| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Section of Endocrinology and Metabolism, University of Illinois College of Medicine (A.B., E.S.-F., T.G., D.M., A.B.S.), Chicago, Illinois 60612; Michael Reese Hospital (E.S.-F.,T.G., A.B.S.), Chicago, Illinois 60616; and Radiation Epidemiology Branch, National Cancer Institute (E.R.), Bethesda, Maryland 20892
Address all correspondence and requests for reprints to: Arthur B. Schneider, M.D., Ph.D., Section of Endocrinology and Metabolism, University of Illinois, 1819 West Polk Street, MC 640, Chicago, Illinois 60612. E-mail: abschnei{at}uic.edu
Abstract
Thyroid cancers detected by screening irradiated individuals are often small and of uncertain clinical significance. We retrospectively analyzed the effect of screening in a cohort of 4296 individuals exposed to radiation as children in the 1940s and 1950s and followed by us from 1974 until the present. We compared the thyroid cancers diagnosed before 1974 (122 cases, routine care) with the cancers found in subjects screened by us after 1974 (172 cases, screened), using cancer recurrence as the end point. Screening included a thyroid scan or, more recently, thyroid ultrasound. As expected, many of the cancers found by screening were very small (52% were <10 mm), but the range of tumor sizes overlapped those found by routine care. The recurrence rate was significantly lower in the cases found by screening, but when the comparison was limited to cancers 10 mm or larger, no difference in the recurrence rates was seen. This would suggest that the lower recurrence rate observed for small thyroid cancers detected at screening was due to earlier diagnosis rather than more effective treatment. By univariate analysis, four factors were associated with an increased risk of recurrence of small (<10 mm) thyroid cancers: short latency (i.e. a shorter time interval between the radiation exposure and the first thyroid surgery), lymph node metastases present at diagnosis, multifocal cancers, and higher radiation dose. In a multivariate analysis combining the four risk factors, only short latency was significant. As thyroid cancers that escape detection by routine means should be diagnosed at screening, and both large and small thyroid cancers have the potential to recur, screening may be of value, but only if groups with a sufficiently high prevalence of thyroid cancer can be identified to offset the adverse effects of unnecessary treatment due to false positive results.
IT IS WELL known that exposure to external radiation for therapeutic purposes, particularly during childhood for benign head and neck conditions, is associated with an enhanced risk of thyroid cancer (1, 2, 3, 4). It has been estimated that in the U.S. there are over 1 million individuals who received such exposure. The association with internal 131I exposure is less clear; however, studies conducted after the Chernobyl accident indicate that childhood exposure increases the incidence of thyroid cancer (5, 6, 7). As a result of the accident, an entire population in a large geographic area was placed at risk. These findings have increased concern among other people who have lived in the vicinity of nuclear facilities or atomic weapons testing grounds because iodine isotopes were released into the atmosphere. Radiation doses to the thyroid resulting from the U.S. atomic weapons tests were high for milk-drinking young children who lived near the Nevada test site in the early 1950s (8).
These exposures have raised an important public health issue regarding the medical value of screening for thyroid disease, particularly nodular disease (9). The goal of the work described here is to evaluate thyroid screening in an irradiated cohort and the clinical behavior of small cancers detected by screening.
Subjects and Methods
Radiation-exposed cohort
Between the years 1939 and 1962, 4296 patients, less than 16 yr of age, received radiation treatments for benign conditions in the head and neck area at Michael Reese Hospital (Chicago, IL). Prior publications describe this cohort and the clinical observations that have been made (3, 10). The medical records of these patients were located, and a screening or follow-up program was initiated in 1974. At that time, an effort was made to contact all irradiated patients. When contacted, the individuals were notified of the radiation exposure, and a detailed medical history with emphasis on thyroid diseases was obtained. They were invited to Michael Reese Hospital for clinical evaluation, which included a physical examination of the head and neck area and thyroid imaging. The imaging was, until recently, a thyroid scan using 99mTc-pertechnetate and a pinhole collimator; since 1993 ultrasound has been used. When abnormalities were detected, the subjects were referred to their own physicians. Follow-up information was obtained by subsequent correspondence and/or visits to the program. If thyroid surgery was performed, surgical and pathology reports as well as histological slides were reviewed when possible. During the course of this follow-up, the histological classification of well differentiated thyroid cancer changed (11). As the slides for most of the cases had been returned to the original institutions, cancers were reclassified based on medical records. Mixed papillary-follicular cancers were reclassified as papillary cancer. Follicular cancers were reclassified as follicular variant of papillary carcinoma when the original pathologist or our reviewing pathologist mentioned the presence of nuclear inclusions. Other follicular cancers were not reclassified because insufficient information was available from the reports, and pathology slides could not be reviewed. There were no medullary or anaplastic carcinomas or thyroid lymphomas.
Study groups and comparisons
To date, 382 individuals in the exposed cohort have been diagnosed with thyroid cancer. Two groups of thyroid cancer patients are compared in this report. The first had thyroid cancers detected before the beginning of the screening program, i.e. before 1974 and is referred to as the routine group (n = 122). The second had thyroid cancers diagnosed after 1974 as part of a Michael Reese Hospital screening program and is referred to as the screened group (n = 172). Another 88 cases of thyroid cancer, diagnosed after 1974 outside of Michael Reese Hospital, are not included in the comparison of the routine with screened group, but are included in the analysis of the behavior of small thyroid cancers (see below).
Each case of thyroid cancer was characterized by sex (female = 1,
male = 2), age at radiation treatment (years), radiation dose to
the thyroid (Gy, gray), age at time of thyroid cancer surgery (years),
and latency (time from radiation treatment to first surgery) as well as
the pathological findings at the time of the initial surgery [tumor
size (millimeters), lymph node involvement (absent = 0,
present = 1), bilaterality (unilateral = 1, bilateral =
2), invasiveness (none = 0, into capsule = 1, into soft
tissues = 2), multifocality (unifocal = 1, multifocal =
2)], the initial treatment [extent of surgery (
50% removed =
1, >50% removed = 2), use of radioactive iodine for completion
(not used = 0, used = 1)], and subsequent evidence of
recurrence (yes/no). The size of the thyroid cancer was taken as the
largest dimension of the largest intrathyroidal focus of the cancer.
Cancers less than 10 mm were classified as small.
Recurrences of thyroid cancer that were confirmed by surgery were reviewed as described above. Other recurrences were detected by 131I scans. Focal uptake outside the thyroid bed and/or uptake in the thyroid bed after ablative (>75 mCi) treatment was considered confirmation of recurrence.
Statistical analysis
Statistical analyses of between-group comparisons for discrete
variables were carried out by 
2 analysis, and
those for continuous variables were performed using t test.
Recurrences were analyzed by Kaplan-Meier plots and were compared using
the log-rank test (12). A difference at the level of
P < 0.05 was considered significant. Univariate and
multivariate comparisons of recurrences between groups were carried out
by the proportional hazards method of Cox (13). For each
covariate the estimate of the relative risk and its 95% confidence
interval were determined. When the confidence interval excluded 1.0,
the risk was considered significant. The analyses were implemented
using NCSS 6.0 (Kaysville, UT), SigmaPlot 4.0 (SPSS, Inc.,
Chicago, IL) for the Kaplan Meier plots and Epicure (14)
for univariate and multivariate proportional hazards analyses.
Results
Effects of screening
The characteristics of the 122 thyroid cancers in the routine
group and the 172 in the screened group are summarized in Table 1
. Before the initiation of screening
more of the cancers occurred in women (54.1%); afterward more occurred
in men (60.5%). This represented a shift toward the gender composition
of the total cohort (
55% of the cohort are men). The ages at first
radiation treatment and the thyroid doses were very similar in the two
groups. Age at cancer surgery and latency were greater in the screened
group, because, by design, all of their cancers were diagnosed in 1974
and later.
|
A frequency histogram of the sizes of the thyroid cancers in the
routine and screened groups and the sizes of the thyroid cancers that
subsequently recurred are shown in Fig. 1. Over 50% of the screened patients had
small cancers (<10 mm), and almost all of these did not recur.
Recurrences were more frequent in the routine group compared with
the screened group (Fig. 2, upper
panel). The plots separated at about 3 yr after surgery and
continued to diverge so that 24 yr after surgery, slightly over 85% of
the patients diagnosed at screening had not developed a recurrence,
whereas only 72% of the routine group remained recurrence free.
|
|
, lower panel). When only these larger cancers were
considered, very little difference between the screened and routine
groups was observed. Behavior of small (<10 mm) thyroid cancers
Among the 382 thyroid cancers, tumor size of the original cancer was not available for 31 cases. Among the remaining 351 cases, 151 of the original cancers were less than 10 mm in greatest dimension. One cancer was found at autopsy, and no follow-up information was available for another, leaving 149 small cancers available for analysis.
The frequency of recurrence among the 149 subjects with small thyroid
cancers was significantly less (P = 0.0005) than that
in the subjects with larger thyroid cancers (Fig. 3). In total, only 11 of the 149 cases
recurred, all in patients with papillary thyroid cancer. A comparison
of the characteristics of the patients with and without recurrences
revealed several potentially explanatory factors, although due to the
small number of recurrences, few were statistically significant (Table 2). The individuals with recurrence had
received a larger dose of radiation to the thyroid, and their cancers
occurred at an earlier age and after a shorter latency (the latter
reaching statistical significance). Recurrences were more frequent in
cases with lymph node involvement, bilateral involvement, multifocal
involvement, and invasion (with lymph node involvement reaching
statistical significance). Recurrences were not associated with the
extent of surgery or the use of radioactive iodine for completion.
Table 3 shows the clinical findings in
the 11 recurrent cases. At last follow-up (mean, 16.2 yr; range,
2.4–42.1 yr), none had known residual thyroid cancer or had died of
thyroid cancer.
|
|
|
). Among the cancers larger than 10 mm,
size was the only significant factor, whereas several factors were
significantly related to recurrences in the small cancers. Among the
general characteristics of the population, higher dose, younger age at
cancer surgery, shorter latency, and detection by routine care were
significant predictors of recurrence. When the latter three highly
correlated time-dependent variables (latency, age at cancer surgery,
and detection by routine care vs. screening) were analyzed
together, latency was the dominant factor and was used in the
multivariate analyses. Among the pathological characteristics, lymph
node involvement and multifocal involvement were significant predictors
of recurrence. In a multivariate analysis (Table 5) of the general characteristics, dose
and latency remained significant factors. In a multivariate analysis of
the two pathological characteristics, neither was significant. When the
four characteristics were analyzed, only latency was a significant risk
factor for the recurrence of small thyroid cancers. The estimated
relative risk for latency was very similar in the univariate and
multivariate analyses. The final multivariate analysis indicates that
the risk of recurrence falls by about 14% for each year separating the
radiation exposure from the first thyroid cancer surgery.
|
|
In this paper we examined the behavior of small thyroid cancers
and whether the clinical outcomes for thyroid cancer are improved by
screening. The findings demonstrate again that many radiation-related
thyroid cancers are not diagnosed during routine medical care, but are
discovered by screening (10). Early detection by screening
appeared to improve the outcome by reducing the frequency of recurrence
in the screened group. However, another likely interpretation is that
the clinical course of the thyroid cancers was not altered by
screening, but, rather, because the cancers were diagnosed earlier,
there appeared to be a longer time interval before recurrences occurred
(lead bias). As the clinical outcomes for the screened and routine
cases of the larger (
10 mm) thyroid cancers did not differ, the
apparent benefit of screening appears to be due in large part to the
detection of small thyroid cancers.
There are two observations, however, that indicate that thyroid cancer screening may be of benefit. First, some large cancers that escaped detection among persons receiving routine care were diagnosed at screening. We recently showed that about half of ultrasound-detected thyroid nodules larger than 15 mm were not palpable in this cohort (15). Similar to the higher recurrence rate seen for large thyroid cancers in nonradiation-exposed patients, in this irradiated cohort thyroid cancer size was the predominant risk factor for recurrence among the cancers that were 10 mm or larger. In addition, screening detected a large number of smaller thyroid cancers as discussed below. They also can recur and require further therapy.
Although many reports describe thyroid cancers found as a result of screening, including recent reports from the Chernobyl area (7), there has been very little attention to documenting the potential benefits of screening compared with routine care. The most direct attempt to evaluate thyroid screening was carried out in Japan; Ishida et al. (16) screened 152,651 women for thyroid cancer, followed them for 7 yr, and then screened 3,557 of them again. At the initial screening thyroid cancers were detected. When the cancers found by screening were compared with those diagnosed in their out-patient clinic (unscreened group), the cancers found by screening were smaller in size and less likely to have lymph node metastases. The cumulative 7-yr survival rate was 97.6% in the screened group compared with 90.3% in the out-patient group. However, as reviewed by Eden et al. (17), potential biases in the selection of the groups make it difficult to interpret these data.
Observations by Mazzaferri and Jhiang (18) suggest that delayed treatment of thyroid cancer is associated with increased mortality. In their 30-yr follow-up study of 1355 thyroid cancer patients, they found that the likelihood of cancer death increased as the time interval between when the cancer became clinically evident and when it was finally removed increased. If screening leads to prompt evaluation and, when necessary, surgery, it is likely to have a beneficial effect on outcome. The observations of Mazzaferri and Jhiang, however, are limited by the fact that their analyses were retrospective, and the reasons for the delayed surgeries were not determined.
The value of screening can only be established definitively by a randomized prospective study. Retrospective analyses, such as the ones presented here, cannot be interpreted unambiguously. The potential biases are especially relevant when analyzing thyroid cancer, because most cases progress very slowly, and some may not progress at all.
The behavior of small thyroid cancers in radiation- exposed patients is not as well described as it is in other patients. The data presented here highlight the similarities as well as some differences. The frequency of recurrences was lower among the patients with small cancers, but it was not negligible. In some instances the recurrences of the small cancers required further surgery.
Although it is evident that screening will detect small thyroid cancers, it is less clear whether there is a benefit to detecting these microcarcinomas. Microcarcinomas have a favorable prognosis, but there have been reports of patients who have died of their cancer. Two large studies of thyroid microcarcinomas show that not all microcarcinomas are cured at the time of first surgery (19, 20). Hay et al. (19) reported follow-up of as long as 48 yr (median, 16 yr) for 400 cases of papillary thyroid microcarcinoma. There were 2 deaths due to thyroid cancer as a result of the incident cancers. There were a total of 27 recurrences, with a 6% cumulative rate at 20 yr after the initial surgery. Most of the recurrences were observed within 10 yr of the initial surgery. Recurrences were associated with lymph node metastases at the initial surgery and lobectomy as the initial procedure. Yamashita and Noguchi (20) followed 1734 patients with microcarcinomas for a mean interval of 11.2 yr. There were 31 recurrences (1.8%, not corrected for varying length of follow-up), and 4 patients died of recurrent thyroid carcinoma. They observed that invasion of metastatic thyroid cancer through the lymph node capsule found at the time of initial surgery was the only significant risk factor for recurrence.
This is the first report analyzing the risk factors for recurrence of small thyroid cancers in irradiated patients. For the small thyroid cancers in our cohort, four factors were found to be associated with recurrences. 1) The increased risk of recurrence associated with short latency may be a marker of younger age at diagnosis, a well known risk factor for thyroid cancer recurrence (18, 21). Because of our study design, younger age at diagnosis, shorter latency, and discovery by routine care were highly correlated. Short latency may also be an indicator of inherently faster growth and earlier metastasis, possibly as a result of radiation-related somatic mutations. Despite this risk, mortality from thyroid cancer arising at a young age is rare. 2) The presence of metastases in lymph nodes was a risk factor. It is included in most, but not all, prognostic staging systems for recurrence of thyroid cancer (22). 3) The observation that multifocal involvement is a risk factor for recurrence has not been made previously. Multifocal cancers are seen frequently in radiation-related cases (10), and it would be expected that the possibility of incomplete surgical removal would be increased in these cases. Also, multicentricity may be a sign of increased susceptibility to the effects of radiation, making new primaries more likely. 4) Why radiation dose is associated with increased risk of recurrence of small thyroid cancers, an observation that has not been made previously, is not clear. It seems likely that higher radiation exposure would increase the possibility of new primary cancers, especially small ones. It is possible that future work using the molecular characteristics of the cancers will be able to differentiate between cancer recurrences and new primary cancers.
In summary, the current study shows that thyroid screening of irradiated individuals identifies many small thyroid cancers associated with a very good prognosis. Some small cancers recur, especially those occurring with a short latency. To the extent that small cancers are included in the evaluation of a screening program, its benefits may be overestimated. However, screening also detects larger cancers, where a clinical benefit is more likely. Only a controlled prospective study will be able to prove clinical benefit, but the logistics of such a study make it unlikely that one will ever be performed.
Footnotes
This work was supported in part by NCI Grant RO1-CA-21518.
Received February 7, 2001.
Accepted April 18, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  D. Mihailescu, E. Shore-Freedman, S. Mukani, J. Lubin, E. Ron, and A. B. Schneider Multiple Neoplasms in an Irradiated Cohort: Pattern of Occurrence and Relationship to Thyroid Cancer Outcome J. Clin. Endocrinol. Metab., July 1, 2002; 87(7): 3236 - 3241. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
