. Tumor size was analyzed as a continuous variable and also as a categoric variable: ≤2 cm, 2–4 cm, and >4 cm, based on the size classification of the TNM system sponsored by the American Joint Committee on Cancer (AJCC).15 Lymph node involvement was categorized according to the anatomic site of metastasis: none; ipsilateral, bilateral/contralateral/midline cervical; mediastinal; and distant lymph nodes.
Statistical Analysis
Simple summary statistics were used to describe the demographic and clinical landscape. The Student t test or analysis of variance (ANOVA) were used for continuous variable comparisons, and the chi-square test was used for categoric variables.
Survival time was calculated as the time (in years) from diagnosis until death, last follow-up, or December 31, 2002, whichever came first. Because SEER does not report the cause of death for individual patients, overall mortality was used for the survival analysis. For the univariate analysis, the Kaplan-Meier method and log-rank test were used to identify significant prognostic factors. For the multivariate analysis, the Cox proportional hazards regression model was used for variables identified as significant in the univariate analysis. The strength of association between each predictor variable and survival was expressed as a hazards ratio (HR), which is presented along with a 95% confidence interval (95% CI). All tests were 2-sided, and a P-value <.05 was considered statistically significant. All data were analyzed with Statistical Package for the Social Sciences (SPSS) software (version 11.5; SPSS Inc., Chicago, IL).
RESULTS
Summary Statistics
A total of 1252 patients were identified in our dataset as having been given the diagnosis of MTC between 1973 and 2002. Their demographic, clinical, and pathologic characteristics are reported in Table 1. Of these patients, the majority (87%) were white and 57% were married; 60% were female. The mean age at diagnosis was 50.1 years (median, 51 years; range, 3–95 years), with no difference in age noted between the genders. Patients with more advanced disease based on tumor stage were found to be significantly older (P < .0001) (Fig. 1), and they were more often male (P < .0001). The number of incident cases of MTC were distributed relatively evenly across our 3 diagnosis eras (1973–1982, 25%; 1983–1992, 37%; and 1993–2002, 38%).
Age at diagnosis of medullary thyroid cancer based on Surveillance, Epidemiology, and End Results (SEER) stage. Boxes represent the median and interquartile ranges.
| Patient characteristics | No. | Percent | |
|---|---|---|---|
| Demographic characteristics | Gender | ||
| Male | 506 | 40 | |
| Female | 746 | 60 | |
| Age at diagnosis, y (mean ± SD) | 50 ± 18 | ||
| <40 | 379 | 30 | |
| 40–65 | 586 | 47 | |
| >65 | 287 | 23 | |
| Race | |||
| White | 1089 | 87 | |
| Nonwhite | 163 | 13 | |
| Marital status | |||
| Married | 717 | 57 | |
| Single/divorced/widowed | 433 | 35 | |
| Unknown | 102 | 8 | |
| Vital status | |||
| Alive | 852 | 68 | |
| Dead | 400 | 32 | |
| Year of diagnosis | |||
| 1973–1982 | 317 | 25 | |
| 1983–1992 | 457 | 37 | |
| 1993–2002 | 478 | 38 | |
| Clinical characteristics | No. of primary tumors/patients | ||
| 1 | 1088 | 87 | |
| 2 | 139 | 11 | |
| >2 | 25 | 2 | |
| Surgery (n = 643) | |||
| None | 34 | 5 | |
| Lobectomy | 91 | 14 | |
| Subtotal/total thyroidectomy without lymph node dissection | 235 | 37 | |
| Subtotal/total thyroidectomy with limited lymph node dissection | 113 | 18 | |
| Total thyroidectomy with radical or modified radical lymph node dissection | 170 | 26 | |
| Radiation therapy | 221 | 18 | |
| External beam radiation | 148 | 12 | |
| Radioisotopes | 48 | 4 | |
| Combination of external beam and radioisotopes | 8 | 0.6 | |
| Other radiation | 17 | 1.4 | |
| Survival time, y (mean ± SD) | 8.6 ± 7.6 | ||
| Minimum, y | 0 | ||
| Maximum, y | 29.7 | ||
| Pathologic characteristics | SEER stage | ||
| Localized | 595 | 48 | |
| Regional | 437 | 35 | |
| Distant | 166 | 13 | |
| Unstaged | 54 | 4 | |
| Tumor size, cm (mean ± SD) | 2.8 ± 2.4 | ||
| Minimum (cm) | 0 | ||
| Maximum (cm) | 33 | ||
| Extent of tumor (1988–2002) (n = 762) | |||
| Within thyroid capsule | 562 | 45 | |
| Outside thyroid capsule | 200 | 16 | |
| Lymph node metastases (1998–2002) (n = 594) | |||
| None | 368 | 62 | |
| Ipsilateral | 121 | 20 | |
| Bilateral, contralateral, or midline | 49 | 8 | |
| Mediastinal | 36 | 6 | |
| Distant | 20 | 4 | |
| No. of regional lymph nodes examined (1998–2002) (n = 428) (mean ± SD) | 12 ± 19 | ||
| Minimum | 0 | ||
| Maximum | 90 | ||
| No. of positive lymph nodes (1998–2002) (n = 428) (mean ± SD) | 4 ± 8 | ||
| Minimum | 0 | ||
| Maximum | 75 | ||
- SD indicates standard deviation; SEER, Surveillance, Epidemiology, and End Results Program.
MTC was the only malignancy diagnosed in 1088 patients (87%). Thirty patients had concurrent papillary or follicular thyroid carcinoma. Additional malignancies were identified in the following sites: breast (33 patients), gastrointestinal tract (29 patients), prostate (26 patients), lung (14 patients), hematologic malignancy (13 patients), melanoma of the skin (11 patients), urinary bladder (9 patients), uterus (8 patients), pancreas (4 patients), and ovary (3 patients).
Nearly half of patients (48%) were diagnosed with localized disease, whereas 35% had tumor extending beyond the thyroid into surrounding tissues or regional lymph node metastases, and 13% had MTC metastatic to distant organs. The average size of medullary thyroid tumors was 2.8 cm (range, 0–3.3 cm). Histologic grade was not reported for 92.2% of patients. Since 1998 SEER has provided data regarding the extent of tumor and lymph node metastases, as well as the number of regional lymph nodes examined. Based on this subset of data, 45% of patients had tumor confined to the thyroid capsule, whereas 16% patients had tumor extension beyond the thyroid capsule. Of the 594 patients for whom valid information was available with regard to lymph node metastases, 62% had no lymph node involvement; 20% had ipsilateral cervical lymph nodes; 8% had bilateral, contralateral, or midline cervical lymph nodes; 6% had mediastinal lymph nodes; and 4% had distant lymph nodes. On average, 12 regional lymph nodes were examined at the time of surgery (range, 0–90 lymph nodes), and 4 lymph nodes, on average, were found to be positive for cancer.
The type of surgery performed was documented in 643 patients (51%). Within this group, 37% of patients underwent total/subtotal thyroidectomy with no lymph node dissection; 26% underwent total thyroidectomy with a radical/modified radical lymph node dissection; 18% underwent total/subtotal thyroidectomy with a limited lymph node dissection; 14% underwent lobectomy; and no surgery was performed in 5% of patients. When stratified by disease stage, subtotal/total thyroidectomy was performed in 85%, 89%, and 69.5% of patients with localized, regional, and distant stage MTC, respectively. Radical/modified radical lymph node dissection was performed in 14.4%, 58.6%, and 41.7% of patients with localized, regional, and distant stage disease, respectively.
Adjuvant radiation therapy was administered to 18% of patients. This included external beam radiation (12%), radioisotopes (4%), a combination of the 2 (0.6%), and other radiation (1.4%). Of the patients who received adjuvant radiation, 54.4% had regional stage disease, 26% had distant stage disease, and 19.5% demonstrated localized stage MTC. Moreover, 85% of these patients underwent total thyroidectomy, and 50% of these patients had a modified radical or radical lymphadenectomy.
As of December 31, 2002, 68% of patients in our cohort were still alive. The mean survival time after the diagnosis of MTC was 8.6 years (range, 0–29.6 years). Overall survival differed by SEER stage (Fig. 2). Patients with tumors confined to the thyroid gland had a 10-year survival rate of 95.6%, whereas patients with regional stage disease had an overall survival rate of 75.5%. Patients with distant metastases at diagnosis had a poor prognosis, with only 40% surviving 10 years.
Ten-year, disease-specific survival by Surveillance, Epidemiology, and End Results (SEER) stage for patients with histologically confirmed medullary thyroid cancer. SEER, 1973–2002.
Univariate Analysis
Twelve demographic, clinical, and pathologic prognostic variables were studied by univariate analysis (Table 2). Statistically significant predictors of improved overall survival included female gender, younger age at diagnosis, localized stage of disease, well or moderately differentiated histologic grade, smaller tumor size (when size was treated as a continuous variable), tumor confined to within the thyroid capsule, and lack of lymph node involvement. Patients with concurrent tumors had significantly worse survival than patients with MTC only (P < .0005). Overall, patients with MTC who underwent surgery survived longer than those who had no surgery (P < .0001), and the more extensive the surgery, the better the patients' overall survival; patients who underwent total thyroidectomy were found to have the best survival, followed by those who underwent lobectomy (P < .0001). When stratified by disease stage (Fig. 3), all patients with localized MTC underwent surgery, and no significant difference in survival was observed between those who underwent lobectomy compared with patients who underwent total/subtotal thyroidectomy (Fig. 3A). Of note, 2 patients did not undergo surgery and survived <1 month. In patients with regional stage disease, total thyroidectomy was associated with the best survival, followed by lobectomy. The prognosis was worst in patients who underwent no surgery at all (Fig. 3B). In patients with distant disease, surgery was associated with improved survival (P < .0001). However, no statistically significant difference in survival was observed between patients who underwent lobectomy versus those who underwent total thyroidectomy (Fig. 3C). When different lymph node dissections were compared, there was no significant difference in survival noted based on the extent of lymphadenectomy performed. Patients who received radiation therapy had a worse prognosis than patients who had no adjuvant therapy (P < .0001). There was no association noted between survival and marital status, race, and diagnostic era.
Overall survival by Surveillance, Epidemiology, and End Results (SEER) stage and surgery performed in patients with histologically confirmed medullary thyroid cancer. SEER, 1973–2002. (A) Localized medullary thyroid cancer. No significant difference was observed between lobectomy and total thyroidectomy. (B) Regional medullary thyroid cancer. Patients who underwent surgery were found to have a had better prognosis than patients who had no surgery (P < .0001). Total/subtotal thyroidectomy was associated with a better prognosis than lobectomy. (C) Distant medullary thyroid cancer. Patients who underwent surgery were found to have a better prognosis than patients who had no surgery (P < .0001). No significant difference was observed between lobectomy and total thyroidectomy.
| Variables | Hazards ratio(95% CI) | P |
|---|---|---|
| Gender | ||
| Male | 1 | Reference |
| Female | 0.53 (0.43–0.64) | <.0001 |
| Age at diagnosis, y | ||
| <40 | 1 | Reference |
| 40–65 | 2.88 (2.01–3.96) | <.0001 |
| >65 | 8.72 (6.32–12.03) | <.0001 |
| Race | ||
| White | 1 | Reference |
| Black | 1.09 (0.75–1.67) | NS |
| Other | 0.6 (0.37–0.97) | .03 |
| No. of primary tumors/patient | ||
| 1 | 1 | Reference |
| >1 | 1.65 (1.23–2.1) | <.0001 |
| SEER stage | ||
| Localized | 1 | Reference |
| Regional | 2.57 (1.85–3.58) | <.0001 |
| Distant | 7.21 (5.27–9.87) | <.0001 |
| Histologic grade | ||
| Well differentiated | 1 | Reference |
| Moderately differentiated | 1.11 (0.49–2.49) | NS |
| Poorly differentiated | 4.18 (2.73–6.4) | <.0001 |
| Undifferentiated | 3.59 (2.17–5.93) | <.0001 |
| Tumor size, cm | ||
| ≤2 | 1 | Reference |
| 2–4 | 0.8 (0.6–1.1) | NS |
| >4 | 1.3 (0.9–1.9) | NS |
| Extent of tumor | ||
| Within thyroid capsule | 1 | Reference |
| Outside thyroid capsule | 6.52 (4.72–9) | <.0001 |
| No. of positive lymph nodes | ||
| 0 | 1 | Reference |
| ≥1 | 1.79 (1.15–2.79) | .01 |
| Surgery | ||
| None | 1 | Reference |
| Lobectomy | 0.18 (0.1–0.33) | <.0001 |
| Total thyroidectomy | 0.09 (0.06–0.15) | <.0001 |
| Lymph node dissection | ||
| None | 1 | Reference |
| Limited | 0.2 (0–1.4) | NS |
| Modified radical/radical | 0.7 (0.3–1.7) | NS |
| Surgery | ||
| None | 1 | Reference |
| Lobectomy/isthmectomy | 0.18 (0.1–0.33) | <.0001 |
| Subtotal/total thyroidectomy without lymph node dissection | 0.08 (0.05–0.14) | <.0001 |
| Subtotal/total thyroidectomy with limited lymph node dissection | 0.07 (0.04–0.14) | <.0001 |
| Subtotal/total thyroidectomy with radical or modified radical lymph node dissection | 0.12 (0.72–0.22) | <.0001 |
| Radiation therapy | ||
| None | 1 | Reference |
| Yes | 2.6 (2.07–3.27) | <.0001 |
- SEER indicates Surveillance, Epidemiology, and End Results Program; 95% CI, 95% confidence interval; NS, not significant.
Multivariate Analysis
Variables identified as having a significant association with survival on univariate analysis were studied in a Cox proportional hazards model (Table 3). Because the majority of patients did not have tumor differentiation reported, this variable was not included in the final model. Age at diagnosis remained a strong predictor of survival, with the risk of dying increased by 5.2% for each additional year of age. Overall, prognosis was especially poor in patients age >65 years. Survival was strongly associated with stage of disease; compared with patients with local disease, patients with regional stage disease had a 2.69 times greater risk of dying and patients with distant disease had a 4.47 times greater risk of dying. Extent of surgery also was a significant independent predictor of survival; after controlling for all other variables, patients who underwent surgery fared better than comparable patients without surgery, and patients with more extensive surgery survived longer than those who undergone lobectomy only. Finally, adjuvant radiation therapy was an independent predictor of decreased survival (P < .05) after controlling for age, stage of disease, and extent of surgery.
| Variables | Hazards ratio(95% CI) | P |
|---|---|---|
| Age, y | ||
| <40 | 1 | Reference |
| 40–65 | 2.35 (1.31–4.22) | .004 |
| >65 | 6.55 (3.69–11.63) | <.0001 |
| Stage | ||
| Localized | 1 | Reference |
| Regional | 2.69 (1.73–4.18) | <.0001 |
| Distant | 4.47 (2.58–7.74) | <.0001 |
| Surgery | ||
| None | 1 | Reference |
| Lobectomy | 0.47 (0.25–0.91) | .04 |
| Subtotal/total thyroidectomy without node dissection | 0.29 (0.15–0.56) | <.0001 |
| Subtotal/total thyroidectomy with limited node dissection | 0.17 (0.08–0.35) | <.0001 |
| Total thyroidectomy with radical or modified node dissection | 0.25 (0.13–0.47) | <.0001 |
| Radiation therapy | ||
| None | 1 | Reference |
| Yes | 1.65 (1.11–2.45) | .01 |
- SEER indicates Surveillance, Epidemiology, and End Results Program; 95% CI, 95% confidence interval.
DISCUSSION
To our knowledge, the current study offers an analysis of the largest contemporaneous population-based sample of prognostic factors in patients with MTC studied within the last decade. The benefit of the SEER cancer registry is that it allows evaluation of a large cross-section of the population of patients with cancer diagnoses while providing long-term follow-up without inherent institutional biases such as referral, selection, treatment preferences, or available surgical expertise.
The incidence of MTC does not appear to have increased over the last 29 years, despite the development of better diagnostic and genetic screening tools. This finding is in contradiction to the results of the French Study Group, who analyzed 899 patients with MTC from 1959–1996 and found that half of their patients were diagnosed after 1990.16 The strongest independent clinical predictor of survival was stage at diagnosis. We used the SEER staging system. This finding has been substantiated in multiple clinical and population-based studies.17 Extent of surgical resection was also found to be an independent predictor of outcome. Patients who underwent any type of surgery, including lobectomy alone, had a significantly improved survival compared with patients who underwent no surgery at all (only 5% of our cohort). Overall, patients who underwent total thyroidectomy fared significantly better than those who underwent a lesser resection. However, patients with localized disease and patients with distant metastases did not benefit from a more extensive thyroid surgery. This may be explained by the fact that patients with localized disease will have the MTC extirpated by any adequate level of resection, whereas patients with distant metastases will not benefit in overall survival by a more aggressive thyroid resection because they have incurable distant disease. Local palliation was not evaluated in this analysis. Patients with regional disease, who represented the majority of patients, gained a significant survival benefit from a total thyroidectomy, but the extent of additional lymphadenectomy did not appear to impart additional benefit with regard to survival.
This finding is controversial because to our knowledge most treatment guidelines to date recommend at the minimum a limited central lymphadenectomy for MTC, given the fact that MTC often has lymph node metastases to the central neck and ipsilateral lateral neck.18, 19 Several authors have shown in prospective studies the high rates of lymph node metastases found in MTC. Miyauchi et al.20 prospectively followed 15 nonhereditary MTC patients. Thirteen of these patients underwent lobectomy with ipsilateral neck dissection, achieving biochemical cure. This study showed that in select patients with unilateral tumors, lobectomy with ipsilateral lymphadenectomy was successful. The SEER data does not contain information regarding hereditary or biochemical results. Therefore, we could not identify in our database a similar cohort of patients for analysis. Some authors have recommended aggressive lymphadenectomy with primary bilateral modified radical neck dissections to attempt to completely extirpate all cervical and mediastinal disease, improve survival, and establish cure.21 Fifty-one percent of patients in our study underwent either lobectomy or total thyroidectomy, without lymphadenectomy of any kind. Moley and DeBenedetti22 studied 73 patients with palpable MTC prospectively and found that >75% had associated lymph node metastases, which were often not apparent to the surgeon in the operating room. They recommended routine central and bilateral functional neck dissections in all patients with palpable MTC. Many clinical series have shown that lymphadenectomy can facilitate normalization of serum calcitonin levels, which has been considered an indicator of cure. However, we did not find lymphadenectomy to have a significant benefit for long-term survival.23 Studies have shown that many patients will have prolonged survival despite elevated serum levels of calcitonin in the absence of documentable disease. This may in part explain our findings.24
We were unable to measure the effect of specific surgical techniques, such as central lymphadenectomy, unilateral or bilateral modified lymphadenectomy, radical lymphadenectomy, or mediastinal lymphadenectomy, because this degree of surgical detail is not available in the SEER database. Another limitation of this study is that our analysis could not include biochemical evaluation with serum calcitonin or carcinoembryonic antigen (CEA) levels, which have been shown to be significantly associated with outcome in some studies.25 Rates of recurrence and reoperation are not available through SEER. Because MTC is largely a surgical disease, this information would be useful. Finally, SEER does not provide information regarding familial MTC, such as multiple endocrine neoplasia (MEN)-2A or 2B, familial medullary thyroid cancer, or associated genetic RET mutations. Therefore, we were unable to independently evaluate any prognostic factors related to genotype.
The strongest sociodemographic predictors of survival were age <40 years and female gender. Patients age >65 years of age with MTC had decreased survival. In prior studies, age has been alternatively studied as a continuous and a categorical variable.26 In our multivariate analysis, we found that each additional year of age increased the risk of dying by 5.2% in a continuous fashion. To our knowledge the etiology of the association between age and survival is still unclear. The SEER data allowed us to determine that older patients were more likely to have concomitant cancers, such as breast, gastrointestinal, and prostate malignancies. These patients also had a significantly reduced mean survival when compared with patients without synchronous cancers. This was the only index of comorbid conditions available for review through the SEER database. One may postulate that with increasing age, patients are more likely to suffer from other medical problems, poorer nutrition, impaired healing, and decreased immune response and physiologic reserves.
The relation between age, gender, tumor stage, and histology is a complex one. These factors were found to be strong independent predictors of survival in our series, yet they may be interrelated. Men and older patients presented with more advanced stages of MTC at the time of diagnosis. Tumor differentiation could not be examined in our multivariate analysis because pathologic detail was available for only a minority of patients in SEER. Tumor differentiation played a significant role on univariate analysis, with poorly differentiated or undifferentiated histologic grade found to be associated with poor survival. Tumor characteristics also played a significant role in patient outcome. Tumor multicentricity and advanced local extent/invasion were found to be associated with decreased survival on univariate analysis, but they were not found to be significant predictors of survival on multivariate analysis. Unlike some studies that have shown that up to 66% of sporadic MTC and up to 86% of familial MTC may be multicentric and involve both thyroid lobes, we found that only 13% of all patients in our cohort had >1 primary tumor in the thyroid.10
Adjuvant radiation in the form of external beam, radioisotope treatment, or a combination of these was performed in 18% of our patients. Several studies have documented some benefit from radiation for advanced disease, whereas others have shown no response to adjuvant treatment.26 Adjuvant radioisotope treatment has no utility in MTC unless patients have concomitant papillary or follicular thyroid malignancies.23 Radioimmunotherapy for metastatic MTC using anti-CEA antibodies has been attempted in restricted clinical trials.27 We found that the addition of any type of radiation was associated with decreased survival, even for stage-matched patients with regional and distant disease. Fifty-seven percent of patients with MTC in our cohort who received radiation had significant extrathyroidal extension and local invasion. Information regarding adjuvant chemotherapy is not collected by SEER, but other groups have shown little efficacy of traditional chemotherapy in MTC, and experience with newer tyrosine kinase inhibitors is limited.26
In conclusion, the current study represents a large population-based analysis of patients with MTC with 29 years of follow-up. Stage of disease and age at time of diagnosis appear to be the strongest predictors of overall survival. Surgical intervention less than total thyroidectomy was found to be associated with decreased survival. Extent of lymphadenectomy should be directed based on clinical and pathologic factors. Adjuvant radiation is associated with decreased survival for patients with regional and distant stage MTC. Despite technological advances in diagnostic modalities, refinement of imaging techniques, and the development of genetic screening over the last decade, to our knowledge there has been no survival improvement over nearly 30 years of treatment for MTC. This may be explained in part by the fact that most patients with thyroid disease in the U.S. see general surgeons without expertise in endocrine surgery, let alone a rare pathologic entity such as MTC. Sosa et al.28 examined outcomes from thyroidectomy based on surgeon and institutional experience. They found that for patients with the diagnosis of thyroid cancer and for patients undergoing complex (subtotal and total) thyroidectomy, clinical and economic outcomes were improved if thyroid procedures were performed by surgeons with significant endocrine surgical experience, as measured by case volume. We propose that healthcare providers be astute to the diagnosis of MTC, and consider the referral of patients with MTC to high-volume thyroid surgeons.
