Endocrine and Immunogenetic Testing in Individuals with Type 1 Diabetes and 21-Hydroxylase Autoantibodies: Addison’s Disease in a High-Risk
http://www.100md.com
《临床内分泌与代谢杂志》
Abstract
Individuals with type 1 diabetes mellitus (T1D) at risk for Addison’s disease (AD) can be identified with RIAs for autoantibodies to the adrenal antigen 21-hydroxylase (21-OHAA). Screening individuals with T1D for 21OH-AA shows a relatively high prevalence of positive autoantibodies (1.4%, 38 of 2696 subjects). After detection of 21-OHAA, individuals were evaluated with endocrine testing, including baseline cortisol, ACTH, and plasma renin activity and low (1 μg) and high (250 μg) dose cortrosyn stimulation. Typing for DR and DQ alleles and for the major histocompatibility complex class I-related chain A (MICA) gene polymorphisms was performed. Six individuals were diagnosed with AD; five were identified on initial endocrine evaluation. Follow-up over 2.9 yr yielded one additional diagnosis of AD. Endocrine testing showed a correlation between baseline ACTH and peak cortisol (r = –0.61; P < 0.0001), baseline and peak cortisol (r = 0.70; P < 0.0001), and stimulated cortisol after low- and high-dose testing (r = 0.92; P < 0.0001). DR3-DQ2/DR4-DQ8 with DRB1* 0404 was associated with expression of 21-OHAA. At 2 yr, individuals homozygous for MICA5.1 had AD-free survival of 60% compared with 100% AD-free survival in those who were not homozygous for MICA5.1. Homozygosity for MICA5.1 may increase progression to overt AD among 21-OHAA-positive individuals.
Introduction
TYPE 1 DIABETES (T1D) is associated with Addison’s disease (AD). Autoimmunity associated with AD [21-hydroxylase autoantibodies (21-OHAA)] may be identified before the development of clinical symptoms (1). Factors influencing the time course of development of adrenal autoimmunity and AD in the setting of T1D are not well characterized. We postulate that factors known to be associated with adrenal autoimmunity, such as MICA (major histocompatibility complex class I-related chain A) polymorphisms (2, 3), will influence the development of clinical disease in those who are already 21-hydroxylase autoantibody positive.
The association of AD and autoantibodies to adrenal cortical cells has been recognized for several decades (4). The cytochrome P450 enzyme 21-hydroxylase has been identified as a major antigen for these autoantibodies (5, 6), and production of the autoantibodies is highly sensitive and specific for AD (7, 8, 9). RIAs have been developed that reliably detect 21-OHAA (10). Studies using these assays to screen populations of T1D for autoimmunity have found a prevalence of positive autoantibodies in approximately 1–2% of individuals with T1D (11, 12, 13).
Genetic susceptibility for AD has been linked to the major histocompatibility complex on chromosome 6 (MHC). Class II histocompatibility leukocyte antigen (HLA) alleles have been linked to AD in cohorts with and without T1D. Two groups have previously reported that the genotype DRB1*0404/DQ8 with DRB1*0301/DQ2 was at an increased frequency in individuals with AD alone and in those with AD with T1D (12, 14).
A second locus within the MHC, the MHC class I-related MICA, has been linked to AD (2, 3). Polymorphisms of exon 5 are based on the number of GCT triplet repeats within an exon (alleles A4, A5, A6, and A9) and a single nucleotide insertion (allele 5.1) (15). This additional nucleotide (allele 5.1) results in a premature stop codon and truncated MICA protein lacking its membrane binding domain. Homozygosity for MICA5.1 was associated with an 18-fold increase in odds for AD, and the excess risk occurred even with identical class II DR and DQ genotypes (3). Given this strong association, we postulated that expression of the polymorphism MICA5.1 will influence the development of AD in a population that is 21-OHAA positive.
The diagnosis of AD can be made on clinical grounds, by typical symptoms of adrenal insufficiency such as fatigue, weight loss, salt craving (16), and, in T1D, unexplained hypoglycemia (17) or decrease in insulin dose (18). Routine screening of individuals with T1D has identified individuals who are antibody positive before the development of overt symptoms of AD. The natural history of the progression of the endocrine defect in individuals with 21-OHAA has been evaluated in the past and proposed to follow a typical pattern, progressing through stages starting with increased production of renin and ultimately leading to low basal cortisol levels (1). The progression of the endocrine defect in individuals with adrenal autoimmunity and T1D has not been studied in detail.
The Barbara Davis Center for Childhood Diabetes has evaluated approximately 3000 individuals with T1D and their relatives for 21-OHAA. Those individuals with positive autoantibodies have been followed prospectively with cortrosyn stimulation testing for the development of endocrine derangements typical of AD. In this report we describe the HLA DRB1 and MICA typing and endocrine testing for these individuals.
Subjects and Methods
Subjects
Individuals with T1D and their relatives were recruited to participate in studies assessing the coexistence of organ-specific autoimmunity after providing informed consent and in a protocol that was approved by the combined institutional review board at University of Colorado Health Sciences Center. We have previously reported the results of 21OH-AA screening in individuals with T1D (11, 12). Since this time, a total of 2696 patients with T1D seen in the diabetes clinic at the Barbara Davis Center have been screened at least once with 21-OHAA, eight (0.3%) individuals had preexisting AD at the time of screening and are not included in this analysis. Six hundred fourteen individuals with T1D were screened more than one time, with an average of 2.47 autoantibody tests (range, 2–8). Thirty-eight (1.4%) were found to be 21-OHAA positive on one or more 21OH-AA screens. An additional 453 relatives were screened, and nine (1.9%) were positive for 21-OHAA.
Endocrine testing was performed in the General Clinical Research Center. Our combined institutional review board approved the protocol. Individuals were tested after providing informed consent. At least some form of endocrine testing was performed in 38 21-OHAA-positive individuals. Twenty subjects were seen once, 15 had two visits, two had three visits, and one individual was evaluated four times, for a total of 60 visits. Each individual with endocrine testing had an average of 1.6 visits.
Of the 38 individuals, 20 (53%) were female. Thirty-two (84%) had T1D; the remaining six were nondiabetic relatives of individuals with T1D. In those with T1D, the median age of diagnosis of diabetes was 7.2 yr (range, 0.5–28.8 yr), median age of initial screening for 21-OHAA was 14.3 yr (range, 3.1–47.3), and the median age at first detection of positive 21-OHAA was 15.7 yr (range, 6–47.3). In the group of relatives, the median age at first 21-OHAA screening was 21.8 yr (range, 6.25–78), and the median age of first positive 21-OHAA was 23.2 yr (range, 6.25–78).
21-OHAA RIA
21-OHAA were measured by a method previously described (10, 19), and the 21-hydroxylase cDNA clone was provided by Drs. Falorni (University of Perugia, Perugia, Italy) and Lernmark (University of Washington, Seattle, WA). In brief, 21-hydroxylase cDNA was transcribed and translated in vitro with a commercially available kit (Promega Corp., Madison, WI). [35S]Methionine was incorporated, and 20,000 cpm of the labeled product were incubated with 2 μl serum overnight. Protein A-Sepharose (Amersham Biosciences, Piscataway, NJ) was used to capture autoantibody-bound radioactivity in 96-well filtration plates (Millipore Corp., Bedford, MA). Scintillation fluid was added directly to the 96-well plate for counting with a Top-Counter (Packard, Downers Grove, IL). Autoantibodies are reported as an index relative to a positive control sample. Positivity for 21-OHAA was defined as exceeding the highest index of 241 normal controls (an index >0.149).
HLA and MICA typing
HLA-DQB and DRB genotyping were performed using PCR probe-based genotyping kits (Dynal Biotech, Lafayette Hill, PA). DNA amplification was performed in a 9600 Thermal Cycler (PerkinElmer, Norwalk, CT). The amplified DNA was added to strips containing immobilized sequence-specific oligonucleotides for DQB1 or DRB1. The HLA-linked microsatellite marker MICA was determined with a fluorescent-based method as reported previously (2, 20).
Endocrine testing
The individuals were admitted to the General Clinical Research Center and placed in a supine position, and an iv catheter was inserted 30 min before baseline laboratory work. At baseline, blood was drawn for autoantibodies, plasma renin activity (PRA), ACTH, cortisol, and HLA genotyping. At time zero, 1 μg cortrosyn (low-dose cortrosyn stimulation) was given iv; 30 min later, serum was obtained for cortisol measurement. Then a dose of 250 μg cortrosyn (high-dose cortrosyn stimulation) was given iv, and cortisol levels were determined 30 and 60 min later. The highest cortisol level after stimulation was designated the peak level.
Blood for ACTH and cortisol was stored as plasma at –20 C, and analysis was performed at the University of Colorado Health Sciences Center Endocrinology Laboratory. PRA was also performed in this laboratory. The normal range for ACTH was less than 5–60 pg/ml (1.1–13.2 pmol/liter); that for PRA was 0.2–3.0 ng/ml·h. For the purpose of calculating the sensitivity and specificity of the different components of the endocrine testing, a peak cortisol level after 250 μg cortrosyn of less than 18 μg/dl (496.6 nmol/liter) or a fasting cortisol level of 3 μg/dl (82.8 nmol/liter) or less was used as the gold standard for the diagnosis of AD (21).
Statistical analyses
Statistical analyses were performed using SAS 8.0 (SAS Institute, Cary, NC). For normally distributed variables, descriptive statistics are reported as the mean and SD. Otherwise, the median and range are reported. Comparisons for continuous variables were made with t test for normally distributed variables, and nonparametric tests were used for nonnormally distributed variables. The association between the different components of the endocrine testing performed was examined using correlation analysis. The same analyses were performed only in those individuals with AD to determine whether the correlation was attributable to those with abnormal results. Survival analysis was performed for DRB1 and MICA typing from the time of the first positive 21-OHAA test to the development of AD or the most recent visit. Results were considered statistically significant at P < 0.05.
Results
21-OHAA
Forty-seven individuals were identified with positive autoantibodies. Eight had autoantibody testing performed on one visit only. The remaining 39 had multiple autoantibody tests performed; three were positive on their initial screen and then became negative, two with T1D and one relative. The levels of 21-OHAA were lower in the group that became negative (median, 0.17) than in those that remained positive (median, 1.08; P = 0.01). One individual was positive, then negative, and is now positive again. The remaining 35 were tested multiple times, and once positive, remained positive.
Thirty-five of the individuals with T1D were positive on their first screening for 21-OHAA, nine as new-onset diabetics. Three diabetic individuals (of 2696) were negative on their first test and then positive on subsequent screening 1.5, 4 and 6.25 yr after initial testing. Seven relatives were positive on their first screening. Two relatives (of 453) were initially negative and then positive on subsequent screens, 2.2 and 5.0 yr later. Thus, a total of five individuals converted to positivity after an initial negative autoantibody test.
Endocrine testing
Table 1 shows the results of the 60 endocrine tests performed in 36 individuals with positive 21-OHAA. Fifty-two tests were performed in 30 individuals who were not diagnosed with AD; eight tests were performed in six individuals diagnosed with AD. Baseline ACTH and stimulated cortisol levels differed significantly between the two groups. There was no difference in age at endocrine testing, coexistence of T1D, baseline PRA, or baseline cortisol.
Individuals with abnormal endocrine testing [defined by a peak cortisol <18 μg/dl (496.6 nmol/liter) or a baseline cortisol 3 μg/dl (82.8 nmol/liter)] are depicted in Table 2. Five individuals were identified as having abnormal endocrine testing at their first evaluation. One individual was diagnosed with Addison’s disease by baseline samples alone, with a baseline cortisol of 0.99 μg/dl (27.31 nmol/liter) and an ACTH of 1031 pg/ml (226.8 pmol/liter). The remaining six individuals had cortrosyn stimulation testing performed and showed elevated ACTH at baseline and decreased baseline and peak cortisol values. Individual 212144 was originally evaluated at 6.75 yr of age and had an ACTH of 78 pg/ml (17.16 pmol/liter), with a peak cortisol of 17 μg/dl (469.03 nmol/liter); repeat evaluation showed an increase in the ACTH to 195 pg/ml (42.9 pmol/liter) and a decrease in peak cortisol to 11 μg/dl (303.49 nmol/liter). Of note, all three individuals with AD in whom PRA was determined had an elevated PRA. Individual 201602 has been evaluated with cortrosyn stimulation testing four times over a 4-yr period. At the time of his abnormal testing, he was using inhaled corticosteroids and had low ACTH and peak cortisol levels, consistent with exogenous suppression from the inhaled steroids.
Sensitivity and specificity
Individual 201602, who was using inhaled corticosteroids at the time of his abnormal testing, was excluded from the analysis. A baseline ACTH level elevated above the normal range (60 pg/ml; 13.2 pmol/liter) was a sensitive and specific indicator of AD. One hundred percent (six of six) of the subjects with AD had elevated ACTH levels, and 98% (45 of 46) of those without AD had normal ACTH levels. The low-dose cortrosyn test was also sensitive and specific for AD, i.e. 100% (three of three) of those with AD had a cortisol level less than 18 μg/dl (496.6 nmol/liter) after stimulation, and 78% (25 of 32) of those without AD had a cortisol peak of 18 μg/dl (496.6 nmol/liter) or greater. For comparison and by definition, after high-dose cortrosyn testing, five of five individuals with AD had a peak cortisol level less than 18 μg/dl (496.6 nmol/liter), and 38 of 38 without AD had a peak cortisol level more than 18 μg/dl (496.6 nmol/liter).
HLA and MICA typing
Table 3 shows the DRB1 and DQ subtypes and MICA typing for each individual in our study. Similar to reports of patients with AD, there was a higher proportion of the DRB1*0404 subtype in 21-OHAA-positive patients with the DR3-DQ2/DR4-DQ8 genotype (29.7%; 13 of 47) compared with T1D controls (11%; 62 of 578) (2) and nondiabetic controls (0.7%; 109 of 15,547; P < 0.0001) (2). In particular, there was an excess of DRB1* 0404 and a decrease in DRB1*0401 compared with patients with T1D. However, DRB1*0404 had no effect on progression to AD in this 21-OHAA-positive group.
None (of nine) of the 21-OHAA-positive nondiabetic relatives had the high diabetes risk HLA genotype DR3-DQ2, DR4-DQ8, compared with 21 of 38 individuals with type 1 diabetes (P = 0.003).
DNA was available for MICA typing in 44 individuals. All six individuals with AD were MICA5.1 homozygous compared with 11 of 38 (29%) of the individuals positive for 21-OH AA without AD (P < 0.0005). Homozygosity for MICA5.1 was associated with a shorter AD-free survival, with approximately 40% of MICA5.1 homozygotes developing AD by 2 yr compared with 0% without homozygosity for this allele (P = 0.013; Fig. 1).
Correlation
Peak cortisol was designated the highest level after cortrosyn stimulation. The logarithmically transformed ACTH level and baseline cortisol level were significantly associated with peak cortisol (r = –0.61; P < 0.0001 and r = 0.71; P < 0.0001, respectively). Baseline and peak cortisol levels were positively correlated (r = 0.7; P < 0.0001). In addition, the cortisol level after the low-dose cortrosyn stimulation test was correlated with the peak cortisol level (r = 0.92; P < 0.0001; Fig. 2). No relationship was noted between renin and peak cortisol. Evaluation of these relationships in individuals with abnormal results revealed a stronger association between log ACTH and peak cortisol (r = –0.89; P = 0.006).
Change in endocrine parameters over time
Follow-up of individuals with positive 21-OHAA allowed evaluation of the change in endocrine parameters over time. For those who had more than one endocrine test (n = 18), the average follow-up time was 2.9 yr. Figure 3 shows the peak cortisol levels over time from the first endocrine evaluation. As can be seen, the individual who developed AD upon subsequent testing experienced a decrease in peak cortisol over the 4-yr follow-up period.
Discussion
The ability to screen for adrenal autoimmunity with 21-OHAA has identified individuals who are at risk for the development of AD before clinical presentation. We report that 21-OHAA was present in 1.4% of patients with T1D and in 1.9% of their nondiabetic relatives (12).
The natural history of adrenal autoimmunity and factors contributing to progression to overt AD in patients with T1D continue to be defined. In the past, groups have identified characteristics such as higher 21-OHAA index (22); HLA haplotypes, specifically DRB1*0404-DQ8 and DRB1*0301-DQ2 (1, 12); and younger age (23) as risk factors for the development of clinical AD in the group of autoantibody-positive individuals.
In this report we confirm and extend previous studies indicating that HLA genotypes are associated with the expression of adrenal autoimmunity. Individuals who are 21-OHAA positive are much more likely to have the genotype DR3-DQ2/DR4-DQ8 DRB1*0404 than either normal subjects or diabetic controls. However, this high-risk HLA genotype was not associated with an increased progression to AD in the population that was already 21OH-AA positive.
Previous studies have shown that MICA5.1 is related to the presence of AD and adrenal autoimmunity (2, 3, 24). These reports have shown this association in populations that are 21-OHAA positive and in those with AD. The effect of this polymorphism on the development of clinical disease in a population that is autoantibody positive has not been performed. Our report suggests that homozygosity for MICA5.1 may influence the progression to AD in the group of individuals who are 21-OHAA positive. One hundred percent (six of six) of the individuals diagnosed with AD in our group were homozygous for MICA5.1 compared with approximately 30% in the group that was not diagnosed with AD. The association is especially interesting in light of the function of the MICA gene. It encodes a protein that is a ligand for the NKG2D T cell receptor. MICA can activate this receptor on natural killer cells. This interaction may be important for thymic T cell maturation (25). MICA5.1 is a variant that disrupts the coding sequence (premature stop codon) and therefore may affect the ability to activate its receptor and disrupt T cell maturation in the thymus.
Given the rarity of subjects with AD and adrenal autoimmunity, small numbers of autoantibody-positive subjects have been followed for the development of AD. Betterle et al. (1) followed 17 subjects with adrenal autoantibodies and no initial evidence for clinical AD for a mean of 3.2 yr. Seven (41%) of these subjects developed AD over the period of the study. Based on these results, Betterle et al. (1) proposed a sequence of endocrine abnormalities that occurs during the natural history of AD with the initial abnormality and elevated PRA. De Bellis et al. (26) evaluated 20 individuals with positive autoantibodies. Eleven of the subjects remained persistently positive for adrenal autoantibodies, and two (18%) developed clinical AD; the initial abnormality identified was an elevated PRA level. Laureti et al. (27) reported endocrine testing in 11 individuals with 21-OHAA. Five (45%) of these individuals showed a pathological response to cortrosyn stimulation (low and high doses). Four of five of these individuals had elevated baseline ACTH levels.
One study suggested a yearly incidence of AD in those with adrenal autoimmunity of approximately 20% (1) in a group that had an average follow-up of 3.2 yr. In our cohort, the majority of individuals (five of six) who were diagnosed with AD were identified at their initial endocrine evaluation. Therefore, we may have missed the preclinical period marked by elevation of PRA alone in these individuals. Also supporting the fact that the individuals diagnosed with AD in this study may have been mildly symptomatic is the lower insulin dose and lower hemoglobin A1c levels observed before the diagnosis of AD, which increased to pre-AD levels after institution of therapy. We do have a group of individuals with isolated elevated PRA who will require additional follow-up. All of the individuals in our study showed elevated ACTH and PRA at the time of diagnosis of AD.
AD and T1D are known to be part of autoimmune polyendocrine syndromes 1 and 2. None of our patients showed hypoparathyroidism, chronic mucocutaneous candidiasis, or other autoimmune diseases associated with autoimmune polyendocrine syndrome 1. Therefore, these individuals probably represent autoimmune polyendocrine syndrome 2. Our patient population confirmed the known association with thyroid disease, in that 38% of those with 21OH-AA had autoimmune thyroid disease.
Endocrine evaluation of adrenal autoantibody-positive individuals has led to reports that their endocrine parameters may fluctuate over time, such that some who appeared to have early evidence of adrenal insufficiency marked by the elevation of PRA and/or abnormal results in cortrosyn stimulation testing have resolved their endocrine abnormalities and not developed clinically significant AD while simultaneously becoming negative for adrenal cortical autoantibodies (26). In our study we observed resolution of autoantibodies in a small subset of individuals. These individuals were not tested with endocrine tests, so we cannot comment on their adrenal function at the time of the positive autoantibody measurements. Of note, these individuals tended to have lower 21-OHAA levels than the group that remained persistently positive. This may be a reflection of borderline or transient autoimmunity that ultimately resolves and never leads to clinically significant disease, or it may reflect the false-positive rate associated with the cutoff for the assay. The cutoff for positivity in the present study was defined as exceeding the highest index for 241 normal controls. Using this criterion, we have over 90% confidence that the false positive rate is less than 1%.
Endocrine evaluation of adrenal function includes baseline hormone measures of ACTH, PRA, and cortisol and cortrosyn stimulation testing (16, 28). Stimulation testing can be performed using either 1 or 250 μg cortrosyn; the traditional method of measurement is with the higher dose. However, this higher dose of cortrosyn may mask mild adrenal insufficiency by exposing the adrenal glands to overwhelming amounts of ACTH. Therefore, some have argued that the low-dose cortrosyn test is more reliable for the diagnosis of primary adrenal insufficiency (27, 29). Our study suggests that the low-dose cortrosyn test has good sensitivity (100%) for AD, but with a specificity of approximately 80% there is the potential for diagnosing someone with adrenal insufficiency who would have passed standard cortrosyn testing. The clinical significance of this disparity is unclear.
The exact frequency of screening for 21-OHAA in individuals with T1D remains an area of active research and controversy. Our current practice is to screen for 21-OHAA at diagnosis of T1D and then every 2 yr if the first test is negative. If the autoantibodies are positive (which in our laboratory indicates that the same sample has been measured and found to be positive in a confirmatory assay), we believe it is important to confirm positivity in an independent sample. Currently, on a research basis for 21-OHAA-positive individuals, we measure morning baseline ACTH, cortisol, and PRA levels (supine position) and perform high (1 μg) and low (250 μg) dose cortrosyn stimulation testing annually. On a nonresearch basis, a normal annual basal ACTH measurement in the absence of symptoms may be sufficient screening if the patients receive education concerning the symptoms of AD and Addisonian crisis. If an individual becomes symptomatic at any time, we evaluate with baseline cortisol, ACTH, and/or stimulation testing depending upon the situation. We generally use a cutoff of stimulated cortisol less than 18 mg/dl as positive for adrenal insufficiency in the presence of elevated ACTH. The autoantibody tests are repeated at the time of endocrine testing. If an individual converts to negative autoantibodies, we recommend repeating the autoantibody tests annually for several years and performing stimulation testing if the patient becomes symptomatic and/or again positive for the autoantibodies.
ACTH levels are highly predictive of AD and correlate very well with peak cortisol levels in our study. The majority of this correlation is due to the elevated levels in those subjects who have evidence for adrenal insufficiency. Our study has demonstrated that baseline ACTH levels greater than 60 pg/ml (13.2 pmol/liter) are 100% sensitive and 98% specific for the diagnosis of adrenal insufficiency. Thus, we propose that screening of T1D individuals with 21-OHAA can be performed with baseline cortisol and ACTH measurements, with stimulation testing performed in those with abnormal results. Larger series and longer follow-up by multiple groups should allow more robust estimates of ideal endocrine evaluation and more robust life-table analysis of progression to AD, particularly for those who are 21-OHAA positive and MICA5.1 homozygous.
Acknowledgments
We acknowledge the General Clinical Research Centers for their contribution in performing the endocrine testing and the endocrine laboratory at the University of Colorado Health Sciences Center for performing the endocrine assays.
Footnotes
This work was supported by NIH Grants DK-32083, DK-32493, A146374, and DK-57516; General Clinical Research Centers Grants RR-00051 and RR-00069; and Juvenile Diabetes Research Foundation/Larson Wilkins Pediatric Endocrine Fellowship 13-2002-444 (to J.M.B.).
First Published Online October 13, 2004
Abbreviations: AD, Addison’s disease; HLA, class II histocompatibility leukocyte antigen; MHC, major histocompatibility complex on chromosome 6; MICA, major histocompatibility complex class I-related chain A; 21-OHAA, 21-hydroxylase autoantibodies; PRA, plasma renin activity; T1D, type 1 diabetes mellitus.
Received May 14, 2004.
Accepted September 28, 2004.
References
Betterle C, Scalici C, Presotto F, Pedini B, Moro L, Rigon F, Mantero F 1988 The natural history of adrenal function in autoimmune patients with adrenal autoantibodies. J Endocrinol 117:467–475
Park YS, Sanjeevi CB, Robles D, Yu L, Rewers M, Gottlieb PA, Fain P, Eisenbarth GS 2002 Additional association of intra-MHC genes, MICA and D6S273, with Addison’s disease. Tissue Antigens 60:155–163
Gambelunghe G, Falorni A, Ghaderi M, Laureti S, Tortoioli C, Santeusanio F, Brunetti P, Sanjeevi CB 1999 Microsatellite polymorphism of the mhc class I chain-related (mic-a and mic-b) genes marks the risk for autoimmune Addison’s disease. J Clin Endocrinol Metab 84:3701–3707
Anderson J, Goudie RB, Gray KG, Timbury GC 1957 Autoantibodies in Addison’s disease. Lancet 272:1123–1124
Winqvist O, Karlsson FA, Kampe O 1992 21-Hydroxylase, a major autoantigen in idiopathic Addison’s disease. Lancet 339:1559–1562
Baumann-Antczak A, Wedlock N, Bednarek J, Kiso Y, Krishnan H, Fowler S, Smith BR, Furmaniak J 1992 Autoimmune Addison’s disease and 21-hydroxylase. Lancet 340:429–430
Falorni A, Laureti S, Nikoshkov A, Picchio ML, Hallengren B, Vandewalle CL, Gorus FK, Tortoioli C, Luthman H, Brunetti P, Santeusanio F 1997 21-Hydroxylase autoantibodies in adult patients with endocrine autoimmune diseases are highly specific for Addison’s disease. Clin Exp Immunol 107:341–346
Betterle C, Volpato M, Pedini B, Chen S, Smith BR, Furmaniak J 1999 Adrenal-cortex autoantibodies and steroid-producing cells autoantibodies in patients with Addison’s disease: comparison of immunofluorescence and immunoprecipitation assays. J Clin Endocrinol Metab 84:618–622
Soderbergh A, Winqvist O, Norheim I, Rorsman F, Husebye ES, Dolva O, Karlsson FA, Kampe O 1996 Adrenal autoantibodies and organ-specific autoimmunity in patients with Addison’s disease. Clin Endocrinol (Oxf) 45:453–460
Falorni A, Nikoshkov A, Laureti S, Grenb?ck E, Hulting A-L, Casucci G, Santeusanio F, Brunetti P, Luthman H, Lernmark ? 1995 High diagnostic accuracy for idiopathic Addison’s disease with a sensitive radiobinding assay for autoantibodies against recombinant human 21-hydroxylase. J Clin Endocrinol Metab 80:2752–2755
Brewer KW, Parziale VS, Eisenbarth GS 1997 Screening patients with insulin-dependent diabetes mellitus for adrenal insufficiency. N Engl J Med 337:202
Yu L, Brewer KW, Gates S, Wu A, Wang T, Babu S, Gottlieb PA, Freed BM, Noble J, Erlich HA, Eisenbarth GS 1999 DRB1*04 and DQ alleles: expression of 21-hydroxylase autoantibodies and risk of progression to Addison’s disease. J Clin Endocrinol Metab 84:328–335
Peterson P, Salmi H, Hyoty H, Miettinen A, Ilonen J, Reijonen H, Knip M, ?kerblom HK, Krohn K 1997 Steroid 21-hydroxylase autoantibodies in insulin-dependent diabetes mellitus. Childhood Diabetes in Finland (DiMe) Study Group. Clin Immunol Immunopathol 82:37–42
Myhre AG, Undlien DE, Lovas K, Uhlving S, Nedrebo BG, Fougner KJ, Trovik T, Sorheim JI, Husebye ES 2002 Autoimmune adrenocortical failure in Norway: Autoantibodies and HLA class II associations related to clinical features. J Clin Endocrinol Metab 87:618–623
Ota M, Katsuyama Y, Mizuki N, Ando H, Furihata K, Ono S, Pivetti-Pezzi P, Tabbara KF, Palimeris GD, Nikbin B, Davatchi F, Chams H, Geng Z, Bahram S, Inoko H 1997 Trinucleotide repeat polymorphism within exon 5 of the MICA gene (MHC class I chain-related gene A): allele frequency data in the nine population groups Japanese, Northern Han, Hui, Uygur, Kazakhstan, Iranian, Saudi Arabian, Greek and Italian. Tissue Antigens 49:448–454
Oelkers W 1996 Adrenal insufficiency. N Engl J Med 335:1206–1212
McAulay V, Frier BM 2000 Addison’s disease in type 1 diabetes presenting with recurrent hypoglycaemia. Postgrad Med J 76:230–232
Armstrong L, Bell PM 1996 Addison’s disease presenting as reduced insulin requirement in insulin dependent diabetes. BMJ 312:1601–1602
Colls J, Betterle C, Volpato M, Prentice L, Smith BR, Furmaniak J 1995 Immunoprecipitation assay for autoantibodies to steroid 21-hydroxylase in autoimmune adrenal diseases. Clin Chem 41:375–380
Park YS, Lee H, Eisenbarth GS, Sanjeevi CB 2001 2001 MICA polymorphism is associated with type 1 diabetes in the Korean population. Diabetes Care 24:33–38
Grinspoon SK, Biller BM 1994 Clinical review 62: laboratory assessment of adrenal insufficiency. J Clin Endocrinol Metab 79:923–931
Laureti S, De Bellis A, Muccitelli VI, Calcinaro F, Bizzarro A, Rossi R, Bellastella A, Santeusanio F, Falorni A 1998 Levels of adrenocortical autoantibodies correlate with the degree of adrenal dysfunction in subjects with preclinical Addison’s disease. J Clin Endocrinol Metab 83:3507–3511
Betterle C, Volpato M, Rees Smith B, Furmaniak J, Chen S, Zanchetta R, Greggio NA, Pedini B, Boscaro M, Presotto F 1997 II. Adrenal cortex and steroid 21-hydroxylase autoantibodies in children with organ-specific autoimmune diseases: markers of high progression to clinical Addison’s disease. J Clin Endocrinol Metab 82:939–942
Bilbao JR, Martin-Pagola A, Perez de Nanclares G, Calvo B, Vitoria JC, Vazquez F, Castano L 2003 HLA-DRB1 and MICA in autoimmunity: common associated alleles in autoimmune disorders. Ann NY Acad Sci 1005:314–318
Hue S, Monteiro RC, Berrih-Aknin S, Caillat-Zucman S 2003 Potential role of NKG2D/MHC class I-related chain A interaction in intrathymic maturation of single-positive CD8 T cells. J Immunol 171:1909–1917
De Bellis A, Bizzarro A, Rossi R, Amoresano Paglionico V, Criscuolo T, Lombardi G, Bellastella A 1993 Remission of subclinical adrenocortical failure in subjects with adrenal autoantibodies. J Clin Endocrinol Metab 76:1002–1007
Laureti S, Arvat E, Candeloro P, Di Vito L, Ghigo E, Santeusanio F, Falorni A 2000 Low dose (1 μg) ACTH test in the evaluation of adrenal dysfunction in pre-clinical Addison’s disease. Clin Endocrinol (Oxf) 53:107–115
Dorin RI, Qualls CR, Crapo LM 2003 Diagnosis of adrenal insufficiency. Ann Intern Med 139:194–204
Laureti S, Falorni A 2002 Reliability of the ACTH low dose test in the evaluation of adrenal insufficiency. J Postgrad Med 48:251–252(Jennifer M. Barker, Akane)
Individuals with type 1 diabetes mellitus (T1D) at risk for Addison’s disease (AD) can be identified with RIAs for autoantibodies to the adrenal antigen 21-hydroxylase (21-OHAA). Screening individuals with T1D for 21OH-AA shows a relatively high prevalence of positive autoantibodies (1.4%, 38 of 2696 subjects). After detection of 21-OHAA, individuals were evaluated with endocrine testing, including baseline cortisol, ACTH, and plasma renin activity and low (1 μg) and high (250 μg) dose cortrosyn stimulation. Typing for DR and DQ alleles and for the major histocompatibility complex class I-related chain A (MICA) gene polymorphisms was performed. Six individuals were diagnosed with AD; five were identified on initial endocrine evaluation. Follow-up over 2.9 yr yielded one additional diagnosis of AD. Endocrine testing showed a correlation between baseline ACTH and peak cortisol (r = –0.61; P < 0.0001), baseline and peak cortisol (r = 0.70; P < 0.0001), and stimulated cortisol after low- and high-dose testing (r = 0.92; P < 0.0001). DR3-DQ2/DR4-DQ8 with DRB1* 0404 was associated with expression of 21-OHAA. At 2 yr, individuals homozygous for MICA5.1 had AD-free survival of 60% compared with 100% AD-free survival in those who were not homozygous for MICA5.1. Homozygosity for MICA5.1 may increase progression to overt AD among 21-OHAA-positive individuals.
Introduction
TYPE 1 DIABETES (T1D) is associated with Addison’s disease (AD). Autoimmunity associated with AD [21-hydroxylase autoantibodies (21-OHAA)] may be identified before the development of clinical symptoms (1). Factors influencing the time course of development of adrenal autoimmunity and AD in the setting of T1D are not well characterized. We postulate that factors known to be associated with adrenal autoimmunity, such as MICA (major histocompatibility complex class I-related chain A) polymorphisms (2, 3), will influence the development of clinical disease in those who are already 21-hydroxylase autoantibody positive.
The association of AD and autoantibodies to adrenal cortical cells has been recognized for several decades (4). The cytochrome P450 enzyme 21-hydroxylase has been identified as a major antigen for these autoantibodies (5, 6), and production of the autoantibodies is highly sensitive and specific for AD (7, 8, 9). RIAs have been developed that reliably detect 21-OHAA (10). Studies using these assays to screen populations of T1D for autoimmunity have found a prevalence of positive autoantibodies in approximately 1–2% of individuals with T1D (11, 12, 13).
Genetic susceptibility for AD has been linked to the major histocompatibility complex on chromosome 6 (MHC). Class II histocompatibility leukocyte antigen (HLA) alleles have been linked to AD in cohorts with and without T1D. Two groups have previously reported that the genotype DRB1*0404/DQ8 with DRB1*0301/DQ2 was at an increased frequency in individuals with AD alone and in those with AD with T1D (12, 14).
A second locus within the MHC, the MHC class I-related MICA, has been linked to AD (2, 3). Polymorphisms of exon 5 are based on the number of GCT triplet repeats within an exon (alleles A4, A5, A6, and A9) and a single nucleotide insertion (allele 5.1) (15). This additional nucleotide (allele 5.1) results in a premature stop codon and truncated MICA protein lacking its membrane binding domain. Homozygosity for MICA5.1 was associated with an 18-fold increase in odds for AD, and the excess risk occurred even with identical class II DR and DQ genotypes (3). Given this strong association, we postulated that expression of the polymorphism MICA5.1 will influence the development of AD in a population that is 21-OHAA positive.
The diagnosis of AD can be made on clinical grounds, by typical symptoms of adrenal insufficiency such as fatigue, weight loss, salt craving (16), and, in T1D, unexplained hypoglycemia (17) or decrease in insulin dose (18). Routine screening of individuals with T1D has identified individuals who are antibody positive before the development of overt symptoms of AD. The natural history of the progression of the endocrine defect in individuals with 21-OHAA has been evaluated in the past and proposed to follow a typical pattern, progressing through stages starting with increased production of renin and ultimately leading to low basal cortisol levels (1). The progression of the endocrine defect in individuals with adrenal autoimmunity and T1D has not been studied in detail.
The Barbara Davis Center for Childhood Diabetes has evaluated approximately 3000 individuals with T1D and their relatives for 21-OHAA. Those individuals with positive autoantibodies have been followed prospectively with cortrosyn stimulation testing for the development of endocrine derangements typical of AD. In this report we describe the HLA DRB1 and MICA typing and endocrine testing for these individuals.
Subjects and Methods
Subjects
Individuals with T1D and their relatives were recruited to participate in studies assessing the coexistence of organ-specific autoimmunity after providing informed consent and in a protocol that was approved by the combined institutional review board at University of Colorado Health Sciences Center. We have previously reported the results of 21OH-AA screening in individuals with T1D (11, 12). Since this time, a total of 2696 patients with T1D seen in the diabetes clinic at the Barbara Davis Center have been screened at least once with 21-OHAA, eight (0.3%) individuals had preexisting AD at the time of screening and are not included in this analysis. Six hundred fourteen individuals with T1D were screened more than one time, with an average of 2.47 autoantibody tests (range, 2–8). Thirty-eight (1.4%) were found to be 21-OHAA positive on one or more 21OH-AA screens. An additional 453 relatives were screened, and nine (1.9%) were positive for 21-OHAA.
Endocrine testing was performed in the General Clinical Research Center. Our combined institutional review board approved the protocol. Individuals were tested after providing informed consent. At least some form of endocrine testing was performed in 38 21-OHAA-positive individuals. Twenty subjects were seen once, 15 had two visits, two had three visits, and one individual was evaluated four times, for a total of 60 visits. Each individual with endocrine testing had an average of 1.6 visits.
Of the 38 individuals, 20 (53%) were female. Thirty-two (84%) had T1D; the remaining six were nondiabetic relatives of individuals with T1D. In those with T1D, the median age of diagnosis of diabetes was 7.2 yr (range, 0.5–28.8 yr), median age of initial screening for 21-OHAA was 14.3 yr (range, 3.1–47.3), and the median age at first detection of positive 21-OHAA was 15.7 yr (range, 6–47.3). In the group of relatives, the median age at first 21-OHAA screening was 21.8 yr (range, 6.25–78), and the median age of first positive 21-OHAA was 23.2 yr (range, 6.25–78).
21-OHAA RIA
21-OHAA were measured by a method previously described (10, 19), and the 21-hydroxylase cDNA clone was provided by Drs. Falorni (University of Perugia, Perugia, Italy) and Lernmark (University of Washington, Seattle, WA). In brief, 21-hydroxylase cDNA was transcribed and translated in vitro with a commercially available kit (Promega Corp., Madison, WI). [35S]Methionine was incorporated, and 20,000 cpm of the labeled product were incubated with 2 μl serum overnight. Protein A-Sepharose (Amersham Biosciences, Piscataway, NJ) was used to capture autoantibody-bound radioactivity in 96-well filtration plates (Millipore Corp., Bedford, MA). Scintillation fluid was added directly to the 96-well plate for counting with a Top-Counter (Packard, Downers Grove, IL). Autoantibodies are reported as an index relative to a positive control sample. Positivity for 21-OHAA was defined as exceeding the highest index of 241 normal controls (an index >0.149).
HLA and MICA typing
HLA-DQB and DRB genotyping were performed using PCR probe-based genotyping kits (Dynal Biotech, Lafayette Hill, PA). DNA amplification was performed in a 9600 Thermal Cycler (PerkinElmer, Norwalk, CT). The amplified DNA was added to strips containing immobilized sequence-specific oligonucleotides for DQB1 or DRB1. The HLA-linked microsatellite marker MICA was determined with a fluorescent-based method as reported previously (2, 20).
Endocrine testing
The individuals were admitted to the General Clinical Research Center and placed in a supine position, and an iv catheter was inserted 30 min before baseline laboratory work. At baseline, blood was drawn for autoantibodies, plasma renin activity (PRA), ACTH, cortisol, and HLA genotyping. At time zero, 1 μg cortrosyn (low-dose cortrosyn stimulation) was given iv; 30 min later, serum was obtained for cortisol measurement. Then a dose of 250 μg cortrosyn (high-dose cortrosyn stimulation) was given iv, and cortisol levels were determined 30 and 60 min later. The highest cortisol level after stimulation was designated the peak level.
Blood for ACTH and cortisol was stored as plasma at –20 C, and analysis was performed at the University of Colorado Health Sciences Center Endocrinology Laboratory. PRA was also performed in this laboratory. The normal range for ACTH was less than 5–60 pg/ml (1.1–13.2 pmol/liter); that for PRA was 0.2–3.0 ng/ml·h. For the purpose of calculating the sensitivity and specificity of the different components of the endocrine testing, a peak cortisol level after 250 μg cortrosyn of less than 18 μg/dl (496.6 nmol/liter) or a fasting cortisol level of 3 μg/dl (82.8 nmol/liter) or less was used as the gold standard for the diagnosis of AD (21).
Statistical analyses
Statistical analyses were performed using SAS 8.0 (SAS Institute, Cary, NC). For normally distributed variables, descriptive statistics are reported as the mean and SD. Otherwise, the median and range are reported. Comparisons for continuous variables were made with t test for normally distributed variables, and nonparametric tests were used for nonnormally distributed variables. The association between the different components of the endocrine testing performed was examined using correlation analysis. The same analyses were performed only in those individuals with AD to determine whether the correlation was attributable to those with abnormal results. Survival analysis was performed for DRB1 and MICA typing from the time of the first positive 21-OHAA test to the development of AD or the most recent visit. Results were considered statistically significant at P < 0.05.
Results
21-OHAA
Forty-seven individuals were identified with positive autoantibodies. Eight had autoantibody testing performed on one visit only. The remaining 39 had multiple autoantibody tests performed; three were positive on their initial screen and then became negative, two with T1D and one relative. The levels of 21-OHAA were lower in the group that became negative (median, 0.17) than in those that remained positive (median, 1.08; P = 0.01). One individual was positive, then negative, and is now positive again. The remaining 35 were tested multiple times, and once positive, remained positive.
Thirty-five of the individuals with T1D were positive on their first screening for 21-OHAA, nine as new-onset diabetics. Three diabetic individuals (of 2696) were negative on their first test and then positive on subsequent screening 1.5, 4 and 6.25 yr after initial testing. Seven relatives were positive on their first screening. Two relatives (of 453) were initially negative and then positive on subsequent screens, 2.2 and 5.0 yr later. Thus, a total of five individuals converted to positivity after an initial negative autoantibody test.
Endocrine testing
Table 1 shows the results of the 60 endocrine tests performed in 36 individuals with positive 21-OHAA. Fifty-two tests were performed in 30 individuals who were not diagnosed with AD; eight tests were performed in six individuals diagnosed with AD. Baseline ACTH and stimulated cortisol levels differed significantly between the two groups. There was no difference in age at endocrine testing, coexistence of T1D, baseline PRA, or baseline cortisol.
Individuals with abnormal endocrine testing [defined by a peak cortisol <18 μg/dl (496.6 nmol/liter) or a baseline cortisol 3 μg/dl (82.8 nmol/liter)] are depicted in Table 2. Five individuals were identified as having abnormal endocrine testing at their first evaluation. One individual was diagnosed with Addison’s disease by baseline samples alone, with a baseline cortisol of 0.99 μg/dl (27.31 nmol/liter) and an ACTH of 1031 pg/ml (226.8 pmol/liter). The remaining six individuals had cortrosyn stimulation testing performed and showed elevated ACTH at baseline and decreased baseline and peak cortisol values. Individual 212144 was originally evaluated at 6.75 yr of age and had an ACTH of 78 pg/ml (17.16 pmol/liter), with a peak cortisol of 17 μg/dl (469.03 nmol/liter); repeat evaluation showed an increase in the ACTH to 195 pg/ml (42.9 pmol/liter) and a decrease in peak cortisol to 11 μg/dl (303.49 nmol/liter). Of note, all three individuals with AD in whom PRA was determined had an elevated PRA. Individual 201602 has been evaluated with cortrosyn stimulation testing four times over a 4-yr period. At the time of his abnormal testing, he was using inhaled corticosteroids and had low ACTH and peak cortisol levels, consistent with exogenous suppression from the inhaled steroids.
Sensitivity and specificity
Individual 201602, who was using inhaled corticosteroids at the time of his abnormal testing, was excluded from the analysis. A baseline ACTH level elevated above the normal range (60 pg/ml; 13.2 pmol/liter) was a sensitive and specific indicator of AD. One hundred percent (six of six) of the subjects with AD had elevated ACTH levels, and 98% (45 of 46) of those without AD had normal ACTH levels. The low-dose cortrosyn test was also sensitive and specific for AD, i.e. 100% (three of three) of those with AD had a cortisol level less than 18 μg/dl (496.6 nmol/liter) after stimulation, and 78% (25 of 32) of those without AD had a cortisol peak of 18 μg/dl (496.6 nmol/liter) or greater. For comparison and by definition, after high-dose cortrosyn testing, five of five individuals with AD had a peak cortisol level less than 18 μg/dl (496.6 nmol/liter), and 38 of 38 without AD had a peak cortisol level more than 18 μg/dl (496.6 nmol/liter).
HLA and MICA typing
Table 3 shows the DRB1 and DQ subtypes and MICA typing for each individual in our study. Similar to reports of patients with AD, there was a higher proportion of the DRB1*0404 subtype in 21-OHAA-positive patients with the DR3-DQ2/DR4-DQ8 genotype (29.7%; 13 of 47) compared with T1D controls (11%; 62 of 578) (2) and nondiabetic controls (0.7%; 109 of 15,547; P < 0.0001) (2). In particular, there was an excess of DRB1* 0404 and a decrease in DRB1*0401 compared with patients with T1D. However, DRB1*0404 had no effect on progression to AD in this 21-OHAA-positive group.
None (of nine) of the 21-OHAA-positive nondiabetic relatives had the high diabetes risk HLA genotype DR3-DQ2, DR4-DQ8, compared with 21 of 38 individuals with type 1 diabetes (P = 0.003).
DNA was available for MICA typing in 44 individuals. All six individuals with AD were MICA5.1 homozygous compared with 11 of 38 (29%) of the individuals positive for 21-OH AA without AD (P < 0.0005). Homozygosity for MICA5.1 was associated with a shorter AD-free survival, with approximately 40% of MICA5.1 homozygotes developing AD by 2 yr compared with 0% without homozygosity for this allele (P = 0.013; Fig. 1).
Correlation
Peak cortisol was designated the highest level after cortrosyn stimulation. The logarithmically transformed ACTH level and baseline cortisol level were significantly associated with peak cortisol (r = –0.61; P < 0.0001 and r = 0.71; P < 0.0001, respectively). Baseline and peak cortisol levels were positively correlated (r = 0.7; P < 0.0001). In addition, the cortisol level after the low-dose cortrosyn stimulation test was correlated with the peak cortisol level (r = 0.92; P < 0.0001; Fig. 2). No relationship was noted between renin and peak cortisol. Evaluation of these relationships in individuals with abnormal results revealed a stronger association between log ACTH and peak cortisol (r = –0.89; P = 0.006).
Change in endocrine parameters over time
Follow-up of individuals with positive 21-OHAA allowed evaluation of the change in endocrine parameters over time. For those who had more than one endocrine test (n = 18), the average follow-up time was 2.9 yr. Figure 3 shows the peak cortisol levels over time from the first endocrine evaluation. As can be seen, the individual who developed AD upon subsequent testing experienced a decrease in peak cortisol over the 4-yr follow-up period.
Discussion
The ability to screen for adrenal autoimmunity with 21-OHAA has identified individuals who are at risk for the development of AD before clinical presentation. We report that 21-OHAA was present in 1.4% of patients with T1D and in 1.9% of their nondiabetic relatives (12).
The natural history of adrenal autoimmunity and factors contributing to progression to overt AD in patients with T1D continue to be defined. In the past, groups have identified characteristics such as higher 21-OHAA index (22); HLA haplotypes, specifically DRB1*0404-DQ8 and DRB1*0301-DQ2 (1, 12); and younger age (23) as risk factors for the development of clinical AD in the group of autoantibody-positive individuals.
In this report we confirm and extend previous studies indicating that HLA genotypes are associated with the expression of adrenal autoimmunity. Individuals who are 21-OHAA positive are much more likely to have the genotype DR3-DQ2/DR4-DQ8 DRB1*0404 than either normal subjects or diabetic controls. However, this high-risk HLA genotype was not associated with an increased progression to AD in the population that was already 21OH-AA positive.
Previous studies have shown that MICA5.1 is related to the presence of AD and adrenal autoimmunity (2, 3, 24). These reports have shown this association in populations that are 21-OHAA positive and in those with AD. The effect of this polymorphism on the development of clinical disease in a population that is autoantibody positive has not been performed. Our report suggests that homozygosity for MICA5.1 may influence the progression to AD in the group of individuals who are 21-OHAA positive. One hundred percent (six of six) of the individuals diagnosed with AD in our group were homozygous for MICA5.1 compared with approximately 30% in the group that was not diagnosed with AD. The association is especially interesting in light of the function of the MICA gene. It encodes a protein that is a ligand for the NKG2D T cell receptor. MICA can activate this receptor on natural killer cells. This interaction may be important for thymic T cell maturation (25). MICA5.1 is a variant that disrupts the coding sequence (premature stop codon) and therefore may affect the ability to activate its receptor and disrupt T cell maturation in the thymus.
Given the rarity of subjects with AD and adrenal autoimmunity, small numbers of autoantibody-positive subjects have been followed for the development of AD. Betterle et al. (1) followed 17 subjects with adrenal autoantibodies and no initial evidence for clinical AD for a mean of 3.2 yr. Seven (41%) of these subjects developed AD over the period of the study. Based on these results, Betterle et al. (1) proposed a sequence of endocrine abnormalities that occurs during the natural history of AD with the initial abnormality and elevated PRA. De Bellis et al. (26) evaluated 20 individuals with positive autoantibodies. Eleven of the subjects remained persistently positive for adrenal autoantibodies, and two (18%) developed clinical AD; the initial abnormality identified was an elevated PRA level. Laureti et al. (27) reported endocrine testing in 11 individuals with 21-OHAA. Five (45%) of these individuals showed a pathological response to cortrosyn stimulation (low and high doses). Four of five of these individuals had elevated baseline ACTH levels.
One study suggested a yearly incidence of AD in those with adrenal autoimmunity of approximately 20% (1) in a group that had an average follow-up of 3.2 yr. In our cohort, the majority of individuals (five of six) who were diagnosed with AD were identified at their initial endocrine evaluation. Therefore, we may have missed the preclinical period marked by elevation of PRA alone in these individuals. Also supporting the fact that the individuals diagnosed with AD in this study may have been mildly symptomatic is the lower insulin dose and lower hemoglobin A1c levels observed before the diagnosis of AD, which increased to pre-AD levels after institution of therapy. We do have a group of individuals with isolated elevated PRA who will require additional follow-up. All of the individuals in our study showed elevated ACTH and PRA at the time of diagnosis of AD.
AD and T1D are known to be part of autoimmune polyendocrine syndromes 1 and 2. None of our patients showed hypoparathyroidism, chronic mucocutaneous candidiasis, or other autoimmune diseases associated with autoimmune polyendocrine syndrome 1. Therefore, these individuals probably represent autoimmune polyendocrine syndrome 2. Our patient population confirmed the known association with thyroid disease, in that 38% of those with 21OH-AA had autoimmune thyroid disease.
Endocrine evaluation of adrenal autoantibody-positive individuals has led to reports that their endocrine parameters may fluctuate over time, such that some who appeared to have early evidence of adrenal insufficiency marked by the elevation of PRA and/or abnormal results in cortrosyn stimulation testing have resolved their endocrine abnormalities and not developed clinically significant AD while simultaneously becoming negative for adrenal cortical autoantibodies (26). In our study we observed resolution of autoantibodies in a small subset of individuals. These individuals were not tested with endocrine tests, so we cannot comment on their adrenal function at the time of the positive autoantibody measurements. Of note, these individuals tended to have lower 21-OHAA levels than the group that remained persistently positive. This may be a reflection of borderline or transient autoimmunity that ultimately resolves and never leads to clinically significant disease, or it may reflect the false-positive rate associated with the cutoff for the assay. The cutoff for positivity in the present study was defined as exceeding the highest index for 241 normal controls. Using this criterion, we have over 90% confidence that the false positive rate is less than 1%.
Endocrine evaluation of adrenal function includes baseline hormone measures of ACTH, PRA, and cortisol and cortrosyn stimulation testing (16, 28). Stimulation testing can be performed using either 1 or 250 μg cortrosyn; the traditional method of measurement is with the higher dose. However, this higher dose of cortrosyn may mask mild adrenal insufficiency by exposing the adrenal glands to overwhelming amounts of ACTH. Therefore, some have argued that the low-dose cortrosyn test is more reliable for the diagnosis of primary adrenal insufficiency (27, 29). Our study suggests that the low-dose cortrosyn test has good sensitivity (100%) for AD, but with a specificity of approximately 80% there is the potential for diagnosing someone with adrenal insufficiency who would have passed standard cortrosyn testing. The clinical significance of this disparity is unclear.
The exact frequency of screening for 21-OHAA in individuals with T1D remains an area of active research and controversy. Our current practice is to screen for 21-OHAA at diagnosis of T1D and then every 2 yr if the first test is negative. If the autoantibodies are positive (which in our laboratory indicates that the same sample has been measured and found to be positive in a confirmatory assay), we believe it is important to confirm positivity in an independent sample. Currently, on a research basis for 21-OHAA-positive individuals, we measure morning baseline ACTH, cortisol, and PRA levels (supine position) and perform high (1 μg) and low (250 μg) dose cortrosyn stimulation testing annually. On a nonresearch basis, a normal annual basal ACTH measurement in the absence of symptoms may be sufficient screening if the patients receive education concerning the symptoms of AD and Addisonian crisis. If an individual becomes symptomatic at any time, we evaluate with baseline cortisol, ACTH, and/or stimulation testing depending upon the situation. We generally use a cutoff of stimulated cortisol less than 18 mg/dl as positive for adrenal insufficiency in the presence of elevated ACTH. The autoantibody tests are repeated at the time of endocrine testing. If an individual converts to negative autoantibodies, we recommend repeating the autoantibody tests annually for several years and performing stimulation testing if the patient becomes symptomatic and/or again positive for the autoantibodies.
ACTH levels are highly predictive of AD and correlate very well with peak cortisol levels in our study. The majority of this correlation is due to the elevated levels in those subjects who have evidence for adrenal insufficiency. Our study has demonstrated that baseline ACTH levels greater than 60 pg/ml (13.2 pmol/liter) are 100% sensitive and 98% specific for the diagnosis of adrenal insufficiency. Thus, we propose that screening of T1D individuals with 21-OHAA can be performed with baseline cortisol and ACTH measurements, with stimulation testing performed in those with abnormal results. Larger series and longer follow-up by multiple groups should allow more robust estimates of ideal endocrine evaluation and more robust life-table analysis of progression to AD, particularly for those who are 21-OHAA positive and MICA5.1 homozygous.
Acknowledgments
We acknowledge the General Clinical Research Centers for their contribution in performing the endocrine testing and the endocrine laboratory at the University of Colorado Health Sciences Center for performing the endocrine assays.
Footnotes
This work was supported by NIH Grants DK-32083, DK-32493, A146374, and DK-57516; General Clinical Research Centers Grants RR-00051 and RR-00069; and Juvenile Diabetes Research Foundation/Larson Wilkins Pediatric Endocrine Fellowship 13-2002-444 (to J.M.B.).
First Published Online October 13, 2004
Abbreviations: AD, Addison’s disease; HLA, class II histocompatibility leukocyte antigen; MHC, major histocompatibility complex on chromosome 6; MICA, major histocompatibility complex class I-related chain A; 21-OHAA, 21-hydroxylase autoantibodies; PRA, plasma renin activity; T1D, type 1 diabetes mellitus.
Received May 14, 2004.
Accepted September 28, 2004.
References
Betterle C, Scalici C, Presotto F, Pedini B, Moro L, Rigon F, Mantero F 1988 The natural history of adrenal function in autoimmune patients with adrenal autoantibodies. J Endocrinol 117:467–475
Park YS, Sanjeevi CB, Robles D, Yu L, Rewers M, Gottlieb PA, Fain P, Eisenbarth GS 2002 Additional association of intra-MHC genes, MICA and D6S273, with Addison’s disease. Tissue Antigens 60:155–163
Gambelunghe G, Falorni A, Ghaderi M, Laureti S, Tortoioli C, Santeusanio F, Brunetti P, Sanjeevi CB 1999 Microsatellite polymorphism of the mhc class I chain-related (mic-a and mic-b) genes marks the risk for autoimmune Addison’s disease. J Clin Endocrinol Metab 84:3701–3707
Anderson J, Goudie RB, Gray KG, Timbury GC 1957 Autoantibodies in Addison’s disease. Lancet 272:1123–1124
Winqvist O, Karlsson FA, Kampe O 1992 21-Hydroxylase, a major autoantigen in idiopathic Addison’s disease. Lancet 339:1559–1562
Baumann-Antczak A, Wedlock N, Bednarek J, Kiso Y, Krishnan H, Fowler S, Smith BR, Furmaniak J 1992 Autoimmune Addison’s disease and 21-hydroxylase. Lancet 340:429–430
Falorni A, Laureti S, Nikoshkov A, Picchio ML, Hallengren B, Vandewalle CL, Gorus FK, Tortoioli C, Luthman H, Brunetti P, Santeusanio F 1997 21-Hydroxylase autoantibodies in adult patients with endocrine autoimmune diseases are highly specific for Addison’s disease. Clin Exp Immunol 107:341–346
Betterle C, Volpato M, Pedini B, Chen S, Smith BR, Furmaniak J 1999 Adrenal-cortex autoantibodies and steroid-producing cells autoantibodies in patients with Addison’s disease: comparison of immunofluorescence and immunoprecipitation assays. J Clin Endocrinol Metab 84:618–622
Soderbergh A, Winqvist O, Norheim I, Rorsman F, Husebye ES, Dolva O, Karlsson FA, Kampe O 1996 Adrenal autoantibodies and organ-specific autoimmunity in patients with Addison’s disease. Clin Endocrinol (Oxf) 45:453–460
Falorni A, Nikoshkov A, Laureti S, Grenb?ck E, Hulting A-L, Casucci G, Santeusanio F, Brunetti P, Luthman H, Lernmark ? 1995 High diagnostic accuracy for idiopathic Addison’s disease with a sensitive radiobinding assay for autoantibodies against recombinant human 21-hydroxylase. J Clin Endocrinol Metab 80:2752–2755
Brewer KW, Parziale VS, Eisenbarth GS 1997 Screening patients with insulin-dependent diabetes mellitus for adrenal insufficiency. N Engl J Med 337:202
Yu L, Brewer KW, Gates S, Wu A, Wang T, Babu S, Gottlieb PA, Freed BM, Noble J, Erlich HA, Eisenbarth GS 1999 DRB1*04 and DQ alleles: expression of 21-hydroxylase autoantibodies and risk of progression to Addison’s disease. J Clin Endocrinol Metab 84:328–335
Peterson P, Salmi H, Hyoty H, Miettinen A, Ilonen J, Reijonen H, Knip M, ?kerblom HK, Krohn K 1997 Steroid 21-hydroxylase autoantibodies in insulin-dependent diabetes mellitus. Childhood Diabetes in Finland (DiMe) Study Group. Clin Immunol Immunopathol 82:37–42
Myhre AG, Undlien DE, Lovas K, Uhlving S, Nedrebo BG, Fougner KJ, Trovik T, Sorheim JI, Husebye ES 2002 Autoimmune adrenocortical failure in Norway: Autoantibodies and HLA class II associations related to clinical features. J Clin Endocrinol Metab 87:618–623
Ota M, Katsuyama Y, Mizuki N, Ando H, Furihata K, Ono S, Pivetti-Pezzi P, Tabbara KF, Palimeris GD, Nikbin B, Davatchi F, Chams H, Geng Z, Bahram S, Inoko H 1997 Trinucleotide repeat polymorphism within exon 5 of the MICA gene (MHC class I chain-related gene A): allele frequency data in the nine population groups Japanese, Northern Han, Hui, Uygur, Kazakhstan, Iranian, Saudi Arabian, Greek and Italian. Tissue Antigens 49:448–454
Oelkers W 1996 Adrenal insufficiency. N Engl J Med 335:1206–1212
McAulay V, Frier BM 2000 Addison’s disease in type 1 diabetes presenting with recurrent hypoglycaemia. Postgrad Med J 76:230–232
Armstrong L, Bell PM 1996 Addison’s disease presenting as reduced insulin requirement in insulin dependent diabetes. BMJ 312:1601–1602
Colls J, Betterle C, Volpato M, Prentice L, Smith BR, Furmaniak J 1995 Immunoprecipitation assay for autoantibodies to steroid 21-hydroxylase in autoimmune adrenal diseases. Clin Chem 41:375–380
Park YS, Lee H, Eisenbarth GS, Sanjeevi CB 2001 2001 MICA polymorphism is associated with type 1 diabetes in the Korean population. Diabetes Care 24:33–38
Grinspoon SK, Biller BM 1994 Clinical review 62: laboratory assessment of adrenal insufficiency. J Clin Endocrinol Metab 79:923–931
Laureti S, De Bellis A, Muccitelli VI, Calcinaro F, Bizzarro A, Rossi R, Bellastella A, Santeusanio F, Falorni A 1998 Levels of adrenocortical autoantibodies correlate with the degree of adrenal dysfunction in subjects with preclinical Addison’s disease. J Clin Endocrinol Metab 83:3507–3511
Betterle C, Volpato M, Rees Smith B, Furmaniak J, Chen S, Zanchetta R, Greggio NA, Pedini B, Boscaro M, Presotto F 1997 II. Adrenal cortex and steroid 21-hydroxylase autoantibodies in children with organ-specific autoimmune diseases: markers of high progression to clinical Addison’s disease. J Clin Endocrinol Metab 82:939–942
Bilbao JR, Martin-Pagola A, Perez de Nanclares G, Calvo B, Vitoria JC, Vazquez F, Castano L 2003 HLA-DRB1 and MICA in autoimmunity: common associated alleles in autoimmune disorders. Ann NY Acad Sci 1005:314–318
Hue S, Monteiro RC, Berrih-Aknin S, Caillat-Zucman S 2003 Potential role of NKG2D/MHC class I-related chain A interaction in intrathymic maturation of single-positive CD8 T cells. J Immunol 171:1909–1917
De Bellis A, Bizzarro A, Rossi R, Amoresano Paglionico V, Criscuolo T, Lombardi G, Bellastella A 1993 Remission of subclinical adrenocortical failure in subjects with adrenal autoantibodies. J Clin Endocrinol Metab 76:1002–1007
Laureti S, Arvat E, Candeloro P, Di Vito L, Ghigo E, Santeusanio F, Falorni A 2000 Low dose (1 μg) ACTH test in the evaluation of adrenal dysfunction in pre-clinical Addison’s disease. Clin Endocrinol (Oxf) 53:107–115
Dorin RI, Qualls CR, Crapo LM 2003 Diagnosis of adrenal insufficiency. Ann Intern Med 139:194–204
Laureti S, Falorni A 2002 Reliability of the ACTH low dose test in the evaluation of adrenal insufficiency. J Postgrad Med 48:251–252(Jennifer M. Barker, Akane)