| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Metabolic Diseases Branch (W.H., M.C.S., W.F.S., L.S.W., S.K.A., C.M., L.J.-N., S.J.M.) and Digestive Diseases Branch (F.G., R.T.J.), National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland 20892; and Department of Community Medicine and Statistics (G.R.H.), West Virginia University School of Medicine, Morgantown, West Virginia 26506
Address all correspondence and requests for reprints to: Stephen J. Marx, M.D., Building 10, Room 9C-101, National Institutes of Health, 9000 Rockville Pike, Bethesda, Maryland 20892. E-mail: stephenm{at}intra.niddk.nih.gov.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Results in cases from two kindreds were that 93% showed parathyroid adenoma, 40% pituitary tumor (always prolactinoma), and 27% enteropancreatic endocrine tumor. The latter included 10% insulinoma, 7% nonfunctioning islet tumor, but only 10% gastrinoma. Compared with prior large series, this lower prevalence of gastrinoma (10% vs. 42%, P < 0.01) and higher prevalence of prolactinoma (40% vs. 22%, P < 0.01) define this variant. Many possible biases of retrospective analyses were excluded as possible explanations. Previously sequenced DNA showed no characteristic MEN1 mutation in these two kindreds and in a third, reported previously; the lack of any shared MEN1 mutation also excluded common ancestry for MEN1 among the three kindreds. The causes for differences between this variant and typical MEN1 are unknown.
In conclusion, this variant shows more frequent prolactinoma and less frequent gastrinoma than typical MEN1; the variant is reproducible among kindreds. MEN1 carriers in such families should have periodic monitoring adjusted for the expected penetrance of tumors.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
A variant of MEN1 that is reproducible among families seems possible. In fact, an MEN1 variant with unusually high prevalence of prolactinoma was reported in four large and seemingly independent kindreds, originating around the Burin Peninsula of Newfoundland (9, 10, 11, 12, 13, 14, 15). Hence, this has been referred to as the prolactinoma, Burin, or Newfoundland variant of MEN1. Affected members of all four Newfoundland families with MEN1 were recently shown to share not only the same MEN1 germline mutation but also the same 11q13 haplotype (i.e. the DNA marker pattern about the MEN1 gene’s chromosomal locus) (12, 14). This implies a nonrandom pattern, specifically their sharing the same segment of the one copy of chromosome 11, carrying their MEN1 mutation; therefore, their repeating MEN1 phenotype is associated with each family’s descent from a common ancestor with the same MEN1 mutation. Therefore, these four families have more recently been treated together as one huge MEN1 kindred (12, 14, 15). This one and three other less large kindreds are the best candidates as repeats of the same clinical variant of MEN1 (16, 17, 18, 19). Still, the validity of this variant is not universally accepted. It was not mentioned in several relevant articles (8, 19, 20), and its interpretation was questioned in another (21).
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Studies were approved by the Institutional Review Board of the National Institute of Diabetes and Digestive and Kidney Diseases. Each subject gave informed consent.
Ascertainment of carriers
The MEN1 carrier is defined as a subject virtually certain to have an MEN1 mutation, whether silent or expressed. In general, this can be deduced by testing DNA, tumor expressions, or carrier pattern in a family tree (for a silent but obligatory transmitter). Because DNA testing was not done for carrier ascertainment in this study, some asymptomatic carriers (particularly younger than age 15 yr) could not be recognized. Screening for the MEN1 carrier state in asymptomatic first-degree relatives of patients with MEN1 was generally by serum calcium (sometimes ionized), PTH, and prolactin (22). Carrier ascertainment did not include imaging studies, such as computed tomography (CT) and/or magnetic resonance imaging (MRI).
Periodic surveillance of known carriers
The highly suspected or proven carriers of the MEN1 trait were evaluated periodically (every 6–24 months) and in more detail than for carrier ascertainment. Obligate carriers without expression of MEN1 and family members without sufficient data available were not labeled herein as affected. Carriers were monitored as inpatients at the NIH Clinical Center, if possible, or another center. For those patients and relatives with probable MEN1 and without evaluation at NIH, information was obtained from medical records and/or telephone interviews. Per standard guidelines, symptoms, biochemical analyses, and imaging were used for their periodic evaluations (22).
Criteria for diagnosing a tumor
Hyperparathyroidism (HPT) was defined as hypercalcemia and elevated PTH; alternately, the finding of parathyroid tumor(s) at surgery established the trait. HPT in an MEN1 carrier was considered diagnostic of parathyroid adenoma(s) (23). Prolactinoma or other hormone-secreting pituitary tumor was defined as the repeated documentation in serum of an elevated prolactin or other pituitary hormone that could not be attributed to another cause. Zollinger Ellison syndrome (ZES) was defined, with the subject not receiving acid-blocking drugs, as an elevated fasting gastrin and an elevated basal acid output (above 15 mEq/h in nonoperated patients or above 5 mEq/h if acid-reduction surgery had been performed). ZES in an MEN1 carrier was considered diagnostic of gastrinoma(s). Insulinoma was defined as fasting hypoglycemia with remission after removal of an islet cell tumor. Nonfunctional islet tumor was an islet mass without an excess state for an islet hormone. Foregut carcinoid was defined by histology.
Prior reports of typical MEN1
All reports meeting each of the following criteria were included: publication after 1985, a large kindred or multiple kindreds of MEN1, usable information about the prevalence of all three of the major MEN1-related tumors, and unlikely case overlap with another report, including those with the three prolactinoma families.
Statistics
Averages were reported as mean ± 1 SEM. The one-sample t test compared the prevalence of MEN1 or an MEN1-related tumor with prior reports. Prevalence from a subtraction (i.e. pituitary tumor other than prolactinoma) was calculated only from publications documenting both parameters. P < 0.05 was considered significant. Penetrance by age was cumulative or a Kaplan-Meier life-table (24).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Kindred A.
Kindred A (Fig. 1A
) was initially reported in 1973 with hypercalcemia in nine members among three generations (16). Five of the nine had had parathyroid surgery before the initial report. No case had gastrinoma or prolactinoma recognized at this time. Three of the nine showed other endocrine tumors. This included one with a pituitary macroadenoma causing panhypopituitarism; in retrospect, that tumor was a prolactinoma and was thus reclassified. A 13-yr-old boy had Cushing’s syndrome. Subtotal adrenalectomy showed bilateral adrenocortical hyperplasia and led to long-term cure (16). NIH evaluation at age 27 yr showed stable and long-term normal adrenocortical function and no evidence of a pituitary tumor. After the initial report, eight additional cases with MEN1, including seven with HPT were newly identified. Five of these eight also showed prolactinoma; average age at prospective diagnosis of prolactinoma in this group was 30 yr (range 14–41 yr). Gastrinoma was diagnosed at age 40 yr in one member. During the 30-yr follow-up, three MEN1 subjects died at 63–70 yr from causes not attributable to MEN1. One died from soft tissue sarcoma metastatic from the shoulder; the possible relation to leiomyoma inherent in MEN1 was not studied (25).
|
) was first reported in 1983, with six MEN1 cases in one generation (17). Each presented with prolactinoma symptoms and HPT at about the same time, with MEN1 onset from 14 to 26 yr. No subject had gastrinoma, attributable in part to their young ages at time of the report. Among these, two developed gastrinoma during follow-up; one later developed also a large meningioma attributable to MEN1 and/or past radiation to the pituitary region (26). Six new cases of MEN1 with HPT were identified during follow-up. One newly identified but archival case of MEN1 with only insulinoma was identified retrospectively and lacked other details. Two affected subjects in kindred B died during follow-up at age 64 and 70 yr of causes not attributable to MEN1. MEN1 carriers and their ages
Among the 162 known family members in the two kindreds, 30 subjects expressing MEN1 endocrine tumors were ultimately identified; these included 17 in kindred A and 13 in kindred B (Fig. 1). There were 17 males and 13 females. The average age at diagnosis of MEN1 was 31 yr (range 12–53 yr), and average age at most recent follow-up or death was 51 ± 3 yr. Eighteen affected members have been admitted to the NIH Clinical Center, including 16 admitted two or more times.
Tumor prevalence overall and its age dependency
The overall prevalences of all identified endocrine tumors are listed (Table 1). The cumulative prevalences of MEN1 among ultimately known MEN1 carriers by age 20, 30, 40, and 50 yr were 20, 40, 70, and 90%, respectively. At the same ages, the cumulative prevalences of a selected tumor type in MEN1 carriers were the following: for parathyroid adenoma(s) 10, 37, 63, and 87%; for prolactinoma 10, 27, 33, and 40%. None of the enteropancreatic endocrine tumors or foregut carcinoids were identified before age 30 yr.
|
). The midpoint for age of onset of HPT or of prolactinoma were similar (age 25–30 yr) (P not significant) and one to two decades earlier than those for gastrinoma or bronchial carcinoid (age 40–50 yr) (P < 0.01). The penetrance of prolactinoma and gastrinoma plateaued at 48 and 16%, respectively, after age 50 and 40 yr, respectively.
|
HPT was ultimately identified in 28 of 30 affected subjects, including 17 males and 11 females. Six (22%) of the 28 patients presented with symptoms of kidney stone and/or renal colic. The average age at onset of HPT symptoms (i.e. nephrolithiasis) was 29 ± 4 yr (range 19–41 yr, n = 6 subjects); the average age at diagnosis of HPT was 32 ± 2 yr (range 12–53 yr) (Fig. 2A). Twenty-one had one or more parathyroid operations; multiple parathyroid adenomas were identified in each case.
Features of prolactinoma
Among the 30 subjects with MEN1, prolactinoma was identified in 12 (40%), seven males and five females. Ten presented with symptoms of oligo/amenorrhea, galactorrhea, impotence, headache, or visual impairment, alone or in combination. Panhypopituitarism was identified before treatment in two. Two tumors were identified only by periodic surveillance with CT and/or MRI. Nine had a pituitary mass, imaged by CT and/or MRI. Microadenoma was ultimately diagnosed in six and macroadenoma in six. The average age at onset of prolactinoma symptoms was 22 yr (range 12–39 yr, n = 10 subjects); the average age at diagnosis of prolactinoma was 28 yr (range 14–48 yr). All 12 had high serum prolactin before treatment; no high serum GH or IGF-I was identified.
The distributions of prolactinoma differed between the two families after partitioning each family into two cohorts, related to inclusion in the initial report (Table 2). This may have been due in part to prolactinoma evolving in a sequence that differed between the two families. It was also due to dramatic improvements in the availability and quality of prolactin RIA between 1973 and 1983, the dates that separated these two initial reports (27).
|
Gastrinoma. Gastrinoma had not been reported initially in either kindred, and it developed subsequently in surprisingly few members of each. Gastrinoma(s) was ultimately identified in only three of 30 affected subjects; these were two males and one female. Their age at onset of symptoms of ZES was 21 and 35 yr and unknown in one subject. The age at diagnosis of ZES was 38–40 yr. The main manifestations of ZES were abdominal pain in two subjects, diarrhea in two, and heartburn in one. Each of the three subjects at the time of diagnosis of ZES had high serum gastrin level of 360, 519, and 611 pg/ml (normal < 100 pg/ml) and elevated basal acid output of 21, 30, and 37 mEq/h. Abdominal imaging by CT in subjects with ZES identified a pancreatic mass in two of three subjects. A pancreatic mass removed at surgery was positive for gastrin in each of the two operated subjects; one of the two masses was also positive for insulin. ZES was not improved by gastrinoma surgery in either subject.
Islet tumor other than gastrinoma. Symptomatic insulinoma was identified in three subjects including two males and one female; the age at diagnosis was 38 and 42 yr and unknown. In addition, a 1- to 2-cm nonfunctioning pancreatic islet tumor was identified surgically in two males at 46 and 58 yr.
Foregut carcinoid tumor. Foregut carcinoid was identified in two males and three females. The ages at the diagnosis were 39, 41, 57, and 61 yr and unknown. The carcinoids were three bronchial, two gastrointestinal carcinoids, and none thymic. All five subjects were asymptomatic. Bronchial carcinoid presented in three females as a 1- to 2-cm pulmonary nodule, identified at screening by CT and/or MRI. Two of these three are in remission post surgery; the third developed metastases in bone. One male had duodenal carcinoid identified at endoscopy; the other had gastric carcinoid (details not available).
Two cohorts based on case report in the initial publication
The initial reports of the two families contained 15 pooled MEN1 cases. These 15 cases and their tumors, recognized in the initial report, were defined as cohort 1 (Table 2). Cohort 1 contained all subjects with MEN1 tumors that could be biased from (a) inclusion in the initial report of each family and (b) exclusion of any new tumors at later follow-up in those initially reported cases. Only two gastrinomas were later identified in these subjects (Table 2). This emphasizes the age dependency of gastrinoma and the fact that delayed diagnosis of gastrinoma even after 20–30 yr of follow-up did not contradict the overall impression of rare gastrinoma in these two families.
Another 15 subjects not overlapping with cohort 1 were newly identified as MEN1 carriers after the two initial publications and then followed up prospectively for as long as that family’s 20- to 30-yr follow-up period allowed. These were defined as cohort 2, designed to be a group of subjects who had been less or even not at all influenced by selection around the initial report (Table 2). Despite differing selection biases, cohorts 1 and 2 showed remarkably similar tumor patterns (Table 2 and Fig. 2B).
Comparisons with prior reports of typical MEN1
The cumulative prevalences of parathyroid adenoma(s) in these two MEN1 variant families are comparable with those previously reported in typical MEN1 (1, 2, 3, 4, 5, 6, 7, 8) (Table 3). In contrast, the prevalences of prolactinoma in these two families are higher than in prior reports of typical MEN1. Surprisingly, no pituitary tumor other than prolactinoma was seen in either family (0 vs. 14% in prior, P < 0.05) (Table 3).
|
). In fact, the difference in prevalence of enteropancreatic tumor after excluding gastrinoma (17 vs. 25%) was not significant.
The low prevalence of gastrinoma in these two families had been noted at their initial reports (16, 17). This reflected, only in part, a bias of young age in these archival reports. The low prevalence (1 of 15 MEN1 cases) of gastrinoma even in cohort 2, composed of MEN1 cases not in the initial report (Table 2 and Fig. 2), is not attributable to subsequent identification of MEN1 cases and in a cohort that was unusually young; for example, 10 of 15 subjects in cohort 2 were older than 40 yr at most recent follow-up (Fig. 2B). The prevalence of gastrinoma was atypically low, particularly considering the high age of the MEN1 carriers at the most recent follow-up. In fact, the difference from typical MEN1 was more striking for gastrinoma than for prolactinoma.
Prolactinoma variant of MEN1 in prior reports
Two other previously reported large families should be considered as possibly showing the prolactinoma variant of MEN1 (see above) (15, 19). First, Green et al. (15) reported a preliminary but extensive study in follow-up of the extended kindred with the prolactinoma variant of MEN1 from Newfoundland. Among the 78 patients expressing MEN1, 94% had parathyroid adenoma(s), 34% pituitary tumor (32% prolactinoma and only 2% other), 10% pancreatic endocrine tumor (4% gastrinoma, 6% other), and 9% foregut carcinoid (5% lung, 3% thymus, 1% other). Analyses from this extended family first defined and then elaborated the tumor pattern in the prolactinoma variant (9, 10, 11, 12, 13, 14, 15). And the pattern is remarkably close to that in either kindred A or kindred B.
Second, Waterlot et al. (19) reported a large kindred with 18 MEN1 cases. The prevalences were parathyroid adenoma(s) 89%, pituitary tumor 39% (prolactinoma 33%, other 6%), pancreatic tumor 28% (presumably all endocrine but not stated). Gastrinoma frequency was not reported. In fact, many affected patients in this family were young enough to make expression of gastrinoma unlikely, including in the setting of typical MEN1; for example, 15 of the 18 affected patients were younger than age 40 yr. Diagnosis of the prolactinoma variant is therefore possible but not established in this large family.
Two other families have been labeled as prolactinoma variant of MEN1 (28, 29). These families are too small for a secure diagnosis of the prolactinoma variant of MEN1.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Bias from small family size. One small and unusual MEN1 family could show a pattern of tumors due to chance combinations among more than 20 tumor types, inherent in MEN1 (23). This may explain why several reports have called attention to an unusual tumor pattern (a so-called variant of MEN1) of uncertain generality. Unusual patterns in one small family include the following: several ACTH-producing pituitary tumors in one MEN1 kindred (30), several insulinomas in an MEN1 kindred (4, 31), or several foregut carcinoids in one MEN1 kindred (32).
Even large families and large collections of subjects and families have not previously helped to define a variant of MEN1 that is reproducible among families (1, 2, 3, 4, 5, 6, 7, 8, 23). This is in part because no repeating variant has differed strikingly from typical MEN1. However, in this report, the high number of affected subjects in three MEN1-variant families have helped establish the reproducibility of a subtle variant.
Bias from young age.
Because gastrinoma, most other islet tumors, or foregut carcinoid typically begins in MEN1 about 15 yr later than parathyroid tumor, the average age of affected members in almost any MEN1 kindred at the time of data analysis must influence that kindred’s tumor pattern at that time (5, 22, 23). This age dependency of tumor pattern is also evident in the two families herein (Fig. 2). It was addressed here by examining overall age dependency of tumor penetrance and age dependency in two well-defined cohorts within the two families.
At the same time, lack of helpful data about age in prior reports compromised comparisons to kindreds A and B. Average patient ages differed among reports mainly because of higher fractions of younger, asymptomatic subjects in some series (5).
Two important findings about age stand out: first, high average age (51 yr) at follow-up herein did not account for the high frequency of prolactinoma in the MEN1 variant because most prolactinomas were diagnosed at younger ages (Fig. 2). Second, despite high final age among carriers and among unaffected members, gastrinoma was less frequent than in each large report of typical MEN1 (Fig. 2).
Bias from unequal surveillance. Members of two families may have been ascertained and monitored differently from prior reports. However, the methods and criteria for clearcut diagnosis of prolactinoma do not differ greatly among centers. And gastrinoma was infrequent in this series despite even more intense surveillance for gastrinoma than usual in other academic centers (33). On the other hand, the low prevalence of gastric carcinoid herein among subjects with the prolactinoma variant may reflect the lower frequency of gastrinoma and thus the lower use of gastroscopy (33).
Bias from kindred selection. The two kindreds reported in detail herein were drawn from about 200 MEN1 kindreds evaluated at the NIH during this period (Marx, S. J., unpublished observations). The two had been selected long ago for unusual features and continued under scrutiny. Such selection bias was minimized by evaluation of two logical patient cohorts; one cohort excluded all tumors known in the initial report.
Bias from prior publications. Any prior publication can contain aspects of the biases cited above. The highest frequency of prolactinoma in two cited reports was only minimally below that in the prolactinoma variant kindreds herein. In one report, most of the prolactinomas were highly concentrated in two kindreds (3). This or other publications may have included some families with the prolactinoma variant of MEN1 undiagnosed (7). In the second report, the method for measuring total calcium in serum was insensitive for diagnosing HPT, explaining, at least in part, a high fraction with prolactinomas (7). Our conservative citation of both of these reports and the conservative weighting by publication rather than by number of cases still showed prolactinoma more prevalent in the MEN1 variant than in typical MEN1.
One reported series of typical MEN1 with the lowest frequency of gastrinoma also had a low average age of 33 yr at last follow-up; this resulted from inclusion of many asymptomatic subjects ascertained during family screening. That gastrinoma rate of 25% was still higher than in the much older population herein with the MEN1 variant.
Hallmarks of a variant of MEN1, reproducible among families
The prolactinoma variant shows more frequent prolactinoma and less frequent gastrinoma than in typical MEN1. This pattern is clear in the two large U.S. families reported herein and the extremely large and previously reported family from Newfoundland (15). A possibly lower prevalence of nonprolactinoma pituitary tumors and possibly higher prevalence of foregut carcinoid in this variant of MEN1 will need further evaluation. In particular, the seemingly high prevalence of foregut carcinoid in this MEN1 variant may be the result of more detailed surveillance (34, 35).
DNA testing of the MEN1 locus in the germline
MEN1 germline mutations have been found in over 70% of tested MEN1 families (23, 36, 37, 38). No clear genotype-phenotype relation has yet been identified involving the MEN1 gene (23, 36, 37, 38, 39, 40). Preliminary data suggest that, among a minority of all families showing the phenotype of isolated hyperparathyroidism, a genotype is represented by MEN1 missense mutations clustered around codons 255–305 (41, 42). (A missense mutation predicts change of one to two amino acids among the 610 in menin without disrupting the open reading frame, so protein length and linear composition are changed minimally.)
The open reading frame and intron-exon boundaries of the MEN1 gene had been sequenced and reported previously in the germline from each of the two families herein and from a third family, reported previously with the prolactinoma variant of MEN1 (14, 18, 36). Germline MEN1 mutation had been identified in two: kindred B has Y312X (normal tyrosine at codon 312, replaced by a stop or nonsense codon); the Newfoundland kindred has R460X (normal arginine at codon 460, replaced by a stop or nonsense codon). No MEN1 mutation was identified in kindred A; however, the MEN1 trait in kindred A showed significant linkage to the MEN1 locus at chromosome 11q13 (log of the odds score 3.25), indicating almost certainly an unidentified MEN1 mutation in that family (36). These data established that there is no shared MEN1 mutation among any of these three kindreds. For the same reason, these data exclude a founder effect (i.e. common ancestor with the same MEN1 mutation) as an explanation for reproducibility of this MEN1 variant.
The MEN1 mutations identified thus far in two families with the prolactinoma variant of MEN1 predict truncation or absence of the encoded menin protein; this is also the pattern in 80% of identified germline mutations in typical MEN1 (23). Interestingly, these same two mutations had also been reported previously in eight other MEN1 index subjects or kindreds (36, 37, 38, 43, 44, 45, 46, 47). Unfortunately, those reports do not permit evaluation for the prolactinoma variant in any. Family A almost certainly has an unidentified MEN1 mutation. This could be a small MEN1 mutation outside the MEN1 gene regions tested or a large MEN1 deletion invisible to PCR methods (48, 49). Although all three kindreds with the prolactinoma variant results almost certainly have MEN1 mutation, an explanation of the difference from typical MEN1 is unknown.
The two families herein have had MEN1 carrier ascertainment through monitoring of symptoms and biochemical indices. Because one family has an identified germline MEN1 mutation and the other has the MEN1 trait tightly linked to chromosome 11q13, DNA testing for MEN1 mutation or an 11q13 haplotype would be more sensitive alternatives for MEN1 carrier assignment. This approach almost certainly would not change the major conclusions about a variant of MEN1 that is reproducible among families.
Management implications
The overall frequency of this variant among all families and sporadic cases of MEN1 is not known, but it is probably less than 10%. Another reason for infrequent diagnosis of the prolactinoma variant is that recognition currently requires a substantial number of affected subjects (to document frequent prolactinoma) and/or many affected subjects over age 40 yr (to document the rarity of gastrinoma). When the diagnosis of the prolactinoma variant of MEN1 is established, appropriate modifications in management should be initiated. While any of the many tumors of MEN1 seem possible in this variant and while clinical vigilance for each must continue, the tumor surveillance in a family with this variant should be adjusted for the expected tumor penetrances (22).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received August 29, 2003.
Accepted April 16, 2004.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  O Vierimaa, T M L Ebeling, S Kytola, R Bloigu, E Eloranta, J Salmi, E Korpi-Hyovalti, L Niskanen, A Orvola, E Elovaara, et al. Multiple endocrine neoplasia type 1 in Northern Finland; clinical features and genotype phenotype correlation Eur. J. Endocrinol., September 1, 2007; 157(3): 285 - 294. [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 |
