CCN Proteins Are Distinct from, and Should Not Be Considered

0021-972X/01/$03.00/0 The Journal of Clinical Endocrinology & Metabolism Copyright © 2001 by The Endocrine Society

Vol. 86, No. 2 Printed in U.S.A.

LETTERS TO THE EDITOR CCN Proteins Are Distinct from, and Should Not Be Considered Members of, the Insulin-Like Growth Factor-Binding Protein Superfamily To the editor:

including platelet-derived growth factor, nerve growth factor, and transforming growth factor-␤ (2). Cop-1/WISP-2 is unique because it lacks the carboxyl-terminal cysteine knot domain. Each of the modular structural domains is encoded by a separate exon, suggesting that genes of the CCN family arose through exon shuffling of preexisting genes. Sequence similarity with IGFBPs exists only in the N-terminal domain encoded by one exon. From the sequence perspective, CCN proteins are no more related to IGFBPs than to von Willebrand factor, thrombospondin, or growth factor cysteine knots (Refs. 2–5; Fig. 1). Functionally, a number of biologically significant activities have been clearly demonstrated for CCN proteins, none of which has any apparent relationship to IGF binding (12–27). Purified CCN proteins have been demonstrated to mediate and promote cell adhesion, migration, proliferation, and survival (3–5, 12). As matrix-associated, heparin-binding

In a recent letter, Baxter et al. (1) proposed renaming a family of extensively studied proteins, known in the literature as the CCN family (reviewed in Refs. 2–5), as insulin-like growth factor binding proteinrelated proteins (IGFBP-rPs). This proposed name change lacks functional or biological basis, has been suggested unilaterally without consensus or consultation with those working on these proteins, and serves to confuse rather than clarify the literature. The prototypic members of the CCN family (CTGF, CYR61, and

FIG. 1. Schematic representation of CCN proteins compared with Mac25 and IGFBP-1. Following the secretory signal (open oval), CCN proteins exhibit four discrete conserved domains: IGFBP, VWC (von Willebrand factor type C repeat), TSP1 (thrombospondin type 1 repeat), and CT (carboxyl-terminal cysteine knot). A central variable region separate the proteins into two halves. The overall percent amino acid sequence identities and homologies including conservative substitutions (%ID/HO) of the human sequences of each protein compared with CTGF are listed. Within the N-terminal domain, CTGF shows 48 – 60% amino acid sequence identity compared with CCN family members, but only 31–33% compared with Mac25 and IGFBP-1. Outside of the N-terminal domain, CCN proteins, Mac25, and IGFBP-1 share no sequence similarity. proteins (13), CYR61 and CTGF are novel ligands of the integrins ␣V␤3 and ␣IIb␤3 (14 –16), and NOV interacts with fibulin 1C (17), suggesting their involvement in cell adhesion signaling. Both CYR61 and CTGF induce angiogenesis in vivo (15, 18) and chondrogenesis in vitro (19, 20). CTGF is expressed in fibroblasts during wound healing (21, 22) and can induce fibrosis in vivo (22). Furthermore, CTGF has been demonstrated to mediate both the mitogenic and matrigenic activities of transforming growth factor-␤ (23–25). Other studies have revealed that CYR61 promotes tumor growth (18), whereas Cop-1/WISP-2 or Elm-1/WISP-1 can inhibit tumor growth (9, 10). It has also been established that the expression of NOV is abnormal in tumor cells (26, 27) and that expression

NOV) were discovered in our laboratories in the early 1990s (6 – 8). Additional members of the family have been identified, including Elm1/WISP-1, Cop-1/WISP-2, and WISP-3 (Refs. 9 –11; Fig. 1). These highly conserved cysteine-rich proteins share four conserved modular domains with sequence similarities to IGFBP, von Willebrand factor, thrombospondin, and a cysteine knot characteristic of some growth factors Received February 10, 2000. Address correspondence to: Lester F. Lau, Ph.D., Department of Molecular Genetics, 900 South Ashland Avenue, Chicago, Illinois 60607-7170. E-mail:[email protected].

944

LETTERS TO THE EDITOR of an amino-truncated form of NOV is transforming but full-length NOV inhibits fibroblast growth (8), suggesting an involvement of this protooncogene in malignancy. If a unified nomenclature were to be proposed for these multifunctional proteins, such nomenclature should reflect their demonstrated biological activities. The term “IGFBP-related proteins” does not fulfill this need. Is there any functional or biological basis for addressing the CCN proteins as members of the IGFBP superfamily? The only data in existence is that CTGF (28) and NOV (29) bind IGF in vitro with a 100- to 1000-fold lower affinity than authentic IGFBPs. Inasmuch as no IGF binding to NOV was observed under standard ligand blotting assay conditions (27), the low-affinity binding for IGF remains controversial. No published data speak to any potential binding of CYR61 to IGF. Clearly, these proteins cannot compete with the high-affinity IGFBPs that are so abundant in serum. More importantly, to date, there is no demonstrated physiological significance of IGF binding by any member of the CCN family. Thus, proposing to abandon the established names of the CCN proteins and to rename and reclassify them on a speculative basis does not make sense and serves only to divert attention from the carefully documented and published work that has identified specific biological activities of these molecules. The proposal of Baxter et al. (1) to reclassify substantively different molecules under the same rubric is misleading, exemplified in this case by the placement of CCN proteins in the same category as Mac25, a protein homologous to the activin-binding protein follistatin (30). Even a cursory inspection of Fig. 1 reveals that the CCN proteins form a distinct family, separate and apart from Mac25 and IGFBPs. The proposed renaming misleadingly suggests an intimate relationship among CCN proteins Mac25 and IGFBPs that does not exist and implies that the biological activities of CCN proteins function through an IGF-binding activity, which has not been demonstrated in any context. Changes in nomenclature often make good sense in a field where clarity and focus can be served based on accumulated new information. However, this should be done with the consensus of those who work in the field, rather than unilaterally. In this instance, because the very low-affinity binding of IGF by CCN proteins has no demonstrated biological significance, this proposed name change serves no scientific or intellectual purpose. The proposed renaming of the CCN family as IGFBP-rPs simply ignores the multitude of well-documented and established biological activities of these proteins (3–27). The use of superfluous names such as IGFBP-rP serves only to add confusion rather than insight into the functions and activities of this complex and important emerging family of proteins.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

Gary R. Grotendorst, Lester F. Lau, and Bernard Perbal Department of Cell Biology and Anatomy (G.R.G.), University of Miami School of Medicine, Miami, Florida 33136; Department of Molecular Genetics (L.F.L.), University of Illinois College of Medicine, Chicago, Illinois 60607; and Laboratoire d’Oncologie Virale et Mole´culaire (B.P.), UFR de Biochimie, Universite´ Paris 7, 75005 Paris and Unite´ INSERM 515, Hoˆpital SaintAntoine, Paris 75012, France

References 1. Baxter RC, Binoux MA, Clemmons DR, et al. 1998 Recommendations for nomenclature of the insulin-like growth factor binding protein superfamily. J Clin Endocrinol Metab. 83:3213. 2. Bork P. 1993 The modular architecture of a new family of growth regulators related to connective tissue growth factor. FEBS Lett. 327:125–130. 3. Grotendorst GR. 1997 Connective tissue growth factor: a mediator of TGF-␤ action on fibroblasts. Cytokine Growth Factor Rev. 8:171–179. 4. Lau LF, Lam SC-T. 1999 The CCN family of angiogenic regulators: the integrin connection. Exp Cell Res. 248:44 –57. 5. Brigstock DR. 1999 The connective tissue growth factor/cysteine-rich 61/ nephroblastoma overexpressed (CCN) family. Endocr Rev. 20:189 –206. 6. O’Brien TP, Yang GP, Sanders L, Lau LF. 1990 Expression of cyr61, a growth factor-inducible immediate early gene. Mol Cell Biol. 10:3569 –3577. 7. Bradham DM, Igarshi A, Potter RL, Grotendorst GR. 1991 Connective tissue growth factor: a cysteine-rich mitogen secreted by human vascular endothelial cells is related to the SRC-induced immediate early gene product CEF-10. J Cell Biol. 114:1285–1294. 8. Joliot V, Marinerie C, Dambrine G, et al. 1992 Proviral rearrangements and

23.

24.

25.

26.

27.

28.

29.

30.

945

overexpression of a new cellular gene (nov) in myeloblastosis-associated virus type-1 induced nephroblastomas. Mol Cell Biol. 12:10 –21. Hashimoto Y, Shindo-Okada N, Tani M, et al. 1998 Expression of the Elm1 gene, a novel gene of the CCN (connective tissue growth factor, Cyr61/cef10, and nephroblastoma overexpressed gene) family, suppresses in vivo tumor growth and metastasis of K-1735 murine melanoma cells. J Exp Med. 187:289 –296. Zhang R, Averboukh L, Zhu W, et al. 1998 Identification of rCOP-1, a new member of the CCN protein family, as a negative regulator for cell transformation. Mol Cell Biol. 18:6131– 6141. Pennica D, Swanson TA, Welsh JW, et al. 1998 WISP genes are members of the connective tissue growth factor family that are up-regulated in Wnt-1 transformed cells and aberrantly expressed in human colon tumors. Proc Natl Acad Sci USA. 95:14717–14722. Kireeva ML, Mo F-E, Yang GP, Lau LF. 1996 Cyr61, product of a growth factor-inducible immediate-early gene, promotes cell proliferation, migration, and adhesion. Mol Cell Biol. 16:1326 –1334. Yang GP, Lau LF. 1991 Cyr61, product of a growth factor-inducible immediate early gene, is associated with the extracellular matrix and the cell surface. Cell Growth Differ. 2:351–357. Kireeva ML, Lam SC-T, Lau LF. 1998 Adhesion of human umbilical vein endothelial cells to the immediate-early gene product Cyr61 is mediated through integrin ␣V␤3. J Biol Chem. 273:3090 –3096. Babic AM, Chen C-C, Lau LF. 1999 Fisp12/mouse connective tissue growth factor mediates endothelial cell adhesion and migration through integrin ␣V␤3 promotes endothelial cell survival, and induces angiogenesis in vivo. Mol Cell Biol. 19:2958 –2966. Jedsadayanmata A, Chen C-C, Kireeva ML, Lau LF, Lam SC-T. 1999 Activation-dependent adhesion of human platelets to Cyr61 and Fisp12/mouse connective tissue growth factor is mediated through integrin IIb␤. J Biol Chem. 274:24321–24327. Perbal B, Martinerie C, Sainson R, Werner M, He B, Roizman B. 1999 The C-terminal domain of the regulatory protein NOVH is sufficient to promote interaction with fibulin 1C: a clue for a role of NOVH in cell-adhesion signaling. Proc Natl Acad Sci USA. 96:869 – 874. Babic AM, Kireeva ML, Kolesnikova TV, Lau LF. 1998 CYR61, a product of a growth factor-inducible immediate early genes, promotes angiogenesis and tumor growth. Proc Natl Acad Sci USA. 95:6355– 6360. Wong M, Kireeva ML, Kolesnikova TV, Lau LF. 1997 Cyr61, product of a growth factor-inducible immediate early gene, regulates chondrogenesis in mouse limb bud mesenchymal cells. Dev Biol. 192:492–508. Grotendorst GR. Induction of tissue, bone or cartilage formation using connective tissue growth factor. United States Patent 5,837,258, November 17, 1998. Igarashi A, Okochi H, Bradham DM, Grotendorst GR. 1993 Regulation of connective tissue growth factor gene expression in human skin fibroblasts and during wound repair. Mol Biol Cell. 4:637– 645. Frazier K, Williams S, Kothapalli D, Klapper H, Grotendorst GR. 1996 Stimulation of fibroblast cell growth, matrix production, and granulation tissue formation by connective tissue growth factor. J Invest Dermatol. 107:404 – 411. Kothapalli D, Frazier K, Grotendorst GR. 1997 TGF-␤ induces anchorageindependent growth of NRK fibroblasts via the synergistic action of CTGFdependent and CTGF-independent signaling pathways. Cell Growth Differ. 8:61– 68. Kopthapalli D, Hayashi N, Grotendorst GR. 1998 Inhibition of TGF-␤ stimulated CTGF gene expression and anchorage independent growth by elevation of intracellular cAMP. FASEB J. 12:1151–1161. Duncan MR, Frazier KS, Abramson S, et al. 1999 Connective tissue growth factor mediates transforming growth factor-collagen synthesis: down-regulation by cAMP. FASEB J. 13:1774 –1786. Li WX, Martinerie C, Zumkeller W, Westphal M, Perbal B. 1996 Differential expression of novH and CTGF in human glioma cell lines. J Clin Mol Pathol. 49:M91–M97. Chevalier G, Yeger H, Martinerie C, et al. 1998 novH: differential expression in developing kidney and a marker of heterotypic differentiation in Wilms’ tumor. Am J Pathol. 52:1563–1575. Kim HS, Nagalla SR, Oh Y, Wilson E, Roberts CTJ, Rosenfeld RG. 1997 Identification of a family of low-affinity insulin-like growth factor binding proteins (IGFBPs): characterization of connective tissue growth factor as a member of the IGFBP superfamily. Proc Natl Acad Sci USA. 94:12981–12986. Burren CP, Wilson EM, Hwa V, Oh Y, Rosenfeld RG. 1999 Binding properties and distribution of insulin-like growth factor binding protein-related protein 3 (IGFBP-rP3/NovH), and additional member of the IGFBP superfamily. J Clin Endocrinol Metab. 84:1096 –1103. Kato MV, Sato H, Tsukada T, Ikawa Y, Aizawa S, Nagayoshi M. 1996 A follistatin-like gene, mac25, may act as a growth suppressor of osteosarcoma cells. Oncogene. 12:1361–1364.

946

JCE & M • 2001 Vol. 86 • No. 2

LETTERS TO THE EDITOR

Nomenclature of the Insulin-Like Growth FactorBinding Protein Superfamily To the editor: The letter by Grotendorst, Lau, and Perbal raises several important issues that relate to: 1) the biological properties of the insulin-like growth factor-binding proteins (IGFBPs) and the 9⫹ more recently identified proteins termed “IGFBP-related proteins” (IGFBP-rPs); 2) the concept of a protein “superfamily;” and 3) the uses and misuses of nomenclature. Certainly, the proposal of an IGFBP superfamily (1, 2) was not intended to disparage the work of these and other investigators, which has been appropriately acknowledged and referenced, but to attempt to provide some framework for structure/function analysis of this most interesting group of proteins. A nomenclature for the IGFBPs was developed in the 1980s, when it became apparent that there were more than two proteins found in biological fluids associated with IGF peptides (3, 4). The six IGFBPs share the following characteristics: 1) the ability to bind IGF-I and -II with high affinity; and 2) a cysteine-rich structure with high conservation of the Nand C-terminal domains (5). In recent years, select IGFBPs have also been shown to regulate cell proliferation in an IGF-independent manner, associate with cell membranes and be translocated into the nucleus, and associate with a variety of proteoglycans in extracellular matrix (5, 6). The concept of an IGFBP superfamily was proposed in 1997 (7), with the recognition that four additional proteins–Mac25 (8) and the three original CCN proteins (7, 9)–share the cysteine-rich N-terminal domain of the IGFBPs, and bound IGF (although with low affinity; shown for Mac25, CTGF, and NOV). The fact that these proteins might have actions totally unrelated to conventional IGF physiology was felt to be consistent with the increasing evidence that some IGFBPs, as well as the conserved N-terminal domains of proteolyzed IGFBPs, might be, themselves, capable of “IGF-independent” actions, and that the conventional view of IGFBPs as simple carrier proteins for IGF ligands needed to be modified in light of an expanded involvement of IGFBPs in cell proliferation, adhesion, migration, and survival (many of the characteristics that Grotendorst et al. ascribe to the CCN proteins). The current concept of the IGFBP superfamily has already transcended inclusion of CCN proteins, since, in addition to the CCN proteins listed by Grotendorst et al., the superfamily also includes Mac25, L56, and ESM-1, all of which also share the N-terminal domain of the IGFBPs but are otherwise dissimilar to the CCN proteins (2). It is of note that in all six high-affinity IGFBPs and all nine to-date identified IGFBPrPs, the conserved N-terminal domain is encoded by a single exon. Indeed, all of these proteins are modular proteins, in which the Nterminal domain was almost certainly introduced through exon shuffling. Furthermore, an N-terminal domain-like sequence has been identified in the twisted gastrulation protein (Tsg) from Drosophila. A phylogenetic tree of the N-terminal domain of the IGFBP superfamily demonstrates evolutionary conservation of this motif (2). The term “superfamily” was introduced by Dayhoff (10) in 1978, with the relatedness of proteins based determined exclusively on the basis of primary protein structures, and set at less than 50% amino acid similarity for superfamilies (and ⬎50% for “families”). The similarities between the N-terminal domains from human IGFBPs and human IGFBP-rPs ranges from 40 to 57%, consistent with the definition of a superfamily. This view of a superfamily can be expanded to include functional relationships, tissue specificity, and hormonal regulation (much of which is shared by the various members of the IGFBP superfamily). It is certainly reasonable to consider the CCN proteins to be a family within the domain of the IGFBP superfamily; the same is true for the family of six high-affinity IGFBPs. We do agree with Grotendorst et al. that the selection of the terminology “IGFBP” superfamily was arbitrary and certainly reflects our bias as IGF investigators. On the other hand, it is necessary to point out that this nomenclature was selected to specifically underscore the structural relationship among these proteins, the evolutionary conservation of the exon-defined N-terminal domain, and the significant functional relationships among the proteins. Ultimately, nomenclature is always arbitrary. It is not clear, for example, what particular biological insight is Received January 4, 2000. Address correspondence to: Ron G. Rosenfeld, M.D., Department of Pediatrics, Oregon Health Sciences University, 707 SW Gaines Road, Mail Code: CDRCP, Portland, Oregon 97201.

obtained from the names “CYR61” or “NOV,” and the biological actions of CTGF are certainly not limited to connective tissue cells. Mac25 had already had a half dozen names assigned to it, before its inclusion as a member of the IGFBP superfamily (11). If, on the other hand, such names provide value to research, they should, by all means, be preserved; doing so will not undermine or challenge the structural and functional relationships among the various proteins. In the case of the IGFBP superfamily, the proposal was made in an effort to focus attention on the potential biological significance of the remarkably conserved N-terminal IGFBP motif and to encourage investigators from the IGF field, CCN field, and Mac25, L56, and ESM-1 areas to search for insights from each other’s work. We are hopeful that this is a goal that we can share with Grotendorst, Lau, and Perbal. Ultimately, it is far more important to explore what the physiological and evolutionary significance may be for preserved modular domains, rather than quibbling over nomenclature. With great insight into the arbitrariness of nomenclature, Gould (12) wrote: “But classifications are not passive ordering devices in a world objectively divided into obvious categories. Taxonomies are human decisions imposed upon nature—theories about the causes of nature’s order. The chronicle of historical changes in classification provided our finest insight into conceptual revolutions in human thought. Objective nature does exist, but we can converse with her only through the structure of our taxonomic systems. . . . Categories are human impositions upon nature (though nature’s factuality offers hints and suggestions in return).” Ron G. Rosenfeld, Vivian Hwa, and Youngman Oh Department of Pediatrics Oregon Health Sciences University Portland, Oregon 97201

References 1. Baxter RC, Binoux MA, Clemmons DR, et al. 1998 Recommendations for nomenclature of the insulin-like growth factor binding protein superfamily. J Clin Endocrinol Metab. 83:3213. 2. Hwa V, Oh Y, Rosenfeld RG. 1999 The insulin-like growth factor-binding protein superfamily. Endocr Rev. 20:761–787. 3. Ballard FJ, Baxter RC, Binoux M, et al. 1989 On the nomenclature of the IGF binding proteins. Acta Endocrinol [Copenh]. 121:751–752. 4. Ballard FJ, Baxter RC, Binoux M, et al. 1992 Report on the nomenclature of the IGF binding proteins. J Clin Endocrinol Metab. 74:1215–1216. 5. Jones JI, Clemmons DR. 1995 Insulin-like growth factors and their binding proteins: biological actions. Endocr Rev. 16:3–34. 6. Oh Y, Yamanaka Y, Wilson E, et al. 1998 IGF-independent actions of IGFBPs. In: Takano K, Hizuka N, Takahashi S-I, et al., eds. Molecular mechanisms to regulate the activities of insulin-like growth factors. Amsterdam: Elsevier: 123–133. 7. Kim H-S, Nagalla SR, Oh Y, Wilson E, Roberts Jr CT, Rosenfeld RG. 1997 Identification of a family of low-affinity insulin-like growth factor binding proteins (IGFBPs): characterization of connective tissue growth factor as a member of the IGFBP superfamily. Proc Natl Acad Sci USA. 94:12981–12986. 8. Oh Y, Nagalla SR, Yamanaka Y, Kim H-S, Wilson E, Rosenfeld RG. 1996 Synthesis and characetrization of insulin-like growth factor binding protein (IGFBP)-7. J Biol Chem. 271:30322–30325. 9. Burren CP, Wilson EM, Hwa V, Oh Y, Rosenfeld RG. 1999 Binding properties and distribution of insulin-like growth factor binding protein-related protein 3 (IGFBP-rP3)/NovH), an additional member of the IGFBP superfamily. J Clin Endocrinol Metab. 84:1096 –1103. 10. Dayhoff MO. 1978 Atlas of protein sequence and structure. Washington, DC: National Biomedical Research Foundation; 5. 11. Rosenfeld RG, Oh Y. 1998 Editorial: the blind men and the elephant—a parable for the study of insulin-like growth factor binding proteins. Endocrinology. 139:5–7. 12. Gould SJ. 1997 Full house. New York: Crown Publishers, Inc; 39.

Leptin Actions on the Reproductive Axis To the editor: We read with interest the comprehensive editorial by Hileman et al. (1) entitled “Leptin, nutrition, and reproduction: timing is everything,” which appeared in JCEM. The paper extensively revises the pleiotropic actions of leptin with particular emphasis on the involvement of this hormone in the regulation of energy metabolism and of the reproductive axis. The authors underline that “although leptin clearly influences Received July 19, 2000. Address correspondence to: Paolo Magni, M.D., Ph.D., Institute of Endocrinology, University of Milan, 20133 Milan, Italy.

LETTERS TO THE EDITOR reproduction, where leptin acts to exert its effects is not yet resolved.” This remains a still not fully clarified aspect of leptin physiology. According to studies mentioned in the paper, leptin has been found to stimulate both GnRH and LH release from hypothalamic and pituitary explants, respectively (2). These data suggest that leptin may act at the hypothalamic level, but leave open the question of whether it acts directly or indirectly (or both) on GnRH neurons. On this aspect, the authors comment that “to date, leptin receptors have not been identified on GnRH neurons,” according to some studies performed on brain tissue (3). This observation, however, should be revisited in light of our study published in Endocrinology (5). In this paper, we report that leptin receptors are expressed in mouse immortalized GnRH neurons and that these receptors seem to be functional in modulating GnRH secretion (5). These data, thus, suggest the possibility that leptin may also regulate GnRH secretion by acting directly at the hypothalamic level. GT1 immortalized GnRH neurons represent a well-established in vitro system for studying the control of GnRH secretion (6). This model system seems very useful, since the study of the regulation of GnRH secretion is difficult to perform in vivo, due to the fact that GnRH-secreting neurons are quite few and scattered through the hypothalamic region. The utilization of GT1 immortalized GnRH neurons has generated a large body of papers that have appeared in the most important endocrinological journals. Additional studies seem, however, still necessary to fully clarify this issue in vivo and in the different species. Paolo Magni, Luciano Martini, and Marcella Motta Institute of Endocrinology University of Milan 20133 Milan, Italy

References 1. Hileman SM, Pierroz DD, Flier JS. 2000 Leptin, nutrition, and reproduction: timing is everything. J Clin Endocrinol Metab. 85:804 – 807. 2. Yu WH, Kimura M, Walczewska A, Karanth S, McCann SM. 1997 Role of leptin in hypothalamic-pituitary function. Proc Natl Acad Sci USA. 94:1023–1028. 3. Hakansson M-L, Brown H, Ghilardi N, Skoda RC, Meister B. 1998 Leptin receptor immunoreactivity in chemically defined target neurons of the hypothalamus. J Neurosci. 18:559 –572. 4. Finn PD, Cunningham MJ, Pau K-YF, Spies HG, Clifton DK, Steiner RA. 1998 The stimulatory effect of leptin on the neuroendocrine reproductive axis of the monkey. Endocrinology. 139:4652– 4662. 5. Magni P, Vettor R, Pagano C, et al. 1999 Expression of a leptin receptor in immortalized GnRH-secreting neurons. Endocrinology. 140:1581–1585. 6. Mellon P, Windle J, Goldsmith P, Padula C, Roberts J, Weiner R. 1990 Immortalization of hypothalamic GnRH neurons by genetically targeted tumorigenesis. Neuron. 5:1–10

Hashimoto’s Encephalitis To the editor: The article titled “Triple H syndrome: a novel autoimmune endocrinopathy characterized by dysfunction of the hippocampus, hair follicle, and hypothalamic-pituitary-adrenal axis,” published in JCEM (1) mentions Hashimoto’s encephalitis in its introductory paragraphs, which prompted us to write this letter. This is a vague term, describing an association between presence of thyroid antibodies and encephalitic features, published by Seipelt et al. (2) in their article in Journal of Neurology Neurosurgery and Psychiatry (2), which has not been corroborated by any other investigators, to date. This term has been loosely applied here and there and, over time, has become an established diagnosis, which is disturbing. We have been asked about this diagnosis in consultation, from neurologic wards. Extensive literature searches and peer discussions on this topic have failed to document conclusive evidence that this diagnosis exists. Thyroid antibodies have been reported to be commonly prevalent in normal humans, in the range of 3– 4% (3). As such, it would not be uncommon to find them in the context of rare diseases, but a cause and effect relationship should not be described without sufficient scientific evidence. There have been numerous reports, which link the elevated Received October 11, 2000. Address correspondence to: Cary N. Mariash, M.D., Department of Medicine, Medical School, University of Minnesota, 420 Delaware Street SE, Minneapolis, Minnesota 55455.

947

autoantibodies to several conditions like habitual abortions, myasthenia gravis, multiple sclerosis, and hepatitis C; etc. (4 –7). However, this only suggests underlying autoimmunity, not necessarily cause and effect. The patients described in the original and only article on Hashimoto’s encephalitis had neurological symptoms of myoclonus, dementia, cerebellar ataxia suggestive of a neurodegenerative disorder, but lacked Creutzfeld-Jakob disease markers (2). Three of seven patients had antithyroglobulin antibodies, while a different three of seven had microsomal antibodies. The authors reported that all of them responded dramatically to steroids, leading to full recovery, but later explain that their improvement was only temporary with residual symptoms and fluctuating courses. We do not feel compelled to believe that these patients had, indeed, developed an encephalitis from the thyroid autoantibodies, and the authors themselves have reported that they failed to demonstrate any thyroid autoantibody or antigen in the cerebrospinal fluid studies. There are no published articles describing an increased incidence of encephalitis in established cases of autoimmune thyroid disease. There have been several reports of high prevalence of thyroid autoantibodies or thyroid dysfunction among patients with chronic conditions, like hepatitis C. However, a study from Sardinia, Italy, where hepatitis C is endemic, did not show an association between hepatitis C virus and thyroid autoimmunity (7). One should exercise reasonable judgment before making valid assumptions, and, so far, there is no evidence establishing a causative relationship between thyroid autoantibodies and encephalitis. We believe that the diagnosis “Hashimoto’s encephalitis” does not exist and henceforth be restricted from common usage until definite proof has been found that thyroid antibodies do cause encephalitis. Gopinath S. Sunil and Cary N. Mariash Division of Endocrinology and Diabetes Medical School University of Minnesota Minneapolis, Minnesota 55455

References 1. Farooqui I.S, Keston Jones M, Evans M, O’Rahilly S, Hodges JR. 2000 Triple H syndrome: a novel autoimmune endocrinopathy characterized by dysfunction of the hippocampus, hair follicle, and hypothalamic-pituitary-adrenal axis. J Clin Endocrinol Metab. 85:2644 –2648. 2. Seipelt M, Zerr I, Nau R, et al. 1999 Hashimoto’s encephalitis as a differential diagnosis of Creutzfeldt-Jakob disease. J Neurol Neurosurg Psychiatry. 66:172–176. 3. Vanderpump MBJ, Tunbridge WMG, French JM, et al. 1995 The incidence of thyroid disorders in the community: a twenty-year follow-up of the Whickham survey. Clin Endocrinol. 43:55– 68. 4. Kiprov DD, Nachtigall RD, Weaver RC, Jacobson A, Main EK, Garovoy MR. 1996 The use of intravenous immunoglobulin in recurrent pregnancy loss associated with combined alloimmune and autoimmune abnormalities. Am J Reprod Immunol. 36:228 –234. 5. Weissel M, Mayr N, Ziethofer J. 2000 Clinical significance of autoimmune thyroid disease in myasthenia gravis. Endocrinol Diabetes. 108:63– 65. 6. Sakuma R, Fujihara K, Sato N, Mochizuki H, Itoyama Y. 1999 Optic-spinal form of multiple sclerosis and anti-thyroid autoantibodies. J Neurol. 246:449 – 453. 7. Loviselli A, Oppo A, Velluzzi F, et al. 1999 Independent expression of serological markers of thyroid autoimmunity and hepatitis virus C infection in the general population: results of a community-based study in northwestern Sardinia. J Endocrinol Invest. 22:660 – 665.

Paraneoplastic Hypercortisolism as a Risk Factor for Severe Infections in Patients with Malignant Diseases To the editor: The work by Sarlis et al. (1) confirms that paraneoplastic hypercortisolism constitutes a risk factor for severe bacterial and opportunistic infections even when patients with small cell lung cancer (SCLC) are excluded. They suggest that inclusion of SCLC cases, as in our earlier paper (2), cloud the question because “these patients have tumor-related Received January 21, 2000. Address correspondence to: Rene Vassilopoulou-Sellin, The University of Texas M.D. Anderson Cancer Center, Division of Medicine, Section of Endocrine Neoplasia and Hormonal Disorders, 1515 Holcombe Boulevard, Box 015, Houston, Texas 77030-4095.

948

JCE & M • 2001 Vol. 86 • No. 2

LETTERS TO THE EDITOR

features that could confound the association.” Indeed, patients with carcinoid or other neuroendocrine tumors are clinically different from patients with SCLC, and both groups are at increased risk. I would like to emphasize two clinical points about these patients: 1) the superimposition of additional immune compromise, such as follows the application of myelosuppressive cytotoxic chemotherapy creates a highly lethal combination (even in non-SCLC patients). This can be avoided, since 2) correction of hypercortisolism can be achieved promptly and safely with the administration of metyrapone (typically 2– 4 g in divided doses will normalize serum cortisol within 48 h, if not sooner). At our institution, patients with paraneoplastic hypercortisolism begin metyrapone promptly, whether the treatment plan includes cytotoxic chemotherapy, surgery, or conservative surveillance; this provides the patients with gratifying clinical improvement and decreases complications. When cytotoxic chemotherapy seems to be an urgent priority (as in most patients with SCLC), our colleagues in medical oncology try to avoid myelosuppressive regimens until cortisol has been normalized for a few weeks. Rena Vassilopoulou-Sellin, M.D. University of Texas, M.D. Anderson Cancer Center Houston, Texas 77030-4095

References 1. Sarlis NJ, Chanock SJ, Nieman LK. 2000 Cortisolemic indices predict severe infections in Cushing’s syndrome due to ectopic production of adrenocorticotropin. J Clin Endocrinol Metab. 85:42– 47. 2. Dimopoulos MA, Fernandez JF, Samaan NA, Holoye PY, VassilopoulouSellin Rena. 1992 Paraneoplastic Cushing’s syndrome as an adverse prognostic factor in patients who die early with small cell lung cancer. Cancer. 69:66 –71.

Is Hydrocortisone Clearance 50% Slower in the Evening Than in the Morning? To the editor: The paper entitled “Metabolic effects of short-term elevations of plasma cortisol are more pronounced in the evening than in the morning” by Plat et al. (1) was of great interest to me, and the authors’ findings are extremely important. I would like to comment, however, on the hydrocortisone clearance (CL) findings and the method of its estimation. In their study, the authors define hydrocortisone CL as “the time required to achieve a 50% reduction from the maximum difference in plasma cortisol levels between the hydrocortisone and placebo conditions,” and they find hydrocortisone CL to be 50% slower in the evening than in the morning (156 ⫾ 15 min vs. 109 ⫾ 13 min, P ⬍ 0.003; page 3087, paragraph 5). CL describes the efficiency of irreversible elimination of a drug from the body. It is actually defined as “the volume of blood cleared of the drug per unit time,” and the units are, thus, volume per time, usually liters per hour (L/h) or milliliters per minute (mL/min). Another definition of CL is that it is the constant relating the concentration of the drug in the plasma to the elimination rate: Elimination rate (mg/h) ⫽ CL (L/h) ⴱ plasma drug concentration (mg/L) It is apparent that for a given CL the elimination rate varies directly with the plasma drug concentration (2). The best method of calculating CL is to give a single iv dose of the drug (to ensure 100% bioavailability of the drug) and draw frequent blood samples for measurement of the plasma drug concentrations until they reach undetectable levels. Then CL can be calculated by dividing the dose administered iv by the area under the drug concentration vs. Received May 18, 2000. Address correspondence to: Dr. Evangelia Charmandari, Pediatric and Reproductive Endocrinology Branch, National Institute of Child Health and Human Development, National Institutes of Health, 10 Center Drive, Building 10, Suite 9D42, Bethesda, Maryland 20892-1583. E-mail: [email protected].

time curve (AUC) from time ⫽ 0 min to infinite time (2– 4):

CL 共L/h兲⫽dose (iv) (mg)/AUC0 –inf (mg ⴱ h/L) Alternatively, when the volume of distribution (V) and half-life (t1/2) of a drug have been estimated, CL can be calculated using the formula:

t1/2 ⫽ 0.693 ⴱ V/CL, where 0.693 is the natural log of 2 (3, 5). Based on the above principles of pharmacokinetics it becomes clear that the results that the authors obtained on the hydrocortisone CL may not be reliable, because they assumed CL to be the time required to achieve 50% reduction from the maximum difference in plasma cortisol concentrations achieved after administration of hydrocortisone and placebo. But do they refer to hydrocortisone half-life rather than CL, hence the units are minutes and not L/h or mL/min? If this is the case, then half-life should be estimated as the time required for the plasma cortisol concentrations achieved in individual patients to fall by half. Plasma cortisol concentrations should be natural log (loge) transformed before further statistical analysis. The relationship between time and the transformed data can be described by linear regression. Then, half-life can be calculated by dividing 0.693 (loge 2) by the slope of the regression line that represents the elimination constant rate (5). Like CL, half-life is also most accurately estimated following iv administration of the drug. I would be interested in having the authors’ views and explanation for their estimation and findings. Dr. Evangelia Charmandari London Centre for Paediatric Endocrinology University College London London W1N 8AA, United Kingdom

References 1. Plat L, Leproult R, L’Hermite-Baleriaux M, et al. 1999 Metabolic effects of short-term elevations of plasma cortisol are more pronounced in the evening than in the morning. J Clin Endocrinol Metab. 84:3082–3092. 2. Birkett DJ. 1998 Clearance. In: Pharmacokinetics made easy. Roseville, Australia: McGraw-Hill Australia Ltd.; 1– 8. 3. Rowland M, Tozer TN. 1980 Hepatic clearance and elimination. In: Clinical pharmacokinetics: concepts and applications. Philadelphia: Lea & Febiger; 65–78. 4. Rowland M, Tozer TN. 1980 Clearance and renal excretion. In: Clinical pharmacokinetics: concepts and applications. Philadelphia: Lea & Febiger; 48 – 64. 5. Birkett DJ. 1998 Half life. In: Pharmacokinetics made easy. Roseville, Australia: McGraw-Hill Australia Ltd.; 16 –24.

Author’s Response: Is Hydrocortisone Clearance 50% Slower in the Evening Than in the Morning? To the editor: We thank Dr. Charmandari for her positive comments regarding our article entitled “Metabolic effects of short-term elevations of plasma cortisol are more pronounced in the evening than in the morning.” The objective of the study described in this article was not the measurement of hydrocortisone clearance, and, therefore, the protocol was not designed to determine hydrocortisone clearance according to the textbook definitions of clearance accurately quoted by Dr. Charmandari. Nevertheless, when faced with the remarkable morning vs. evening difference in the rate of decline of cortisol levels posthydrocortisone illustrated in Fig. 5 of our article, we felt compelled to quantify this difference. Thus, we estimated the “apparent rate of hydrocortisone clearance” as “the time to achieve a 50% reduction from the maximum difference in plasma cortisol levels between the hydrocortisone and placebo conditions.” The word “apparent” was, unfortunately, omitted from the definition provided in the last paragraph of page 3087 but was repeatedly used in the last paragraph of the Results section on page 3089. We agree with Dr. Received November 15, 2000. Address correspondence to: Eve Van Cauter, M.D., Department of Medicine—MC 1027, University of Chicago, 5841 South Maryland Avenue, Chicago, Illinois 60637.

LETTERS TO THE EDITOR Charmandari that our calculation would have been better referred to by the term “apparent half-life” rather than “apparent clearance.” We also agree that a rigorous study examining diurnal variations in hydrocortisone clearance using a protocol specifically designed for that purpose is warranted. Eve Van Cauter Department of Medicine University of Chicago Chicago, Illinois 60637

Usefulness of the Diagnostic Total Body Scan in Differentiated Thyroid Cancer To the editor: In the paper by Cailleux et al. (1), in the January 2000 issue of the journal, 256 patients with differentiated thyroid cancer were reported. A total or near-total thyroidectomy was done, followed by 3.7 GB radioiodine for remnant ablation a month later. None had uptake outside the thyroid bed at this time. Thereafter, T4 suppression was instituted, and 6 –12 months later a ‘control’ total body scan (TBS) was performed after T4 withdrawal with 2–5 mCi radioiodine. There was no uptake at all in 236 patients and low uptake in 20 patients. Serum thyroglobulin was done at the time of T4 withdrawal and was more than 10 ng/mL in 15 patients (group A), 1–10 ng/mL in 31 patients (group B), and undetectable in 210 patients (group C). There was uptake in the thyroid bed on the control TBS in 3 group A patients, 2 group B patients, and 15 group C patients. Of the eight total recurrences, five occurred in group A, one in group B, and two in group C (Table 2 in the paper). The authors report that there seems to be no relationship between detectable thyroglobulin level and the presence of uptake in the thyroid bed on the control (diagnostic) TBS. However, there was no analysis of the relationship between failure of remnant ablation (uptake on diagnostic TBS) and recurrence. It is well known that postoperative ablation of functioning thyroid tissue reduces the long-term risk of recurrence and death. This beneficial effect results from the destruction of potentially malignant cells or occult multifocal disease that may occur in up to 30% of patients with papillary tumors (2– 8). The authors conclude on the lack of usefulness of the TBS based on the correlation with thyroglobulin levels and not based on recurrence. It is well known that the specificity of thyroglobulin as a tumor marker is increased if the remnant is ablated (9, 10). Although they did not indicate recurrence in relation to uptake on the control TBS, it is clear from the data in Table 2 of the paper that of the eight recurrences a maximum of six could have had uptake on the diagnostic TBS. The range, therefore, for scan-positive patients with recurrence is from zero to six patients of eight. If a 2 ⫻ 2 table analysis is done with ablation status as the risk factor and recurrence as the disease status, the odds of recurrence given a positive control TBS is 50, 26, 14, 8, 4, and 2 times that given a negative control TBS assuming that of the patients with recurrences, six, five, four, three, two, and one patient, respectively, had a positive scan. The odds ratios remain statistically significant even if just three patients had a positive scan and if less than three were positive there was still a trend toward recurrence in scan-positive patients. Therefore, Cailleux et al. (1) do not have much grounds for concluding that the diagnostic (control) TBS had “poor usefulness” unless none of the patients with recurrences had a positive scan, and, even then, this would have to be confirmed in a bigger trial. The key question for the authors, therefore, is how many of these eight recurrences had a positive diagnostic (control) TBS? Suhail A. R. Doi Division of Endocrinology and Metabolic Medicine Mubarak Hospital Jabriya 32052, Kuwait

References 1. Cailleux AF, Baudin E, Travagli JP, Ricard M, Schlumberger M. 2000 Is diagnostic iodine-131 scanning useful after total thyroid ablation for differentiated thyroid cancer? J Clin Endocrinol Metab. 85:175–178.

Received June 12, 2000. Address correspondence to: Suhail A. R. Doi, MRCP, Division of Endocrinology and Metabolic Medicine, Mubarak Hospital, P.O. Box 43787, Jabriya 32052, Kuwait.

949

2. Mazzaferri EL, Jhiang SM. 1994 Long-term impact of initial surgical and medical therapy on papillary and follicular thyroid cancer. Am J Med. 97:418 – 428. 3. Samaan NA, Schultz PN, Hickey RC, et al. 1992 The results of various modalities of treatment of well differentiated thyroid carcinomas: a retrospective review of 1599 patients. J Clin Endocrinol Metab. 75:714 –720. 4. Simpson WJ, Panzarella T, Carruthers JS, Gospodarowicz MK, Sutcliffe SB. 1988 Papillary and follicular thyroid cancer: impact of treatment in 1578 patients. Int J Radiat Oncol Biol Phys. 14:1063–1075. 5. Tubiana M, Schlumberger M, Rougier P, et al. 1985 Long-term results and prognostic factors in patients with differentiated thyroid carcinoma. Cancer. 55:794 – 804. 6. DeGroot LJ, Kaplan EL, McCormick M, Straus FH. 1990 Natural history, treatment and course of papillary thyroid carcinoma. J Clin Endocrinol Metab. 71:414 – 424. 7. Mazzaferri EL. 1991 Treating differentiated thyroid carcinoma: where do we draw the line? Mayo Clin Proc. 66:105–111. 8. Wong JB, Kaplan MM, Meyer KB, Pauker SG. 1990 Ablative radioactive iodine therapy for apparently localized thyroid carcinoma. A decision analytic perspective. Endocrinol Metab Clin North Am. 19:741–760. 9. Sisson JC. 1983 Applying the radioactive eraser: I-131 to ablate normal thyroid tissue in patients from whom thyroid cancer has been resected. J Nucl Med. 24:743–745. 10. Goolden AW. 1985 The indications for ablating normal thyroid tissue with 131I in differentiated thyroid cancer. Clin Endocrinol (Oxf). 23:81– 86.

Authors’ Response: Usefulness of the Diagnostic Total Body Scan in Differentiated Thyroid Cancer To the editor: The letter by Dr. Doi gives us the opportunity to clarify several points concerning the follow-up of thyroid cancer patients (1): 1. A residual uptake in the thyroid bed cannot be considered as failure of ablation when it is visible but so low that it cannot be quantified (that is ⬍0.3% of the administered activity); furthermore, such a low uptake cannot be considered as an evidence of disease when there is no abnormality at neck palpation or ultrasonography, and when serum thyroglobulin (Tg) level is undetectable. 2. Beneficial effects of 131I ablation have been demonstrated only in patients at high risk of recurrence or cancer-related death (2, 3). In our Institution, ablative doses of 131I are given only in patients who may benefit from this treatment modality. 3. Remnant ablation increases the specificity of serum Tg measurement. However, even after total thyroidectomy only, as performed in low-risk patients, serum Tg level has a high prognostic value (4). This study also demonstrates that a low uptake in the thyroid bed cannot be considered as the origin of circulating Tg. 4. Uptake in the thyroid bed was visible in 20 patients, but no recurrence occurred in the thyroid bed. Among these 20 patients, only 1 had a clinical recurrence during the subsequent follow-up, and among the other 236 patients without visible uptake in the thyroid bed, seven recurrences occurred; a total body scan (TBS) performed at the time of recurrence with a high dose of 131I in six patients showed uptake at the site of the recurrence in four. As stated in the text, there was no focus of uptake outside the thyroid bed on any of the 256 diagnostic 131I TBS, and these scans were, therefore, considered negative; they did not even have any predictive value on the risk of subsequent recurrence. We, therefore, maintain our statement that diagnostic control 131I total body scanning has a poor usefulness. 5. Finally, this study confirms the high predictive value of scanning after the administration of a large activity of 131I, even if it is given for ablation (5, 6). In this series of 256 patients with no focus of uptake outside the thyroid bed on this high-dose TBS, the control diagnostic TBS only confirmed previous findings. Our data suggest that 131I TBS should be performed only if there is some suspicion of residual disease, that is in general an elevated serum Tg level, and when a TBS is performed, a high dose of 131I should be given.

Received November 15, 2000. Address correspondence to: Martin Schlumberger, M.D., University Paris-Sud, Institut Gustave Roussy, 39, Rue C. Desmoulins, Villejuif Cedex 94 805, France.

950

JCE & M • 2001 Vol. 86 • No. 2

LETTERS TO THE EDITOR A. F. Cailleux, E. Baudin, and M. Schlumberger Institut Gustave Roussy 94805 Villejuif Cedex, France

References 1. Cailleux AF, Baudin E, Travagli JP, Ricard M, Schlumberger M. 2000 Is diagnostic iodine-131 scanning useful after total thyroid ablation for differentiated thyroid cancer? J Clin Endocrinol Metab 85:175–178. 2. Schlumberger MJ. 1998 Papillary and follicular thyroid carcinoma. N Engl J Med. 338:297–306. 3. Wartofsky L, Sherman SI, Gopal J, Schlumberger M, Hay ID. 1998 Therapeutic controversy: the use of radioactive iodine in patients with papillary and follicular thyroid cancer. J Clin Endocrinol Metab. 83:4195– 4203. 4. Schlumberger M, Baudin E. 1998 Serum thyoglobulin determination in the follow-up of patients with differentiated thyroid carcinoma. Eur J Endocrinol. 138:249 –252. 5. Tenenbaum F, Corone C, Schlumberger M, Parmentier C. 1996 Thyroglobulin determination and postablative iodine-131 total body scan after total thyroidectomy for differentiated thyroid carcinoma in patients with no evidence of disease. Eur J Cancer. 32A:1262. 6. Sherman SI, Tielens ET, Sostre S, Wharam Jr MD, Ladenson PW. 1994 Clinical utility of posttreatment radioiodine scans in the management of patients with thyroid carcinoma. J Clin Endocrinol Metab. 78:629 – 634.

Dietary Protein and Fiber Intake and Sex HormoneBinding Globulin To the editor: Recently, Longcope et al. (1) reported findings on the importance of dietary components in regulating the circulating levels of sex hormonebinding globulin (SHBG). The authors analyzed a large cross-sectional sample of data from the Massachusetts Male Aging Study. After controlling for a number of confounding variables (e.g. sex steroids, age, BMI) fiber intake was found to be significantly positively correlated to serum SHBG concentration, whereas protein intake showed a clear negative association with SHBG. Longcope et al. (1) proposed the following plausible mechanism by which protein intake can be a controlling factor on SHBG concentration: protein ingestion is known to increase insulin secretion, and insulin in turn has been shown to inhibit hepatic SHBG production. However, dietary carbohydrate intake, another stimulus for insulin release, did not show a significant association with SHBG. Therefore, the authors argued that the relationship of protein intake to SHBG levels probably involves more than only an effect on insulin. But from their data it was not possible to identify what that could be. Recently, we have shown in a controlled experimental diet study that a lactovegetarian low protein diet, which is characterized by a diminished overall insulin secretion, can reduce adrenocortical activity as reflected in a reduced urinary dehydroepiandrosterone sulfate and free cortisol excretion (2). Because glucocorticoids have an independent potential to suppress circulating SHBG levels (3, 4), the reduction in adrenocortical activity with low protein intake could be an additional causal factor involved in the regulation of SHBG. With regard to the role of insulin as a regulator of circulating SHBG, it is of interest that apart from dietary protein also dietary fiber intake affects the insulin secretory dynamics. Not only low protein diets but also high fiber intakes attenuate postmeal insulin secretory peaks (5, 6). Thus, it is intelligible why a vegetarian nutrition, although providing very high carbohydrate intakes, can result in very low overall insulin secretion rates (2). Our corresponding diet study (comprising 5-day diet periods) proved, indeed, long enough to detect clear effects of nutrition on adrenal corticosteroid output but not long enough to see clear changes in SHBG levels (2). Therefore, controlled dietary long-term intervention studies are required to validate whether the discussed diet effects represent sustained changes in insulin secretion, adrenocortical activity, and, finally, hepatic SHBG production. Taken together, the increase in circulating SHBG with low protein and high fiber intakes observed by Longcope et al. (Ref. 1 and other authors) could be a result of the combined effects of reduced insulin secretion and Received February 29, 2000. Address correspondence to: Thomas Remer, Ph.D., Research Institute of Child Nutrition, Forschungsinstitut fuer Kindererneahrung, Heinstueck 11, 44225 Dortmund, Germany.

reduced glucocorticoid secretion. With regard to practice, reductions in serum SHBG levels will lead to elevated bioavailable testosterone (and vice versa). However, recommendations to decrease SHBG with high protein diets in elderly men (who have age-related low levels of bioavailable testosterone) are problematic because high protein diets can induce marked increases in renal acid loads (7) and renal net acid excretion capacity is reduced in aging subjects (8). Thomas Remer Research Institute of Child Nutrition 44225 Dortmund, Germany

References 1. Longcope C, Feldman HA, McKinlay JB, Araujo AB. 2000 Diet and sex hormone-binding globulin. J Clin Endocrinol Metab. 85:293–296. 2. Remer T, Pietrzik K, Manz F. 1998 Short-term impact of a lactovegetarian diet on adrenocortical activity and adrenal androgens. J Clin Endocrinol Metab. 83:2132–2137. 3. Davis SR, Burger HG. 1996 Androgens and the postmenopausal woman. J Clin Endocrinol Metab. 81:2759 –2763. 4. Hautanen A, Sarna S, Pelkonen R, Adlercreutz H. 1993 Serum sex hormonebinding globulin, cardiovascular risk factors, and adrenal cortisol responses to dexamethasone and corticotropin: Metabolism. 42:870 – 874. 5. Ranganathan S, Champ M, Pechard C, et al. 1994 Comparative study of the acute effects of resistant starch and dietary fibers on metabolic indexes in men. Am J Clin Nutr. 59:879 – 883. 6. Reiser S. 1987 Metabolic effects of dietary pectins related to human health. Food Technol. 41:91–99. 7. Remer T, Manz F. 1995 Potential renal acid load of foods and its influence on urine pH. J Am Diet Assoc. 95:791–797. 8. Frassetto LA, Morris RC, Sebastian A. 1996 Effect of age on blood acid-base composition in adult humans: role of age-related renal functional decline. Am J Physiol. 271:F1114 –F1122.

Author’s Response: Dietary Protein and Fiber Intake and Sex Hormone-Binding Globulin To the Editor: We thank Dr. Remer for his comments on our article. Glucocorticoids can increase insulin resistance of peripheral tissues and lead to higher insulin levels. The decrease in glucocorticoid secretion that he noted could have resulted in decreased resistance to insulin, lower insulin levels, and, hence, higher sex hormone-binding globulin levels. Christopher Longcope Department of Obstetrics and Gynecology University of Massachusetts Medical School Worcester, Massachusetts 01655

Metabolic Abnormalities in Patients with Adrenal Incidentaloma To the editor: Rossi et al. (1) recently reported in the journal an unexpectedly high prevalence of subtle autonomous cortisol secretion, associated with high occurrence of hypertension, diabetes mellitus, elevated lipids, and diffuse obesity in patients with adrenal incidentaloma. The occurrence of diabetes was higher in the group of subclinical Cushing’s syndrome (CS) patients (50% vs. 26%). Glucose intolerance was further diagnosed in 2 of 12 CS patients. Arterial hypertension occurred in 91.8% of patients with subclinical CS and in 34% of the remaining patients. In the final part of the manuscript, the authors claim that “. . .the high prevalence of hypertension and metabolic disorders found in our study is to be confirmed by others. . . .” Rossi et al. (1) diagnosed type 2 diabetes mellitus and glucose intolerance by means of fasting glucose values (2). In the last years, we coordinated a multicenter study of 64 consecutive patients harboring incidental adrenal adenomas (3). The prevalence of glucose intolerance or diabetes mellitus was determined through an oral glucose tolerance test. The prevalence of altered glucose tolerance was 39 of 64 Received October 27, 2000. Address correspondence to: Christopher Longcope, M.D., Department of Obstetrics and Gynecology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, Massachusetts 01655.

LETTERS TO THE EDITOR (61%), well above the prevalence of type 2 diabetes mellitus described for a population of similar age in our area (4). Two-hour serum glucose levels were significantly higher in patients with arterial hypertension. In the last 3 yr, we have followed 62 consecutive patients with adrenal incidentaloma using a similar protocol in our single center. In this series (34 men and 28 women; mean age, 59 ⫾ 10 yr; body mass index, 28 ⫾ 4 kg/m2), we have found that 66% of the patients exhibited altered glucose tolerance, in a proportion very similar to that found in the multicenter study (24% of the patients showed glucose intolerance, and 42% diabetes mellitus), and 48% had arterial hypertension. According to our experience, we proposed that patients with incidental adrenal tumours should be tested for glucose tolerance (3). Up to 90% of patients with cortisol hypersecretion are glucose intolerant or show hyperinsulinemia, both basally and following stimulation by glucose (5, 6). Since 1995, we (3, 7–10) and others (11) have elaborated several hypotheses on the relationship between disturbed glucose tolerance in adrenal incidentaloma. In fact, limited data have associated a higher prevalence of nonhypersecretory adrenocortical adenomas in diabetes mellitus (2- to 5-fold increased prevalence of adrenal adenomas among diabetic patients over nondiabetics) (12, 13), obesity, and arterial hypertension (reviewed in Ref. 14), all components of the metabolic syndrome, in which insulin resistance is believed to play a key role (15). Only the oral glucose tolerance test unravels glucose intolerance in patients with adrenal incidentaloma because even in CS severe fasting hyperglycemia is uncommon (5). We agree with Rossi et al. (1) in that metabolic consequences of more subtle but likely, prolonged effects of hypercortisolism on the cardiovascular system in subclinical CS might also be present (1). Jose´-Manuel Ferna´ndez-Real, Jose Gonzalbez, and Wifredo Ricart Unitat de Diabetologia, Endocrinologia i Nutricio´ University Hospital of Girona “Dr Josep Trueta” 17007 Girona, Spain

References 1. Rossi R, Tauchmanova L, Luciano A, et al. 2000 Subclinical Cushing’s syndrome in patients with adrenal incidentaloma: clinical and biochemical features. J Clin Endocrinol Metab. 85:1440 –1448. 2. Expert Committee. 1997 Report of the Expert Committee on the diagnosis and classification of diabetes mellitus. Diabetes Care. 20:1183–1197. 3. Ferna´ndez-Real JM, Ricart W, Simo´ R, et al. 1998 Study of glucose tolerance in consecutive patients harbouring incidental adrenal tumours. Clin Endocrinol. 49:53– 61. 4. Castell C, Tresserras R, Serra J, Goday A, Lloveras G, Salleras L. 1999 Prevalence of diabetes in Catalonia (Spain): an oral glucose tolerance test-based population study. Diabetes Res Clin Pract. 43:33– 40. 5. Modigliani E, Strauch G, Luton JP. 1970 Effect of glucose and arginine on insulin secretion in Cushing’s syndrome. Diabetologia. 6:8 –12. 6. Urbanic RC, George JM. 1981 Cushing’s disease—18 years’ experience. Medicine. 60:14 –24. 7. Ferna´ndez-Real JM, Ricart W. 1996 Relationship between fasting insulin resistance index and adrenal sensitivity to ACTH in incidental adrenal adenoma. Endocrinologia. 43(Suppl 1):7– 8. 8. Ricart W, Ferna´ndez-Real JM. 1995 Hyperinsulinemia and hyperglycemia in incidental adrenal adenoma. Endocrinologia. 42(Suppl):18 –19. 9. Ricart W, Ferna´ndez-Real JM. 1996 Carbohydrate metabolism and ACTH secretion as tools of study of incidental adrenal adenomas. Endocrinologia. 43:2–5. 10. Ferna´ndez-Real JM, Ricart W, Simo´ R. 1994 Preclinical Cushing’s syndrome: report of three cases and literature review. Horm Res. 41:230 –235. 11. Reincke M, Fa␤nacht M, Va¨th S, Mora P, Allolio B. 1996 Adrenal incidentalomas: A manifestation of the metabolic syndrome? Endocrine Res. 22:757–761. 12. Russi S, Blumenthal HT, Gray SH. 1945 Small adenomas of the adrenal cortex in hypertension and diabetes. Arch Intern Med. 76:284 –291. ¨ stberg G, Ho¨kfelt B. 1968 On the prevalence of adrenocortical 13. Hedeland H, O adenomas in an autopsy material in relation to hypertension and diabetes. Acta Med Scand. 84:211–214. 14. Kloos RT, Gross MD, Francis IR, Korobkin M, Shapiro B. 1995 Incidentally discovered adrenal masses. Endocr Rev. 16:460 – 484. 15. Reaven GM. 1988 The Banting Lecture 1988: role of insulin resistance in human disease. Diabetes. 37:1595–1607.

Received July 11, 2000. Address correspondence to: J. M. Ferna´ndezReal, M.D., Department of Endocrinology, Hospital de Girona, Carretera de Francia s/n, 17007 Girona, Spain.

951

Authors’ Response: Metabolic Abnormalities in Patients with Adrenal Incidentaloma To the editor: In a single center study on incidental adrenal adenoma published recently (1), we reported on unexpectedly high occurrence of cardiovascular risk factors in subclinical Cushing’s syndrome (CS), among which hypertension (92%), obesity (50%), type 2 diabetes mellitus (DM) (42%), and alteration in serum lipid concentrations (50%). Their frequency was higher than that reported in the Italian multicentric study published just 2 months before ours (2), where hypertension, diabetes and obesity were reported in 41%, 8%, and 38% of subclinical CS patients, respectively. Moreover, frequencies found in our study were higher than those reported in CS (3, 4), and it is well known that patients with CS exhibit frequently glucose intolerance but less commonly overt DM (3, 4). Although the patients with subclinical CS were generally older than those with overt CS (1, 2), we hypothesized an overestimation of cardiovascular risk factors in our center, related to the small number of subjects evaluated. However, in 24 subclinical CS patients evaluated until now, the impaired glucose tolerance was found in similar proportion (62.5%) to those previously reported (1). Ferna´ndez-Real et al. (5) had previously found a frequency of impaired glucose tolerance in adrenal incidentalomas similar to ours (58.3%), in both a multicentric (61%) and a single center study (66%). In the study by Ferna´ndez-Real et al. (5), the diagnosis of DM was performed according to the National Diabetes Data Group (NDDG) (6) by the oral glucose tolerance test (OGTT). The NDDG criteria considered fasting glucose values above 140 mg/dL (conversion factor to SI units, 0.056) as suggestive of DM. On the other hand, fasting glucose values above 126 mg/dL were diagnostic of DM in our study, according to the new criteria of the American Diabetes Association (7). Some patients who were not affected by DM according to the first criteria (NDDG), should be diabetic on the basis of the new one (American Diabetes Association). However, 3 of 12 subclinical CS patients (1) were already on treatment for DM, 1 by multiple insulin administration and 2 by glucose-lowering agents; in two patients DM was first diagnosed during the evaluation for incidentaloma, on the basis of more than two fasting glucose values above 126 mg/dL. Their values were between 126 and 140 mg/dL and would, thus, not be considered as diabetic with the NDDG criteria. Nevertheless, the best diagnostic approach to DM is still a matter of debate. The gold standard has been reported to be the OGTT, but this test is not perfectly reproducible and its results are liable to be influenced by a variety of factors, including diet, previous fasting duration, the time of day the test is performed, and physical activity (8 –10). Some authors have found fasting glucose concentration to be more stable (11). However, in the above mentioned study (1), the OGTT was performed in nondiabetic patients and glucose impaired tolerance was detected in two of them. Moreover, four of five patients with DM had a positive family history for type 2 DM, as described previously for CS. Recent evidence suggests that incidental adrenal adenomas are characterized by variable alteration in cortisol secretion rates, degree of autonomous cortisol production being a continuum from slightly abnormal to completely pathological and biochemically similar to the pattern of overt CS (1, 12). The diagnosis of subclinical CS represents an arbitrary separation of the category of subjects with a greater degree of autonomous cortisol production. In conclusion, we agree with Ferna´ndez-Real et al. (5) that all nondiabetic patients with incidental adrenal tumors should be tested for glucose tolerance, to decide the best therapeutic strategy. Riccardo Rosse and Libuse Tauchmanova` Department of Molecular and Clinical Endocrinology and Oncology University “Federico II” in Naples Naples 80131, Italy Received October 27, 2000. Address correspondence to: Riccardo Rossi, M.D., Department of Endocrinology, Oncology, and Molecular, Universita´ di Napoli “Federico II,” piazza degli Artisti n.17, Napoli 80129, Italy.

952

LETTERS TO THE EDITOR References

1. Rossi R., Tauchmanova` L, Luciano A, et al. 2000 Subclinical Cushing’s syndrome in patients with adrenal incidentaloma: clinical and biochemical features. J Clin Endocrinol Metab. 85:1440 –1448. 2. Mantero F, Terzolo M, Arnaldi G, et al. 2000 A survey on adrenal incidentaloma in Italy. J Clin Endocrinol Metab. 85:637– 644. 3. Ross EJ, Linch DC. 1982 Cushing’s syndrome-killing disease: discriminatory value of signs and symptoms aiding early diagnosis. Lancet. ii:646 – 649. 4. Urbanic RC, George JM. 1981 Cushing’s disease—18 years experience. Medicine. 60:14 –24. 5. Ferna´ndez-Real JM, Ricart W, Simo` R, et al. 1998 Study of glucose tolerance in consecutive patients harbouring incidental adrenal tumours. Clin Endocrinol. 49:53– 61. 6. National Diabetes Data Group. 1979 Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes. 28:1039 –1057.

JCE & M • 2001 Vol. 86 • No. 2

7. Expert Committee. 1997 Report of the Expert Committee on the diagnosis, and classification of diabetes mellitus. Diabetes Care. 20:1183–1197. 8. Olefsky JM, Reaven GM. 1974 Insulin and glucose responses to identical oral glucose tolerance tests performed forty-eight hours apart. Diabetes. 23:449 – 453. 9. Kosaka K, Mizuno Y, Kuzuga T. 1966 Reproducibility of the oral glucose tolerance test and the rice-meal test in mild diabetes. Diabetes. 15:901–904. 10. Wilkerson HLC, Butler FK, Francis JOS. 1960 The effects of prior carbohydrate intake on the oral glucose tolerance test. Diabetes. 9:386 –391. 11. Rushforth NB, Miller M, Bennett PH. 1979 Fasting and two-hour post-load glucose levels for the diagnosis of diabetes. Diabetologia. 16:373–379. 12. Sartorio A, Conti A, Ferrero S, et al. 1998 Evaluation of markers of bone and collagen turnover in patients with active and preclinical Cushing’s syndrome and in patients with adrenal incidentaloma. Eur J Endocrinol. 138:146 –152.