| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Department of Internal Medicine, Division of Diabetology and Endocrinology (E.M.E., B.B.), and Department of Occupational and Environmental Medicine (Z.M., L.H.), University Hospital, S-221 85 Lund, Sweden
Address all correspondence and requests for reprints to: Dr. Eva Marie Erfurth, Department of Internal Medicine, Division of Diabetology and Endocrinology, University Hospital, S-221 85 Lund, Sweden. E-mail: eva-marie.erfurth{at}med.lu.se
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In a recent meta-analysis it was concluded that serum IGF-I was a similarly strong risk factor for prostate cancer as testosterone (18). Compared with normal subjects, hypopituitary patients receiving conventional hormone treatment, but without GH substitution, have significantly lower mean serum IGF-I and IGFBP-3 levels (19).
Today, information on the risk for secondary neoplasms in patients with hypopituitarism receiving conventional hormone treatment but without GH substitution is lacking. Regarding tumor mortality in hypopituitary patients there have been three studies published, and the results are contradictory (20, 21, 22). The drawback with mortality studies, however, is that only lethal tumors are included, and the diagnostic accuracy is lower than in incidence studies.
As an increasing number of adults with GH deficiency are subjected to GH substitution, there is a pressing need for establishing baseline tumor incidence data in this group of patients. As the majority of adult patients with hypopituitarism and demonstrated GH deficiency have a history of pituitary tumor, we considered this group an appropriate study group.
The main aim of the present study was to assess the risk for a second tumor in a cohort of 328 patients with hypo-pituitarism treated for pituitary tumor. The patients were receiving conventional hormone treatment, but without GH substitution.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During the period 1946–1988, 477 subjects were operated for a pituitary tumor at the Department of Neurosurgery, Lund University Hospital (Lund, Sweden). One hundred and thirty-three of these were excluded from the study base for various reasons, including acromegaly, Cushing’s disease, death during the first postoperative month, insufficient information in medical records, or GH treatment in childhood. This initial cohort of 344 patients with pituitary insufficiency has previously been investigated with respect to overall and cardiovascular mortality (22). All could be identified with a unique 10-digit personal code. Sixteen patients had died before 1958, when the Swedish Cancer Registry was started, and were therefore excluded from the final study cohort, which thus consisted of 328 (123 females) patients.
The observation period started at the date of diagnosed pituitary insufficiency or, if this occurred earlier than surgery, the date of surgery plus 1 month. The median age at the start of observation was 54.5 yr (range, 3.1–81.8; 16 of the patients were 18 yr or younger at inclusion). The observation period continued until December 31, 1992. Date of death, emigration, or a second tumor diagnosis was used as individual end point of follow-up if occurring before 1992. One hundred and seventy-seven of the patients died during the observation period, and 2 (0.6%) emigrated. The median age of the deceased patients was 71.8 yr (range, 3.5–91.3). The remaining 149 patients were still alive and living in Sweden at the end of the observation period. The median age of these subjects at the end of follow-up was 62.6 yr (range, 12.8–85.9).
The characteristics of this cohort are shown in Table 1
. Transcranial operations were performed
in 304 (93%) patients; the rest were transsphenoidal. Radiotherapy was
given to 290 (88%) of the patients, and 283 of them were treated
within 6 months postoperatively. Pituitary insufficiency was diagnosed
in 67.1% at 6 months after surgery, in 76.8% after 1 yr, in 82.3%
after 2 yr, and in 89.0% after 5 yr.
|
Information on malignant tumors for the period 1958–1992, coded according to the International Classification of Diseases, seventh revision, was obtained through linkage with the Swedish Cancer Registry. Expected cancer incidence for the same period was calculated by means of the EPILUND cohort program using cause-, county-, calendar year-, sex-, and 5-yr age group-specific rates. These rates were calculated from incidence rates for malignant tumors obtained from the Swedish Cancer Registry and population counts obtained from Statistics Sweden. The Swedish Data Inspectorate approved the study.
Endocrine evaluation
Pituitary insufficiency was defined as deficiency in at least one axis of the anterior lobe. Endocrine evaluation varied over the years, and pituitary function was assessed according to tests used at each time period. These were described in detail previously (22). Before the late 1970s, when serum PRL assays became available, it is possible that some PRL-producing tumors were classified as nonfunctional tumors. At the start of the follow-up period, 63 patients had complete pituitary failure, defined as insufficiency of gonads, thyroid, and adrenal glands. The remaining 265 patients were classified as having partial pituitary insufficiency. GH secretion was evaluated postoperatively in 62 patients using the insulin tolerance test or the glucagon test (1 mg glucagon given sc). Of these subjects only 1 had a normal response (36 mIU/L); all of the others had a clearly deficient GH response (<10 mIU/L).
Hormone replacement
The details on cortisone and thyroid and gonadal hormone replacement were reported previously (22). At the start of follow-up, 43 of the 169 hypogonadal men had no testosterone treatment, which constituted 21% of all males in the cohort. Seventeen of the 33 hypogonadal females who were 45 yr of age or younger had no estrogen replacement. These 17 females constituted 14% of the entire female cohort.
Statistical methods
The 95% confidence intervals (CIs) for cause-specific standardized cancer incidence ratios (SIRs) were calculated by treating the observed number as a Poisson variable, or as a normal variable if the observed value was greater than 15. The term significant indicates that the 95% CI for SIR does not include 1.00.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
).
Astrocytoma was diagnosed at autopsy 24 yr after operation for
pituitary adenoma, and meningeoma was diagnosed 19 yr after operation;
both patients had been postoperatively irradiated. Two endocrine tumors
(1 adrenal and 1 parathyroid tumor, part of multiple endocrine neoplasm
syndromes) had occurred, compared with the 0.50 expected (SIR, 4.00;
95% CI, 0.48–14.5). The tumors were diagnosed 2 and 9 yr after the
initial operation. Only 2 prostate cancers had been diagnosed, compared
with the 5.82 expected (SIR, 0.34; 95% CI, 0.04–1.24). The overall
tumor incidence was somewhat lower than expected (SIR, 0.85), but the
95% CI did not exclude unity (0.59–1.21). When excluding brain and
endocrine tumor diagnoses, the SIR for remaining tumor diagnoses
decreased to 0.77, but it did still not differ significantly from unity
(0.52–1.13).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The patients have been subjected to regular health control programs, which may have resulted in somewhat earlier tumor diagnoses in some cases compared with the general population. The Swedish Cancer Registry covers, however, almost all incident malignant tumors and some specific types of other tumors. Thus, it is unlikely that the health controls more than very marginally affected the comparison with the general population.
In the present study the initial cohort comprised a consecutive series of patients operated for pituitary tumors in one department, recruited from a specific catchment area, and selection bias was therefore avoided. Patients with acromegaly or Cushing’s disease were, however, subsequently excluded due to the well known increases in mortality in cardiovascular (24, 25) and malignant diseases (24) for these patients. The cohort was thus very homogenous, and the cause of pituitary insufficiency was due to the pituitary tumor and its treatment.
Brada et al. (23) observed no increased incidence for tumors located outside the cranial cavity in a cohort of patients irradiated for pituitary tumors. However, patients with acromegaly were not excluded and no information on hypo-pituitarism was provided, which makes comparison with the present study difficult. Popovic et al. (26) claimed that patients with nonfunctioning pituitary adenomas, among whom only 50% had hypopituitarism, have an increased incidence of malignant tumors. This finding should also be interpreted with caution, as the tumor diagnoses among the patients were obtained by clinical follow-ups, whereas the incidence rates used for comparisons were obtained from the Serbian Tumor Registry. In this registry a high proportion of diagnoses was based on death certificates alone, with a low level of histological verifications, which indicates underascertainment of malignant tumor cases in the general population (27). In contrast to the Serbian Registry, there is an almost total ascertainment of incident malignant tumor cases in the Swedish Cancer Registry, from which the incidence rates used in our study were obtained. Another advantage with the present study is that the general population in the catchment area from which the patients were recruited constituted the reference population. This is a more relevant reference population than the national one, as there are regional differences in the incidence of neoplasms. The cohort was geographically very stable, as only 2 of the 177 deceased patients died outside the catchment area.
Besides the tumor incidence studies discussed above (23, 26), 3 tumor mortality studies on patients with hypopituitarism have been performed (20, 21, 22). In the study of Bülow et al. (22), mainly based on the present cohort, no increased mortality in malignant diseases was observed (21 observed vs. 22 expected cases). The relative risk was not differential with respect to sex (unpublished data). In contrast, in the 2 other studies death in all malignant diseases was lower than expected in males; however, this was based on only 3 observed vs. 10.1 expected cases and 3 vs. 5.7 expected cases, respectively (20, 21). In the study by Bates et al. (21) an increase in tumor mortality in women was shown, again based on low numbers, 9 observed vs. 3 expected cases. The major drawbacks with tumor mortality studies compared with tumor incidence studies are that only lethal tumors are included and that the diagnostic accuracy is lower than in incidence studies based on data from high quality tumor incidence registers.
It can be hypothesized that long-standing preoperative or postoperative unsubstituted estrogen or androgen deficiency might lower the risk for breast and prostate cancer (28, 29, 30). However, the present study gave no indication of a decreased risk for breast cancer. On the other hand, a tendency for decreased risk was observed for prostate cancer, but the low numbers hamper a firm evaluation. Prostate cancer is androgen dependent (29, 30), and androgen depletion very rapidly induces apoptosis in the prostate (31). The male GH-deficient patients had during 11% of their person-years under risk not been substituted with androgens, which might marginally have contributed to the present finding. Another cause of androgen deficiency might have had a larger impact on the risk for prostate cancer. As a negative association has been shown between serum levels of IGF-I and sex hormone-binding globulin (32), it is possible that the lower serum IGF-I levels may have induced a decrease in the bioavailable testosterone in the normo- and hypogonadal men.
GH deficiency is an early finding in pituitary failure after surgery or radiotherapy for pituitary adenomas (33, 34, 35). However, as all patients in the present cohort had been operated upon, and 88% had received radiotherapy, it is a very high probability that they already had GH deficiency at the time of documented pituitary insufficiency or at least shortly thereafter (35).
In previous epidemiological studies the association between serum IGF-I and cancer risk was found in middle-aged subjects (5, 6, 7, 8, 12, 13). The changes in serum IGF-I levels throughout life mimic those in GH secretion and decline gradually with advancing age (36). About 50% of GH-deficient patients between 40 and 60 yr of age have serum IGF-I levels below the normal range, and the remaining subjects have IGF-I levels within the low normal range (37). In the present study, with a very high probability of GH deficiency and with a median age at operation of 53 yr, serum IGF-I levels can thus be presumed to have been within the low to subnormal range. Chan et al. (6) found a lowered relative risk for prostate cancer among men with low serum IGF-I levels. Our observation of a slight decrease in prostatic cancer risk gives some indirect evidence supporting the finding of Chan et al. (6). However, our study design does not allow a direct evaluation of the relative importance of IGF-I and androgen levels on prostate cancer risk.
In conclusion, in a cohort of 328 patients with hypopituitarism after treatment of a pituitary tumor, a tendency for a decreased overall tumor risk was noted, especially when excluding brain and endocrine tumors. Considering specific diagnoses, a tendency of decreased risk was observed for prostate cancer, but the low numbers hamper a firm evaluation. Despite this limitations the present results may serve as a baseline for follow-up studies of tumor risk among adult patients with pituitary insufficiency supplemented with GH.
|
Footnotes |
|---|
Received February 1, 2000.
Revised June 23, 2000.
Revised October 4, 2000.
Accepted November 18, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S Dubois, S Guyetant, P Menei, P Rodien, F Illouz, B Vielle, and V Rohmer Relevance of Ki-67 and prognostic factors for recurrence/progression of gonadotropic adenomas after first surgery Eur. J. Endocrinol., August 1, 2007; 157(2): 141 - 147. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Minniti, D. Traish, S. Ashley, A. Gonsalves, and M. Brada Risk of Second Brain Tumor after Conservative Surgery and Radiotherapy for Pituitary Adenoma: Update after an Additional 10 Years J. Clin. Endocrinol. Metab., February 1, 2005; 90(2): 800 - 804. [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 |
