Abstract
Background:
Death from well-differentiated thyroid cancer (WDTC) is rare, and over the past century there has been a trend away from local recurrence as the primary cause of death. The objective of our study was to report the cause of death from thyroid cancer in patients with WDTC treated with curative intent with surgery ± adjuvant radioactive iodine.
Methods:
An institutional database of 1811 patients with WDTC treated surgically for WDTC between 1986 and 2005 was analyzed and identified 165 (9.4%) who had died. Case records were studied to determine the cause of death in each patient.
Results:
Of the 165 deaths, 17 (10%) patients were confirmed to have died of thyroid cancer and 6 (4%) died of an unknown cause but had thyroid cancer present at the time of last follow-up. The remaining 142 (86%) died from other causes and were considered free of thyroid cancer at their last follow-up. We therefore identified only 23 cause-specific deaths from the entire cohort (1.3%). Of the 17 patients known to have died of thyroid cancer, all had distant recurrence. Ninety-four percent had pulmonary metastases. Of these, 47% also had bony metastasis at the time of death. Two patients had recurrent disease in the neck at the time of death, but both also had distant disease. Of the six patients (4%) who died of unknown causes but had thyroid cancer at last follow-up, four (67%) had distant disease alone, one (17%) had local and regional recurrence, and one had local and distant recurrence at last follow-up.
Conclusion:
After successful resection of WDTC, we report a low disease-specific death rate (1.3%). In contrast to earlier reports, death caused by central compartment disease in this recent series is very rare, with metastatic disease accounting for almost all fatalities.
Introduction
The long clinical course of patients with thyroid cancer, combined with low death rates, makes analysis of the specific cause of death in patients who die of thyroid cancer difficult. The objective of our study was therefore to analyze the cause of death from WDTC in patients treated with curative intent with surgery ± adjuvant radioiodine (RAI). In this select group of patients, all patients were considered free of disease after initial treatment. The study had the advantage of analyzing a large series of patients (1811 patients with WDTC) treated at a single institution with a long period of follow-up.
Materials and Methods
After approval from the institutional review board, and with waiver of consent, the records of a consecutive series of 1811 patients who had surgery for WDTC between 1986 and 2005 at Memorial Sloan Kettering Cancer Center (MSKCC) were reviewed. This database comprised patients who had their primary surgical treatment at MSKCC; patients who underwent initial treatment elsewhere before referral were excluded from the database.
Of the 1811 patients, 59 patients were excluded: 2 (0.1%) due to unresectable disease at the time of surgery and a further 57 (3%) due to the presence of distant metastases at presentation. This left 1752 patients available for analysis. All these patients had no evidence of macroscopic disease in the thyroid bed or cervical lymph node basins after surgery at MSKCC. In addition, there was no evidence of distant disease on postsurgical imaging in any of these patients. As such, all patients were considered free of disease after initial treatment at MSKCC. The median follow-up for this group was 99.5 months (range 4–290 months). From this group we identified 165 (9.4%) patients who have died. This represents the cohort available for study in our report.
All patients were treated according to our institutional approach to the management of WDTC (5). An individualized approach to treatment is adopted, with low-risk patients who present with uni-nodular disease being offered thyroid lobectomy and high-risk patients or those with nodular disease in the contralateral lobe being offered total thyroidectomy. Lymph node dissection in the central and lateral compartment is undertaken only in cases where preoperative investigations or intraoperative assessment raises the suspicion or confirms metastatic disease. Radio-iodine ablation is offered to high-risk patients after total thyroidectomy. Post-treatment follow-up is provided by both the endocrine and surgical departments, and includes clinical examination, ultrasonography, and thyroglobulin measurement. Throughout the study period, treatment and follow-up patterns evolved with increased use of ultrasonography and thyroglobulin measurement in particular.
Data were collected on patient demographics, surgical details including extent of both thyroid and neck surgery, and the presence of gross extra-thyroid extension or residual disease on completion of surgery. Pathological details included tumor histology, size, and presence of extra-thyroid spread, number of nodes collected and number of positive nodes, size of positive nodes, and presence of extracapsular spread. Postoperative details included use of radioiodine or external beam radiotherapy. Outcomes data included local, regional, or distant recurrence. The presence of local or regional recurrence after treatment was based on cytological or histopathological evidence in disease. Distant disease was determined by imaging studies, including radioiodine uptake scans and computed tomography scans, or cytological and histopathological evidence where available. Biochemical evidence of recurrence was not accepted as definitive, as the use of thyroglobulin measurement was not routine practice during the early part of the study period. The date of death and cause of death was ascertained from hospital records, death certificates, and updated from the social security death index.
Results
The entire cohort of 1752 patients had a median age of 46 years. Twenty-seven percent were men and 73% were women. The T-stage distribution at presentation was T1 48%, T2 19%, T3 27%, and T4 6%. Sixty-eight percent were considered N0 after initial treatment, 16% N1a, and 16% N1b. Therefore, in our thyroid cancer database one-third of patients had advanced T stage at presentation and one third had positive neck disease.
Of 165 patients who died, the median age at presentation of patients was 65 years (range 24–94 years). Sixty-eight patients (41%) were men and 97 (59%) were women. One hundred forty-two (86%) patients died of unknown cause and were considered free of thyroid cancer at their last follow-up (Table 1). Twenty-three patients died directly from thyroid cancer or had recurrent thyroid cancer at the time of death. Of these 23 patients, 17 (10%) patients were confirmed to have died of thyroid cancer and 6 (4%) died of an unknown cause but with thyroid cancer present at the time of last follow-up. These patients are considered to have died from disease. The median survival of these 23 patients was 66 months (range 4–273 months) and the disease specific mortality was 1.3% (23/1752). The patient and tumor characteristics of these 23 patients are shown in Table 1.
Patients are stratified into those who died of thyroid cancer and those who died of other causes but with thyroid cancer at last follow-up.
Deaths due to thyroid cancer (n = 17)
The majority of patients were over 45 years of age and had papillary thyroid carcinoma. Over two-thirds of these patients had T3 or T4 disease at presentation, and after histological analysis of excised tissues, 64% were classified as pN1b. All patients were considered either intermediate risk (29%) or high risk (71%).
The site of disease at the time of death was recorded for all 17 patients. Ninety-four percent of patients had pulmonary metastases at the time of death, 47% had bone metastasis (in addition to pulmonary metastases in all but one case), and two patients (12%) had recurrent disease in the central or lateral neck as well as pulmonary metastases at the time of death (Table 2). Thirteen patients (76%) had disease at multiple sites when they died, three patients (18%) had only pulmonary metastases, and one patient (6%) had only bone metastases. No patients had disease in the neck alone when they died of disease.
The cause of death in 15 patients (88%) was distant metastases alone. The remaining two patients (12%) died of aspiration pneumonia. Both patients had central neck recurrence associated with vocal cord palsy, and both also had pulmonary metastases (Table 3). Of the two patients with central neck disease and pulmonary metastases, one had unresectable local recurrence for which he received external beam radiation. This patient went on to develop a trachesophageal fistula causing a fatal aspiration pneumonia. The second patient developed a nodal central neck mass, which caused paralysis of the recurrent laryngeal nerve, contributing to fatal aspiration pneumonia.
Deaths with evidence of thyroid cancer at last follow-up (n = 6)
Six patients (4%) died of unknown causes but had evidence of recurrent thyroid cancer at the time of last follow-up. Three patients (50%) had pulmonary metastases alone, one (17%) had bone metastases alone, one patient had central and lateral neck recurrence, and the remaining patient had both central neck recurrence and pulmonary metastases. Therefore, five of the six patients had distant metastases at the time of death (Table 2).
Discussion
Cause-specific mortality from thyroid cancer is rare. Difficulties in analyzing this group of patients arise from the slow progression of disease requiring long periods of follow-up. Despite this, a number of groups have reported on the cause of death in patients managed over the past century (6 –13). In all of these studies, death due to local recurrence was an important cause. For example, Tollefsen in 1964 reported on 70 patients who died of thyroid cancer from a cohort of 700 patients (6). Of these 70 patients, 40% died of locoregional recurrence, 52% died of distant metastases, and 7% a combination of local and distant recurrence. Similar results were reported by Smith et al. (1988) at the Mayo clinic. In the Mayo study, there were 56 lethal papillary carcinomas from 859 patients (7%) who underwent primary treatment in the Mayo Clinic (10). Thirty-six percent of these patients died of locally recurrent disease, and 35% died of causes relating to pulmonary metastases. Since then, a series of more recent studies have shown a continued trend away from local recurrence and toward distant metastases as the cause of death from WDTC (10 –13). The most recent series reported by Ronga et al. in 2002 described 83 deaths from over 1900 patients (4%). In this series locally recurrent disease was present in 12% of cases, neck disease in 13%, and pulmonary or bone metastases were present in 49% (13).
Other than the Mayo group (10), previous authors have not been specific in their inclusion criteria for studies. Most groups report on consecutive series of patients treated in a single institution without mention of preoperative staging, or previous treatment for thyroid cancer. It is likely that these series include patients who presented with recurrent disease, distant metastases, and unresectable lesions that have undergone initial management before presentation within that institution.
The objective of our study was to determine the cause of death in the select group of patients with WDTC who have been treated with curative intent with surgery and postoperative RAI and are considered to be free of disease after initial treatment. In this group of patients, we report an extremely low death rate due to thyroid cancer: only 23 patients from 1752 (1.3%). This figure reflects improvements in thyroid cancer management. However, it also reflects the selection criteria in the cohort of patients whom we have studied. The most striking observation is the extremely low rate of neck recurrence as cause of death. Rather, the cause of death in the majority of patients was distant metastatic disease. In our selected surgical series, 23 patients died of disease, and all were considered either intermediate or high risk (5). At the time of death, only four (17%) had disease present in the central neck and in only two (9%) was central neck disease known to have contributed to the cause of death.
In contrast to earlier series, distant metastases were present in all but one of the cause-specific fatalities. This change in the pattern of recurrent disease is partly due to a more aggressive surgical approach to WDTC in our institution. This involves total thyroidectomy but also resection of all surrounding tissues with gross invasion, including strap muscles, recurrent laryngeal nerve, trachea, esophagus, and larynx to achieve no gross residual disease. Increased use of radioiodine in recent years in comparison to historical reports may also have contributed to this shift in disease recurrence.
In conclusion, after complete surgical resection of WDTC and RAI treatment, we report a low disease-specific death rate of 1.3%. In contrast to earlier reports, death due to central compartment disease is very rare. Metastatic lung disease accounts for almost all fatalities.
Footnotes
Acknowledgments
The authors I.J.N., I.G., and S.G.P. had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The authors M.M.W. and F.L.P. were involved in data collection. The authors J.P.S., A.R.S., and R.M.T. were involved in article editing.
Disclosure Statement
The authors have no conflicts of interest to declare.