X/97/$03.00/0 Vol. 82, No. 11 Journal of Clinical Endocrinology and Metabolism Copyright 1997 by The Endocrine Society

Transcription

1 X/97/$03.00/0 Vol. 82, No. 11 Journal of Clinical Endocrinology and Metabolism Printed in U.S.A. Copyright 1997 by The Endocrine Society Pathological Tumor-Node-Metastasis (ptnm) Staging for Papillary and Follicular Thyroid Carcinomas: A Retrospective Analysis of 700 Patients KEH-CHUAN LOH, FRANCIS S. GREENSPAN, LAUREN GEE, THEODORE R. MILLER, AND PETER P. B. YEO Division of Endocrinology, Metabolic Research Unit and Department of Medicine (K.-C.L., F.S.G., P.P.B.Y.), the Department of Epidemiology and Biostatistics (L.G.), and the Department of Pathology (T.R.M.), University of California, San Francisco, California ABSTRACT The TNM classification (tumor-node-metastasis) was adopted by the American Joint Committee on Cancer and the International Union against Cancer a decade ago to avoid heterogeneity of prognostic classification schemes used for differentiated thyroid cancers. To date, however, clinical data based on this classification are lacking. We retrospectively evaluate the prognosis of 700 patients (208 men and 492 women) with papillary (89%) and follicular (11%) thyroid cancers according to the pathological TNM (ptnm) staging system, treated over a 25-yr period ( ). Patients who received primary treatment at our center constituted 87.4% of the cases; the majority underwent total thyroidectomy, followed by 131 I ablative therapy in high risk groups, as standard treatment. Clinical and follow-up data were obtained from the medical records and our cancer registry. Disease-free and cancer-specific survival data were analyzed by Kaplan-Meier product limit estimates and Cox proportional hazard models. Patient distribution by the ptnm system were: stage I, 516 patients; stage II, 57 patients; stage III, 104 patients; and stage IV, 23 patients. Over a mean SE follow-up of yr, the overall cancer recurrence and mortality rates were 20.5% and 8.4%, respectively. However, the respective cancer recurrence and mortality rates were distinctly different in the various ptnm stages: 15.4% and 1.7% in stage I, 22% and 15.8% in stage II, 46.4% and 30% in stage III, and 66.7% and 60.9% in stage IV tumors. Using actuarial survival plots, a clear separation in both disease-free survival and cancer-specific survival was noted among all the stages (P ). Risk factors THE PROPER staging and management of thyroid cancer are of great interest and concern. Differentiated thyroid cancer constitutes the commonest endocrine malignancy, and because of its generally favorable course, various tumorstaging schemes have been formulated to improve risk group assignment of affected individuals (1 8). Until recently, the proliferation of numerous classifications created the problem that there was no generally accepted staging scheme for thyroid cancer. In the past decade, the American Joint Committee on Cancer and the TNM Committee of the International Union against Cancer have agreed on acceptable rules for a staging system in cancer of the thyroid gland (9, 10). As Received June 4, Revision received July 23, Accepted August 1, Address all correspondence and requests for reprints to: Dr. Keh-Chuan Loh, Department of General Medicine, Tan Tock Seng Hospital, Moulmein Road, Singapore , Republic of Singapore. keh_chuan_loh@notes.ttsh.gov.sg. analyses showed a significant association between all the prognostic variables used in TNM staging (age, tumor size, extent of primary tumor, and presence of nodal or distant metastases) and the observed end points of recurrence or death from thyroid cancer. After correcting for TNM stages, the risk of cancer recurrence was halved in female compared to male patients, whereas this was 1.7-fold higher in multifocal than unifocal tumors. Conversely, cancer mortality was 3.4- fold higher in follicular than papillary thyroid cancer. In the analysis of effect of primary treatment among 492 patients with tumor more advanced than the T1N0M0 category, patients who underwent less extensive surgery (lobectomy or subtotal thyroidectomy) had a 2.5-fold risk of cancer recurrence (P ) and a 2.2-fold risk of death (P 0.01) compared to those who underwent total or near-total thyroidectomy. Patients not treated with 131 I ablation had a 2.1-fold greater risk of cancer recurrence (P ) than those given 131 I ablation, although no difference was noted in deaths from thyroid cancer. Based on our data, the ptnm classification is useful in distinguishing patients with different prognostic outcomes. However, the small patient numbers in ptnm stages other than stage I precludes us from evaluating its usefulness as a guide for therapy. Until prospective data could be accrued from controlled treatment trials, we support the standard practice of total thyroidectomy followed by 131 I ablative therapy (if focal iodide uptake was noted) in patients with papillary thyroid cancer more advanced than the T1N0M0 category or of multicentric nature and in the majority of patients with follicular thyroid cancer. (J Clin Endocrinol Metab 82: , 1997) shown in Table 1, the TNM system relies on assessment of three components: the extent of primary tumor (T), the absence or presence of regional lymph node metastases (N), and the absence or presence of distant metastatic lesions (M). This classification may be either clinical, based on evidence acquired before definitive treatment, or pathological (ptnm), when intraoperative and surgical pathological data are available. An interesting feature of the TNM staging system compared to other classifications is the age factor. Regardless of the T and N categories, all patients under 45 yr without distant metastases (M0) are classified as stage I, whereas those with distant metastases (M1) belong only to stage II. Despite the current widespread acceptance of the ptnm classification for thyroid cancer staging, there is limited published clinical data on its prognostic utility (11 13). We reviewed our patients with papillary and follicular thyroid cancer over a 25-yr period using the ptnm staging system. Our analysis specifically examines the value of 3553

2 3554 LOH ET AL. JCE&M 1997 Vol 82 No 11 TABLE 1. TNM staging: American Joint Committee on cancer stage classification for thyroid cancer Stage ptnm classification in predicting disease-free survival and cancer-specific survival. We also performed risk factor analysis on various prognostic variables by using multivariate models. In view of the unsettled controversy over the optimal primary treatment regimen for thyroid cancer, we secondarily examined the effect of treatment on prognosis in 492 patients with tumors more advanced than the T1N0M0 category. Patient selection Subjects and Methods Among 736 patients treated for papillary or follicular thyroid cancer (including the Hurthle cell variant) at the University of California-San Francisco (UCSF) Medical Center during the period from , 700 patients with complete data were studied retrospectively. These included 612 patients (87%) who received primary treatment at our institution and 88 patients (13%) who were referred for cancer recurrence. Another 36 patients treated for recurrence were excluded due to insufficient data regarding the initial treatment and/or ptnm staging. The study cohort consisted of 208 male and 492 female patients (male/female ratio 1:2.4). The mean se age at diagnosis was yr; the values for male and female patients were and yr, respectively. Histological classification All specimens obtained at surgery or outside histology slides were reviewed by a senior pathologist (T.R.M.) or an associate. The histological classification was made according to the WHO criteria (14). Six hundred and twenty patients (89%) were classified with papillary thyroid cancer, and 80 patients (11%) had follicular thyroid cancer; the latter group included 8 patients with the Hurthle cell variant of follicular cancer. Primary treatment Papillary or follicular Age 45 yr Age 45 yr Medullary (any age) Anaplastic (any age) I M0 T1 T1 II M1 T2 3 T2 4 III T4 or N1 N1 IV M1 M1 Any T, N or M T (tumor): T1, 10 mm; T2, mm; T3, 40 mm; T4, extrathyroidal. N (node): N0, node negative; N1, node positive. M (metastasis): M0, no distant metastasis; M1, distant metastasis. Total or near-total thyroidectomy has been used as a standard treatment in our institution for differentiated thyroid cancer since Macroscopically suspicious nodes were excised, and modified radical neck dissections were performed in patients with confirmed nodal metastases. This is usually followed by a 30- to 50-mCi out-patient dose radioactive 131 I ablation of residual thyroid tissue 6 12 weeks after operation if focal uptake is detected in the thyroid bed on a 2- to 3-mCi 131 I diagnostic scan. Patients with unifocal papillary cancers 1 cm or less in size (T1N0M0 category) were not given 131 I treatment. In these cases, lobectomy was considered adequate if there were no risk factors, such as radiation exposure or family history of thyroid cancer. Patients with residual tumor or distant metastases were treated with a therapeutic dose ranging from mci 131 I. Total body iodide scanning was repeated after 6 12 months, and reablation with 131 I was performed if there was persistent uptake. All patients were placed on l-t 4 suppressive therapy to maintain subnormal or unmeasurable serum TSH levels, depending upon their disease status. However, the primary treatment regimen among the patients treated elsewhere was heterogeneous, and many had less extensive surgery or no 131 I ablation. TABLE 2. Distribution of patients by ptnm staging between primary treatment cases (primary group) and patients referred for cancer recurrence (secondary group) Stage Primary group Secondary group Total No. % No. % No. % I II III IV Total P for comparison between groups, by Wilcoxon rank sum test. Observed end point Prognostic outcomes were obtained from follow-up examinations and the UCSF Cancer Registry. Follow-up duration was calculated from the time of last evaluation or the time of death. The study end point was either cancer recurrence or death from thyroid cancer. Tumor recurrence was defined as new evidence of loco-regional disease or distant metastases occurring more than 6 months after successful primary therapy. Hence, this was evaluated only in patients who had undergone a potentially curative operation followed by successful 131 I ablative therapy in those with residual tumor. Data analysis Data were analyzed using SAS system 6.11 statistical software (SAS Institute, Cary, NC). Time-dependent variables were analyzed by the Cox proportional hazard models and the Kaplan-Meier product limit estimates of survival curves (15, 16). The primary operation was used as the entry date in survival models; statistical correction (left truncation up to the time of referral to UCSF) for disease-free survival was performed in patients referred for recurrent disease. A log-rank test was used for comparison of survival curves. The observed differences are assumed statistically significant if the probability of chance occurrence is P Results Clinical data according to ptnm staging Table 2 shows the distribution of patients according to ptnm staging: stage I, 516 patients; stage II, 57 patients; stage III, 104 patients; and stage IV, 23 patients. In the group referred for recurrent disease, there were proportionately fewer patients with stage I but more patients with stage III and IV tumors (P ). Distributions of age, sex, and tumor type in the respective ptnm stages are presented in Table 3. The mean age at diagnosis increased with more advanced tumor stages, although this was partly due to the age criteria used in the staging system (Table 1). The distributions by gender and tumor histology were different in the various stages (P ). There was greater female to male preponderance in stage I compared to other stages. Conversely, the predominance of papillary to follicular cancer was markedly reduced in stage II and was not observed in stage IV tumors, respectively. Table 4 shows the type of primary surgery and adjuvant treatment administered to patients in the respective ptnm stages. Primary curative surgery was attempted in 97.7% of the total patients. Total or near-total thyroidectomy was performed in 551 patients (78.7%), whereas 133 patients (19%) had lobectomy or subtotal thyroidectomy, and 16 patients (2.3%) had noncurative surgical debulking. Postoperative adjuvant treatment with 131 I ablation was administered in

3 TNM STAGING FOR THYROID CARCINOMA patients (60.9%), whereas 253 patients (36.1%) received no adjuvant treatment. The remaining 21 patients (3%) were treated by external radiation, immunotherapy, or systemic chemotherapy. The majority of patients receiving adjuvant therapy other than 131 I ablation had locally advanced disease or distant metastases with poor avidity for iodide uptake. TABLE 3. Clinical characteristics of the 700 patients by ptnm staging Stage I (n 516) Stage II (n 57) ptnm Stage III (n 104) Stage IV (n 23) Overall (n 700) Age in yr Mean SE Range Sex a Male Female Ratio 1:3.2 1:1.4 1:1 1:1.9 1:2.4 Tumor type a Papillary Follicular Ratio 12.6:1 1.9:1 9.4:1 0.9:1 7.8:1 a P for comparison between stages by Fisher s exact test. Outcome by ptnm staging Treatment outcome and follow-up data are summarized in Table 5. The median follow-up interval was 10.6 yr, with a mean se of yr. Initial cure, as defined by serum thyroglobulin concentrations less than 4 ng/dl and/or negative total body iodide scans, was achieved in more than 90% of patients after primary treatment of thyroid cancer; this fell from 96.7% in stage I to 87.7% in stage II, 80.8% in stage III, and 26.1% in stage IV tumors. However, 131 of 639 patients (20.5%) with initial cure had recurrence during follow-up; this increased progressively from 15.4% in stage I, to 22% in stage II, 46.4% in stage III, and 66.7% in stage IV tumors. The overall cancer-specific mortality rate was 8.4%; this increased from 1.7% in stage I, to 15.8% in stage II, 30% in stage III, and 60.9% in stage IV tumors. Time to recurrence was evaluated in patients with recurrent disease as the interval from primary treatment that resulted in cure to the first episode of cancer recurrence. As shown in Table 5, the time to recurrence was similar in stage I III tumors, but was markedly shorter in stage IV tumors. Among patients who succumbed to thyroid cancer, the time to death progressively decreased from stage I IV tumors. Table 6 characterizes deaths from thyroid cancer by ptnm staging. One fifth of the patients who had died from thyroid cancer were less than 45 yr of age at the time of primary evaluation and treatment of their malignancy, whereas 62% of the overall patients were in this age category. The youngest fatality had stage I papillary thyroid cancer diagnosed and treated at age 13 yr, he died 10 yr later from widespread peritoneal metastases. Young patients who succumbed to thyroid cancer were often found to have extrathyroidal invasion at presentation, as this constituted 5 of the 7 deaths in stage I patients who presented before 45 yr of age. However, the vast majority of young individuals with extrathyroidal invasion at diagnosis demonstrated an otherwise benign course. An analysis of our data revealed that of the 64 patients who presented below 45 yr of age with extrathyroidal invasion but not distant metastasis (hence stage I), 5 patients (7.8%) died from thyroid cancer eventually, whereas 43 patients (67%) were alive and disease free at the last follow-up. Conversely, of the 57 patients who presented at age 45 yr or greater with extrathyroidal invasion but not distant metastasis (hence stage III), 24 patients (42%) had died from thyroid cancer, whereas only 14 patients (24.6%) were alive and disease free at last follow-up. Besides an overrepresentation of male patients among fatalities from all tumor stages, the absolute number of males who succumbed also outnumbered the females in stages I III. Although patients with follicular thyroid cancer constituted only 11% of the total cases, this histological type accounted for 42% of the cancer deaths. With regard to the major cause of cancerspecific deaths, 30% died from locally invasive tumor, whereas 70% died of distant metastases. A small proportion of patients with locally invasive tumor at presentation eventually succumbed to local effects of tumor invasion into vital structures, including massive hemoptysis, asphyxia, and venous obstruction. Conversely, distant metastases, commonly to the lungs, bones, or brain, accounted for most deaths from the other tumor stages. Other sites of distant metastases noted were the pituitary, adrenal, liver, pericardium, peritoneum, and skin. There were no predisposition of age, sex, or the major cause of death between patients who died from papillary and follicular thyroid cancers. Data on disease-free survival and cancer-specific survival for the 700 patients were tabulated by Kaplan-Meier productlimit estimates of survival. Figure 1 shows the disease-free survival of patients according to ptnm staging; similar plots for cancer-specific survival are illustrated in Fig. 2. These survival curves demonstrate distinctly different prognostic outcomes between the different tumor stages (P , by log-rank test). Analysis of prognostic factors Table 7 shows the risk ratio (RR) and 95% confidence interval of different prognostic variables for tumor recurrence obtained by Cox proportional hazards modelling. After adjusting for age and gender, analysis of the variables used in TNM staging (tumor size, extent of primary tumor, and presence of nodal or distant metastases) uniformly depicted a significant association with tumor recurrence. Both tumor type and tumor focality (variables not used in TNM staging) also demonstrated an association with tumor recurrence; the risks were higher in follicular than papillary thyroid cancers and in multifocal than unifocal tumors. With reference to ptnm stage I tumors, the RR of recurrence was 2.5 in stage II, 5.6 in stage III, and 32 in stage IV tumors. In view of the heterogeneous patient factors among different tumor stages, the effects of sex, tumor histology, and tumor focality were reevaluated after adjustment for ptnm staging: both gender and tumor focality remained as significant predictors of cancer recurrence, whereas tumor type did not confer an independent risk. Similar Cox modelling for the prognostic variables predicting thyroid cancer mortality are presented in Table 8.

4 3556 LOH ET AL. JCE&M 1997 Vol 82 No 11 After adjusting for age and gender, analysis of the variables used in TNM staging similarly depicted a significant association between each variable and mortality. Patients with T2 tumors, however, showed a nonsignificant trend toward an increased mortality over T1 tumors. With regard to tumor type and tumor focality, a significantly higher mortality risk was noted in patients with follicular compared to papillary thyroid cancers, whereas no association was noted between tumor focality and mortality. With reference to ptnm stage I tumors, the RR of death from thyroid cancer was 5.5 in stage II, 11 in stage III, and 28 in stage IV tumors. The increased mortality risk associated with follicular thyroid cancer remained after correction for ptnm staging. Effect of treatment on outcome Table 9 examines the effect of the extent of primary surgery and the use of 131 I ablative therapy on prognostic outcome. As the small patient numbers in stages II IV preclude statistically meaningful evaluation of treatment according to tumor stage, patients with tumors more advanced than T1N0M0 were considered together (n 492) in the respective Cox models. Patients with T1N0M0 tumors were excluded, as the majority of them often do well with relatively minimal therapy. With regard to surgical treatment, the combined TABLE 4. Treatment characteristics of the 700 patients by ptnm staging Stage I (n 516) Stage II (n 57) ptnm Stage III (n 104) Stage IV (n 23) Overall (n 700) Primary surgery Total tx/nt tx st tx/lobec Others Adjuvant treatment Iodine None Others tx, Thyroidectomy; nt, near-total; st, subtotal; lobec, lobectomy. results of operations performed within 6 months of initial assessment are grouped together to indicate the extent of thyroidectomy, if these procedures constituted the intended primary surgical intervention. In the cohort evaluated, patients who had less extensive thyroid surgery (subtotal thyroidectomy or lobectomy) showed an increased risk of recurrence and death, respectively, compared to subjects with more extensive thyroid surgery (total or near-total thyroidectomy). Conversely, those undergoing noncurative debulking surgery constituted a minority of patients with intrinsically poor prognosis and, therefore, are inappropriate for comparison. The disease-free and cancer-specific survival for patients with extensive surgery (total or near-total thyroidectomy) vs. limited surgery (lobectomy or subtotal thyroidectomy) are depicted in Figs. 3 and 4, respectively, showing a clear separation of treatment outcome between the two surgical treatment subgroups. With regard to postoperative 131 I ablative therapy, patients who were treated within 12 months of primary surgery with the intent to ablate normal functioning thyroid TABLE 6. Breakdown of cancer mortality in the various ptnm stages (respective cells depicting the distribution by the number of deaths) Stage I (n 9) Stage II (n 9) ptnm Stage III (n 27) Stage IV (n 14) Overall (n 59) Age 45 yr 7 5 NA NA yr Sex Male Female Tumor histology Papillary cancer Follicular cancer Major cause of death Local invasion Distant metastases NA, Not applicable. TABLE 5. Treatment outcome of the 700 patients by ptnm staging Stage I (n 516) Stage II (n 57) ptnm Stage III (n 104) Stage IV (n 23) Overall (n 700) Follow-up duration Median (yr) Mean (yr) SE (yr) Cancer recurrence No. of patients (%) 77 (15.4) 11 (22) 39 (46.4) 4 (66.7) 131 (20.5) Time to recurrence Median (yr) Mean (yr) SE (yr) Cancer deaths No. of patients (%) 9 (1.7) 9 (15.8) 27 (30) 14 (60.9) 59 (8.4) Time to death Median (yr) Mean (yr) SE (yr) P comparing the proportions with initial cure, cancer recurrence, and cancer deaths, respectively, among the different stages by Wilcoxon rank sum test.

5 TNM STAGING FOR THYROID CARCINOMA 3557 FIG. 1. Disease-free survival for the 700 patients by ptnm staging. P among all stages. FIG. 2. Cancer-specific survival (patients who did not succumb to thyroid cancer) for the 700 patients by ptnm staging. P among all stages. tissue and/or treat residual disease are considered in the 131 I-treated category. In the cohort evaluated, patients without postoperative 131 I ablation had a significantly higher risk of recurrence of, but not death from, thyroid cancer. Like subjects undergoing noncurative surgery, individuals receiving other adjuvant modalities (external radiation, immunotherapy, or systemic chemotherapy) constituted a noncomparable group with an inherently poor prognosis. Figure 5 shows a significantly improved disease-free survival in the subgroup with compared to that without 131 I ablative therapy. Although the cancerspecific survival rates are not statistically different between the two treatment subgroups, a trend toward reduced survival is noted beyond 20 yr of follow-up in the subgroup without 131 I ablative therapy (Fig. 6). Discussion Our long follow-up interval gives an adequate perspective to study the prognostic value of the ptnm staging system. The overall cancer-specific mortality in our patients with papillary and follicular thyroid cancers was 8.4% at a mean follow-up period of 11.3 yr. This overall figure is very similar to those of several other series with comparable follow-up intervals (4, 12, 17). Nevertheless, mortality from thyroid cancer varied significantly across the different ptnm stages, ranging from 1.7% in stage I to 60.9% in stage IV tumors. Our series revealed that the majority of thyroid cancer patients, excluding those with distant metastases at diagnosis, developed recurrence within the first 5 yr of diagnosis. This finding is consistent with other series and emphasizes the need for frequent follow-up examinations regardless of low risk categories, especially during the first 5 yr after primary surgery (11, 18). Despite being a relatively benign disease, a continual decline in both disease-free survival and cancer-specific survival is noted in all tumor stages at successive follow-up intervals. This underscores that surveillance should be life-long, but may be performed less frequently

6 3558 LOH ET AL. JCE&M 1997 Vol 82 No 11 TABLE 7. Risk factors for cancer recurrence by Cox models Variable Risk ratio 95% CI P value Variables adjusted for age and sex Tumor type Papillary cancer 1 Reference Follicular cancer Tumor size T1: 10 mm 1 Reference T2: mm T3: 40 mm T4: extrathyroidal invasion Tumor focality Unifocal 1 Reference Multifocal Nodal metastases N0: negative 1 Reference N1: positive Distant metastases M0: negative 1 Reference M1: positive ptnm staging Stage I 1 Reference Stage II Stage III Stage IV Variables adjusted for ptnm stages Sex Male 1 Reference Female Tumor type Papillary cancer 1 Reference Follicular cancer NS (0.70) Tumor focality Unifocal 1 Reference Multifocal when a decade or more has passed with no evidence of relapse. Using the Cox model with adjustment for age and gender, the variables used in TNM staging (namely tumor size, presence of extrathyroidal invasion, and initial nodal or distant metastatic lesions) are found to be important prognostic factors for recurrence or death from thyroid cancer. With regard to the primary tumor size, none of the patients with an intrathyroidal tumor of 1 cm or less died of cancer, an observation similarly reported by others (6, 19, 20). Conversely, patients with extrathyroidal invasion (T4) had more than 3-fold risk of recurrence and death, respectively, compared to those with large intrathyroidal (T3) tumors, similar to the findings reported by DeGroot et al. (4). We found a 4-fold risk of recurrence and a 2.5-fold risk of cancer-specific death in patients with regional lymph node metastases. The presence of cervical lymph node metastases has been variously reported to be associated with an unchanged (4, 8, 12, 13, 19), worse (1, 6), or even better (21) survival; the discrepancies in earlier reports may be contributed by their correlation with other prognostic factors. Mazzaferri and co-workers had earlier noted an increased risk of recurrence, but not mortality, in patients with nodal metastases (19); however, their more recent analysis of a larger patient cohort showed significantly higher 30-yr cancer recurrence and mortality rates in subjects with bilateral cervical or mediastinal lymph node metastases regardless of tumor histology (6). Using matched pair analysis for similar prognostic risk factors, Hughes and co-workers found that nodal disease carried an increased risk of recurrence and a tendency toward lower 20-yr survival in patients 45 yr or older (22). Most investigators currently agree that regional nodal metastases indicate more extensive disease, and that affected patients have a tendency toward regional nodal recurrence (4, 12, 22 24). After correction for ptnm staging, there was a lower risk of recurrence, but not death, from thyroid cancer in the female patients, similar to the observation reported by DeGroot and co-workers (4). Although some investigators identified gender as a strong independent predictor of survival (1, 8, 12), others failed to show gender as a consistent independent prognostic factor (19, 20, 23). In a large population-based study by Gilliland and co-workers, the investigators found that males had a 50% greater risk of death than females after adjusting for other factors in the multivariate models (25). However, as the absolute differences between relative survival for males and females were small, the investigators concluded that gender is not a strong predictor of survival. We found a higher risk of recurrence, but not death, in patients with multifocal tumors, even after correction for ptnm stages. The biological potential of multifocal thyroid cancer is uncertain, as most studies do not report an association between tumor multicentricity and prognosis (4, 8, 13, 23). However, Mazzaferri and Jhiang (6) reported increased cancer mortality rates in patients with three or more foci of papillary or follicular thyroid cancers, although this association was lost in multivariate analysis. Until more data are available, it is believed prudent to manage patients with

7 TNM STAGING FOR THYROID CARCINOMA 3559 TABLE 8. Risk factors for thyroid cancer deaths by Cox models Variable Risk ratio 95% CI P value Variables adjusted for age and sex Tumor type Papillary cancer 1 Reference Follicular cancer Tumor size T1: 10 mm 1 Reference T2: mm NS (0.50) T3: 40 mm T4: extrathyroidal invasion Tumor focality Unifocal 1 Reference Multifocal NS (0.18) Nodal metastases N0: negative 1 Reference N1: positive Distant metastases M0: negative 1 Reference M1: positive ptnm staging Stage I 1 Reference Stage II Stage III Stage IV Variables adjusted for ptnm stages Sex Male 1 Reference Female NS (0.13) Tumor type Papillary cancer 1 Reference Follicular cancer TABLE 9. Effect of treatment on prognosis in 492 patients with tumors more advanced than T1N0M0 category Treatment variable Risk ratio 95% CI P value Cancer recurrence Extent of surgery Total/near-total thyroidectomy 1 Reference Lobectomy/subtotal thyroidectomy Debulking surgery Adjuvant therapy Iodine ablation 1 Reference None Other modalities Deaths from thyroid cancer Extent of surgery Total/near-total thyroidectomy 1 Reference Lobectomy/subtotal thyroidectomy Debulking surgery Adjuvant therapy Iodine ablation 1 Reference None NS (0.76) Other modalities multifocal tumors by total or near-total thyroidectomy and 131 I remnant ablative therapy (6, 20, 26). In our series, the increased risk of cancer-specific death noted with follicular thyroid cancer persisted after correction for ptnm staging. The prognostic differences between papillary and follicular thyroid cancer have not been resolved; some investigators have reported a less favorable outcome in patients with follicular cancer (3, 27), others considered the distinction to be of mere academic interest (20, 23, 28). Gilliland et al. (25) found that patients with follicular cancer have lower survival rates than those with papillary cancer, but they noted that the prognosis is more strongly determined by tumor staging and other factors than by tumor histology. Mazzaferri and Jhiang (6) likewise noted a higher mortality in patients with follicular thyroid cancer, but this significance was lost when subjects who presented with distant metastases at diagnosis were excluded. To help resolve this enigma and to provide a more precise risk group assignment, patients with papillary and follicular thyroid cancer are best analyzed as separate cohorts in the ptnm classification. Like all other thyroid cancer staging systems, the ptnm classification has limitations, as a minority of patients in the low risk group will die of thyroid cancer. In our evaluation,

8 3560 LOH ET AL. JCE&M 1997 Vol 82 No 11 FIG. 3. Disease-free survival in patients with tumors other than T1N0M0 category, comparing the extensive surgery group (total or near-total thyroidectomy) vs. the limited surgery group (subtotal thyroidectomy or lobectomy). P between treatment groups. FIG. 4. Cancer-specific survival in patients with tumors other than T1N0M0 category, comparing the extensive surgery group (total or near-total thyroidectomy) vs. the limited surgery group (subtotal thyroidectomy or lobectomy). P 0.01 between treatment groups. the majority of case fatalities from stage I and II tumors occurred in patients who were diagnosed and treated for thyroid cancer well before age 45 yr. Four of the cancer deaths in our series (three papillary and one follicular thyroid cancers) occurred in individuals less than 30 yr of age at the time of diagnosis. Although most of the young subjects who died from stage I tumor had locally advanced disease, the vast majority of the young cohort with extrathyroidal invasion at diagnosis demonstrated a benign course. Indeed, the challenge would be to identify the risk factors that can effectively select out young patients with advanced disease who will show poor outcome from the majority with relatively good prognosis. Despite the limitations of ptnm classification, our retrospective analysis of a large number of patients shows that this is helpful in defining groups of patients with significantly different probability of cancer recurrence and mortality. The earlier reports on this classification, however, failed to clearly separate patients with stage I and II tumors with regard to prognostic outcome (11 13), a limitation commonly noted in other staging systems (1, 4, 5). A comparison study on five different prognostic classification schemes (EROTC, AGES, DeGroot s, AMES, and TNM) by DeGroot and associates on their institutional series indicated that the TNM system is the most satisfactory for stratifying patients by risks (29). As the small patient numbers precludes us from analyzing the outcome of treatment by ptnm stages, we could not assess the utility of this classification in guiding therapy to avoid overaggressive treatment in patients with excellent prognosis. Until more data become available, our analysis of patients with differentiated thyroid cancer more advanced than the T1N0M0 category supports the current practice to manage these individuals more aggressively. In many studies, recurrence rates are evidently higher after a partial, compared to a total, thyroidectomy, even after adjustment for extent of disease (6, 13, 17, 19, 29). The impact of recurrence on survival is underscored by follow-up results indicating that up to 40 50% of patients who die of thyroid cancer do

9 TNM STAGING FOR THYROID CARCINOMA 3561 FIG. 5. Disease-free survival in patients with tumors other than the T1N0M0 category, comparing the 131 I ablation group vs. the no 131 I ablation group (excluding patients receiving other types of adjuvant therapy). P between treatment groups. FIG. 6. Cancer-specific survival in patients with tumors other than the T1N0M0 category, comparing the 131 I ablation group vs. the no 131 I ablation group (excluding patients receiving other types of adjuvant therapy). P 0.76 (NS) between treatment groups. so because of recurrent disease in the thyroid bed or central compartment of the neck (18). Furthermore, total thyroidectomy facilitates more effective 131 I ablative therapy and the use of serum thyroglobulin levels as tumor marker for cancer recurrence during follow-up. Results from our analysis also support the use of adjuvant 131 I ablative therapy in most patients who have undergone total or near-total thyroidectomy. Similarly, Mazzaferri s group found a significantly lower tumor recurrence rate in patients with larger ( 1.5 cm) tumors treated with 131 I ablation and T 4 suppression compared to those treated with T 4 suppression alone, notwithstanding that patients given 131 I ablative therapy had more adverse risk factors, such as invasive tumors and cervical nodal metastases (19, 30). Although improvement in cancer-specific survival was not observed in our patient cohort with 131 I ablative therapy, Mazzaferri and Jhiang noted improvement in both cancer recurrence and mortality rates when the 131 I-treated cohort was evaluated at a longer interval of 30 yr (6). Based on their findings, they concluded that total or near-total thyroidectomy, followed by radioiodine ablation and life-long thyroid hormone suppression, are indicated in subjects if their primary thyroid tumors are larger than 1.5 cm or are associated with multicentricity, local invasion, or metastases (6). DeGroot and co-workers also found that 131 I ablation was associated with a significant reduction of recurrences independent of the extent of surgery, and they supported postoperative 131 I ablation of residual thyroid tissue in low risk papillary thyroid cancer patients with tumors greater than 1 cm (4, 29). However, controversies still exist regarding the benefit of thyroid remnant ablation, with others reporting no improvements (13, 31, 32). Patients with follicular thyroid cancer in our series are not evaluated separately, as they constituted only 11% of the study cohort. However, based on the 3.4-fold higher cancer mortality risk obtained by multivariate analysis, it is prudent to consider total thyroidectomy as standard surgery in most patients with follicular thyroid cancer, followed by 131 I ablation if focal iodide uptake is evident on diagnostic scanning (26, 33).

10 3562 LOH ET AL. JCE&M 1997 Vol 82 No 11 In conclusion, our data support the use of ptnm staging system for differentiated thyroid cancer. Although it may be cautioned that the prognostic value of ptnm staging system could vary among groups of patients, its potential in providing risk stratification deserves further reporting from other centers. Until better predictors of tumor behavior are available, the ptnm classification is shown to be useful for prognostication, and its widespread use will facilitate the exchange of information between centers. Acknowledgment The authors are grateful to Dr. Fang Dong for statistical assistance. References 1. Byar DP, Green SB, Dor P, et al A prognostic index for thyroid carcinoma. A study of the E.O.R.T.C. Thyroid Cancer Cooperative Group. Eur J Cancer. 15: Hay ID, Grant CS, Taylor WF, McConahey WM Ipsilateral lobectomy versus bilateral lobar resection in papillary thyroid carcinoma: a retrospective analysis of surgical outcome using a novel prognostic scoring system. Surgery. 102: Cady B, Rossi R An expanded view of risk-group definition in differentiated thyroid carcinoma. Surgery. 104: DeGroot LJ, Kaplan EL, McCormick M, Straus FH Natural history, treatment, and course of papillary thyroid carcinoma. J Clin Endocrinol Metab. 71: Hay ID, Bergstralh EJ, Goellner JR, Ebersold JR, Grant CS Predicting outcome in papillary thyroid carcinoma: development of a reliable prognostic scoring system in a cohort of 1779 patients surgically treated at one institution during 1940 through Surgery. 114: Mazzaferri EL, Jhiang SM Long-term impact of initial surgical, and medical therapy on papillary and follicular thyroid cancer. Am J Med. 97: Shaha AR, Loree TR, Shah JP Prognostic factors and risk group analysis in follicular carcinoma of the thyroid. Surgery. 118: Shaha AR, Shah JP, Loree TR Risk group stratification and prognostic factors in papillary carcinoma of thyroid. Ann Surg Oncol. 3: Hermanek P, Sobin LH, eds UICC: TNM classification of malignant tumors, 4th ed. Berlin: Springer-Verlag. 10. Beahrs OH, Henson DE, Hutter RVP, Myers M, eds American Joint Commision on Cancer: mauual for staging of cancer, 3rd ed. Philadelphia: Lippincott 11. Kukkonen ST, Haapiainen RK, Franssila KO, Sivula AH Papillary thyroid carcinoma: the new, age-related TNM classification system in a retrospective analysis of 199 patients. World J Surg. 14: Salvesen H, Njolstad PR, Akslen LA, Albrektsen G, Soreide O, Varhaug JE Papillary thyroid carcinoma: a multivariate analysis of prognostic factors including an evaluation of the p-tnm staging system. Eur J Surg. 158: Hay ID Papillary thyroid carcinoma. Endocrinol Metab Clin North Am. 19: Headinger CHR, ed WHO: International histological classification of tumors. Berlin: Springer-Verlag. 15. Cox DR Regression models and life tables. J R Stat Soc. 34: Kaplan EL, Meier P Nonparametric estimation from incomplete observations. J Am Stat Assoc. 53: Segal K, Friedental R, Lubin E, Shvero J, Sulkes J, Feinmesser R Papillary carcinoma of the thyroid. Otolaryngol Head Neck Surg. 113: Gagel RF, Goepfert H, Callender DL Changing concepts in the pathogenesis and management of thyroid carcinoma. Cancer J Clin. 46: Mazzaferri EL, Young RL Papillary thyroid carcinoma: a 10 year follow-up report of the impact of therapy in 576 patients. Am J Med. 70: Segal K, Raveh E, Lubin E, Abraham A, Shvero J, Feinmesser R Welldifferentiated thyroid carcinoma. Am J Otolaryngol. 17: Cady B, Sedgwick CE, Meissner WA, Brokwatter JR, Romagosa V, Werber J Changing clinical pathologic, therapeutic, and survival patterns in differentiated thyroid carcinoma. Ann Surg. 184: Hughes CJ, Shaha AR, Shah JP, Loree TR Impact of lymph node metastasis in differentiated carcinoma of the thyroid: a matched-pair analysis. Head Neck. 18: Simpson WJ, McKinney SE, Carruthers JS, Gospodarowicz MK, Sutcliffe SB, Panzarella T Papillary and follicular thyroid cancer: prognostic factors in 1,578 patients. Am J Med. 83: Clark OH Predictors of thyroid tumor aggressiveness. West J Med. 165: Gilliland FD, Hunt WC, Morris DM, Key CR Prognostic factors for thyroid carcinoma. A population-based study of 15:698 cases from the surveillance, epidemiology and end results (SEER) program Cancer. 79: Tezelman S, Clark OH Current management of thyroid cancer. Adv Surg. 28: Franssila KO Is the differentiation between papillary and follicular thyroid carcinoma valid? Cancer. 32: Ontai S, Straehley CJ The surgical treatment of well-differentiated carcinoma of the thyroid. Am Surgeon. 51: DeGroot LJ, Kaplan EL, Straus FH, Shukla MS Does the method of management of papillary thyroid carcinoma make a difference in outcome? World J Surg. 18: Mazzaferri EL Papillary thyroid carcinoma: factors influencing prognosis and current therapy. Semin Oncol. 14: Grebe SK, Hay ID Thyroid cancer nodal metastases: biologic significance and therapeutic considerations. Surg Oncol Clin North Am. 5: Simpson WJ, Panzarella T, Carruthers JS, Gospodarowicz MK, Sutcliffe SB Papillary and follicular thyroid cancer: impact of treatment in 1578 patients. Int J Radiat Oncol Biol Physiol. 14: Singer PA, Cooper DS, Daniels GH, et al Treatment guidelines for patients with thyroid nodules and well-differentiated thyroid cancer. Arch Intern Med. 156: