| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
CLINICAL CASE SEMINAR |
Metabolic Research Laboratories (D.B.S., R.K.S., C.A., S.O.), Institute of Metabolic Science, Department of Medicine (M.R.C., P.A.L., G.J.M.A., A.N.C., K.G.C.S.), Cambridge Institute for Medical Research, and Department of Pathology (S.T.), University of Cambridge, Addenbrooke’s Hospital, Cambridge CB2 0QQ, United Kingdom; Department of Medical Biochemistry and Immunology (B.P.M.), School of Medicine, Cardiff University, Cardiff CF10 3AT, United Kingdom; Clinical Endocrinology Branch (E.K.C., P.G.), National Institute of Diabetes, Digestive and Kidney Diseases, Bethesda, Maryland 20892-2560; Wellcome Trust Clinical Research Facility (P.R.-B., P.R.M.), Addenbrooke’s Hospital, Cambridge CB2 0SP, United Kingdom; Medway Maritime Hospital (I.S.), Gillingham, Kent ME7 5NY, United Kingdom; Department of Haematological Medicine (G.J.M.), Kings College London, London WC2R 2LS, United Kingdom; and Institute of Normal Human Morphology (I.M., S.C.), University of Ancona, I-60131 Ancona, Italy
Address all correspondence and requests for reprints to: Dr. D. B. Savage or Dr. R. K. Semple, Metabolic Research Laboratories, Institute of Metabolic Science, University of Cambridge, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 0QQ, United Kingdom. E-mail: dbs23{at}medschl.cam.ac.uk, or rks16{at}cam.ac.uk, respectively.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objective: We now describe an immunologically distinct form of acquired generalized lipodystrophy, with evidence of activation of the classical complement pathway (low C4) and autoimmune hepatitis.
Patients and Research Design: Three unrelated patients with acquired lipodystrophy and low complement C4 levels are described. In vitro analysis of the complement pathway was undertaken to determine the reason for the low C4 complement levels. Biopsies were obtained from liver, bone marrow, and adipose tissue for histological analysis.
Results: All three patients manifested near-total lipodystrophy, chronic hepatitis with autoimmune features, and low C4 complement levels. Additional autoimmune diseases, including severe hemolytic anemia, autoimmune thyroid disease, and polyneuropathy, were variably present. Detailed studies of complement pathways suggested constitutive classical pathway activation.
Conclusions: Although the previously described syndrome, which typically results in a cephalad pattern of partial lipodystrophy, results from activation of the alternative complement pathway, this form, in which lipodystrophy is generalized, is associated with activation of the classical pathway. Future therapeutic approaches to these disorders may benefit from being tailored to their distinct immunopathogenesis.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Biochemical analysis
Insulin, leptin, and adiponectin were measured using customized AutoDELFIA (Wallac Oy, Turku, Finland) immunoassays as previously described (14, 15).
Complement assays
C3a/C3a(desArg) was measured in a commercial ELISA (Quidel Corp., San Diego, CA). Terminal complement complex (TCC) was measured in an ELISA using the TCC neoantigen-specific monoclonal antibody ae11 (Hycult biotechnology, Uden, The Netherlands) as capture and a horseradish peroxidase-labeled anti-C6 monoclonal antibody (in house) as detection, as described in Ref. 16 . Standard for the TCC assay was highly purified TCC. The normal range for each assay was established in house by measuring EDTA plasma samples from 30 healthy donors; the upper limit of this range is used to identify pathological activation. The CP hemolytic complement titer (CP-CH50) was measured in the fluid phase using antibody sensitized sheep erythrocytes as target. The alternative pathway (AP) hemolytic complement titer (AP-CH50) was measured in the fluid phase using unsensitized rabbit erythrocytes as target. Hemolytic activity in hemolytic units (HUs) is calculated relative to a standard serum sample.
C3 nephritic factors were identified based upon their capacity to stabilize the AP C3 convertase in vitro, as described in Ref. 17 . C4 nephritic factors were identified based upon CP C3 convertase stabilizing activity using a modified version of a published assay (18). In brief, ELISA plates were coated with human IgG, blocked with albumin, then incubated (15 min, 20 C) with C3-deficient human serum diluted in veronal-buffered saline containing calcium and magnesium ions (VBS++) to deposit the CP convertase C4b2a. After a brief wash in cold VBS++, wells were incubated (5, 15, or 30 min, 20 C) with IgG prepared from patient or control serum diluted in VBS++. Wells were washed in cold VBS++ and incubated (60 min on ice) with horseradish peroxidase-labeled polyclonal antihuman C2 to detect residual C4b2a complex. After washing in cold VBS++, plates were developed using ortho-phenylenediamine substrate and read in a spectrophotometer.
Histopathology
The first patient underwent a surgical biopsy of residual anterior abdominal wall sc fat, which was processed as described previously (19).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
A 53-yr-old Indian woman with a 2-yr history of oral lichen planus developed jaundice. Laboratory investigation revealed an elevated prothrombin time (18.6 sec, normal range 11–14), severe hypoalbuminemia (6 g/liter, normal range 30–51), extreme polyclonal hypergammaglobulinemia (Table 1
), and weakly positive antimitochondrial and antinuclear (double-stranded DNA) antibodies. Liver histology showed active cirrhosis; the periportal infiltrate was characterized by lymphocytes, plasma cells, and polymorphonuclear neutrophils, and the features were consistent with autoimmune hepatitis. She was treated with corticosteroids with rapid recovery of liver function. There was no evidence of steatosis. Three months later hemolytic anemia with a positive direct antiglobulin test developed, followed by diabetes mellitus after initiation of oral corticosteroids. At this point, fat loss was also noted on her limbs and, to a lesser extent, her abdomen. Her course was complicated by poor glycemic control (despite large doses of insulin), and increasing axillary, nuchal, and flexural skin pigmentation. Physical examination revealed scleral icterus, extensive acanthosis nigricans, and predominantly limb lipodystrophy (in this manuscript we consider femorogluteal fat as part of limb fat) (Fig. 1 and Table 2). Abdominal sc fat loss was patchy with some apparently preserved areas. Hyperinsulinemia, hyperglycemia, and dyslipidemia were found, with serum leptin levels toward the lower end of the reference range and undetectable plasma adiponectin (Table 1). Severe polyclonal hypergammaglobulinemia and weakly positive antinuclear antibody persisted with selective suppression of complement factor C4 (Table 1 and supplemental Fig. 1, which is published as supplemental data on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org). No anti-insulin receptor antibodies were detected.
|
|
|
). Immunohistochemistry confirmed the residual nature of these lipid droplets because they stained negatively for perilipin (only viable adipocytes are immunoreactive to perilipin) (19). Electron microscopy (EM) revealed the presence of adipocyte remnants near the residual lipid droplets and numerous macrophages laden with lipid droplets (Fig. 2). Together, these light microscopy, immunohistochemistry, and EM findings suggest that the dead adipocytes surrounded by lipid reabsorbing macrophages (some in the form of giant multinucleated cells) represent crown-like structures (CLSs) previously described by Cinti and colleagues (19, 20) in adipose tissue samples from obese mice and humans.
|
). CP CH50, a functional assay of classical and terminal complement pathways, was persistently low (720 HU, normal range 1000–1500), whereas AP CH50 (170 HU, normal range 80–200) was normal, and TCC levels (12.6 µg/liter, normal range < 0.5 µg/liter serum) and C3a (desArg) levels (0.95 µg/liter, normal range < 0.2) were markedly elevated, providing strong evidence of ongoing complement activation. Complement component levels were otherwise normal (data not shown), and there was no evidence of either C3 or C4 nephritic factor activity.
Case history 2
A 19-yr-old woman was referred with acne, hirsutism, and oligomenorrhea. At 10 yr old, she had been noted to have reduced adipose tissue on her arms and legs. At that time her C3 level was normal, but C4 was low at 0.08 g/liter (0.10–0.40). Examination confirmed moderate hirsutism without virilization, flexural acanthosis nigricans, and limb lipodystrophy. At this stage abdominal sc fat was preserved. Laboratory investigation revealed severe insulin resistance. Serum leptin was lower than body mass index (BMI) and sex-matched controls, whereas adiponectin was undetectable (Table 1).
Rosiglitazone was introduced, but over the subsequent month, there was progressive weight loss, worsening liver function and dyslipidemia, and no improvement in her lipodystrophy. Although serum alanine aminotransferase improved after withdrawal of the thiazolidinedione, severe hemolytic anemia developed with a strongly positive direct antiglobulin test (hemoglobin nadir 3.7 g/dl). Further investigation revealed end-stage liver disease and portal hypertension. Shortly afterwards, diabetes mellitus was diagnosed and insulin started.
Glucocorticoid therapy was begun, sustaining the hemoglobin around 7 g/dl, however, lipodystrophy progressed to become generalized over the next year. Complement C3 was persistently normal, whereas C4 remained undetectable (Table 1). Despite iv Ig and plasmapheresis, severe hemolysis persisted. A detailed autoantibody screen, including antinuclear antibody, antineutrophil cytoplasmic antibody, and anti-C1q, was negative, but IgG was elevated at 18.4 g/liter. Despite the introduction of rituximab and further plasmapheresis, her condition deteriorated, and she died while awaiting a liver transplant.
Case history 3
A 5-yr-old boy presented with a history of recurrent infections and failure to thrive. His parents had also noted that he had become very thin with minimal sc fat. His birth weight was 3.1 kg. Generalized lipodystrophy was formally recognized at age 7 yr. At this time IgG was elevated, and C4 barely detectable, whereas C3 levels were within the normal range (Table 1). Antinuclear factor, antidouble-stranded DNA, and antimitochondrial antibodies were undetectable. At the age of 10 yr, he presented with jaundice and was noted to have a significantly elevated aspartate aminotransferase (397 IU/liter), for which no specific cause was identified. At this time he was also noted to have acanthosis nigricans, impaired glucose tolerance, and very high insulin levels (Table 1).
He subsequently remained lipodystrophic and insulin resistant, but no further evidence of hepatitis or other autoimmune disease was found. Between 5 and 18 yr old, immunological assessment revealed persistently nearly undetectable C4 with normal C3, high levels of Igs, and elevated C1q binding of serum, evidence for the presence of circulating immune complexes. CH50 assay of classical and terminal complement pathways was also persistently low, consistent with high complement turnover. Reevaluated at the age of 32 yr old, he had normal C3 and C4 levels (1.3 and 0.17 g/liter, respectively). However, his lipodystrophy was unchanged, and leptin and adiponectin were both nearly undetectable (Table 1).
Additional cases
Gorden and colleagues (21, 22) have assembled a large cohort of patients with lipodystrophy. Within this cohort, 13 patients were classified as having AGL. Four of these patients had normal C3 and low C4 complement levels (supplemental Table 1, which is published as supplemental data on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org). Interestingly, three of these patients also had autoimmune hepatitis, and none of them had autoimmune glomerulonephritis.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
We have identified a subset of AGL in which complement factor C4, but not C3, is very low. In all cases, the pattern of lipodystrophy was distinct from that associated with C3 nephritic factor, involving both legs and upper body, and progressing toward generalized lipodystrophy. Plasma leptin levels were consistent with the degree of fat loss, but interestingly, adiponectin levels were exceedingly low (undetectable) in all three patients, despite the preservation of some fat in patients 1 and 2. Plasma adiponectin is well known to be suppressed in states of severe insulin resistance, including those forms associated with partial lipodystrophy, except in cases due to insulin receptor mutations or insulin receptor antibodies (25). In keeping with this, the extreme hypoadiponectinemia seen in cases 1 and 2, despite preservation of some adipose tissue and detectable leptin levels, is most likely a consequence of the metabolic stress imposed upon residual adipocytes, as evidenced by the large size of residual adipocytes and the presence of CLSs surrounding dead/dying adipocytes (Fig. 2). However, the ability of adiponectin to bind complement factor C1q and activate the CP (26) raises the possibility that it may additionally have a pathogenic role in the autoimmune adipocyte destruction, and that disproportionately low levels may in part reflect its consumption in this process.
Congenitally low C4 levels are associated with an increased risk of autoimmune disease, particularly systemic lupus erythematosus (SLE) (27). However, the fact that C4 levels fluctuated in response to therapy in patient 1, and that they eventually normalized in patient 3, argues against low C4 levels being the primary abnormality. Instead, low CP CH50 and marked elevation of terminal complex components in patient 1, together with persistently elevated circulating immune complexes and low CH50 in patient 3 suggest that activation of the classical complement pathway ultimately depletes C4 levels. The occurrence of multiple autoimmune diseases in susceptible individuals is a well-described phenomenon, and is likely to reflect an underlying genetic predisposition to the breakdown of self-tolerance (28, 29). In such individuals, these susceptibility genes are likely to affect important, proximal tolerance mechanisms leading to a variety of autoimmune phenotypes. Chronic hepatitis with autoimmune features was present in all except one of the cases we know of with acquired lipodystrophy and low C4 levels. Although nonalcoholic fatty liver disease is almost universally present in generalized lipodystrophy (30), liver disease was noted to precede lipodystrophy in some of the cases, and histological findings, including the presence of inflammatory cell infiltrates in patient 1 in the absence of any steatosis, were very suggestive of autoimmune disease. In contrast to C3 nephritic factor-associated lipodystrophy, an antibody/immune-complex-mediated glomerulonephritis was not apparent, nor was there glomerular complement deposition in any of the patients with low C4 levels. Although proteinuria was present in all the patients in whom it was measured, this is a common feature of generalized lipodystrophy, most commonly accounted for by focal segmental glomerulosclerosis (22).
All three patients had a low serum C4 but had no evidence of a "C4 nephritic factor" to account for this. C4 consumption by activation of the CP is the likely cause of these abnormalities, and this in turn is very likely to be driven by complement fixation by IgG. Thus, the primary abnormality driving disease in all three of these patients appears to be abnormal IgG production. In addition to low C4, all patients had evidence of hypergammaglobulinemia, and two of severe hemolytic anemia. Abnormal IgG production was particularly apparent in patient 1, who we were able to analyze in more detail. She had striking accumulation of both B cells and plasma cells in her adipose tissue (Fig. 2). The level of plasma cell infiltration in her marrow (10%) was at a level normally only seen in multiple myeloma, though her plasma cells were clearly polyclonal and not malignant. Widespread abnormalities of Ig production, together with complement consumption, are a feature of SLE. Recently, gene expression in peripheral blood mononuclear cells of SLE patients has been examined using microarrays, and a number of transcription "signatures" characteristic of the disease have been defined (31, 32). To determine whether patient 1 exhibited SLE-like transcriptional changes, we performed microarrays both on peripheral blood mononuclear cells and on purified white cell subsets, as previously described (33). Patient 1 clustered between the SLE patients and controls (supplemental Fig. 2). She did not display evidence of the interferon 
-related signature that is characteristic of SLE but did show evidence of the "plasmablast signature" commonly seen in patients with lupus. Thus, an examination of RNA transcription in patient 1 demonstrates that she has a disease characterized by marked B-cell dysregulation, but distinct from SLE. The identification of dysregulated B cells as an important factor underlying disease pathogenesis allows a more focused approach to therapy. Therefore, it is interesting to note that patient 1’s disease responded to a combination of plasma exchange to remove IgG, and B-cell depletion. As more effective plasma cell-depleting therapies become available, these may well be useful in this condition.
In patient 1, immunosuppression was successful in stabilizing the polyneuropathy and hemolytic anemia clinically, though laboratory evidence of disease activity persists. We are not aware of any reports describing recovery of adipose tissue depots in acquired lipodystrophy, so we monitored patient 1 closely for changes in adipose tissue mass. Initially, her fat mass did appear to increase [11.7% by dual x-ray absorptiometry (DXA) to 14.7% over 6 months], but subsequent measurements over 2 yr (Fig. 1) have not revealed any further changes, and clinical assessment of her fat distribution suggests that areas where fat was lost have not recovered. In our view these data suggest that local pre-adipocytes are also lost in this disease and that adipose tissue precursors (possibly mesenchymal stem cells) do not reconstitute lost fat depots.
In summary, we describe a subtype of AGL associated with activation of the classical complement pathway and low C4 complement levels. This disease cluster is frequently associated with autoimmune hepatitis and is distinct from a form of acquired partial lipodystrophy in which C3 nephritic factor constitutively activates the alternate complement pathway. Although the fact that adipocytes produce several components of the complement cascade tantalizingly hints at a primary role for adipose tissue or an adipocyte or adipokine antibody in this disorder, the pathogenesis remains incompletely understood. Aggressive immunosuppression may be required to limit progressive autoimmune disease in some cases. Recognition that these distinct subtypes of acquired lipodystrophy have very different prognoses and treatment algorithms should alert physicians to the need to measure complement levels in patients with acquired lipodystrophy and aid subsequent clinical monitoring.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online October 14, 2008
1 D.B.S. and R.K.S. contributed equally to this work. 
Abbreviations: AGL, Acquired generalized lipodystrophy; AP, alternative pathway; BMI, body mass index; CLS, crown-like structure; CP, classical pathway; DXA, dual x-ray absorptiometry; EM, electron microscopy; HU, hemolytic unit; SLE, systemic lupus erythematosus; TCC, terminal complement complex.
Received August 4, 2008.
Accepted October 8, 2008.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
and human metabolic disease. J Clin Invest 116:581–589[CrossRef][Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
