Anaplastic thyroid carcinoma†
Treatment outcome and prognostic factors
Presented in part at the 76th Annual Meeting of the American Thyroid Association, Vancouver, Canada, September 29–October 3, 2004.
Abstract
BACKGROUND
Anaplastic thyroid carcinoma (ATC) is rare but is one of the most aggressive human malignancies. Several prognostic factors have been observed in patients with ATC, and some experts advocate aggressive multimodal therapy in selected patients. However, it is unclear whether such an approach significantly improves survival. The authors analyzed prognostic factors and treatment outcomes in patients with ATC reported in the National Cancer Institute's Surveillance, Epidemiology, and End Results data base.
METHODS
The cohort consisted of 516 patients with ATC reported to 12 population-based cancer registries between 1973 and 2000. Demographic, pathologic, and treatment data were used for univariate and multivariate survival analyses.
RESULTS
The mean patient age at diagnosis was 71.3 years, and there were 171 men and 345 women. Eight percent of patients had intrathyroidal tumors, 38% had extrathyroidal tumors and/or lymph node invasion, and 43% of patients had distant metastasis. The average tumor size was 6.4 cm (range, 1–15 cm). Sixty-four percent of patients underwent surgical resection of their primary tumor, and 63% received external beam radiotherapy. The overall cause-specific mortality rate was 68.4% at 6 months and 80.7% at 12 months. Univariate analysis showed that age < 60 years, female gender, intrathyroidal tumor, external beam radiotherapy, surgical resection, and combined surgical resection of tumor and radiotherapy were associated with a lower cause-specific mortality. On multivariate analysis, only age < 60 years, an intrathyroidal tumor, and the combined use of surgical and external beam radiation therapy were identified as independent predictors of lower cause-specific mortality.
CONCLUSIONS
Although most patients with ATC had an extremely poor prognosis, patients < 60 years old with intrathyroidal tumors survived longer. Surgical resection with external beam radiotherapy for ATC was associated with lower cause-specific mortality. Cancer 2005. © 2005 American Cancer Society.
Anaplastic thyroid carcinoma (ATC), one of the most aggressive human malignancies, is associated with an almost uniformly rapid and lethal clinical course. Although < 2% of all thyroid carcinomas are ATC, it accounts for 14–39% of thyroid carcinoma deaths.1, 2 Some investigators have advocated multimodal therapy for patients with ATC, including surgical resection or debulking, radiotherapy, and chemotherapy.3-16 However, to our knowledge, few published data have shown any significant survival benefit from multimodal therapy.17, 18
The aggressive nature and rarity of ATC make it difficult to determine patient outcome, especially in single-institution studies with small cohorts and short follow-up. In fact, ATC is classified as Stage IV thyroid carcinoma according to the American Joint Committee on Cancer, regardless of the tumor size or the presence of lymph node or distant metastasis.19 Furthermore, assessing response to therapeutic approaches in patients with ATC is complicated by its aggressive clinical course and the inability to clinically distinguish differences in the disease course itself from small but real treatment effects. We evaluated the outcome and the effect of treatment on the survival of patients with ATC reported in the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) database.
MATERIALS AND METHODS
Study Design and Data Source
A retrospective cohort design was used to analyze data from the SEER 12-registry database. The study was exempted from review by the University of California—San Francisco Committee on Human Research. The database consists of cases diagnosed from 1973 through 2000 from the states of Connecticut, New Mexico, Hawaii, Iowa, and Utah and from the metropolitan areas of Atlanta, Detroit, San Francisco-Oakland, and Seattle-Puget Sound. The database also includes cases diagnosed from 1992 through 2000 from the Los Angeles and San Jose-Monterey metropolitan areas and the Alaska Native region. The SEER database includes information on demographics, pathology, treatment, and survival.
Variable Definitions
Patients with anaplastic thyroid carcinoma were selected from the SEER database for public use (SEER 12-registry database).20 Tumors that were diagnosed only at autopsy or by death certificate and tumors without pathologic confirmation of ATC were excluded from the survival analyses. The cut-off date for follow-up was December 31, 2000. Classification of stage of disease at diagnosis, tumor size, treatment, and race/ethnicity followed SEER definitions.21 Disease stage was defined as local (confined to the thyroid gland), regional (extension into adjacent tissue or lymph node involvement), distant (metastasis), and unstaged.
Data Analysis
We used the SEER Stat software (version 5.0.20) to abstract patients with ATC into the StatView statistical software package (version 4.51; SAS Institute Inc., Cary, NC) according to the International Classifications of Diseases for Oncology, third revision codes for site and histology (C739 and 8020/8021). The data are reported as means ± standard deviations or as numbers and percentages. For the univariate survival analysis, the log-rank test was used to compare the Kaplan–Meier events. For the multivariate survival analysis, a Cox proportional hazards model was developed by forward, stepwise regression for all predictor variables that were identified as significant in the univariate analysis. The specific variables studied were age (per decade), gender, race/ethnicity, era of diagnosis (per decade), tumor size (in 1-cm increments from 1 cm to 12 cm), extent of disease (local, regional, and distant), surgical resection of tumor, and radiation treatment. It was assumed that the observed differences were statistically significant if the probability of chance occurrence was < 0.05.
RESULTS
In total, 516 patients with ATC were reported from 1973 to 2000 to the SEER Program (Table 1). The majority of patients were white (84.6%) and female (66.9%). ATC was the first primary tumor in 90.3% of patients. Only 39 patients (7.5%) had an intrathyroidal ATC, whereas 194 patients (37.6%) had extrathyroidal invasion and/or regional lymph node metastasis, and 222 patients (43.0%) had distant metastasis. Half of all patients with ATC underwent surgical resection, and 326 patients (63.2%) underwent radiation treatment (Table 2), which consisted of external beam radiation in 305 patients (59.1%) and radioisotopes in 12 patients (2.3%) (Table 2). Most patients were treated with external beam radiation after surgical resection. Unfortunately, data regarding chemotherapy for ATC are not collected in the SEER database.
| Characteristic | No. of patients |
|---|---|
| Demographics | |
| Male:female ratio | 171:345 |
| Age (yrs) | |
| Mean ± SD | 71.3 ± 12.7 |
| Range | 15–95 |
| Race | |
| White | 437 |
| Black | 26 |
| Othera | 51 |
| Unknown | 2 |
| Era of diagnosis and treatment | |
| 1973–1980 | 83 |
| 1981–1990 | 123 |
| 1991–2000 | 310 |
| Pathology | |
| Diagnostic confirmation | |
| Positive histology | 454 |
| Positive cytology | 58 |
| Clinical | 1 |
| Radiographic | 2 |
| Unknown | 1 |
| First malignant primary tumor | |
| Yes | 466 |
| No | 50 |
| No. of primary malignancies | |
| One | 453 |
| Two | 54 |
| Three | 6 |
| Four | 3 |
| Tumor size (cm)b | |
| Mean ± SD | 6.4 ± 2.6 |
| Range | 1.0–15.0 |
| SEER extent of diseasec | |
| Local | 39 |
| Regional | 194 |
| Distant | 222 |
| Unstaged | 61 |
- SD: standard deviation; SEER: Surveillance, Epidemiology, and End Results Program.
- a Other refers to 2 Native Americans and 49 Asian and Pacific Islanders.
- b Tumor size data were collected beginning in 1988 and were available for 37% of patients.
- c Specific sites of distant metastasis are not reported in the SEER database.
| Treatment modality | No. of patients |
|---|---|
| Surgical resectiona | |
| Yes | 253 |
| No | 141 |
| Unknown | 122 |
| Extent of thyroidectomya | |
| Total thyroidectomy | 12 |
| Subtotal or near total thyroidectomy | 37 |
| Lobectomy and/or isthmusectomy | 23 |
| Removal of less than a lobe | 8 |
| Thyroidectomy, NOS | 2 |
| Surgery, NOS | 171 |
| Radiation treatment | |
| External beam | 305 |
| Radioisotopeb | 12 |
| Unspecified | 9 |
| None | 177 |
| Unknown | 13 |
| External beam radiation sequence to surgical resection | |
| Before surgery | 4 |
| After surgery | 147 |
| Before and after surgery | 4 |
| No surgery | 143 |
| Unknown sequence | 7 |
- NOS: not otherwise specified.
- a No details concerning the extent of surgical resection (e.g., radical en bloc, limited, debulking with respect to contiguous structures) are included in the Surveillance, Epidemiology, and End Results (SEER) database except for the extent of thyroidectomy used, which was collected beginning in 1989.
- b The type of radioisotope used is not specified in the SEER database, but these patients most likely represent radioiodine ablation. These patients were censored in the survival analysis when evaluating the effect of external beam radiation treatment.
The overall cause-specific mortality was 69.4% at 6 months and 80.7% at 12 months (Fig. 1). On univariate analysis, patient age < 60 years, gender, primary tumor size (≥ 5cm), SEER extent of disease, surgical resection, external beam radiation, era of diagnosis, and combined surgical resection with external beam radiation were significant prognostic factors; however, race/ethnicity, first malignant primary tumor, and primary tumor size were not (Table 3). On multivariate analysis, age < 60 years, SEER extent of disease, and combined surgical resection with external beam radiation treatment were independent prognostic factors (Table 3, Fig. 2). In a subgroup analysis, combined surgical resection with external beam radiation decreased the cause-specific mortality rate significantly in patients with regional and distant disease, but not in patients with only intrathyroidal ATC (Fig. 3).
Overall cause-specific mortality in patients with anaplastic thyroid carcinoma. The mean cause-specific survival was 12 months (median, 3 months).
| Variable | Univariate P value | Multivariate HR (95% CI) |
|---|---|---|
| Age (< 60 yrs) | 0.0009a | 0.482 (0.268–0.867)a |
| Gender (female) | 0.0028a | 1.089 (0.746–1.590) |
| Race/ethnicityb | 0.0682 | NA |
| First primary malignancy | 0.7859 | NA |
| Tumor size (≥ 5 cm)c | 0.0203a | 1.245 (0.854–1.816) |
| SEER stage | < 0.0001a | |
| Local | 0.572 (0.366–0.893)a | |
| Regional | 0.831 (0.611–1.130) | |
| Distant | 1.492 (1.113–2.000)a | |
| Surgical resection | < 0.0001a | 0.779 (0.312–1.946) |
| External beam radiation | 0.0064a | 0.534 (0.147–1.940) |
| Combined surgical resection with external beam radiation | < 0.0001a | 0.722 (0.587–0.889)a |
- HR: hazard ratio; 95% CI: 95% confidence interval; NA: not applicable; SEER: Surveillance, Epidemiology, and End Results Program.
- a Significant prognostic variable.
- b Survival analysis compared between whites, blacks, and others.
- c Tumor size increased by 1-cm increments from 1 cm to the first significant value.
Cause-specific mortality in patients with anaplastic thyroid carcinoma by (A) age, (B) Surveillance, Epidemiology, and End Results extent of disease, and (C) combined surgical resection with external beam radiation.
Cause-specific survival in patients with anaplastic thyroid carcinoma (excluding patients with local disease only) who treated with combined surgical resection and external beam radiation.
DISCUSSION
Multimodal therapy for patients with ATC has been advocated by some investigators, because most patients present with metastatic disease and die within months of diagnosis. The evidence that supports such an approach, however, is based on single-institution studies of select groups of patients.5, 7, 9, 10, 13, 16, 22-26 Our current study of the population-based SEER data demonstrates that patients < 60 years old who have intrathyroidal ATC have a better prognosis compared with older patients who have distant metastasis. Surgical resection with external beam radiation in patients with regional and distant ATC was associated with lower cause-specific cancer mortality.
The mean patient age at diagnosis and the gender distribution in the current report were similar to those reported in other, smaller cohort studies.5, 16, 24, 27-30 Intrathyroidal ATC was uncommon, and most patients presented with regional disease and distant metastasis. Coexisting thyroid diseases such as differentiated thyroid carcinoma and multinodular goiter are common in patients with ATC and some studies suggested that these factors may influence the clinical course of patients with ATC.28-32 Unfortunately, because the SEER database does not contain information regarding the presence of other thyroid diseases, we could not evaluate the effect of these factors. Although tumor histology was not reexamined in the current study, and patients with medullary thyroid carcinoma, thyroid non-Hodgkin lymphoma, and the insular variant of follicular thyroid carcinoma may be diagnosed incorrectly with ATC, we do not believe this accounts for all 10.6% of patients who survived for > 2 years.27 This is because the misdiagnosis of ATC has been uncommon since the 1970s, when immunohistochemical techniques were implemented, and the SEER database consists of patients with ATC reported since 1973.
Many studies have documented the lethal course of patients with ATC, but the important prognostic factors that determine survival and the effect of multimodal treatment remain unclear. Patient age, gender, tumor size, extent of disease, leukocytosis, presence of acute local symptoms, coexisting multinodular goiter and well differentiated thyroid carcinoma, surgical resection, and multimodal therapy all reportedly influence patient survival according to some (but not all) studies.11, 16, 23, 25, 28-30, 32, 33 In the current study, which we believe includes the largest cohort analyzed to date, we found that only patient age at diagnosis and the presence of intrathyroidal ATC were independent predictors of lower cause-specific mortality. There was a 28.3% difference in the cause-specific mortality between patients < 60 years and patients ≥ 60 years, and there was a 44.9% difference in the cause-specific mortality at 1-year follow-up between patients who had distant metastasis and patients who had intrathyroidal ATC. Although it was reported previously that tumor size (microscopic foci or size < 5–6 cm) was a predictor of mortality, it was not found to be an independent predictor of lower cause-specific mortality in the current study.25, 26, 30, 31
In patients with ATC, the effects of surgical resection, radiation treatment, and chemotherapy on local control or improved survival are unclear. Although some studies have shown no survival benefit from resection, radiation therapy, or chemotherapy, other studies suggested that multimodal therapy may be beneficial in some patients with ATC.5, 7, 9, 10, 13, 16, 22-26 A major confounding factor in determining the effect of treatment on outcome has been selection bias, because patients who undergo surgical resection and radiation treatment often have less extensive disease.18 Although it is impossible to control for such factors in a retrospective study, our subgroup comparison of patients with regional and distant ATC still demonstrates a significant survival advantage for patients who underwent both surgical resection and external beam radiation therapy. Given the small number of patients who had intrathyroidal ATC, it was not surprising to find no significant decrease in the cause-specific mortality rate with combined surgical resection and external beam radiation. Unfortunately, no data regarding the precise extent of surgical resection and chemotherapy are collected in the SEER database. One of the most studied chemotherapy agents in patients with ATC has been the radiosensitizing agent doxorubicin, but no significant survival benefit has been observed.14, 34 In a multicenter Phase II trial that evaluated the effects of paclitaxel in 19 evaluable patients, a 53% response rate to treatment was reported with no significant effect on survival.35 Although an effect from chemotherapy could not be excluded as a confounding factor in the current study, recent studies using paclitaxel monotherapy or a combination of cisplatin, bleomycin, 5-fluorouracil, and cyclophosphamide resulted in no significant survival benefit.36-38
Patients < 60 years old with intrathyroidal ATC appear to have a better prognosis than older patients with distant metastasis. In the current study, surgical resection with external beam radiation in patients with ATC was associated with a lower cause-specific mortality.
