THESIS This is to certify that the thesis entitled RADIOIMMUNOASSAY OF CANINE INSULIN, C-PEPTIDE AND GLUCAGON: A COMPARISON AND EVALUATION OF COMMERICAL RIA KITS presented by I ‘ Steven L . ’Stockham has been accepted towards fulfillment of the requirements for M. S . degree in Pathology Major professor Date November 4, 1380 0-7 639 ‘:.\ .31,ij r, . ‘\ fl“? .a'l‘clll/ " V 'II/ 1f 54.. OVERDUE FINES: 25¢ per day per it... RETURNIQ LIQRARY MATERIALS: Place in book atom to move charge from c1 mutton records RADIOIMMUNOASSAY OF CANINE INSULIN, C-PEPTIDE AND GLUCAGON: A COMPARISON AND EVALUATION OF COMMERCIAL RIA KITS BY Steven L. Stockham, D.V.M. Diplomate, A.C.V.P. A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Pathology 1980 ABSTRACT RADIOIMMUNOASSAY OF CANINE INSULIN, C-PEPTIDE AND GLUCAGON: A COMPARISON AND EVALUATION OF COMMERICAL RIA KITS BY Steven L. Stockham, D.V.M. Diplomate, A.C.V.P. Evaluation criteria for five commercial insulin RIA kits included precision, dilutional parallelism, sensitivity, and comparison of canine IRI concentrations in four control sera. One insulin RIA kit (Corning Medical) had consistent acceptable precision, good dilutional parallelism, and adequate sensitivity. Other RIA kits had relatively poorer performance in either dilutional parallelism or precision. No two RIA kits quantitated the same IRI quantities in all four control sera. Canine sera failed to adequately displace tracer from anti-synthetic human C-peptide antibody (CalbiochemrBehring Corporation). This finding suggested lack of species cross-immunoreactivity and was consistent with reported findings in other species. Two commercial glucagon RIA kits had poor precision in quantitating low canine IRG concentrations. One RIA kit was modified and appeared to increase sensitivity, but precision remained inadequate. To Marcia Your love, understanding, and encouragement have made this possible. 11 ACKNOWLEDGEMENTS I wish to express my gratitude to Dr. J.D. Krehbiel, Acting Chairman of the Department of Pathology, and Dr. Raymond F. Nachreiner for their advice and guidance during my course of study and preparation of this thesis. Appreciation is also extended to Drs. William Schall and George Padgett for serving on my committee. I am also thankful to Mrs. Joanne Kivela, Ms. Phyllis Holmes, and Ms. Marcia Carr for their assistance in RIA techniques and to Mrs. Paula Archer for assistance in sample collection. This project nor associated training would not have been possible without the support of Drs. Harold Tvedten, Douglas Weiss, and co-residents. Thesis preparation would not have been possible without outstanding clerical assistance from Ms. Cathy Powers. This project was supported in part by NIH Grant RRO1173, Animal Health Diagnostic Laboratory and Veterinary Clinical Center, Michigan State University. iii II. III. TABLE OF CONTENTS INTRODUCTION. . . . . . . . . . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . . . . . . . A. History . . . . . . . . . . . . . . . B. Radioimmunoassay Evaluation . . . . . . . . . . C. Canine Immunoreactive Insulin Quantitation. . . . D. Canine Insulin Molecule Compared to Other Species E. C-peptide Radioimmunoassays . . . . . . . . . . F. Glucagon Radioimmunoassays and Heterogeneity of Canine Glucagon. . . . . . . . . . . . . . . . MATERIALS AND METHODS . . .i. . . . . . . . . . . . . A. Experiment 1 - Screening Evaluation of Commercial Insulin Radioimmunoassays. . . . . 1. Commerical Insulin Radioimmunoassays Evaluated. 2. Sources of Canine and Commercial Control Sera 3. Criteria for Evaluation and Comparison of RIA Kits. . . . . . . . . . . . . . . . . . . . . 4. Data Reduction. . . . . . . . . . . . . . . . B. Experiment 2 - Screening Evaluation of Commercial 1. 2. 3. 4. C-peptide Radioimmunoassay . . . . RIA Kit Evaluated . . . . . . . . . . . . . . Sources of Samples. . . . . . . . . . . . . . Criteria for Evaluation of Kit. . . . . . . . Data Reduction. . . . . . . . . . . . . . . iv Page 12 12 12 13 14 17 17 17 18 18 18 Experiment 3 - Screening Evaluation of Commerical Glucagon Radioimmunoassays . . . . . 1. Commercial Glucagon Radioimmunoassays Evaluated 2. Sources of Control Plasma . . . . . . . . . . . 3. Criteria for Evaluation of RIA Kits . . . . . . 4. Data Reduction. . . . . . . . . . . . . . . . . Experiment 4 - Modifications of Commercial Glucagon Radioimmunoassay . . . . . . . . . . 1. Commercial Glucagon RIA Kit Evaluated . . . . . 2. Modifications of Glucagon RIA Kit . . . . . . . 3. Sources of Samples. . . . . . . . . . . . . . . 4. Criteria for Evaluation of Modified Glucagon RIA Kit 0 o o o o o o o 0' o o o o o o o o o o 0 IV. RESULTS A. Experiment 1 - Screening Evaluation of Commercial Insulin Radioimmunoassays. . . . . . B. Experiment 2 - Screening Evaluation of Commercial C-peptide Radioimmunoassay . . . . . C. Experiment 3 - Screening Evaluation of Commerical Glucagon Radioimmunoassay. . . . . . D. Experiment 4 - Modification of Commercial Glucagon Radioimmunoassay . . . . . . . . . . V. DISCUSSION. . . . . . . . . . . . . . . . . . . . . . . VI. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . VII . APPENDICES C O O I O O I O O I C O O O O O O O O O O O O A. C. Components, Performance Characteristics, and Binding Parameters. . . . . . . . . . . . . . . . . Immunoreactive Insulin Reference Values; Fasting and High Dose Intravenous Glucose Tolerance Test. . Stability of Canine Immunoreactive Insulin. . . . . VIII 0 REFERENCES 0 O O O I O O O O O O O O O O O O O O O O O 0 VIII 0 VITA. . 0 O O O O O O O O O O I O O O O O O O O O O O O Page 18 18 19 19 20 20 20 20 21 22 23 28 28 31 37 46 48 51 58 60 68 Table 10 11 12 13 A1 A2 LIST OF TABLES Dilutional parallelism assessment of insulin RIA kits by linear regression. . . . . . . . . . . . . . Dilutional parallelism assessment of insulin RIA kits by comparison of corrected concentration to undiluted concentration . . . . . . . . . . . . . . . . . . . Intra-assay and inter-assay precision of insulin RIA kits Q 0 O 0 O O O ' I O O ‘ O O O O O O O O O O O O 0 Sensitivity assessment of insulin RIA kits. . . . . . Control sera immunoreactive insulin (IRI) concentrations. . . . . . . . . . . . . . . . . . . Comparison of mean immunoreactive insulin (IRI) concentrations by Tukey's test (Gill, 1978) . . . . Human C-peptide RIA evaluation. . . . . . . . . . . . Intra-assay and inter-assay precision of glucagon RIA kits O .1 O O O O O 0 O 0 O O I O O O O I O O O O 0 Sensitivity assessment of commercial glucagon RIA kits 0 0 O O O 0 0 O O O O ‘ O O O l O O O O O a O O 0 Binding parameters of original and modified glucagon RIA O O O O O O O O O O O O O O O ‘ O O O O O O O O O Immunoreactive glucagon quantitated in original and mOdified glucagon RIA o o o o o o o o o o o a o a o o Reproducibility of standard curves in modified glucagon RIA (50% tracer and 71% Ab1) . . . . . . . . Immunoreactive glucagon quantitated in modified glucagon RIA (50% tracer and 71% Ab1) . . . . . . . . Components and lot numbers of evaluated commercial radioimmunoassays . . . . . . . . . . . . . . . . . . Performance characteristics stated in insulin RIA kit package inserts . . . . . . . . . . . . . . . . . . . vi Page 23 24 25 26 27 27 29 30 31 32 33 35 36 48 49 Table Page A3 RIA kit binding parameters in replicated assays . . . . . . 50 81 Reference population laboratory data summary. . . . . . 54 82 HD-IVGTT reference values . . . . . . . . . . . . . 55 33 HD-IVGTT reference values . . . . . . . . . . . . . . . 56 B4 Quality control data; IRI quantitation for HD-IVGTT reference values. . , . . . . . . . . . . . . . . . . . 57 C1 Stability of canine immunoreactive insulin (time and temperature) C O O O O O O O O O O O O O O O O O O O O l O O 59 vii I . INTRODUCTION Insulin and glucagon have been considered primary hormonal regulators of glucose homeostasis (Unger, 1971). Diabetes mellitus and functional beta cell islet ne0plasms (insulinomas) have been the most frequently reported canine disorders relating to abnormal insulin activity (Capen and Martin, 1969; Kramer and Wilson, 1978; Njoku et al., 1972; Schall and Cornelius, 1977). Documentation of clinical disorders relating to abnor- mal glucagon activity in dogs has been limited. Experimental evidence has indicated that diabetes mellitus is a syndrome associated with carbohydrate intolerance and with or without hypoinsulinemia. Hyperglucagonemia (Blazquez et al., 1977) hyper- somatropinemia (Pierluissi and Campbell, 1980), hypersomatostatinemia (Krejs et al., 1979; Schusdziarra et al., 1977), and elevated glucocor- ticoids (Emmer et al., 1971) have been shown to be involved in diabetes mellitus. However, most of these reports involved human diseases or experiments with dogs. Spontaneous canine diabetes mellitus has been assumed to be a syndrome of hypoinsulinemia and roles of other hormones have not been adequately evaluated (Schall and Cornelius, 1977). In addition to above hormones, epinephrine (Levine and Raft, 1970) and soma- ‘tomedins (Phillips and Vassilopoulou-Sellin, 1980) have roles in car- ibohydrate metabolism. Felig et a1. (1976), Gerich et a1. (1976), and Samols and weir (1979) have reported relationships and regulations of these hormones. Tentative diagnoses of canine insulinomas have been based on demonstraton of Whipple's triad (Caywood et al., 1979). Demonstration of concurrent hyperinsulinemia has been used to confirm diagnoses (Johnson, 1977). Connecting peptide (C-peptide) quantitations have been reported to be useful in evaluating hypoglycemic and hyperglycemic disorders of man (Horwitz et al., 1976; Rubenstein et al., 1977; Starr et al., 1979). Similar studies have not been reported in canine disorders. Very few of above articles would have been written had it not been for radioimmunoassay (RIA) development. Canine immunoreactive insulin (IRI) and immunoreactive glucagon (IRG) quantitations have generally utilized radioimmunoassays developed in research laboratories. These "private" assays have hindered RIA availability, created variations in inter-laboratory results, and decreased clinical applications of canine IRI and canine IRG quantitations. Commercial RIA kits have proven useful for quantitating several canine hormones. However, most RIA kits have been developed to quan- titate human hormones. It has been necessary to evaluate these assays for their ability to quantitate respective canine hormones. The purpose of this project was to evaluate commercial insulin, glucagon, and C-peptide RIA kits. Those assays which reliably quantitate respective canine polypeptides could then be utilized for clinical and research investigations. Hormone quantitations could also be offered to veterinarians and other investigators on a reference laboratory basis. Complete hormonal evaluation of spontaneous canine endocrine disorders may lead to a better management of affected dogs, understanding of pathophysiologic mechanisms, and possibly identification of animal models for human diseases. II. LITERATURE REVIEW A. History Radioimmunoassay (RIA) has become an important form of clinical chemr istry in the last two decades. Classic investigations by Rosalyn S. Yalow, Ph.D. and the late Solomon A. Berson, M.D. led to RIA development. Their demonstration of insulin antibodies in insulin-treated humans (Berson et al., 1956) was not readily accepted by the scientific come munity. Yalow (1974) reported that the Journal 2: Clinical Investigation would not allow the word "antibody" in their original article but would accept the term "insulin binding globulin". Berson and Yalow (1959) demonstrated that their human antibody to bovine and porcine insulins could react with preparations of bovine, porcine, equine, and ovine crystalline insulins. First RIA utilization for clinical investigations was with guinea pig antibody to bovine insulin (Yalow and Berson, 1959). Within the next year, Yalow and Berson (1960) reported insulin RIA utilization to quantitate immunoreac- tive insulin (IRI) during glucose tolerance tests and in subjects with functional islet cell tumors. In the 1960's, most RIA development and utilization involved in-house reagent preparation. Today, over 200 biologically important substances can be quantitated with commercial RIA kits (Sweeny, 1979). B. Radioimmunoassay Evaluation As for any quantitative analytical procedure, an RIA should reliably quantitate substances of interest (analytes). Four basic criteria must 3 be evaluated for RIA validation: specificity, precision, sensitivity, and accuracy. (Kagan, 1975; Midgley et al., 1969; Moss et al., 1976; Skelly et al., 1973; Walsh, 1978; Zazuco Higa et al., 1974). However, acceptable limits and methods of evaluating these criteria varied among investigators. 1. Specificity Optimal clinical RIA would measure only biologic active hormones and not closely related substances. Midgley et a1. (1969) and Skelley et a1. (1973) reported that antibody specificity for analyte was dependent on purity and species origin of immunogen, presence of immunologic simi- lar substances (precursor hormones, degradation products), presence of unwanted radiolabelled substances, and differences in incubation medium composition. Two methods have been recommended for assessing specificity: 1) demonstration of dilutional parallelism, 2) demonstration that related substances do not influence quantitation of analyte (Hafs et al., 1977; Witherspoon, 1979). 2. Precision Witherspoon (1979) reported that precision is affected by technical steps (pipetting, incubation, vortexing, counting time) and slope of dose-response curve. He also recommended that precision be evaluated with intra-assay and inter-assay coefficients of variation (C.v.) with respective acceptable limits of 5-10 and 10-20. Other authors recomr mended another method referred to as the index of precision or precision coefficient (Aubert, 1970; Midgeley et al., 1969; Skelley et al., 1973). This method was defined as "the ratio between the standard deviation of logit Y for each Y value and the slope of the regression line between values of logit Y and the corresponding log X". 3. Sensitivity Definitions and methods of assessing sensitivity varied among authors. One group defined sensitivity as detection limits and assessed sensitivity by statistical calculations using zero concentration tubes (Ekins, 1975; Midgley et al., 1969; Skelley et al., 1973, Witherspoon, 1979; Zazuco Higa et al., 1974) Another group defined sensitivity as the smallest quantity of antigen which can be measured reliably (Hafs et al., 1977; Kagan, 1975; Witherspoon, 1979). Using this definition, these investigators assessed sensitivity by determining analyte concentrations which displaced tracer by a fixed amount. These analyte concentrations corresponded to a 10% decrease from zero concentration in either the bound/free (B/F) or bound/bound-at-zero (B/BO) partition indices. 4. Accuracy Generally, two methods were recommended for assessing accuracy. Witherspoon (1979) stated that dilutions of a reference standard should give a parallel curve to kit standards. Longley (1976), Moss et al. (1976), Skelley et al. (1973), Walsh (1978), Witherspoon (1979), and Zazuco Higa et al. (1974) described a second method referred to as a recovery procedure. A ”spiked” sample (addition of reference standard) was assayed and 100% recovery implied accuracy. In both methods, a reference standard was necessary. Bangham (1976) and Cotes (1974) discussed availability and utilization of immunoreactive reference standards. World Health Organization‘a) and associated institutions have provided many immunoreactive hormone standards, however canine standards were not included in their literature. C. Canine Immunoreactive Insulin Qpantitation Canine IRI quantitations were determined by many laboratories and by a variety of methods. Evidence of RIA validation for canine serum or plasma was not found in the literature. Porcine insulin was used predo- minately for immunogens, tracers, and standards. Occasionally, bovine insulin was used for immunogen or tracer and human insulin for standards. Species sources of antibody were rarely reported. Fasting canine IRI concentrations varied with methods of separating free and bound tracer. DeFronzo et a1. (1978) reported 18-25 uU IRI/ml using a talc separation described by Rosselin et al. (1966). Juffe et a1. (1977), Kaneko et al. (1977, 1978a, 1978b), and Manny et al. (1977) reported 2 to 18 uU IRI/ml using a dextran-coated charcoal method described by Herbert et al. (1965). Hommel and Fischer (1977) and Orosz (1974) utilized an alcohol separation technique and reported a mean 40 uU IRI/ml and 19-21 pU IRI/ml, respectively. Campbell et a1. (1978) and Matsuyama et al. (1977) reported 38 :_5 p0 IRI/m1 and 32 1'4 uU IRI/m1, respectively, utilizing a pre-precipitation double antibody technique described by Hales and Randle (1963). Chen et a1. (1977) used a commer- cial RIA(b) based on the same technique and reported a mean of 15 uU IRI/m1. Black et al. (1980) and Manns and Martin (1972) utilized (a)WHO International Laboratory for Biological Standards, National Institute for Biological Standards and Control, Holly Hill, Hampsted, London, NW3 6RB, England (b)Amersham Insulin RIA Kit, Amersham Corporation, 2636 S. Clearbrook Dr., Arlington Heights, IL, 60005 post-precipitation double antibody methods and reported 8-22 uU IRI/ml and 37 1’18 uU IRI/ml, respectively. Cherrington et al. (1978), Jennings et al. (1977), and Keller et a1. (1977) utilized a commercial RIA‘C) based on Sephadexo bound antibodies described by Wide and Parath (1966) and reported 13-23 uU IRI/ml. Specific reasons for IRI concentration variations reported in fasting dog serum or plasma were not found. However, comparisons of insulin radioimmunoassays using human sera or plasma have shown that different IRI concentrations could be quantitated in the same sample. Malvano et al. (1974) compared three RIA methods of quantitating human IRI. They reported lower concentrations in dextran-coated charcoal and resin methods (9.4 i 3.6 and 8.7 i 3.1 uU/ml) than in double antibody tech- niques (16.2 i 4.5 UU/ml). Another comparison study with human sera and plasma resulted in different IRI concentrations in solid phase RIA‘C), double antibody RIA(b), and ethanol precipitation RIA (Thorell and Lanner, 1973). D. Canine Insulin Molecule Compared.£g Other Species Smith (1972) reviewed insulin amino acid sequences determined for many species. Canine insulin has two polypeptide chains (A chain - 21 amino acids, B chain - 30 amino acids) connected by two disulfide bridges. Canine, porcine, fin whale, and sperm whale insulin molecules have identical amino acid sequences. Bovine insulin differs from canine insulin by having amino acid substitution at positions 8 and 10 of the A chain. Human insulin differs from canine insulin only at the carboxyl-terminal of the B chain by having one amino acid substitution. (C)Phadebas Insulin Test, Pharmacia Diagnostics, 800 Centennial Avenue, Piscataway, New Jersey, 08854 Guinea pig, a common source of insulin antibodies, has an insulin mole- cule that differs from canine insulin at 8 positions on the A chain and 10 positions on the B chain. Amino acid sequences (primary structure) have been shown to determine tertiary and quaternary structures (or molecular conformation) of pro- teins (Stryer, 1975). However, Berson and Yalow (1961) determined that some insulin antisera were able to distinguish between porcine and sperm whale insulins. Berson and Yalow (1963) reported that human antibodies against porcine insulin reacted with the region of porcine insulin that had identical amino acid sequences to human insulin. These findings led to the conclusion that other species specific factors determine insulin antigenic characteristics (Berson et al., 1964). Definitive evidence of cross-immunoreactivity between canine insulin and other insulins was not found. Cross-reactivity of canine proinsulin with insulin antibodies was not found in the literature; however, human proinsulin has been shown to cross-react with porcine insulin antibodies (Kubasik, 1978) . Individual A and B chains of insulin did not cross react with insulin antibodies (Berson and Yalow, 1959; Starr et al., 1979). E - C-peptide(d) Radioimmunoassayg Canine immunoreactive C-peptide (IRC-P) has been quantitated to a Very limited extent. Kawanishi et al. (1977) reported quantitation of IRC-P in pancreatic venous blood. They indicated that the primary com- Ponents of their assay consisted of synthetic canine C-peptide, rabbit antibody to synthetic canine C-peptide, and iodinated (125D tyrosine conjugated synthetic canine C-peptide. (CDC-peptide - connecting peptide that links A and B chains of insulin in proinsulin molecule. Work with bovine, porcine, and human C-peptides has indicated that species specific antisera is needed for radioimmunoassays (Chance, 1972; Kaneko et al., 1974; Kemmler et al., 1972; Rubenstein and Steiner, 1971) . This requirement was attributed to differences in amino acid content or sequences in each respective species. Kemmler et al. (1972) and Peterson et al. (1972) reported that canine C-peptide has more differen- ces in primary structure than other domestic species when compared to human C-peptide. Cross-immunoreactive studies with canine C-peptide were not found . F. Glucagon Radioimmunoassays and Heterogeneity g; Canine Glucagon Unger et al. (1959) first described a RIA for glucagon. During the last twenty years, many controversies and contradictions have been reported in tissue sources of immunoreactive glucagon, glucagon action, biologic activity, and molecular forms of immunoreactive glucagon. Tissue sources of canine immunoreactive glucagon (IRG), (synonyms: "true glucagon", pancreatic-like glucagon) have been determined by tissue extractions and isolated perfused organs. Highest concentrations were found in pancreas and gastric mucosa (Blazquez et al., 1976, 1977; Lefebvre and Luyckx, 1978; Morita et al., 1976). Smaller concentrations were found in jejunal, duodenal, ileal, and colonic mucosa (Morita et al., 1976; Muller et al.,1978), and brain (Conlon et al., 1979). Difficulties in obtaining purified IRG was exemplified by one report where 105 canine stomachs were used to obtain 1.5 ug of purified gastric IRG (001 et al., 1979). Tissue sources of canine glucagon-like immunoreactivity (GLI) were found in intestine (Muller et al., 1978; Unger et al., 1968), gastric 10 mucosa (Doi et al., 1979; Srikant et al., 1977), and brain (Conlon et al., 1979). Classification of substances as IRG or GLI was established by dilu- tional parallelism studies using antisera generally accepted as pancreatic glucagon specific or non-specific antisera. Conlon et al. (1979), Faloona and Unger (1974), and Heding et al. (1976) concluded that antisera specific for IRG binds with C-terminal region of glucagon. Conlon et al. (1978, 1979) and Heding et al. (1976) reported that GLI lacked the C-terminal region of IRG and antiserum that reacts with GLI and IRG binds with N-terminal regions. Harris et a1. (1979) recommended that antisera for IRG have the following characteristics: 1) should react weakly or not at all with extracts of jejunum; 2) dilution of plasma should yield proportional glu- cagon immunoreactivity; 3) glucagon immunoreactivity should decrease after oral glucose administration; 4) depancreatized animals should have near zero concentrations of glucagon immunoreactivity. The 29 amino acid sequences of human, bovine, and porcine glucagon molecules were reported to be identical (Bromer et al., 1971; Thomsen et al., 1972). Amino acid compositions of rat, rabbit, and porcine glucagon molecules were reported to be identical (Sundby and Markussen, 1971, 1972). Canine glucagon amino acid sequence or composition were not found in the literature nor did any author state that canine glucagon was iden- tical to other species. Srikant et al. (1977) stated that canine and porcine glucagon dilution curves were parallel and thus implied immu- noreactive unity. Radioimmunoassay quantitation of canine glucagon has been limited to experimental studies to evaluate glucagon physiology. In 18 recent 11 articles reporting canine IRG concentrations in peripheral blood, only 2 reported quality control data or validation criteria. Buchanan et a1. (1969) reported their assay had a precision of 216 pg/ml, detection limit of 310 pg/ml, and canine controls had IRG concentrations from 240 to 350 pg/ml. Nilsson and Uvnas-Wallenstein (1977) reported good recovery stu- dies and a detection limit of 2 pg/ml. Reported concentrations of peripheral blood IRG (fasting or basal) ranged from 33 pg/ml (Eigler et al., 1979) to 278 1.25 pg/ml (Buchanan et al., 1969). Considerably more variation has been reported in GLI with concentrations from 74 $.15 pg/ml (Matsuyama et al., 1977) to 123 :_56 mg/ml (Manns and Martin, 1972). The later was the only article that reported IRG or GLI concentrations above 1 ng/ml. Reports stating use of commerical glucagon radioimr munoassays were not found. III. MATERIALS AND METHODS A. Experiment 1 - Screening Evaluations 9f Commercial Insulin Radioimmunoassays 1. Commerical Insulin Radioimmunoassays Evaluated Five commercial insulin radioimmunoassays develOped for quantitation of human IRI were selected from 17 available kits (Sweeny, 1979) because of differences in their separation procedures or published reports of their use to quantitate canine IRI. (Species sources of RIA components and lot numbers of evaluated assays are listed in Table A1. Performance characteristics stated in each RIA kit package insert are listed in Table A2.) Each insulin RIA was run on two different days with reagents from one kit and using duplicate determinations. Existing laboratory instrumentation and computer programs for data reduction necessitated modifications of some manufacturers' assay protocols. The following insulin RIA kits were evaluated: a. Kit A - Insulin RIA Kit; Amersham Corporation This immunoprecipitation assay utilized a double antibody modifica- tion where first and second antibody are complexed prior to their addi- tion to incubation solutions as first described by Hales and Randle (1963). Modifications of the manufacturer's protocol were approved by an Amersham technical representative. Glass tubes were used as reaction and counting tubes instead of polystyrene tubes. A decanting method was used to separate supernatant and precipitate instead of a draining technique. Total count tubes and standards were run in duplicate. 12 13 b. lKit B - CIS Insulin Radioimmunoassay Kit, INSIK - 3—M; International CIS, Damon Diagnostics This solid phase adsorption RIA utilized dextran-coated charcoal (DCC) to selectively adsorb free ligands as described by Herbert et al. (1965). Modifications of the manufacturer's protocol included decanting, counting free tracer, and insertion of a charcoal blank. Reconstituted phosphate buffer (pH 7.7) was titrated to pH 7.4 with 0.1 N HCl. Data reduction was based on bound tracer with the bound partition calculated as follows: Bound cpm = Total counts - Charcoal blank cpm - Free cpm c. Kit C - ImmophaseTM, Insulin 1251 Radioimmunoassay Test System; Corning Medical This solid phase RIA utilized antibodies covalently bound to glass particles. Manufacturer's protocol was modified only by a quick vor- texing of duplicate reaction tubes prior to incubation. d. Kit D - Phadebas® Insulin Test; Pharmacia Diagnostics This solid phase RIA utilized antibodies covalently bound to Sephadex® particles as described by Wide and Parath (1966). Manufacturer's protocol was followed. e. Kit E - RSL Insulin Kit; Radioassay Systems Laboratories, Inc. This immunoprecipitation RIA utilized double antibody principle as described by Soeldner and Slone (1965). Manufacturer's protocol was followed. 2. Sources 9; Canine and Commercial Control Sera Canine insulin high sera (K-9 Ins Hi) and canine insulin low sera (K—9 Ins Lo) controls were collected from four dogs. Two dogs were fasted 30 hours prior to blood collection to produce a control serum with 14 low insulin concentration (K-9 Ins Lo). Two dogs were given 50% Dextrose (1 gm/kg) intravenously 20 minutes prior to initiation of blood collection to produce a control serum with high insulin concentration (K-9 Ins Hi). All dogs were atropinized (atropine sulfate, 0.2 mg/pound body weight) and anesthetized (sodium thiamylal, methoxyfluorane, to effect) prior to catheterization of right carotid arteries. Blood was collected in plain glass tubes and allowed to clot for1-2 hours at room temperature. Sera from each pair of dogs was pooled, mixed, and ali- quoted into appropriately labelled plastic vials. Control sera were fro- zen (-25°C) until the day of analysis. Two levels of commercial lyOphilized control sera (Corning Medical) were used in all assays of this experiment. These control sera contained porcine insulin with reference values of 44 1 12 uU/ml (CS-1) and 137 i 22 uU/ml (CS-2). Reconstituted sera were analyzed on preparation days or within shelf life recommended by manufacturer. 3. Criteria for Evaluation and Comparison 2£.§£§ Kits a. Dilutional Parallelism Dilutional parallelism was evaluated by two methods: linear regression of expected versus measured concentration and comparison of undiluted concentration to corrected concentration. Zero standards of each respective RIA kit were used for diluents. Duplicate determinations of each dilution were made in each replicated assay. Dilutions of 4/5, 3/5, 2/5, 1/5, and 1/10 were made the day of assay according to the following procedure: 15 1. 4/5 dilution: 800 H1 K-9 Ins Hi + 200 ul diluent, vortex 2. 3/5 dilution: 600 ul K-9 Ins Hi + 400 ul diluent, vortex 3. 2/5 dilution: 800 ml K-9 Ins Hi + 1200 ul diluent, vortex 4. 1/5 dilution: 800 ul 2/5 dilution + 800 ul diluent, vortex 5. 1/10 dilution: 800 ul 1/5 dilution + 800 U1 diluent, vortex Linear regression (y=mx + b) of expected versus measured con- centrations were calculated on a programmed calculator‘e). Linear regression figures for replicated assays were averaged for comparison of RIA kits. Mean concentration of K-9 Ins Hi was compared to corrected con- centration of dilutions as per cent of original concentration. b. Precision Duplicate determinations of each control were made at the beginning and end of each set of assayed samples. Intra-assay coefficient of variation (C.V.) for each control was calculated from mean concentrations of duplicates. Mean intra-assay C.V. for each RIA kit was calculated using mean intra-assay C.V. of each control. Inter-assay C.V. for each control was calculated from mean IRI concentration of replicated assays. Mean inter-assay C.V. for each RIA kit was calculated using inter-assay C.V. of each control. Intra-assay C.V. and inter-assay C.V. of less than 10 were selected as acceptable limits, 10-15 as questionable and greater than 15 as unacceptable. c. Sensitivity Ability of each insulin RIA kit to detect small quantities of IRI was assessed by three methods. eModel 1930, Electronic Display Calculator for Statistics; Moore, The Calculator Company, Orange, N.J. 16 1. Sensitivity limit of each assay was defined as twice the standard deviations of zero binding. Transformations of zero tube Cpm to UU/ml were done using logit-log transformation formulas. The following equations were to calculate sensitivity limits: y=mx + b where: y = logit = ln ( U ). 100-U where U = lowest cpm of zero concentration duplicate X 100 mean cpm of zero concentration duplicate or U = B/Bo x 100 b= y-intercept, extrapolated from logit-log transformation of standard curve m = slope, regression coefficient of logit-log transformation of stan- dard curve x = log [uU IRI/ml] represented by cpm difference from mean cpm of zero concentration duplicates. Antilog x equals pU IRI/ml above "0" concentration representing cpm dif- ference from mean cpm of zero concentration duplicates. Corresponding uU IRI/m1 in replicate assay was also calculated. Each uU IRI/ml and corresponding negative uU IRI/ml were utilized as variables in calcu- lating standard deviation; therefore, mean uU IRI/ml would equal zero. 2. The amount of IRI required to give 10% displacement in IB/F dose-response curve was estimated by mathematical interpolation of IB/F versus concentration standard curve. Each standard tube of repli- cated assays was considered as an individual point. Dose response was assumed to be linear between zero concentration and the standard which gave at least, or close to, 10% decrease in tracer binding. Mean uU/ml and standard deviation were calculated from four interpolated concentrations. 17 3. The amount of IRI required to give 10% displacement in B/BO dose-response curve was estimated by graph interpolation or extrapo- lation of logit B/BO versus logarithm of dose response curve. Mean and standard deviation of IRI concentrations were calculated. d. Comparison of Immunoreactive Insulin Concentrations in Four Control Sera Eight determinations of each control concentration of IRI were made. These concentrations represented duplicate tubes at beginning and end of replicated assays. For each RIA kit, mean concentrations and standard deviations were calculated for each control serum. One way analysis of variance was used to calculate f values. Differences in means among kits were analyzed with Tukey's Honest Significant Difference (HSD) test (Gill, 1978). 4. Data Reduction Scintillation data (cpm) were reduced to UU IRI/ml via computer programs(f) using logit-log transformation. B. Experiment 2 - Screening Evaluation.g£ Commercial Cipeptide Radioimmunoassay 1. RIA Kit Evaluated: Kit F - Radioimmunoassay of Human C-peptide; CalbiochemrBehring Corp. This immunoprecipitation assay utilized double antibody principle. Synthetic human C-peptide was used as immunogen and for standards. Tyrosylated synthetic human C-peptide was iodinated with 125Iodine for use as a tracer. Manufacturer's instructions were followed except total counts were determined in individual duplicate tubes. (Component Lot #‘s are listed in Table A1.) (f)Animal Health Diagnostic Laboratory, Endocrine Diagnostic Section, Lansing, Michigan 48909 18 2. Sources of Samples Canine insulin high serum (K-9 Ins Hi) and canine insulin low serum (K-9 Ins Lo) were used for canine samples. Collection of these sera was outlined in Materials and Methods, Experiment 1, page 13. Human sera was collected from a volunteer to serve as a source of human C-peptide. A clinically healthy male, 31 years old, ingested 50% Dextrose (2 g/kg) orally after an 18 hour fast. Fasting, 30 minute, and 60 minute blood samples were collected and sera harvested. Sera were frozen (-20°C) until the day of analysis. 3. Criteria for Evaluation of RIA Kit Based on previous reports of lack of species cross-reactivity (Chance, 1972; Kemmler et al., 1972; Rubenstein and Steiner, 1971) screening evaluation of this assay was limited to a predicted clinical correlation and dilutional parallelism. Sera obtained from fasting and glucose challenged normal humans should have low and high C-peptide concentrations, respectively (Block et al., 1972). Similar results would be expected in the dog based on pancreatic vein concentrations of C-peptide in response to glucose (Kawanishi et al., 1977). Multiple dilutions of K-9 Ins Hi were used to assess dilutional parallelism. 4. Data Reduction Scintillation data cpm were reduced to ng IRC-P/ml via computer programs using logit-log transformations.(f) C. Experiment 3 - Screening Evaluation 9; Commercial Glucagon Radioimmunoassays 1. Commercial Glucagon Radioimmunoassays Evaluated a. Kit G - Radioimmunoassay of Pancreatic Glucagon; Cambridge Nuclear Radiopharmaceutical Corporation 19 This immunoprecipitation RIA utilized double antibody principle. No modification of the manufacturer's protocol was made in this experiment. b. Kit H - RSL Glucagon Kit; Radioassay Systems Laboratories, Inc. This RIA utilized double antibody principle. Addition of duplicate total count tubes was the only modification to the manufacturer's procedure. (Lot numbers of Kit G and Kit H and corresponding reagents used in this experiment are listed in Table A1.) 2. Sources of Control Plasma Canine glucagon plasma #1 (K-9 Glen-1) and canine glucagon plasma #2 (K-9 Glcn-Z) were collected from four dogs as described in Materials and Methods, Experiment 1, page 13. Plasma from fasted dogs was designated as K-9 Glen-1 and glucose challenged dogs as K-9 Glcn-Z. Blood collected in glass tubes containing K3EDTA was immediately mixed with 0.55 Trypsin Inhibiting Units (TIU)(9) per milliliter blood and placed in an ice bath. These tubes remained in ice until centrifuga- tion at 4°C. Plasma was pooled in a chilled glass container and ali- quoted into appropriately labelled glass vials and frozen (-25°C). Two commerical control plasmas‘h) designated as CP-Lo (850 i 450 pg/ml) and as CP-Hi (5,250 i 2000 pg/ml) were assayed in duplicate in replicated runs of Kit G and once in Kit H. 3. Criteria for Evaluation of RIA Kits Evaluation criteria of two glucagon RIA kits included precision, dilutional parallelism, sensitivity, and comparison of immunoreactive glucagon (IRG) concentration in four plasma controls. Criteria were (9)Aprotinin - Sigman Chemical Company, Box 14508, St. Louis, Mo., 63178 (h)Cambridge Nuclear Pharmaceutical Corporation, 575 Middlesex Turnpike, Billerica, Ma., 01865, (Lot #‘s 20 NCLC 1 and Lot #21 HCLC1) 20 calculated and compared as outlined for the insulin RIA kit experiment (Materials and Methods, Experiment 1, page 14). 4. Data Reduction Scintillation data cpm were reduced to pg IRG/ml via computer programs using logit-log transformation.(f) D. Experiment 3.” Modifications 2: Commercial Glucaggn Radioimmunoassay 1. Commercial Glucagon RIA Kit Evaluated Kit G‘i) of Experiment 3 was selected because of two factors: 1) Kit G had higher percent 30 than Kit H(j) (33% versus 27%). 2) Kit G had a 1 day incubation prior to addition of second antibody where Kit H had a 5 day incubation. One day incubation would be a better clinical assay with faster ”turn-around time”. 2. Modifications of Glucagon RIA Kit A technical representative of the RIA kit manufacturer recommended several modifications to increase sensitivity. These recommendations included the following: 1) increase analyte reagent volume; 2) dilute tracer by 50%; 3) dilute first antibody (Ab1) (sheep anti-porcine glucagon); 4) combinations of 1 through 3. a. Stage 1 Modifications 1. Reagent Preparations Normal kit serum was reconstituted with 4.5 ml of kit buffer instead of 10 ml. Tracer was reconstituted with 1.0 ml double distilled water, then 0.5 ml of reconstituted tracer was added to the 4.5 ml of (1)"Radioimmunoassay of Pancreatic Glucagon" Cambridge Nuclear Radiopharmaceutical Corporation, 575 Middlesex Turnpike, Billerica, MA, 01865 (j)RSL Glucagon Kit, Radioassay Systems Laboratories, Inc., Carson, CA, 90746 21 reconstituted normal serum. Additional standards (50 pg/ml, 35 pg/ml, and 70 pg/ml) were prepared by 1/200 dilution of 10,000 pg/ml, 1/10 and 1/5 dilution of 350 pg/ml, respectively, using kit buffer. First anti- body (Ab1) was reconstituted according to the manufacturer's guidelines and then diluted 8/10, 7/10, and 6/10 with kit buffer. The remaining assay components were reconstituted according to directions. 2. Assay procedure Tracer volume was 50 pl instead of 100 “1. All other reagent volumes were as directed by the manufacturer. A set of standards were analyzed with each dilution of Ab1. All time and handling conditions were constant for each set of standards. Determination of bound cpm was by time required to reach 10,000 counts for each tube. b. Stage 2 Modifications 1. Reagent preparation Two vials of kit normal serum were reconstituted. One vial was used to reconstitute tracer, then reconstituted tracer and second vial were combined and mixed. One 100 pg/ml standard was diluted 1/2 and 1/4 to obtain 50 pg/ml and 25 pg/ml standards. First antibody (Ab1) was reconstituted with 14 ml kit buffer to obtain a 5/7 (71%) dilution of Ab1. Other reagents were reconstituted as directed by the manufacturer. 2. Assay Procedure All volumes and procedures were followed as directed by the manufacturer. Determinations of bound cpm were made from time required to reach 10,000 counts in each tube. The first two assays were conducted in duplicate and the third in triplicate. 3. Source of Samples Four canine control plasmas and one commercial control plasma were used. The fasting control plasma (K-9 Glcn-1) and post-dextrose control 22 plasma (K-9 Glen-2) of Experiment 3 were included. Another fasting control plasma was collected from a dog after a 3 day fast and designated as pre-arginine control plasma (Pre-Arg). A post-arginine control plasma (Post-Arg) was collected from this dog after intravenous infusion of 10% L-Arginine solution‘k) at a dosage of 1 gm/kg over 20 minutes. Aprotinin-treated plasma was frozen (-20°C) until analysis. Serum was also collected before and after arginine infusion for glucose and IRI quantitation. Commercial control plasma(h) (850 i 450 pg/ml) was used in Stage 1 and parts of Stage 2. 4. Criteria for Evaluation of Modified Glucagon RIA Kit a. Stage 1 Binding parameters of original RIA (Experiment 3) were compared with results obtained after modifications. Displacement of tracer from BC was assessed with standards and with canine control plasmas. Control plasma IRG concentrations and duplicate C.V.‘s were compared. b. Stage 2 Reproducibility of standard curves was compared by tracer binding parameters and by linear regression data of logit-log transformation. Duplicate and inter-assay C.V.‘s were assessed for the five control plasmas. (k)L-Arginine-HC1 (25 gm) (United States Biomedical Corporation, Cleveland, Ohio) dissolved in 250 ml of 0.9% NaCl(aq), (resulting pH = 5.5) IV. RESULTS A. Experiment 1 - Screening Evaluations 9f Commercial Insulin Radioimmunoassays 1. Specificity By both methods of assessment (Tables 1 and 2), Kit C had the best dilutional parallelism. Kit A had good parallelism except for 1/5 dilution. Measured concentration of IRI in Kit B was consistently lower than the expected concentration. Kit D had good parallelism at 4/5 and 3/5 dilutions, but greater dilutions resulted in increasing ratios of measured versus expected concentrations. Kit E had good parallelism at 4/5 dilution, but greater dilutions yielded decreasing ratios of measured versus eXpected concentration. Table 1. Dilutional parallelism assessment of insulin RIA kits by linear regression Correlation Kit Coefficient (r) Slope (m) Y-Intercept A 0.976 1.088 -2.04 B 0.999 1.250 -1.15 C 0.999 1.006 0.61 D 0.978 1.282 -10.43 E 0.990 0.976 4.92 23 24 Table 2. Dilutional parallelism assessment of insulin RIA kits by comparison of corrected concentration to undiluted concentration Kit Undiluted 4/5 3/5 2/5 1/5 1/10 A 100 95.7 94.8 96.6 145.6 96.0 B 100 82.0 84.4 83.9 90.2 92.4 C 100 98.4 98.3 96.8 90.8 102.0 D 100 96.0 98.0 120.3 148.9 221.2 E 100 98.2 85.7 75.1 70.2 72.5 2. Precision All RIA kits gave acceptable precision if mean intra-assay C.V. and mean inter-assay C.V. are used as the criteria (Table 3). Kit C con- sistently had the best precision for all controls. Reproducible quan- titation of IRI in K9 Ins Lo was questionable in Kits A, B, and D (intra-assay) and Kits A, B, and E (inter-assay). Intra-assay precision was questionable in Kit D with CS-1 sera. Reproducible quantitation of IRI in CS-2 was questionable in Kit 8 (intra-assay), in Kit D (intra-assay), and in Kit A (inter-assay). Precision in all RIA kits was considered acceptable in quantitating IRI in K-9 Ins Hi. 3. Sensitivity Sensitivity limits ranged from 0.8 uU IRI/ml in Kit D to 4.4 uU IRI/ml in Kit A (Table 4). Amount of IRI required to displace tracer by 10% on 25 p.o h.m me. «two p.ow o.m ms am mcH mix v.0 p.m vv «Imo m.m m.mp m.h o.m hp on mcH mix m v.m m.ow mm NImU m.h m.m mm w: mcH mix m.m N.op pm leu o v m.~ P.op v.mp cm 0A mcH mix a m.m m.o mmp Nlmu c.m h.v vb Hm mcH mix ~.o N.v hm Flmo m.m m.p m.m h.m cm on mcH mum U m.m m.vp mmp mlmo m.~ P.h om am mcH mum v.~ v.m Pm wlmo m.b m.op o.op o.mp hp on mcH mum m w.PP m.m m—F Numo p.¢ ~.m mm mm mcH mIM m.h N.m mm leo ~.op m.bp m.m m.NP up on mcH mum < v": Nu: v": N": Amucv com: uwM maowucoo Houucoo mHouucoo Houucoo AHS\D:C HmH Houucoo Had coo: £06m Ham 2mm: nomm :mmz .>.U hummmnuwucH .>.U ammmMImuucH wwwx m n E *Fasting canine sera (K-9 Ins Lo), post-dextrose canine sera (K-9 Ins Hi), low commercial control sera (CS-1), high commercial control sera (CS-2) 28 B. Experiment 2 - Screening Evaluation pf Commercial ijeptide Radioimmunoassay 1. Predicted Clinical Correlation Based on logit-log transformation of scintillation counts, lower concentrations of IRC-P were found in K-9 Ins Hi than K-9 Ins Lo (Table 7). Concentrations of IRC-P in canine sera did not parallel corresponding increase in IRI concentrations. Concentration of IRC-P in human sera approximated reference values in the package insert and paralleled the increase in IRI concentrations (Table 7). 2. Dilutional Parallelism No parallelism was seen in dilutions of K-9 Ins Hi serum. 3. Displacement of Tracer in Undiluted and Diluted Canine Sera Tracer displacement in B/F index ranged from 0.8% to 7.9% and in B/Bo index from 0.5% to 5.2%. These displacements were considered below levels where quantitation can be reliable. C. Experiment 3 - Screening Evaluation pf Commercial Glucagon Radioimmunoassays 1. Specificity Assessment of dilutional parallelism was not possible as the experiment was designed. Kit G had poor precision in dilutions of K-9 Glen-1. Duplicate C.V.‘s ranged from 32.3 to 126.2. Kit H had poor precision and inadequate sensitivity to quantitate IRG in dilutions. Counts per minute (cpm) were higher than zero tubes in thirteen of twenty tubes. Duplicate C.V.‘s for those with "quantitated" IRG ranged from 11.2 to 81.6. 29 Table 7. Human C-peptide RIA evaluation Sample % Decrease in % Decrease in IRC-P IRI‘C) B/F Index‘a) B/Bo(b) ng/ml UU/ml Mean Mean Mean (n=6) (n=8) K-9 Ins Hi Dilutions 80% (n=2) 3.4 2.2 0.20 --- 60% (n=2) 0.8 0.5 0.04 --- 40% (n=2) 7.9 5.2 0.47 --- 20% (n=2) 6.8 4.5 0.40 --- 10% (n=2) 3.1 2.0 0.18 --- Human -Fasting 26.4 19.0 1.8 13 -3o min.(d) 53.6 42.9 5.4 68 -60 min.(d) 63.7 53.3 8.1 140 (a)% decrease B/F = (1- mean B/F) x ‘00 B/F @ 30 (b)% decrease B/BO = %Bo - mean %B/BO (°)Kit C - Experiment 1 (Corning Medical) (d)Time after 2 g dextrose/kg orally 30 2. Precision Neither assay (Kit G nor Kit H) had acceptable precision at IRG concentrations present in controls K-9 Glcn-1 or K-9 Glen-2 (Table 8). Inter-assay precision of CP Lo and CP Hi were greater than acceptable limits in Kit G. Table 8. Intra-assay and inter-assay precision of glucagon RIA kits Kit Control IRG Intra-assay C.V. Inter-Assay C.V. (pg/ml) Mean each Mean Mean each Mean mean Control Canine Control Canine n=2 Controls n=2 Controls n=2 n=2 G K-9 Glcn-1* 46 60.7 48.5 129.0 84.8 K-9 Glen-2* 80 36.3 40.7 CP Lo* 969 ---- 18.2 CP Hi* 6383 ---- 22.2 H K-9 Glcn-1* 7 98 55.5 76.0 67.4 K-9 Glen-2* 33 13 58.7 CP Lo* 67 --- ---- CP Hi* 81 --- ---- *Fasting canine plasma (K-9 Glen-1), post-dextrose canine plasma (K-9 Glen-2), low commercial control plasma (CP-Lo), high commercial control plasma (CP-Hi) 3. Sensitivity In three methods of assessing sensitivity, Kit H appeared to have better sensitivity than Kit G (Table 9). However, sensitivity limits of each kit approached mean IRG concentrations estimated in K9 Glen-1 (Table 8). Calculated concentration of IRG required to give 10% displacement of 31 Table 9. Sensitivity assessment of commercial glucagon RIA kits Kit Sensitivity Limit 10% Decrease in 10% Decrease in 0 + 2 S.D. B/F from Bo B/BO from Bo (Pg/m1) (n=4) (pg/ml) (pg/ml) (mean) (n=4) (mean) (n=2) G 31 73 105 H 6 10 14 tracer in B/F and B/BO dose response curves (Table 8) exceeded the IRG concentration estimated in K-9 Glen-1. 4. Comparison of Immunoreactive Glucagon in Four Control Plasmas Lack of precison (Table 8) prevented valid comparison of IRG concentrations. Kit G appeared to quantitate higher concentrations than Kit H. This observation was pronounced in IRG concentrations measured in commercial control plasmas (CP Lo and CP Hi). D. Experiment 4 - Modifications pf Commercial Glucagon Radioimmunoassay 1. Stage 1 Binding parameters of modified RIA were compared with original RIA (Table 10). Total counts decreased 31% when 50% tracer was used. Little change was seen in non-specific binding. Percent bound at zero (% 80) decreased with greater dilutions of first antibody (Ab1). The % B/Bo in standards also decreased with Ab1 dilutions except for an apparent rever- sal of % B/Bo at 70 and 100 pg/ml. Greater than 10% displacement (% B/BO partition) occurred at 100 pg/ml in 70% Ab1 assay and 50 pg/ml in 32 Table 10. Binding parameters of original and modified glucagon RIA Original Modified Assay - 50% Tracer & Assay 80% Ab1 70% Ab1 60% Ab1 (mean) (mean) (mean) (mean) (n=4) (n=2) (n=2) (n=2) Total counts (cpm) 8594 5970 5970 5970 %BNSBa 2.6 2.7 2.6 3.2 %BO 33.1 30.9 28.3 26.6 %B/BO @ 35 pg/ml ---- 99.1 96.0 91.4 %B/BO @ 50 pg/ml ---- 96.6c 95.0 88.9 %B/BO @ 70 pg/ml ---- 87.6 93.9 84.6 %B/BO @ 100 pg/m1 90.4 95.2c 86.7c 85.3c %B/BO @ 350 pg/m1 74.5 67.9 66.4 62.9 %B/BO @ 1000 pg/ml 51.7 46.3 42.0 41.4 %B/BO @ 3000 pg/ml 39.7 31.4 31.3 29.9 %B/BO @ K-9 Glcn-1 95.6 82.0 85.4 78.7 %B/BO @ K-9 Glcn-2 91.5 76.1 80.9 76.5 %B/Bo @ CP-Lo 58.6 54.8c 53.5 51.3 %B/Bo @ CP-Hi 27.4 26.0c 20.6 22,0 Y-Interceptb 5.84 6.55 6.50 5.17 Regression Co- efficientb -0.800 -0.939 -0.951 -0.785 Correlation Co- efficientb -0.994 -0.945 -0.986 -0.992 a - percent bound due to non-specific binding b - from linear regression of logit vs log transformation c - single determinations 33 60% Ab1 assay. Erroneous binding at 70 and 100 pg/ml standards in 80% Ab1 assay prevented assessment of sensitivity. Both canine control plasmas (K—9 Glcn-1, K-9 Glcn-2) displaced tracer greater than 10% at all Ab1 dilutions, where neither had in the original assay. The Y-intercept and regression coefficients of original assay were between 70% Ab. and 60% Ab1 standard curves. Best correlation coefficient was seen in 60% Ab1 standard curve. Mean concentration of IRG in K9 Glcn-1 were 3, 3.2, and 4.7 times higher (60%, 70%, and 80% Ab1 respectively) than IRG esti- mated in Experiment 3 (Table 11). A similar trend was seen in K-9 Glcn-Z with IRG concentration of 2, 2.5, and 3.9 times higher (60%, 70%, and 80%, respectively) than IRG estimated in Experiment 3. Lower IRG Table 11. Immunoreactive glucagon quantitated in original and modified glugagon RIA x-9 Glen-1a x-9 Glen-2a cp-Loa Cp-Hia (pg/m1)b (pg/m1)b (pg/m1)° (pg/m1)° Original Assay 46(26.1)d 80 (36.4)d 969 (4.7)d 6383 (5.4)d Modified Assays 80% Ab1 213 (3.3) 311 (4.7) 869 (---) 3244 (---) 70% Ab1 147 (25.0) 205 (25.3) 804 (3.1) 3859 (0.4) 60% Ab1 137 (5.8) 160 (10.4) 679 (17.2) 3672 (10.4) fasting canine plasma (K-9 Glen-1), post-dextrose canine plasma (K-9 Glcn-2), low commercial control plasma (CP-Lo), high commercial control plasma (CP-Hi) b - mean, n=8, from Experiment #3 - mean, n=4, from Experiment #3 numbers in paranthesis represents duplicate C.V. On 0 I 34 concentrations were quantitated in both low commercial control (CP-Lo) and high commercial control (CP-Hi). Only half of the duplicate C.V.‘s were less than 10.0 in both original RIA and modified assays. Single determinations of CP-Lo and CP-Hi were made in 80% Ab1 assay because of insufficient control plasma volume. 2. Stage 2 Best correlation coefficent in replicated assays was -0.980 (Table 12). Within this limit of poor correlation, displacement of tracer in standard curves appeared to be reproducible. Percent bound at zero con- centration (% BO) ranged from 33.3% to 40.1%. At 25 pg/ml standard, bound tracer was higher than at zero concentration (102.1%). ”Analyte" in zero tubes was kit buffer. Nine of fourteen duplicates had C.V.‘s of less than 10.0 (Table 13). Only one of five control plasmas had an inter-assay C.V. of less than 10 (K-9 Glen-1). Two controls had greater than 20 inter-assay C.V. Concentration of IRG estimated in a fasting dog (Pre-Arg) and post-arginine infusion (Post-Arg) appeared to be approxi- mately the same. 35 Table 12. Reproducibility of standard curves in modified glucagon RIA (50% Tracer & 71% Ab1) Run 1 Run 2 Run 3 Mean S.D. Total Counts (cpm) 5071 5374 5024 5156 190 ‘BNSBa 5.1 4.0 6.2 5.1 1.1 %B0 40.1 34.4 33.3 35.9 3.7 %B/BO @ 25 pg/ml 101.4 102.7 ---- 102.1 --- %B/BO @ 50 pg/ml 93.2 92.5 91.7 92.5 0.8 %B/BO @ 100 pg/ml 84.8 83.4 84.1 84.1 0.7 %B/BO @ 350 pg/ml 63.7 69.8 69.7 67.7 3.5 %B/Bo @ 1000 pg/ml 39.6 41.2 36.6 39.1 2.3 Y-Interceptb 6.42 6.02 6.13 6.19 0.2 Regression Coef.b -0.996 -0.912 -0.945 -0.951 0.042 Correlation Coef.b -o.974 -0.980 -0.970 -0.975 0.005 a - percent bound due to non-specific binding b - from linear regression of logit vs. log transformation Table 13. Immunoreactive glucagon quantitated in modified glucagon RIA (50% Tracer & 71% Ab1) Samplea Run 1 Run 2 Run 3 Mean Inter-assay (pg/m1) (pg/m1) (pg/ml) (pg/m1) C .v. x-9 Glen-1 180 (1.7)b 125 (0.2)b 185 (19.3)C 163 20.4 K-9 Glen-2 270 (0.2) 242 (17.4) 274 (14.1) 262 6.7 Pre-Arg 156 (2.2) 136 (8.7) 187 (11.1) 160 16.1 Post-Arg 193 (5.8) 148 (12.3) 194 (9.1) 178 14.7 CP-Lo 1056 (4.3) 584 (3.3) --- 820 40.7 a - fasing canine plasma (K~9 Glcn-1), post-dextrose canine plasma (K-9 Glcnv2), fasting canine plasma before arginine infusion (Pre-Arg), canine plasma after arginine infusion (Post-Arg), low commercial control plasma (CP-Lo). b - number in paranthesis represents duplicate C.V. c - number in paranthesis represents triplicate C.V. V. DISCUSSION The purpose of clinical chemistry procedures is to identify normal and abnormal values that can be used to establish presence or absence of disease. If abnormal, data may help define type, severity, and location of lesion. However, accurate clinical chemistry interpretation is depen- dent on the reliability of the procedure used to quantitate substances of interest. Radioimmunoassays cannot be assumed to be valid for a par- ticular species. One must evaluate assays and establish that they can quantitate analytes reliably before medical decisions are made. A. Experiment 1 - Screening Evaluation 2; Commercial Insulin Radioimmunoassays Kit C had the best dilutional parallelism (specificity), best preci- sion (reproducibility), and good sensitivity. These findings suggested that Kit C would be the best among the RIA kits tested for further evaluation, for establishment of reference canine IRI concentrations, and for clinical and research investigations of diseases related to insulin. Precision and sensitivity data obtained in experiment were similar to performance characteristics listed in package insert (Table A2). Accuracy of IRI quantitation was not assessed because canine IRI stan- dards were not available. Kit A had good dilutional parallelism except for one duplicate at 1/5 dilution (189% of expected concentration). Assuming procedural error, dilutional parallelism was similar to Kit C. Precision assessment did 37 38 not meet precision limits set for screening evaluations. Intra-assay and inter-assay C.V.‘s were above 10 at 14 uU/ml. Inter-assay C.V. at 115 uU/ml was also unacceptable. This experiment's precision data was equal to or better than performance characteristics stated in package insert (Table A2). Sensitivity assessment was considered adequate. Kit B quantitated consistently lower concentrations in dilutional parallelism evaluation (82-92.4% of expected). Dilutions were made with kit buffer consisting of a phosphate buffered saline (PBS) with bovine serum albumin. This buffer may have altered binding properties of anti- body (affinity) or altered canine IRI (antigenicity). Precision was not acceptable at 17 uU/ml and at 133 uU/ml. Variability at low con- centrations exceeded manufacturer's stated performance characteristics (Table A2). Sensitivity was adequate. Lack of dilutional parallelism must be considered as a major deficiency and thus specificity of Kit B must be considered questionable. Evaluation of parallelism with insulin- deficient canine serum would be necessary to further evaluate specificity. Kit D had poor dilutional parallelism at 2/5, 1/5, and 1/10 dilutions of K-9 Ins Hi. Diluent was kit buffer consisting of PBS with protein. Sensitivity limit (0.8 uU/ml) suggested adequate sensitivity to quan- titate IRI in respective dilutions (22.4, 11.2, 5.6 uU/ml). Partition index assessment of sensitivity (B/F = 10.6 uU/ml, B/BO = 10.1 uU/ml) indicated that at 1/5 and 1/10 dilutions, quantitation problems may have been due to sensitivity. Individual determinations at 2/5 dilutions were 107.2% and 133.4% of expected. Assuming procedural error in second determination, it may be concluded that Kit D has dilutional parallelism within its range of sensitivity. Intra-assay precision was questionable 39 in three or four controls; however inter-assay precision was good. Assessment of intra-assay precision in this experiment exceeded perfor- mance characteristics stated in package insert (Table A2). However, inter-assay precision was better. In this experiment, Kit D had relati- vely poor performance and could not be recommended for further evaluation. Kit E quantitated progressively lower IRI concentrations with increasing dilutions. Diluent was kit buffer consisting of PBS with guinea pig serum. Based on this observation, Kit E specificity must be questioned even though its package insert stated 100% cross-reactivity with canine insulin. Intra-assay precision was good, but inter-assay precision at low concentrations (17 uU/ml) was unacceptable. Sensitivity was adequate. Performance characteristics were not stated in Kit E package inserts. Kit E could not be recommended for quantitation of canine IRI unless other assessment methods demonstrated specificity. According to statistical comparison of IRI concentrations in four control sera, evaluated radioimmunoassays did not quantitate equal IRI concentrations at all four levels. This finding dictates that reference values for clinical and research investigations must be established for each RIA utilized. Utilization of published reference values obtained with another RIA must not be encouraged. Inter-laboratory comparison of RIA's would be valuable to establish inter-laboratory reproducibility. Even though statistically different, IRI concentrations in low control (14-20 uU/ml) may not be clinically significant. Differences were more apparent at higher concentrations (95-146 uU/ml) and clinically could be significant. Reference values involving elevated IRI con- centration should be established for each investigator's RIA. These 40 reference values may include IRI levels during glucose tolerance tests (intravenous or oral), glucagon tolerance tests, and expected con- centrations found in functional beta cell neoplasms (insulinomas). Indices used to interpret pancreatic release of insulin (insulin-glucose ratios, insulinogenic index) would also require reference values (reference values for Kit C are summarized in Appendix B). If IRI quan- titation is offered by a reference laboratory, information supporting storage and handling of samples should be provided (preliminary studies on stability of canine IRI are found in Appendix C.) Reasons for differences in mean IRI concentrations quantitated among kits may be multiple. These reasons may include different anti-insulin antibodies, efficiency in separation techniques, and variations in calibration of standards. Package insert in Kit C stated that its stan- dards were calibrated with a World Health Organization immunoreactive reference standard. Supplier of Kit B reported‘l) that its standards were also calibrated against the same WHO immunoreactive standard. Other manufacturers did not supply calibration information. Differences in RIA kit standards could be evaluated by assaying standards as unknowns in other RIA's. B. Experiment 3 - Screening Evaluation pf Commerical ijeptide Radioimmunoassays No evidence was found to indicate that Kit F antibody (Ab1) had cross-immunoreactivity with canine C-peptide. Small displacement of tracer in canine controls and in dilutions of canine sera could have two explanations. First, this RIA may not be designed to quantitate very low levels of canine C-peptide, i.e., below sensitivity of assay. (1)(Personal Communication) 41 Sensitivity limit calculated from one assay was 0.28 ng/ml. Sensitivity stated in package insert was 0.8 ng/ml (90% B/BO). As reports of cir- culating peripheral canine C-peptide were not found, this possibility cannot be ruled out. The second and more likely explanation for the minimal tracer displacement is lack of cross-immunoreactivity. Binding of tracer to first antibody (anti-C-peptide) was less in canine sera than in zero con- centration standards. This may be due to greater non-specific binding rather than true competition for antibody binding sites. This explana- tion is consistent with reports stating requirements of species specific antisera (Chance, 1972; Kaneko et al., 1974; Kemmler et al., 1972; Rubenstein and Steiner, 1971). Dr. Altszuler(m) has obtained synthetic canine C-peptide from Japan but has not attempted to develop a canine C- peptide assay. C. Experiment 3 - Screening Evaluation pf Commercial Glucagon Radioimmunoassays As the experiment was designed, it is difficult to reliably evaluate these RIA's. Neither Kit G or Kit H had reproducible quantitation of IRG present in canine controls (K-9 Glcn-1, K-9 Glcn-Z). This lack of preci- sion may have been due to inadequate sensitivity. Kit G precision in commercial controls (CP-Lo and CP-Hi) was inadequate. Concentrations of IRG in four controls appeared to differ markedly. Kit G quantitated 14 times more IRG in CP-Lo and 79 times more IRG in CP-Hi than did Kit H. This degree of variation may be due to differences in standards, (m)(Personal Communication) Dr. Norman Altxzuler, Department of Pharmacology, School of Medicine, 550 First Ave., New York, N.Y. 10016. 42 differences in calibration of standards, or procedural inaccuracies. To my knowledge, a reference immunoreactive standard is not available for calibrating standards and thus variations in IRG quantities among RIA's must be expected. Based on this experiment, neither Kit G nor H could be recommended for clinical or research IRG quantitation. D. Experiment 4 - Modifications pf Commercial Glucagon Radioimmunoassay This experiment was performed in two stages. Stage 1 involved three simultaneous assays with diluted tracer and different dilutions of first antibody (Ab1). Stage 2 involved three replicated assays with diluted tracer and diluted first antibody. First stage modifications appeared to increase the assay's sen- sitivity by increasing initial slope of dose-response curve. With greater first antibody dilutions, the initial decrease in tracer binding became larger. This finding would be an improvement over the apparent inadequate sensitivity found in Experiment 3. Apparent reversal of expected binding in 80% Ab1 at 50 pg/ml and 70 pg/ml cannot be explained to author's satisfaction. Technician error was not a primary factor as counting data and assay set-up were thoroughly checked. Problems may have been related to lack of precision at low dose range. Another problem was an apparent increase in IRG in canine controls and a decrease in commercial controls as compared to original assay quan- titations (Experiment 3). Dilutions of tracer and first antibody should allow analytes greater opportunity to bind to antibody and thus diSplace tracer. However, the effect should be equal for both standard IRG and unknown IRG. Canine controls had been frozen (-25°C) for approximately 3 months between Experiment 3 and 4. Thus, dehydration may have 43 concentrated plasma components. If storage conditions do not explain observed differences, this assay's specificity for canine IRG must be questioned. For a RIA to reliably quantitate, standards and unknown IRG come petition for antibody binding sites must be constant. Observed charges in the modified assay would suggest that antibody-standard analyte interactions were altered differently than antibody - unknown analyte interaction. Modifications may have decreased antibody affinity for analytes in standards and thus canine IRG concentrations appeared higher. On the other hand, modifications may have increased antibody affinity for canine IRG and thus appeared to displace more tracer than in the original assays. Apparent decreasing concentrations of canine IRG with greater dilutions of the first antibody also suggest altered kinetics in analyte- antibody interactions. A major problem in the original assays was lack of precision. Duplicate C.V.‘s for canine controls were better in 80% Ab1 and 60% Ab1 assays. However, mean duplicate C.V.‘s for standards were 1.43, 4.59, 2.13, and 4.59 in original, 80% Ab1, 70% Ab1, and 60% Ab1 assays, respectively. This lack of duplicate reproducibility in standards made construction and interpretation of standard curves difficult. In the second stage of glucagon RIA modification, precision remained a major problem. Duplicate (or triplicate) C.V.‘s for canine controls ranged from 0.2 to 19.3 with only seven of twelve C.V.‘s less than 10.0. Only one of four canine controls had inter-assay C.V. of less than 10.0. Mean C.V. for standards in replicated assays with 71% Ab1 were 4.5, 3.6, and 3.2. This degree of variation in standards made construction of 44 standard curves difficult. This problem was demonstrated by relatively poor correlation coefficients. Removal of standards that were out-liers from best-fit curve could improve linearity, but should not be encouraged. This consistent lack of precision in original and modified RIA's suggests that reliable quantitations of canine IRG would be difficult. If there are a large differences in IRG concentrations between normal dogs and those dogs with expected elevated IRG concentrations, perhaps this commerical glucagon RIA could be used as a semi-quantitative assay and have clinical applications. Another problem arises from the lack of precision, i.e., evaluation of dilutional parallelism. Results from this study indicate that a markedly elevated canine IRG pool would be needed before dilutional parallelism could be re-evaluated. Another problem identified in the second stage (71% Ab1 modification) was an apparent greater tracer binding at 25 pg/ml than at 80. This "hook effect" has been described by Cresto and Yalow (1974), who suggested that excess antibody gives antigen-antibody complexes at low analyte concentrations. An attempt to produce a high canine IRG control plasma by arginine infusion was unsuccessful. Post-arginine infusion concentration of IRG was approximately 18 pg/ml higher than the fasting level. Clinical investigations of glucagon's role in a variety of canine disorders is needed. This commercial glucagon radioimmunoassay or modi- fications cannot be recommended at this time for clinical investigations. Results of this study dictate that additional modifications or investiga- tions may be necessary or that another glucagon RIA should be evaluated. If additional evaluation of this RIA is desired, several factors must be considered. Due to lack of precision, triplicates will probably be 45 required to give best semi-quantitation. This will necessitate increased expense in canine IRG quantitation and increased sample volume. Another approach to assay modification may be to increase analyte volume. This would dictate purchasing increased volumes of standards and require larger sample volumes. Abnormal (elevated) concentrations of canine IRG should be demonstrated to be at least 100% higher than normal concentrations. A control plasma with IRG concentrations at least 300% higher than normal concentrations will probably be needed for assessment of dilutional parallelism. Based on results of one arginine infusion, another method of stimulating glucagon production and release will be needed. It is recommended that a reference laboratory be used to quan- titate canine IRG concentration in clinical cases and future control plasma before any further investigations are considered. VI. SUMMARY AND CONCLUSION Commercial radioimmunoassay (RIA) kits deve10ped for quantitation of human immunoreactive insulin (IRI), immunoreactive C-peptide (IRC-P), and immunoreactive glucagon (IRG) were evaluated for their ability to quan- titate corresponding canine immunoreactive hormones. Evaluation criteria of five insulin RIA kits included precision, dilutional parallelism, sensitivity, and comparison of IRI concentrations in control sera. One RIA kit (Corning Medical) had good dilutional parallelism, consistently good precision, and adequate sensitivity. Other RIA kits evaluated had poorer performance in dilutional parallelism or precision. No two RIA kits quantitated the same IRI concentrations in all four control sera. Reference values for fasting canine IRI, canine IRI response to High Dose Intravenous Glucose Tolerance Test (HD-IVGTT), and stability of canine IRI were determined with Corning's insulin RIA kit. Canine sera caused minor tracer displacement from anti-C-peptide antibody in the human C-peptide RIA kit (CalbiochemrBehring Corporation). This finding would be consistent with a lack of species cross-immuno- reactivity. Two commercial glucagon RIA kits had poor precision in quantitating low concentrations of canine IRG. Modifications of one RIA kit (Cambridge Nuclear Radiopharmaceutical Corporation) appeared to increase sensitivity, but precision remained inadequate. Neither glucagon RIA kit can be recommended for quantitation of canine IRG until additional evaluations are conducted. 46 47 Additional evaluation or utilization of commercial RIA kits are needed. Reference values for IRI during glucagon tolerance tests are needed. A RIA for canine C-peptide would be an asset in evaluation of canine disorders involving insulin. Immunoreactive glucagon deter- minations will be required before glucagon's role in a variety of clini- cal disorders in dogs is understood. APPENDICES APPENDICES Appendix A - Component, Performance Characteristics, and Binding Parameters The components and lot numbers of commercial radioimmunoassays evaluated in Experiments 1, 2, and 3 are listed in Table A1. Performance characteristics stated in insulin RIA kit package inserts are listed in Table A2. Binding parameters of commercial radioimmunoassays in Experiments 1, 2, and 3 are listed in Table A3. Table A1. Components and lot numbers of evaluated commercial radioimmunoassays RIA Kit Tracer Primary Antibody Standards Imunoggn; Antibody A - Species Source bovine bovine, guinea pig human porcine Lot # 228 228 228 B - Species Source porcine porcine; guinea pig human Lot # I/117#3 IA/45, SCO1 B 1K519#7 C - Species Source porcine porcine; guinea pig porcine Lot # 20040 03109 01639A D - Species Source porcine porcine; guinea pig porcine Lot # 2505 1266 U 2919 U E - Species Source porcine porcine; guinea pig porcine Lot # 46 45 43 F - Species Source syn. human syn. human; rabbit syn. human Lot # 02A010 025005 x-92x096 G - Species Source porcine porcine; rabbit porcine Lot # 1209512L 20AFLA4 20AGLC1 H - Species Source porcine porcine; rabbit porcine Lot # 102L 95 104 48 Houucoo some now .>.U c602 may ma oouuommu .>.U .cv 66:8666 no: Loo HE\D: o.o Show mcoaumw>oo tumocmum N an wouooaumu cfiasmca mo unsoam Any masocan Hmwuacw mo ommouomo wm 606560 umcu cfiasmcw mo ucsosm Adv 49 6666.6 0680m0m1o wmo.c commosam mop cqasmcwowm nee. .ossonom. essence see. .6663 essence 111 1111 1111 1111 coop Acmabgv cdasmcfi m wax 6.6 6.6 o... sonna1snusa A03 Hexoa m.~ v 6.6s o 6.66 6 6.6.6 66m v eaasneaono 6 use ocOEuon we nuzowm amass m.o v.m ammmM1umucH mo commando ~.m o.v >666616uucw am.m caaamswoum ADV HE\D: m.. 111 o.mmF@ o.Fm© poop ceases“ 0 pax apo.o v ooaumwm1u aN.o somcosam wo.h cdanmcwoum mom Aumuv caasmsw 6.6 ..o 6.6. sonnn1souaa woo. Aoaaooud. seasons 6.6 6.6 ..6 annnm1snusa see. mosaeon. caasnaa Amy HE\D: m.~ m.nh @ p.o~ w v.0 @ ace. Assess. caannca m uHM o.c~ W c.cN_W o.o~ w.ammmm1uovsa o.~p 6.6. 6.6 stonnnsnusa 111 cc. 8 co e Nm 9 1111 1111 a van HE\D: w HE\D: w HE\D: w Nua>euomou mmowo 1 mosdunnam aua>wuamsom «A.>.Uv coamaooum huaodmaoomm muuomCH mmmxomm vex «Hm caasmcw ca possum noaunauouomumcc oocmauomuwm .Nd manna 50 Ha\umH mm 1 m a u apex Hs\m1omH on 1 m use Hexan o: 1 ms.o.o.m.< massage Sound mamcam 1 m uwx umooxmauv m.mlh.Nm coo.plop N.m h.m~ ¢o¢.NP : o.mp1v.om coo.oploow h.m p.mm vom.w U N.FP1~.mm omlp m.— m.vm mmh.m h n.5P1m.pm oowim.~ m.m w.®m pop.m m m.hm1o.wm omm1m 111 o.PP cmm.mp a m.m~1m.Nm ooVImN.m 111 m.~m www.m U m.mw1m.Pm oomlm 0.0 m.Fv ppm.pp m h.omlm.mm cmp1op N.m m.mN moo.wN 4 Acmmfiv moumocmum Anvmpumvcmum Amucv com: Amucv com: .vucv :60: mo mmcmm om\m w «0 omcmm mmzmw omw AEQOV musaoo Hove? wax Amvmhmmmm owumoaammu :H mumumsmumm unanswn vex mHm .md wands 51 Appendix B - Immunoreactive Insulin Reference Values: Fasting and High Dose Intravenous Glucose Tolerance Test (HD-IVGTT) 1. Insulin Radioimmunoassay Kit Utilized "Immophase"TM, Insulin 1251 Radioimmunoassay Test System (Corning Medical) was used for IRI determinations. Insulin RIA quality control was monitored during HD-IVGTT sample an- alysis by total counts, %Bo, linear regression values of logit-log trans- formation of standard curve, and duplicate, intra-assay, and inter-assay coefficients of variation. Commercial control sera* were assayed in duplicate before and after unknown samples. Canine control sera (K-9 Ins Lo, K-9 Ins Hi) were assayed in duplicate before, after, and between each set of HD-IVGTT samples. 2. Reference Canine Population Twenty clinically healthy dogs, 6 males, 12 females, 2 spayed females ranging from 8 months to 4 years of age, and 10.3 to 36.0 kilogram body weight were used. Breeds represented in this population included Golden Retriever (4), Black Labrador (8), Retriever-Beagle cross (4), Springer Spaniel (2), and mixed breed (2). Fifteen dogs were housed in a research kennel, five dogs were privately owned pets. Reference papulation qualifications included fasting hematologic and selected serum chemistry values within normal reference ranges for adult dogs established by the Clinical Pathology Laboratory, Veterinary Clinical Center, Michigan State University. Reference population hemato- logic and chemistry values listed in Table B1. No dog had histories of *Corning Medical, Corning Glass Works, Medfield, MA, 02052 52 disorders that would have potential for affecting carbohydrate tolerance, i.e., pancreatic, adrenal, or hepatic disease. 3. High-Dose IV-GTT Procedure and Sample Handling Dogs were fasted overnight (15-18 hours). An intravenous Teflon® catheter* was inserted and secured in the right cephalic vein. Pre-IV-GTT, 5, 10, 15, 30, and 60 minute samples were collected via catheter. Ninety and 120 minute samples were collected from either right or left jugular veins via venipuncture. Heparinized saline (1000 USP Na Heparin in 250 ml 0.9% saline) was used for catheter flush both before and after catheter-obtained samples. Approximately one ml of blood was discarded prior to catheter collection of samples. Catheter collected samples were collected within one minute of schedule time. Ninety and 120 minute samples were collected within 1-5 minutes of scheduled time. After collection of pre-IV-GTT samples, 50% dextrose was administered at a dosage of 1 gm/kg body weight (30-60 secs) in the left cephalic vein. Timing of subsequent samples began at the end of dextrose administration. Blood samples for glucose and IRI determination were allowed to clot for 45-75 minutes and centrifuged at 2760 RPM (1000 g) for 10-15 minutes. Harvested sera was divided into two parts, 300 01 for serum glucose and the remainder (greater than 1 ml) for IRI. Serum for glucose determination was assayed the day of collection or stored over- night (4°C). Serum for IRI assays was frozen (-25°C) in apprOpriately labelled plastic vials until the day of analysis. *I.V. Cath, #6768 or #6769; Becton-Dickinson and Company, Rutherford, N.J. 07070 53 4. Results Results are summarized in Tables B1, 32, and B3. Quality control data are summarized in Table B4. Dog 13 was removed from reference values because of marked IRI response during HD-IVGTT. Dog 13's IRI concentrations (from 0 to 120 min) were 32, 137, 168, 208, 104, 43, 31, and 36 uU/ml. Values from 10 minutes to 120 minutes were greater than two standard deviations above respective means (n=20). Glucose disap- pearance constant was 3.04, well within reference range. 5. Calculations 1. Insulin-Glucose Ratio (IRI/G) (Atkins et al., 1979) IRI = [IRI] X 100 G [glucose] 2. Insulinogenic Index (AIRI/AG) (Kaneko et al., 1977) AIRI = [IRI] (5 min.) - [IRI] (0 min.) X 100 A G [Glucose] (5 min.) - [Glucose] (0 min.) 3. Glucose Disappearance Constants (k) (%/min.) (a) k = 1n [Glucose] 15 min. - ln [Glucose] 45 min. (Chen et al., 1977) (b) k = -230 b, where b equals lepe of linear regression of glucose concentration at 15, 30, and 45 min. (Dyck and Moorhouse, 1966; Moorhouse et al., 1964) (c) k = -69.3/t be, where t9? equals time interval for glucose concentration to decrease from 300 to 150 mg/dl on a line drawn between 15 min and 45 min on a log [glucose] versus time graphic plot. (Greve et al., 1973) 54 Table B1. Reference population laboratory data summary‘a) i S.D. Range HEMATOLOGIC Total Protein (g/dl)(b) 6.4 0.5 5.2-7.1 PCV (%) 45 4 36-50 Hemoglobin (g/dl) 16 1 13-18 RBC (x 106/01) 6.68 0.6 5.63-7.86 MCV (f1) 67 2 62-71 MCHC (g/dl) 35.7 0.6 34.8-36.9 Total White Blood Cell Count(/ul) 10,470 3,257 4,600-19,400 Segmented Neutrophils (/ul) 7,332 2,892 3,128-13,968 Non-segmented Neutrophils (/ul) 15 38 0-134 Lymphocytes (/ul) 2,149 652 1,157-3,290 Monocytes (/ul) 353 248 0-837 Eosinophils (/ul) 615 605 0-2,328 Basophils (/ul) 6 25 0-111 Platelets (subjective) - - Adequate CHEMISTRIES BUN (mg/d1) 15 4 9-27 Glucose (mg/d1) 89 6 79-101 Total protein (g/dl)(b) 6.0 0.4 5.1-6.9 Albumin (g/dl) 3.6 0.3 3.2-4.1 Globulin (q/dl) 2.4 0.4 1.8-3.4 Sodium (mEq/L) 146 3 140-150 Potassium (mEq/L) 4.5 0.3 4.2-5.1 Alanine Aminotransferase (IU/L) 24 9 14-45 Alkaline Phosphatase (IU/L)(b) 39 21 10-95 Cortisol (ng/ml) 40 21 12-90 (a)All determinations by Clinical Pathology Laboratory, Veterinary Clini- cal Center, Michigan State University: (J.D. Krehbiel, D.V.M., Ph.D., Diplomats ACVP, Director) except cortisol (Endocrinology Laboratory, AHDL, Michigan State University (R.F. Nachreiner, D.V.M., Ph.D., Director) (b)Two dogs were immature (8 month Golden Retriever, 13 month Great Dane) and had lower total proteins and higher alkaline phosphatases than remainder of population. 55 Table 82. HD-IVGTT reference values Assessment Time (Minutes) Method 0 5 10 15 30 45 60 90 120 Glucose Mean 89 388 337 281 160 99 85 78 87 (mg/dl) S.D. 6 36 36 40 38 24 13 7 9 (n=19) lowest 79 356 294 225 95 63 54 63 65 highest 101 472 426 383 220 155 114 89 102 IRI Mean 24 80 79 74 53 23 21 19 20 uU/ml S.D. 9 28 30 30 23 6 4 4 4 (n=19) lowest 13 42 34 26 20 16 15 12 12 highest 43 132 132 119 102 38 31 25 28 IRI Mean 27 21 24 27 33 24 25 24 23 glucose S.D. 11 8 10 13 12 7 6 5 6 (n=19) lowest 13 10 9 9 13 14 17 16 13 highest 49 36 42 53 51 38 37 39 34 56 Table B3. HD-IVGTT reference values Assessment Method Mean S.D. Lowest Highest Insulinogenic Index (MRI/AG) 19.2 8.0 7.6 33.0 Glucose Disappearance Constants (%/min) 45 - 15 k= -230 b 3.53 0.76 2.27 5.13 k= -69.3/t1@ 3.57 0.79 2.31 4.95 Glucose tbQ (graphic) (min) 20.3 4.5 14 30 57 Table B4. Quality control data; IRI quantiation for HD-IVGTT reference values Standard Curve Mean S.D. Inter-assay Range n=8 C.V. Total Counts (cpm) 9260 799 8.6 7994-10,467 $80 5405 104 205 5203-5605 Regression Coefficient -0.883 0.017 1.9 -0.903 - -0.858 Correlation Coefficient -0.998 0.001 0.1 -0.998 - -0.996 r2 0.996 0.001 0.1 0.992 - 0.996 Precision Control Mean Duplicate C.V. Intra-Assay Inter-Assay LID/m1 C-V. C.V. mean range mean range mean (n=310) (n=8) (n=8) All Samples --- 4.0 0.1-23.2 -- -- ---- (n=64) (n=64) CS-1* 36 5.0 3.1-7.1 3.8 (n=64) (n=64) *Commercial Control Sera, Corning Medical 58 Appendix C - Stability of Canine Immunoreactive Insulin To assess the stability of IRI in different environmental temperatures, aliquots of two canine sera and one canine plasma were placed in four different temperatures for times ranging one day to ten days. Aliquots of canine sera were also thawed and refrozen from one to five times. Results are summarized in Table C1. Immunoreactive insulin appears to be stable up to ten days at 4°C and 12°C. The IRI at low concentration (serum-1) appears to be stable at 24°C. However, high concentrations of IRI (serum-2, plasma) appeared to deteriorate after one day at 24°C. Immunoreactive insulin at 37°C was not stable. No changes in IRI concentrations were found after thawing and refreezing serum up to 5 times. 59 Table C1. Stability of canine immunoreactive insulin (time and temperature) Sample IRI on Day Storage Days Stored at Given Temperature of Analysis* Temperature (00 IRI/ml) (uU/ml) (°C) 1 2 3 5 10 serum-1 22 4 22 21 22 21 22 serum-2 66 4 65 65 67 65 64 plasma 95 4 93 96 93 95 91 serum-1 22 12 23 22 22 22 24 serum-2 68 12 68 68 65 64 62 plasma 89 12 93 90 86 90 89 serum-1 22 24 23 20 21 18 20 serum-2 66 24 62 60 57 51 39 plasma 95 24 92 89 87 79 79 serum-1 22 37 18 16 14 12 10 serum-2 69 37 45 35 31 25 16 plasma 95 37 70 54 45 35 26 Times Thawed and Refrozen before Analysis 1 2 3 4 5 serum-2 66 -25 to 24 62 62 63 64 64 and return * - samples thawed day of analysis of storage or handling samples LI ST OF REFERENCES LI ST 01“ REFERENCES Atkins, C.E., Hill, J.R., Johnson, R.K.: Diabetes Mellitus in the Juvenile Dog: A Report of Four Cases. J.A.V.M.A. 175: 362, 1979. Auburt, M.L.: Critical Study of the Radioimmunological Assay for the Dosage of Polypeptide Hormones in Plasma. J. Nucl. Biol. Med. 14: 1, 1970. Bangham, D.R.: Standardization in Peptide Hormone Immunoassays: Principle and Practice. Clin. Chem. 22: 957, 1976. Berson, S.A., Yalow, R.S.: Species-Specificity of Human Anti-Beef, Pork Insulin Serum. J. Clin. Invest. 38: 2017, 1959. Berson, S.A., Yalow, R.S.: Immunochemical Distinction Between Insulins with Identical Amino Acid Sequences. Nature (London) 191: 1392, 1961. Berson, S.A., Yalow, R.S.: Antigens in Insulin Determinants of Specificity of Porcine Insulin in Man. Science 139: 844, 1963. Berson, S.A., Yalow, R.S., Bauman, A., Rothschild, A., Newerly, K.: Insulin-I131 Metabolism in Human Subjects: Demonstration of Insulin Binding Globulin in the Circulation of Insulin Treated Subjects. J. Clin. Invest. 35: 170, 1956. Berson, S.A., Yalow, R.S., Glick, S.M., Roth, J.: Immunoassay of Protein and Peptide Hormones. Metab. 13: 1135, 1964. Black, H.E., Rosenblum, I.Y., Capen, C.C.: Chemical Induced (Streptozoticin-Alloxan) Diabetes Mellitus in the Dog. Am. J. Pathol. 98: 295, 1980. Blazquez, E., Munoz-Barragan, L., Patton, G.S., Dobbs, R.E., Unger, R.H.: Gastric A-cell Function in Insulin-Deprived Depancreatized Dogs. Endocrin. 99: 1182, 1976. Blazquez, E., Munoz-Barragan, L., Patton, G.S., Orci, L., Dobbs, R.E., Unger, R.H.: Demonstration of Gastric Glucagon Hypersecretion in Insulin-Deprived Alloxan-Diabetic Dogs. J. Lab. Clin. Med. 89: 971, 1977. Block, M.B., Mako, M.M., Steiner, D.F., Rubenstein, A.H.: Circulating C-Peptide Immunoreactivity, Studies in Normal and Diabetic Patients. Diabetes 21: 1013, 1972. 60 61 Bromer, w.w., Boucher, M.E , Koffenberger, J.E.: Amino Acid Sequences of Bovine Glucagon. J. Biol. Chem. 246: 2822, 1971. Buchanan, R.D., Vance, J.E., Dinstl, R., Williams, R.: Effect of Blood Glucose on Glucagon Secretion in Anesthetized Dogs. Diabetes 18: 11, 1969. Campbell, J., Pierluissi, J., Green, G.R.: Somatotrophic Diabetes: Insulin Release Responses to Arginine and Glucagon in Dogs. Diabetologia 15: 205, 1978. Capen, C.C., Martin, S.L.: Hyperinsulinism in DOgs with NeOplasia of Pancreatic Islets. Vet. Path. 6: 309, 1969. Caywood, D.D., Wilson, J.W., Hardy, R.M., Schull, R.M.: Pancreatic Islet Cell Adenocarcinoma: Clinical and Diagnostic Features of Six Cases. J.A.V.M.A. 174: 714, 1979. Chance, R.H.: Amino Acid Sequences of Proinsulin and Intermediates. Diabetes 21 (suppl. 2): 461, 1972. Chen, C.L., Ganguli, S., Kumar, M.S.A., Zenoble, R.D.: Possible Source and Control of Trauma-Induced Glucagonemia in Dogs. Am. J. Vet. Res. 38: 1259, 1977. Cherrington, A.D., Lacey, w.w., Chiasson, J.L.: Effect of Glucagon on Glucose Production During Insulin Deficiency in the Dog. J. Clin. Invest. 62: 644, 1978. Conlon, J.M., Ipp, E., Unger, R.H.: The Molecular Forms of Immunoreactive Glucagon Secreted by the Isolated Perfused Dog Pancreas. Life Sciences 23: 1655, 1978. Conlon, J.M., Samson, W.K., Dobbs, R.E., Orci, L., Unger, R.H.: Glucagon-Like Polypeptides in Canine Brain. Diabetes 28: 700, 1979. Cotes, P.M.: Standards and Reference Reagents for Radioimmunoassay of Peptide Hormones and Related Substances. IN: Radioimmunoassay and Related Procedures in Medicine, Proc. of Symp. on Radioimmunoassay and Related Procedures in Clinical Medicine and Research. Vol. 1, International Atomic Energy Agency, Vienna, 1974, pg. 71. Cresto, J.C., Yalow, R.S.: Anomalous Ascending Standard Curves in the Radioimmunoassay of Insulin, Human Growth Hormone, and ACTH. IN: Radioimmunoassay: Methodology and Applications in Physiology and in Clinical Studies, Georg Thieme Publishers, Stuttgart, 1974, pg. 1. DeFronzo, R.A., Sherwin, R.S., Dillingham, M., Handler, R., Tamborlane, W.V., Felig, P.: Influence of Basal Insulin and Glucagon Secretion on Potassium and Sodium Metabolism. Studies with Somatostatin in Normal Dogs and In Normal and Diabetic Human Beings. J. Clin. Invest. 61: 472, 1978. 62 Doi, R., Prentki, M., Yip, C., Muller, W.A., Jeanrenaud, B., Vranic, M.: Indentical Biologic Effects of Pancreatic Glucagon and Purified Moiety of Canine Gastric Immunoreactive Glucagon. J. Clin. Invest. 63: 525, 1979. Dyck, D.R., Moorhouse, J.A.: A High-Dose Intravenous Gluocse Tolerance Test. J. Clin. Endocr. 26: 1032, 1966. Eigler, N., Sacca, L., Sherwin, R.S.: Synergistic Interactions of Physiologic Increments of Glucagon, Epinephrine, and Cortisol in the Dog, A Model for Stress-Induced Hyperglycemia. J. Clin. Invest. 63: 114,1979. Ekins, R.P.: Radioimmunoassay Design. IN: Radioimmunoassay i2 Clinical Biochemistry, Ed. C.A. Pastenak, Heyden, N.Y., 1975, pg. 3. Emmer, M., Gorden, P., Roth, J.: Diabetes in Association with Other Endocrine Disorders. Med. Clin. N. Amer. 55: 1057, 1971. Faloona, G.R., Unger, R.H.: Glucagon. IN: Methods 2; Hormone Radioimmunoassay. Academic Press, New York, 1974, pg. 317. Felig, P., wahren, J., Sherwin, R., Hendler, R.: Insulin, Glucagon, and Somatostatin in Normal Physiology and Diabetes Mellitus. Diabetes 25: 1091, 1976. Gerich, E.J., Charles, M.A., Grodsky, G.M.: Regulation of Pancreatic Insulin and Glucagon Secretion. Ann. Rev. Phsyiology 76: 353, 1976. Gill, J.L.: Desigp and Analysis 9: Egperiments‘ig the Animal and Medical Sciences. vo1. 1, the Iowa State University Press, Ames, Iowa, 1978, pg. 179. Greve, T., Dayton, A.D., Anderson, N.V.: Acute Pancreatitis with Coexistent Diabetes Mellitus: An Experimental Study in the Dog. Am. J. Vet. Res. 34: 939, 1973. Hafs, H.D., Niswender, G.D., Malven, P.V., Kaltenbach, C.C., Zimbelman, R.G., Condon, R.J.: Guidelines for Hormone Radioimmunoassays. J. Anim. Science. 46: 927, 1977. Hales, C.N., Randle, P.J.: Immunoassay of Insulin with Insulin Antibody Precipitate. Biochem. J. 88: 137, 1963. Harris, V., Paloona, G.r., Unger, R.H.: Glucagon. IN: Methods 9; Hormone Radioimmunoassay, 2nd Ed., Academic Press, New York, 1979, pg. 643. Heding, L.G., Frandsen, E.K., Jacobsen, H.: Structure-Function Relationship: Immunologic. Metab. 25 (suppl): 1327, 1976. Herbert, V., Lau, R., Gottleb, C.W., Bleicher, S.J.: Coated Charcoal Immunoassay of Insulin. J. Clin. Endocr. 25: 1375, 1965. 63 Hommel, H.H., Fischer, U: The Mechanisms of Insulin Secretion after Oral Glucose Administration. V. Portal Venous IRI Concentration in Dogs after Ingestion of Glucose. Diabetologia 13: 269, 1977. Horwitz, D.L., Kuzuya, H., Rubenstein, A.H.: Circulating Serum C-Peptide - A Brief Review of Diagnostic Implications. N. Eng. J. Med. 295: 207, 1976. Jennings, A.S., Cherrington, A.D., Liljenquist, J.R., Keller, U., Lacy, W.W., Chiasson, J.L.: The Roles of Insulin and Glucagon in the Regulation of Gluconeogenesis in the Postabsorptive DOg. Diabetes 26: 847, 1977. Johnson, R.K.: Insulinoma in the Dog. Vet. Clin. N. Amer. 7: 629, 1977. Juffe, A., Nojek, C., Costillo-Olivares, J.L., Equaras, M., Fiquera, D.: Plasma Levels of Glucose and Insulin During Deep Hypothermia with Varying Degrees of Hemodilution in Dogs. J. Cardiovas. Surg. 18: 325, 1977. Kagan, A: Evaluation of Commerical Radioimmunoassay Kits. Semin. in Nucl. Med. 5: 173, 1975. Kaneko, J.J., Mattheeuws, D., Rottiers, R.P., Van Der Stock, J., Vermeulen, A.: The Effect of Urinary Glucose Excretion on the Plasma Glucose Clearances and Plasma Insulin Responses to Intravenous Glu- cose Loads in Unanesthesized Dogs. ACTA Endocrin. 87: 133, 1978(a). Kaneko, J.J., Mattheeuws, D., Rottiers, R.P., Vermeulen, A.: Renal Function, Insulin Secretion, and Glucose Tolerance in Mild Streptozoticin Diabetes in the Dog. Am. J. Vet. Res. 39: 807, 1978(b). Kaneko, J.J., Mattheeuws, D., Rottiers, R.P., Vermeulin, A.: Glucose Tolerance and Insulin Response in Diabetes Mellitus of Dogs. J. Sm. Anim. Pract. 18: 85, 1977. Kaneko, T., Oka, H., Munemura, M., Oda, T., Yamashita, H., Suzuki, 3., Yanaihara, N., Hashimoto, T., Yanaihara, C.: Radioimmunoassay of Human Proinsulin C-Peptide Using Synthetic Human Connecting Peptide. Endocrinol. Japon 21: 141, 1974. Kawanishi, K., Nishina, Y., Goto, A., Ishida, T., Kawamura: Insulin and C-Peptide in Pancreatic venous Blood in Dogs. Horumon to Rinsho 25: 1067, 1977. (In Japanese) Keller, U., Chiasson, J.L., Liljenquist, J.E., Cherrington, A.D., Jennings, A.S., Crofford, O.B.: The Roles of Insulin, Glucagon, and Free Fatty Acids in the Regulation of Ketogenesis in Dogs. Diabetes 26: 1040, 1977. 64 Kemmler, W., Peterson, J D., Rubenstein, A.H., Steiner, D.F.: On the Biosynthesis, Intracellular Transport and Mechanisms of Conversion of Proinsulin to Insulin and C-Peptide. Diabetes 21 (Suppl. 2): 572, 1972. Kramer, J.R., Wilson, R.B.: A Review of Spontaneous Diabetes Mellitus in Animals. Proc. Ann. Mtg., A.C.V.P., "Biochemical Mechanisms and Disease Processes”, 5, 1978. Krejs, G.J., Orci, L., Conlon, J.M., Ravazzola, M., Davis, G.R., Raskin, P., Collins, S.M., McCarthy, D.M., Baetens, D., Rubenstein, A., Aldor, T.A.M., Unger, R.H.: Somatostatinoma Syndrome: Biochemical, Morphologic, and Clinical Features. N. Eng. J. Med. 301: 285, 1979. Kubasik, N.P.: Recent Developments in the Radioimmunoassay of Insulin. IN: Clinical Immunochemistry, Chemical and Cellular Bases and Application i2 Disease, Am. Ass. for Clin. Chem., washington, D.C., 1978, pg. 273. Lefebvre, P.J., Luyckx, A.S.: Glucose and Insulin in the Regulation of Glucagon Release from the Isolated Perfused Dog Stomach. Endocrinology 103: 1579, 1978. Levine, R., Haft, D.F.: Carbohydrate Homeostasis. N. Eng. J. Med. 283: 175, 1970. Longley, W.J.: Proficiency Testing of Radioimmunoassay. Clin. Biochem. 9: 109, 1976. Malvano, R., Zucchelli, G.C., Quesada, T.: Insulin Radioimmunoassay: A Comparison of Three Methods. J. Nucl. Biol. Med. 18: 80, 1974. Manns, J.G., Martin, C.L.: Plasma Insulin, Glucagon, and Nonesterified Fatty Acids in Dogs with Diabetes Mellitus. Am. J. Vet. Res. 33: 981, 1972. Manny, J., Rabinovic, N., Schiller, M.: Insulin Response to Continuous Glucose Load in Endotoxin Shock in the Dog. Surg. Gyn. Obs. 145: 198, 1977. Matsuyama, T., Hoffman, W.H., Breuer, R.I., Zuckerman, L., Lawrence, A.M., Foa, P.P.: Serum Glucose, Insulin, Total, Pancreatic, and Enteric Glucagon in Dogs Before and After Pancreatectomy, in Patients with Gastrectomy and Pyloroplasty, and in Diabetic Children. Effects of Insulin and of Secretin. Materia Med. Polona 2: 124, 1977. Midgley, A.H., Niswender, G.D., Rebar, R.W.: Principles for the Assessment of the Reliability of Radioimmunoassay Methods (Precision, Accuracy, Sensitivity, Specificty). Acta Endocrinol. (Suppl 142): 163, 1969. Moorhouse, J.A., Grahame, G.R., Rosen, N.J.: J. Clin. Endocr. 24: 145, 1964. 65 Morita, 8., Doi, K., Yip, C.C., Vranic, M.: Measurements and Partial Characterization of Immunoreactive Glucagon in Gastro-Intestinal Tissues of Dogs. Diabetes 25: 1018, 1976. Moss, A.J., Dalrymple, G.V., Boyd, G.M.: Practical Radioimmunoassay, The C.V. Mosby Company, St. Louis, 1976. Muller, W.A., Girardier, L., Seydoux, J., Berger, M., Renold, A.E., Vranic, M.: Extrapancreatic Glucagon and Glucagonlike Immunoreactivity in Depancreatized Dogs. J. Clin. Invest. 62: 124, 1978. Nilsson, G., Uvnas-Wallenstein, K: Effect of Teasing and Sham Feeding on Plasma Glucagon Concentrations in Dogs. Acta Physiol. Scand. 100: 298, 1977. Njoku, C.O., Strafuss, A.C., Dennis, S.M.: Canine Islet Cell Neoplasia: A Review. J.A.A.H.A. 8: 284, 1972. Orosz, L.: Effect of Heparin on Insulin Secretion Induced by Glucose in Dogs. Acta Endocrinol. 75: 531, 1974. Peterson, J.D., Nehrlich, S., Oyer, P.E., Steiner, D.F.: Determination of Amino Acid Sequence of the Monkey, Sheep, and Dog Proinsulin C-Peptide by a Semi-micro Edman Degradation Procedure. J. Biol. Chem. 247: 4866, 1972. Phillips, L.S., Vassilopoulou-Sellin, R.: Somatomedins, N. Eng. J. Med. 302: 371, 438, 1980. Pierluissi, J., Campbell, J.: Metasomatotrophic Diabetes and Its Induction: Basal Insulin Secretion and Insulin Release Responses to Glucose, Glucagon, Arginine, and Meals. Diabetologia 18: 223, 1980. Rosselin, G., Assan, R., Yalow, R.S., Berson, S.A.: Separation of Antibody-Bound and Unbound Peptide Hormones Labelled with Iodine-131 by Talcum Powder and Prepicitated Silica. Nature (London) 212: 355, 1966. Rubenstein, A.H., Kuzuya, H., Horwitz, D.L.: Clinical Significance of Circulating C-Peptide in Diabetes Mellitus and Hypoglycemic Disorders. Arch. Intern. Med. 137: 625, 1977. Rubenstein, A.H., Steiner, D.F.: Proinsulin. Ann. Rev. Med. 22: 1, 1971. Samols, E., weir, G.C.: Adrenergic Modulation of Pancreatic A, B, and D Cells. J. Clin. Invest. 63: 230, 1979. Schall, W.D., Cornelius, L.M.: Diabetes Mellitus. Vet. Clin. N. Amer. 7: 613, 1977. 66 Schusdziarra, V., Dobbs, R.E., Harris, V., Unger, R.H.: Immunoreactive Somatostatin Levels in Plasma of Normal and Alloxan Diabetic Dogs. FEBS Let. 81: 69, 1977. Skelley, D.S., Brown, L.P., Besch, P.K.: Radioimmunoassay. Clin. Chem. 19: 146, 1973. Smith, L.F.: Amino Acid Sequences of Insulins. Diabetes 21 (Suppl. 2): 457, 1972. Soeldner, J.S., Slone, D.: Critical Variables in the Radioimmunoassay of Serum Insulin Using the Double Antibody Technic. Diabetes 14: 771, 1965. Srikant, C.B., McCorkle, K., Unger, R.H.: Properties of Immunoreactive Glucagon Fractions of Canine Stomach and Pancreas. J. Biol. Chem. 252: 1847, 1977. Starr, J.I., Horwitz, D.L., Rubenstien, A.H., Mako, M.E.: Insulin, Proinsulin, and C-Peptide. IN: Methods 2: Hormone Radioimmunoassay, 2nd Ed., Academic Press, New York, 1979, pg. 613. Stryer, L.: Biochemistry, W.H. Freeman and Co., San Francisco, 1975, pg. 35. Sundby, F., Markussen, J.: Isolation, Crystallization, and Amino Acid Composition of Rat Glucagon. Horm. Metab. Res. 3: 184, 1971. Sundby, F., Markussen, J.: Rabbit Glucagon: Isolation, Crystallization, and Amino Acid Composition. Horm. Metab. Res. 4: 56, 1972. Sweeny, M: Fifth Annual Radioassay Directory. Lab Wbrld 30: 92, 1979. Thomsen, J., Kristiansen, K., Brunfeldt, K.: The Amino Acid Sequence of Human Glucagon. FEBS Lett. 21: 315, 1972. Thorell, J.I., Lanner, A.: Influence of Heparin-Plasma, EDTA-Plasma, and Serum on the Determination of Insulin with Three Different Radioimmunoassays. Scand. J. Clin. Lab. Invest. 31: 187, 1973. Unger, R.H.: Glucagon Physiology and Pathophysiology. N. Eng. J. Med. 285: 443, 1971. Unger, R.H., Eisentraut, A.M., McCall, M.S., Keller, 8., Lanz, H.C., Madison, L.L.: Glucagon Antobodies and Their Use for Immunoassay of Glucagon. Proc. Soc. Exp. Biol. Med. 102: 621, 1959. Unger, R.H, Ohneda, H., Valverde, I., Eisentraut, A.M., Exton, J.H.: Characterization of the Responses of Circulating Glucagon-Like Immunoreactivity to Intraduodenal and Intravenous Administration of Glucose. J. Clin. Invest. 47: 48, 1968. Walsh, J.H.: Radioimmunoassay Methodology for Articles Published in Gastroenterology: Gastroent. 75: 523, 1978. 67 Wide, L., Parath, J.: Radioimmunoassay of Proteins with the Use of Sephadex-Coupled Antibodies. Biochim. BiOphys. Acta. 130: 257, 1966. Witherspoon, L.R.: Evaluation of Radioimmunoassay Kits. The Ligand Review 1: 28, 1979. Yalow, R.S.: Radioimmunoassay: Its Past, Present and Potential. IN: Progress i2 Analytical Chemistry, Vol. 1L Methods i2 Radioimmunoassay, Toxicology, and Related Areas. Plenum Press, New York, 1974, pg. 1. Yalow, R.S., Berson, S.A.: Assay of Plasma Insulin in Human Subjects by Immunologic Methods. Nature 184: 1648, 1959. Yalow, R.S., Berson, S.A.: Immunoassay of Endogenous Plasma Insulin in Man. J. Clin. Invest. 39: 1157, 1960. Zazuco Higa, O., Torres de Toledo e Souza, I., Whljchenberg, B.L., Pinto, H., Ribeiro Pieroni, R.: Standization of Radioimmunoassay Technique for Determination of Plasma Insulin and Growth Hormone. IN: Radioimmunoassay and Related Procedures in Medicine, Proc. Symp. Radioimmunoassay and Related Procedures in Clinical Medicine and Research. Vol. 1, International Atomic Energy Agency, Vienna, 1974, pg. 291. VITA VITA The author was born in McPherson, Kansas, on August 25, 1948. He received his Bachelor of Science (1970) and Doctorate of Veterinary Medicine (1972) from Kansas State University. After a 2 year tour of duty in the United States Air Force, he was a small animal practitioner in Idaho and Utah. He entered a Pathology Residency and Graduate School at Michigan State University in September, 1977. He successfully completed the American College of Veterinary Pathology Certification Examination for Veterinary Clinical Pathology in September, 1980. The author is a member of the American Veterinary Medical Association, American Animal Hospital Association, American College of Veterinary Pathology, American Society for Veterinary Clinical Pathology, Honor Society of Phi Kappa Phi, Society of Phi Zeta and Delta Tau Delta. His wife, Marcia, is a certified medical technologist (A.S.C.P.). 68