Necrolytic migratory erythema (synonymous with superficial necrolytic dermatitis) is a rare condition first described in humans in 1942.1 Although there are numerous metabolic conditions associated with NME, it has been most commonly affiliated with glucagonoma in humans.2–4 However, this association has rarely been proven in dogs with HCS.4,5 Pathological mechanisms of cutaneous lesions in NME are speculated to reflect depletion of plasma amino acids required for collagen synthesis necessary to repair sites of mechanical trauma (ie, pressure points).2 Histologically, cutaneous lesions are characterized by epidermal edema and degeneration of the stratum granulosum and stratum spinosum, and they develop in areas exposed to mechanical trauma (eg, the muzzle, mucocutaneous junctions, and contact surfaces of appendages including the elbow joints, tibiotarsal joints, and paw pads). Lesions are erythematous and ulcerative, cause pain, and are associated with crusting exfoliation. The most serious cutaneous complication is development of fissured, excruciatingly painful paw pads that are subject to opportunistic infections. A severe distinctive degenerative vacuolar hepatopathy usually parallels development of cutaneous lesions in dogs with HCS (so-called hepatocutaneous hepatopathy), which can progress to liver failure.5 In dogs, HCS associated with cutaneous lesions carries a poor prognosis, commonly with a survival duration of 6 months after initial diagnosis.6,7 Currently, the most effective treatment entails parenteral and oral administration of amino acids. Nevertheless, there is usually only short-lived remission.7 Intravenous administration of amino acids combined with lipid emulsions has been proposed as a therapeutic strategy but has not yet been thoroughly investigated.8,9
The genesis of hypoaminoacidemia in HCS remains enigmatic. Evidence that hyperglucagonemia is a driving mechanism is sparse.10 However, this may be complicated by the lack of a standard diagnostic method for measuring glucagon concentrations because multiple molecular glucagon moieties exist in plasma.11 Consequently, increased plasma concentrations of unmeasured glucagon isoforms have not been rigorously discounted in HCS. Interest in glucagon as a central component of the pathological mechanism derives from experiments in which hyperglucagonemia provokes hypoaminoacidemia, develops in hepatectomized animals, and accompanies liver failure in humans.4 These observations suggest that liver-dependent factors prevent hyperglucagonemia or that the liver modulates glucagon activity or catabolism. However, dogs with cutaneous lesions may not have liver manifestations, and similarly, in the experience of one of the authors (SAC), dogs may have evidence of classic HCS hepatopathy before the manifestation of cutaneous lesions. This, along with a lack of data clarifying pathogenic involvement of hyperglucagonemia in dogs with HCS, has led to the consideration of alternative pathological mechanisms for the development of hypoaminoacidemia.
A small number of gene mutations in humans have been associated with cutaneous lesions similar to those of NME. These disorders are caused by defects in renal amino acid transporters or defects in hydrolysis of imidodipeptides (dipeptides of proline or hydroxyproline) that result in amino acid wasting in the urine. Hartnup disease is 1 example, whereby massive urinary loss of tryptophan culminates in niacin deficiency.12,13 Another example is LPI caused by a defective amino acid transporter for lysine, ornithine, and arginine.14,15 Erythematous skin lesions develop in humans with these disorders, with some patients developing degenerative hepatic lesions.16,17 The purpose of the study reported here was to investigate amino acid concentrations in urine and plasma in a cohort of dogs with HCS to determine whether aminoaciduria is a plausible explanation for hypoaminoacidemia.
Materials and Methods
Animals
Twenty client-owned dogs of various breeds and ages were included in the study; owner consent was obtained for inclusion of all dogs in the study. Analyses were performed on surplus biological samples obtained for routine diagnostic testing and in accordance with Institutional Animal Care and Use Committee guidelines of Cornell University.
Hepatocutaneous syndrome was definitively diagnosed in 18 dogs on the basis of examination of liver biopsy specimens (n = 12), skin biopsy specimens (4), and liver and skin biopsy specimens (2); HCS was presumptively diagnosed in 2 dogs on the basis of classic cutaneous lesions grossly examined by a board-certified veterinary dermatologist, results of clinicopathologic testing, and classic HCS features on hepatic ultrasonographic images.
Amino acid analyses were performed on unused portions of blood and urine obtained for routine diagnostic testing. Urine samples from 3 healthy control dogs (2 dogs with no abnormalities and 1 dog with well-controlled diabetes mellitus; all 3 dogs were > 5 years old) were used for amino acid analyses performed to ascertain the validity of historical reference data. Body condition score18 (scale, 1 to 9) and muscle condition score19 (scale, 0 to 3; 0 = severe muscle wasting, 1 = moderate muscle wasting, 2 = mild muscle wasting, and 3 = normal musculature) were reported by attending clinicians for each of the 20 affected dogs.
Routine diagnostic testing
Routine baseline assessments included a CBC, biochemical analysis, urinalysis, and abdominal ultrasonographic examination. Ultrasonographic examinations were performed or supervised by a board-certified veterinary radiologist or a veterinarian board-certified in internal medicine. Additional diagnostic assessments for the 20 dogs included measurement of fasting bile acid concentration (n = 1), protein C activity (1), and serum glucagon concentrationa (3) and thoracic radiography (4).
Amino acid analysis
For each dog, a heparinized blood sample and urine sample (obtained via free catch, catheterization, or cystocentesis) were obtained within 8 hours of each other and used to measure amino acid concentrations. A mixed urine sample was used to determine urine creatinine concentration; the urine creatinine concentration was used to normalize urine concentrations of amino acids. Plasma was promptly harvested, and both heparinized plasma and urine were stored at −20°C until shipped frozen on dry ice to a veterinary laboratoryb for amino acid analysis. Briefly, 6% sulfosalicylic acid was added (dilution, 1:1 [vol/vol]) to plasma and urine samples for deproteinization; solutions were then centrifuged at 16.1 × g for 25 minutes. Supernatant was filtered through a 0.45-μm syringe-drive polytetrafluoroethylene filter, and pH was adjusted to 2.2 by the use of 0.3 N lithium hydroxide solution. A volume of 50 μL was loaded into an amino acid analyzer.c For reporting purposes, amino acid concentrations in plasma were compared on the basis of mean values of the reference range established at the analytic laboratory.b This permitted expression of concentrations relative to reference range (control) amino acid concentrations, which facilitated graphic depictions. Individual urine concentrations of amino acids, normalized on the basis of the urine creatinine concentration, were reported relative to creatinine-normalized maximum reference concentrations for control dogs20 (ie, calculation of the ratio of the measured amino acid concentration in urine to the maximum reference amino acid concentration in urine).
Categorization of clinical disease severity
Minimal-subclinical, moderate, and severe disease categories were assigned on the basis of specific criteria to enable comparison of amino acid analysis among dogs by use of heat map analysis. Minimal-subclinical disease was defined as only HCS hepatopathy (as determined by results of examination of biopsy specimens) without cutaneous lesions or subtle cutaneous lesions noted retrospectively after diagnosis of HCS hepatopathy. Moderate disease was defined as obvious but not severely painful or extensive skin lesions, patient survival of ≥ 1 month, and improvement of the condition after ≤ 5 amino acid infusions. Severe clinical disease was defined as painful extensive skin lesions that required > 5 amino acid infusions to achieve partial or full remission of cutaneous lesions or that led to euthanasia because of a lack of response to amino acid infusions. Dogs in the severe and minimal-subclinical disease categories were selected for heat map analysis; these dogs had been evaluated and treated by 1 clinician (JPL). One dog with moderate disease was selected for heat map analysis because of serial measurement of amino acid concentrations (at the time of initial diagnosis and during remission of cutaneous lesions) for that dog.
Statistical analysis
The sex distribution of all 20 dogs was compared with the expected equivalent distribution (ie, 50% males and 50% females) by use of a 2 × 2 table. For plasma and urine (normalized on the basis of the urine creatinine concentration) concentrations of amino acids, mean concentration of individual amino acids for all dogs with HCS was compared with the mean ± SD of laboratory or published20 reference values (control values). A 2-way ANOVA with multiple comparisons by use of the Sidak correction was used to compare values between dogs with HCS and reference (control) amino acid concentrations. Parametric statistics were used because reference data were available only as mean and SEM (plasma) or mean and SD (urine). The Grubb test was performed on mean urine concentrations of amino acids to detect outlier mean values. Linear regression analysis was used to investigate the association between urine concentrations of lysine (normalized on the basis of the urine creatinine concentration) and plasma concentrations of lysine. Values were considered significant at P < 0.05. Data analyses and corresponding generation of graphs were performed with commercial software.d An amino acid heat map was generated with spreadsheet software.e
Results
Animals
Various breeds were represented, consisting of Shih Tzu (n = 4), Maltese (3), Schipperke (2), Shetland Sheepdog (2), West Highland White Terrier (2), Australian Shepherd (1), Bichon Frise (1), Chihuahua (1), Cocker Spaniel (1), German Shorthaired Pointer (1), Poodle cross (1), and Staffordshire Bull Terrier (1). Median age was 9.5 years (range, 5 to 14 years), and median body weight was 9.2 kg (range, 2.3 to 38.0 kg). Although there were 15 males (14 castrated males and 1 sexually intact male) and 5 spayed females, the proportions of males and females did not differ significantly (P = 0.19) from the expected proportions of 50% males and 50% females. Definitive (n = 18) and presumptive (2) diagnoses were achieved as described previously. Histopathologic findings in the liver were consistent with HCS hepatopathy in all dogs from which liver biopsy specimens were obtained (n = 14); however, lesion severity differed considerably among dogs. Hypoaminoacidemia was confirmed in each dog with HCS. Ten dogs had marked hyperkeratosis of paw pads and ulcerative lesions involving the mucocutaneous junctions or intertriginous regions. One dog had equivocal hyperkeratosis of the paw pads, a small painful fissure in 1 paw pad, and a mild erythematous lesion on the lateral aspect of each elbow joint. Diagnosis of HCS was made on the basis of results for examination of liver biopsy specimens without concurrent cutaneous lesions for 6 dogs. Additional physical examination findings included peripheral lymphadenopathy (n = 3) and suspected secondary cutaneous bacterial infection (2) in dogs with skin lesions.
Two dogs (an 11-year-old castrated male Australian Shepherd and a 9-year-old castrated male West Highland White Terrier) were euthanized within 2 weeks after diagnosis of HCS, 2 dogs (a 5-year-old castrated male Shih Tzu and a 12-year-old castrated male Shih Tzu) were euthanized within 1 month after diagnosis of HCS, and 2 dogs (a 5-year-old spayed female Schipperke and an 8-year-old spayed female Shih Tzu) were euthanized within 6 months after diagnosis of HCS. One dog (a 14-year-old castrated male Shetland Sheepdog) was euthanized 5 months after diagnosis of HCS, despite a positive response to treatment, because of acute kidney injury of unknown cause. The survival duration for the remaining 13 dogs was 1 to 20 months after diagnosis of HCS.
Routine diagnostic testing
Common clinicopathologic findings included microcytosis (14/16 dogs), anemia (6/16 dogs, with nonregenerative anemia in 5/6 anemic dogs), high serum alanine aminotransferase activity (17/20 dogs), high serum alkaline phosphatase activity (20/20 dogs), low serum albumin concentration (9/18 dogs), and low BUN concentration (3/18 dogs). Only 1 dog was hyperbilirubinemic (total bilirubin concentration, 1.5 mg/dL; reference range, 0 to 0.2 mg/dL). Total serum bile acids concentrations (before a meal and 2 hours after feeding) were measured in 1 anicteric dog and were within reference intervals. Protein C activity (measured in 1 dog) was low (32%; reference range, ≥ 70%). Abdominal ultrasonography (15 dogs) revealed a subjectively normal-sized liver in 5 dogs, large liver in 3 dogs, and small liver in 4 dogs (liver size was not specified for 3 dogs). A diffuse hyperechoic hepatic parenchyma riddled with numerous scattered hypoechoic nodules consistent with a honeycomb or Swiss cheese pattern7 was identified in all dogs evaluated by use of abdominal ultrasonography.
Five dogs developed insulin-resistant diabetes mellitus. Fasted plasma concentration of glucagon measured in 3 dogs (2 nondiabetic dogs and 1 diabetic dog) at initial HCS diagnosis was 154, 150, and 363 pg/mL, respectively. Because there was no validated canine reference range provided by the analytic laboratorya that performed the analyses, these values were compared with published reference ranges21,22 for healthy adult dogs (a maximum of 174 pg/mL or 25 to 250 pg/mL, depending on the publication). Only 1 patient, a diabetic 10-year-old castrated male German Shorthaired Pointer, had an unequivocally high plasma glucagon concentration that might have been associated with a glucagonoma. However, no mass lesions were identified ultrasonographically, and that dog had remission of the lesions and a survival duration > 18 months.
Plasma amino acid profiles
Alterations of amino acids in plasma were depicted as both the measured concentration relative to the mean reference (control) value determined by the analytic laboratory, which also provided information to complete results of a previous study23 on HCS hypoaminoacidemia, and the mean absolute measured concentrations for all HCS dogs (Figure 1). Each dog had generalized hypoaminoacidemia, with concentrations for many individual amino acids ≤ 50% of the mean control values. Most severe and consistent abnormalities (ratio of the median measured concentration to the mean control value) involved glutamine (0.3X), proline (0.2X), cysteine (< 0.1X), and hydroxyproline (< 0.1X). Four amino acids were relatively increased, compared with control mean values, including glutamic acid (1.7X), phenylalanine (1.7X), 3-methylhistidine (1.8X), and l-α-amino-n-butyric acid (2.5X). Mean plasma concentration of amino acids of HCS dogs was significantly lower than the control value for threonine, glutamine, glycine, and alanine (all P < 0.05); proline and hydroxyproline (P < 0.001); and arginine (P = 0.004). Mean concentration for glutamic acid of HCS dogs was significantly (P < 0.001) higher than the control value.
Urine amino acid profiles
All dogs had profound urine wasting (ratio of the median measured concentration to the mean control value) of lysine (5.0X), 1-methylhistidine (2.1X), and proline (1.8X; Figure 2). When all urine amino acids for which there were reference values were included in the statistical analysis, the urine concentration of 1-methylhistidine was the only amino acid significantly (P < 0.001) increased, relative to control values.20 However, this amino acid was detected in notably greater amounts, which warranted outlier status and exclusion (1-methylhistidine concentration vs all other mean amino acid concentrations; P < 0.01). Reanalysis of individual amino acids after the exclusion of 1-methylhistidine confirmed significant (P < 0.001) urine wasting of lysine (Figure 3). Linear regression analysis of the urine concentration of lysine, compared with the plasma concentration of lysine, revealed that the association was not significant (P = 0.38). Urine lysine concentration in 3 dogs without HCS was 72, 85, and 109 nmol/mg of creatinine, which indicated a ratio of the median measured concentration to the mean control value of 0.85. Additionally, proline was not detectable in the urine obtained from any of these 3 dogs.
Association of plasma amino acid profiles and severity of cutaneous disease
Ratios of plasma amino acids examined on the basis of clinical disease severity were summarized by use of a heat map (Figure 4). Two of 3 dogs with severe clinical disease (an 11-year-old castrated male Australian Shepherd and a 9-year-old castrated male West Highland White Terrier) were euthanized within 10 days after diagnosis of HCS because of quality-of-life concerns and development of acute neurologic signs, respectively. The third dog with severe clinical disease (an 8-year-old spayed female Maltese) required > 10 amino acid (and lipid) infusions to achieve complete clinical remission; that dog was alive 12 months after diagnosis of HCS. For 1 dog with moderate disease (an 11-year-old castrated male Poodle cross), amino acid profiles were determined 1 month after diagnosis and 6 months after remission of cutaneous lesions (which was after 2 parenteral infusions of amino acids and institution of dietary management); that dog was in remission at 20 months after diagnosis of HCS. Three dogs with HCS hepatopathy without skin lesions or with only mild cutaneous lesions (an 8-year-old castrated male West Highland White Terrier, a 10-year-old castrated male German Shorthaired Pointer, and an 11-year-old spayed female Bichon Frise) had undergone liver biopsy because of persistent increases in serum liver enzyme activities. Thus, 3 categories were contrasted: severe disease (n = 3), moderate disease progressing to remission (1), and HCS hepatopathy without any or with only minimal cutaneous involvement (3). Relative plasma concentrations of several amino acids appeared to be associated with reduced clinical disease severity or remission status. These plasma amino acids included asparagine, lysine, proline, and the branched-chain amino acids valine, isoleucine, and leucine. Importantly, concentrations of amino acids associated with this pattern all increased after treatment and remission in the dog with moderate disease progressing to remission.
Discussion
For the study reported here, evidence of aminoaciduria purportedly reflected acquired amino acid transporter dysfunction in dogs with HCS. Hypoaminoacidemia has long been identified as a central feature of HCS; however, its pathological mechanisms have eluded investigators for decades, and findings for the present study provided a plausible mechanism. The group of patients in the present study fit demographically with previous descriptions of HCS,23,24 with an over-representation of Shih Tzu, Maltese, Schipperke, West Highland White Terrier, and Shetland Sheepdog; a predominance of middle-aged or older dogs; and a population with a 3:1 proportion of males to females. Thus, we believed that this population of dogs was representative of those with HCS.
We detected discordance between classic dermatologic features and hepatic histologic features in dogs with HCS. The diagnosis for a number of dogs in the present study was made on the basis of examination of liver biopsy samples acquired in response to persistent increases in serum liver enzyme activity, an abnormal honeycomb liver pattern recognized on abdominal ultrasonography, or both. This likely reflected bias associated with the large number of liver biopsy specimens submitted because of the hepatobiliary interest in diagnostic pathology at Cornell University. One patient had a historical diagnosis of copper-associated hepatopathy and had been treated with d-penicillamine for 1 year, but that treatment failed to resolve liver enzyme activity abnormalities. When referred because of the emergence of diabetes mellitus, consideration of HCS prompted acquisition of liver sections from a reference laboratory. Examination of these liver sections by 2 of the authors (SAC and SPM) revealed classic HCS hepatopathy that had been overlooked. Thus, findings for the present study confirmed that HCS hepatopathy can precede development of classic HCS cutaneous lesions, a feature that has seldom been discussed in the veterinary literature on HCS.24 Although HCS is a relatively rare syndrome, we suspect it may be underdiagnosed in dogs that lack classic cutaneous lesions.
Plasma amino acid profiles in dogs with HCS in the present study were consistent with those previously established for this syndrome in dogs. Samples in this study were measured in the same laboratory that performed a seminal study.23 We propose that wasting aminoaciduria most likely reflected an acquired pathological mechanism of HCS hypoaminoacidemia. However, it remains possible that urine amino acid wasting also might have reflected amino acid depletion from some other cause that subsequently impacted membrane transporters and influenced amino acid homeostasis. Further studies will be necessary to resolve this conundrum.
The most prominent plasma amino acid depletions impacted the urea cycle (arginine and ornithine), glutathione synthesis (glutamine, glycine, and cysteine), and collagen synthesis (proline and hydroxyproline). Reductions in cysteine in these patients must be cautiously interpreted because the free form of this amino acid is rapidly oxidized, which yields cystine and adducts with plasma proteins in vitro.25 However, sulfosalicylic was added to plasma samples prior to analysis for both the reference samples and HCS samplesf; thus, the low cysteine values in the cohort of the present study were unlikely to have reflected relevant spurious or technical reductions in the concentrations of this amino acid.
Reduced concentrations of certain amino acids appear to coordinate with suspected metabolomic dysregulations in dogs with HCS. In humans and rodents, defects in the urea cycle can result in microvesicular hepatocellular lipid vacuolation as well as ballooning degeneration associated with glycogen-type vacuolation.26–28 Each of these histologic features accompanies classic HCS hepatopathy.11,24 Reduced availability of proline and hydroxyproline could explain cutaneous lesions by limiting prompt reparative collagen synthesis. Reduced amounts of glutathione precursors would augment oxidative injury resulting from normal as well as abnormal physiologic biochemical events.
Although a mechanistic role of aminoaciduria in the causation of HCS cannot be simply deduced from the findings of the study reported here, a gradation in the severity of hypoaminoacidemia that coordinated with the severity of clinical signs was detected. Dogs with more extensive cutaneous lesions had a greater reduction in the plasma lysine concentration. We were not surprised that the magnitude of lysinuria was not associated with the magnitude of the reduction in plasma lysine concentrations because there are considerable body stores of lysine in structural proteins and a steady flux from these resources, depending on metabolic status and dietary intake.29,30 Protein turnover should be altered in these dogs to compensate for hypoaminoacidemia, as has been reported for dietary lysine deficiency in dogs.31 Considering that all dogs of the present study had marked lysinuria, irrespective of cutaneous lesions or disease severity, it may be a central pathogenic feature of HCS that could serve as an early screening marker for this disorder.
A number of plasma amino acids (glutamic acid, l-α-amino-n-butyric acid, phenylalanine, and 3-methylhistidine) had increased mean values (ratio with reference range values), although these were not significantly increased, except for glutamic acid. Nevertheless, we believed that these increases were likely of biological importance because they may have developed secondary to protein-calorie malnutrition, liver injury, or reductions in concentrations of other amino acids, as has been reported for other conditions with reductions in plasma glutamine concentrations.32 The ratio of plasma concentrations of l-α-amino-n-butyric acid to leucine has been used as a marker for alcoholism in humans, a syndrome complicated by negative nitrogen balance33; however, increased values for this ratio are also associated with other liver disorders and with an unfavorable nutritional status.33 Thus, in dogs with HCS, these changes may reflect altered hepatic metabolism, imbalanced nutritional homeostasis, hyporexia, and anorexia. An increase in the plasma phenylalanine concentration may have reflected hepatic metabolic dysfunction because increases in phenylalanine concentrations have been associated with fibrosis in hepatitis C-associated liver injury and in hepatectomized mice.34,35 However, there was wide variation in clinical and hepatic histologic severity in the cohort of dogs in the study reported here, which suggested that this relationship, if present, is complex.
Aminoaciduria is a feature of a number of diseases in humans that can be congenital or acquired. Urinary losses of amino acids can be restricted to a small number of specific amino acids or can be more generalized, depending on the underlying cause. Clinical manifestations of disorders associated with aminoaciduria depend on the specific amino acid or amino acids lost and the magnitude of aminoaciduria. A limited number of diseases with aminoaciduria or imidodipeptideuria (dipeptides containing at least 1 imino acid [ie, proline or hydroxyproline]) have been associated with cutaneous lesions. Data for the present study excluded a close association with Hartnup disease as a basis for HCS, given the lack of tryptophan loss in urine of the affected dogs.
Renal handling of lysine and proline involves separate amino acid transporters. Lysine, along with arginine, ornithine, and cysteine, is predominantly reabsorbed through the dibasic amino acid transporter. Proline transport is accompanied by hydroxyproline via multiple transporters, which are thought to be specific for these imino acids or shared by glycine.36 There is no transporter common to all amino acids that had increased concentrations in the urine of dogs with HCS. Thus, aminoaciduria may involve a currently unidentified transporter or dysfunction of multiple transporters simultaneously, perhaps by altered metabolic status of epithelial cells in the proximal renal tubules. It also might reflect an as-yet unidentified acquired deficiency that curtails transporter gene expression, protein expression, or protein function. This might be a cumulative effect of depletion of 1 or more amino acids from the plasma.
Although a direct association cannot be made between species, prolidase deficiency in humans shares similar cutaneous features with HCS in dogs.37,38 Prolidase deficiency disrupts recycling of proline and hydroxyproline residues during protein degradation, thereby limiting their availability for collagen synthesis and degradation recycling. In turn, this leads to accumulation of imidodipeptides, which are prominently produced during collagen degradation. In prolidase deficiency, the dipeptides x-proline and x-hydroxyproline are lost in urine, which compromises their availability for collagen synthesis.37,38 Thus, we believe that reduced availability of proline and hydroxyproline for collagen synthesis may be a common element leading to cutaneous lesions in prolidase deficiency and HCS, despite potentially different mechanisms for depletion of amino acids.
Defects in the dibasic amino acid transporter can lead to LPI, whereby urea cycle substrates are diminished as a result of intestinal malabsorption or increased renal losses of arginine, ornithine, and lysine. Similarities exist between LPI in humans and HCS in dogs; however, there are a number of differences that preclude consideration of HCS as a canine version of LPI. Although humans with LPI typically have postprandial hyperammonemia when fed a protein-rich meal, a small number with lysinuria do not have protein intolerance but still have urine wasting of arginine and ornithine.36 To the authors’ knowledge, there are no reported cases of humans in which there has been lysinuria without concomitant renal wasting of arginine and ornithine. Despite these dissimilarities, hepatic histopathologic findings in patients with LPI share some features with those of hepatopathy in dogs with HCS. These include hepatocellular vacuoles and ballooning degeneration, macrovesicular and microvesicular cytosolic lipid vacuoles, proliferative foci of hepatocytes with minor inflammatory infiltrates, and degenerative changes progressing to a nodular liver with no parenchyma. Lesions are not highly fibrotic, and remodeling is degenerative with parenchymal nodules not marginated by fibrosis, as observed for necroinflammatory hepatopathies.16,39 Also, lupus-like cutaneous lesions bearing some similarities to HCS lesions occasionally develop in humans with LPI.37,40 Finally, other hypoaminoacidemias, including proline deficiency, have been reported in LPI patients.17
Because many amino acid transporters are expressed by both renal and intestinal epithelial cells, decreased enteral uptake of amino acids also might contribute to hypoaminoacidemia in dogs with HCS.36 This might explain the better response to parenterally administered, compared with enterally administered, supplemental amino acids detected by the authors (unpublished data). Enteral administration of supplemental amino acids requires supraphysiologic amounts of amino acids in a free form or in highly digestible protein sources to achieve clinical improvement. Another consideration is that cationic amino acid transporters expressed by the dermal epithelium41 might also contribute to nonhealing cutaneous wounds. Finally, lysine deficiency can contribute to acquired niacin deficiency, which results in pellagra-related skin lesions that also share features with cutaneous HCS lesions.42
Segregation of disease phenotype by severity of cutaneous lesions revealed a change commensurate with the severity of depletions of plasma concentrations of amino acids. Greater reductions in branched-chain amino acid concentrations appeared to be associated with increasing severity of HCS. Because branched-chain amino acids are preferentially used as energy sources (gluconeogenesis) during hepatic failure,43 this might have reflected altered hepatic metabolism. Observationally, the association between plasma concentrations of lysine and proline and clinical severity of cutaneous lesions supported an important role for these amino acids in HCS. Therefore, we speculate that aminoaciduria has pathophysiologic importance in HCS. The role of glucagon in this syndrome cannot be dismissed. A study44 on amino acids in 1 human patient with a glucagonoma revealed pathologically increased plasma clearance of amino acids known to become deficient in glucagonoma. Unfortunately, the authors are aware of no relevant studies of aminoaciduria in humans with a glucagonoma.
A limitation of the present study was the relatively small sample size and lack of a concurrent contemporary control population. However, data on plasma concentrations of amino acids were similar to values previously reported for HCS, which suggested that the study population was representative of dogs with HCS. Because the study reported here provided the first results on urine concentrations of amino acids in dogs with HCS, there are no historical references for comparison. However, by evaluating urine concentrations of amino acids by use of maximum published values, it is reasonable to conclude that the consistent aminoaciduria observed represented a biologically relevant component of HCS. Historical control data were used to detect significant differences in plasma and urine concentrations of amino acids. However, we did not monitor for sequential changes in aminoaciduria concordant with treatment response.
Further investigations are warranted to determine specific defects that lead to selected aminoaciduria; the role of aminoaciduria in the pathogenesis, progression, and recovery for dogs with HCS; whether aminoaciduria can be used for disease screening or prognosis; and whether individually tailored nutritional supplementation may better resolve features of HCS, compared with the currently poor response to general administration of supplemental amino acids.
Acknowledgments
The authors thank Drs. Andrea J. Fascetti and Zengshou Yu for technical assistance.
ABBREVIATIONS
HCS | Hepatocutaneous syndrome |
LPI | Lysinuric protein intolerance |
NME | Necrolytic migratory erythema |
Footnotes
The Ohio State University Diagnostic Laboratory, Wexner Medical Center, The Ohio State University, Columbus, Ohio.
Amino Acid Laboratory, College of Veterinary Medicine, University of California-Davis, Davis, Calif.
Biochrom 30, Biochrom Ltd, Cambridgeshire, Cambourne, England.
GraphPad Prism, version 6.0 for Mac, GraphPad Software Inc, San Diego, Calif.
Excel, Mac 2011 and 2016, Microsoft Software Inc, Redmond, Wash.
Zengshou Yu, Davis Amino Acid Laboratory, University of California-Davis, Davis, Calif: Personal communication, 2016.
References
1. Becker SW, Kahn D, Rothman S. Cutaneous manifestations of internal malignant tumors. Arch Derm Syphilol 1942; 45: 1069–1080.
2. Tierney EP, Badger J. Etiology and pathogenesis of necrolytic migratory erythema: review of the literature. MedGenMed 2004; 6: 4.
3. Norton JA, Kahn CR, Schiebinger R, et al. Amino acid deficiency and the skin rash associated with glucagonoma. Ann Intern Med 1979; 91: 213–215.
4. Boden G, Rezvani I, Owen OE. Effects of glucagon on plasma amino acids. J Clin Invest 1984; 73: 785–793.
5. Jacobson LS, Kirberger RM, Nesbit JW. Hepatic ultrasonography and pathological findings in dogs with hepatocutaneous syndrome: new concepts. J Vet Intern Med 1995; 9: 399–404.
6. McEwen BJ. Superficial necrolytic dermatitis (hepatocutaneous syndrome) in a dog. Can Vet J 1994; 35: 53–54.
7. Byrne KP. Metabolic epidermal necrosis-hepatocutaneous syndrome. Vet Clin North Am Small Anim Pract 1999; 29: 1337–1355.
8. Bach JF, Glasser SA. A case of necrolytic migratory erythema managed for 24 months with intravenous amino acid and lipid infusions. Can Vet J 2013; 54: 873–875.
9. Alexander EK, Robinson M, Staniec M, et al. Peripheral amino acid and fatty acid infusion for the treatment of necrolytic migratory erythema in the glucagonoma syndrome. Clin Endocrinol (Oxf) 2002; 57: 827–831.
10. Gross TL, O'Brien TD, Davies AP, et al. Glucagon-producing pancreatic endocrine tumors in two dogs with superficial necrolytic dermatitis. J Am Vet Med Assoc 1990; 197: 1619–1622.
11. Gross TL, Song MD, Havel PJ, et al. Superficial necrolytic dermatitis (necrolytic migratory erythema) in dogs. Vet Pathol 1993; 30: 75–81.
12. Patel AB, Prabhu AS. Hartnup disease. Indian J Dermatol 2008; 53: 31–32.
13. Bröer S, Cavanaugh JA, Rasko JE. Neutral amino acid transport in epithelial cells and its malfunction in Hartnup disorder. Biochem Soc Trans 2005; 33: 233–236.
14. Posey JE, Burrage LC, Miller MJ, et al. Lysinuric protein intolerance presenting with multiple fractures. Mol Genet Metab Rep 2014; 1: 176–183.
15. Ogier de Baulny H, Schiff M, et al. Lysinuric protein intolerance (LPI): a multi organ disease by far more complex than a classic urea cycle disorder. Mol Genet Metab 2012; 106: 12–17.
16. McManus DT, Moore R, Hill CM, et al. Necropsy findings in lysinuric protein intolerance. J Clin Pathol 1996; 49: 345–347.
17. Cimbalistiene L, Lehnert W, Huoponen K, et al. First reported case of lysinuric protein intolerance (LPI) in Lithuania, confirmed biochemically and by DNA analysis. J Appl Genet 2007; 48: 277–280.
18. Laflamme DP. Development and validation of a body condition score system for dogs: a clinical tool. Canine Pract 1997; 22(4):10–15.
19. Michel KE, Sorenmo K, Shofer FS. Evaluation of body condition and weight loss in dogs presented to a veterinary oncology service. J Vet Intern Med 2004; 18: 692–695.
20. Blazer-Yost B, Jezyk PF. Free amino acids in the plasma and urine of dogs from birth to senescence. Am J Vet Res 1979; 40: 832–838.
21. Allenspach K, Arnold P, Glaus T, et al. Glucagon-producing neuroendocrine tumour associated with hypoaminoacidaemia and skin lesions. J Small Anim Pract 2000; 41: 402–406.
22. Nyland TG, Barthez PY, Ortega TM, et al. Hepatic ultrasonographic and pathologic findings in dogs with canine surperficial necrolytic dermatitis. Vet Radiol Ultrasound 1996; 37: 200–205.
23. Outerbridge CA, Marks SL, Rogers QR. Plasma amino acid concentrations in 36 dogs with histologically confirmed superficial necrolytic dermatitis. Vet Dermatol 2002; 13: 177–186.
24. Hall-Fonte DL, Center SA, McDonough SP, et al. Hepatocutaneous syndrome in Shih Tzus: 31 cases (1996–2014). J Am Vet Med Assoc 2016; 248: 802–813.
25. Tôrres CL, Miller JW, Rogers QR. Determination of free and total cyst(e)ine in plasma of dogs and cats. Vet Clin Pathol 2004; 33: 228–233.
26. Badizadegan K, Perez-Atayde AR. Focal glycogenosis of the liver in disorders of ureagenesis: its occurrence and diagnostic significance. Hepatology 1997; 26: 365–373.
27. Yaplito-Lee J, Chow C-W, Boneh A. Histopathological findings in livers of patients with urea cycle disorders. Mol Genet Metab 2013; 108: 161–165.
28. Hautekeete ML, Degott C, Benhamou JP. Microvesicular steatosis of the liver. Acta Clin Belg 1990; 45: 311–326.
29. Miller LL, Bale WF. The use of radioactive lysine in studies of protein metabolism; synthesis and utilization of plasma proteins. J Exp Med 1949; 90: 297–313.
30. Yuile CL, Lamson BG, Miller LL, et al. Conversion of plasma protein to tissue protein without evidence of protein breakdown; results of giving plasma protein labeled with carbon 14 parenterally to dogs. J Exp Med 1951; 93: 539–557.
31. Humbert B, Bleis P, Martin L, et al. Effects of dietary protein restriction and amino acids deficiency on protein metabolism in dogs. J Anim Physiol Anim Nutr (Berl) 2001; 85: 255–262.
32. Kinscherf R, Hack V, Fischbach T, et al. Low plasma glutamine in combination with high glutamate levels indicate risk for loss of body cell mass in healthy individuals: the effect of N-acetyl-cysteine. J Mol Med (Berl) 1996; 74: 393–400.
33. Yudkoff M, Blazer-Yost B, Cohn R, et al. On the clinical significance of the plasma alpha-amino-n-butyric acid:leucine ratio. Am J Clin Nutr 1979; 32: 282–285.
34. Holecek M, Mráz J, Tilšer I. Plasma amino acids in four models of experimental liver injury in rats. Amino Acids 1996; 10: 229–241.
35. Sands CJ, Guha IN, Kyriakides M, et al. Metabolic phenotyping for enhanced mechanistic stratification of chronic hepatitis C-induced liver fibrosis. Am J Gastroenterol 2015; 110: 159–169.
36. Bröer S. Amino acid transport across mammalian intestinal and renal epithelia. Physiol Rev 2008; 88: 249–286.
37. Klar A, Navon-Elkan P, Rubinow A, et al. Prolidase deficiency: it looks like systemic lupus erythematosus but it is not. Eur J Pediatr 2010; 169: 727–732.
38. Trent JT, Kirsner RS. Leg ulcers secondary to prolidase deficiency. Adv Skin Wound Care 2004; 17: 468–472.
39. Shinawi M, Dietzen DJ, White FV, et al. Early-onset hepatic fibrosis in lysinuric protein intolerance. J Pediatr Gastroenterol Nutr 2011; 53: 695–698.
40. Dionisi-Vici C, Hachem ME, Bertinie E. Skin and hair disorders. In: Hoffmann GF, Zschocke J, Nyhan WL, eds. Inherited metabolic diseases: a clinical approach. New York: Springer-Verlag Berlin Heidelberg, 2010; 197–217.
41. Schnorr O, Suschek CV, Kolb-Bachofen V. The importance of cationic amino acid transporter expression in human skin. J Invest Dermatol 2003; 120: 1016–1022.
42. Baker DH. Comparative nutrition and metabolism: explication of open questions with emphasis on protein and amino acids. Proc Natl Acad Sci U S A 2005; 102: 17897–1 7902.
43. Kato M, Miwa Y, Tajika M, et al. Preferential use of branched-chain amino acids as an energy substrate in patients with liver cirrhosis. Intern Med 1998; 37: 429–434.
44. Almdal TP, Heindorff H, Bardram L, et al. Increased amino acid clearance and urea synthesis in a patient with glucagonoma. Gut 1990; 31: 946–948.