AJEBAK 54 (Pt. I) 57-70 (1976)
ISOLATED ISLET TRANSPLANTATION IN EXPERIMENTAL DIABETES bv BRUCE N. GRAY* AND ELTON WATKINS, JR. (From the Sias Surgical Lalwratory. Lahey Clinic Foundation, Boston, Massachusetts, 02146.) {Accepted for publication January 15,1976.) Suininury. Pancreatif islets have been isolated from the exocrine pancreas of inbred rats hy the toIlaKcnase digestion method. Transplantation of isolated islets into tlw> portal venous .system of streptozottKrin-cIiabetic retipient.s resiiltptl in complete ahrogation of the diabetic .'itate as measured by non-fasting scrum glucose levels, 24 h urinary output, rate of weight gain and glucose tolerance test. Transplantation to other sites resulted in less than optimal survival and function of islets. Allogeneif islets, transplnnted across weak histocompatibility barriers, can survive and function for prolonged periods of time when transplant recipients are imniunosuppressed with antiljniphocyte serum (ALS). Recipients of allogcneic islets, after a period of immunosuppression with ALS, become permanently tolerant to the allografted islets and to subsequent skin grafts from similar allogcnpic donors. Allografted islets arc able to prevent the occurrence of dialx-tic renal and ophthalmic changes that IKCUT in control dialn-tic animals which had not undergone transplantation.
INTRODUCTION. Since the introduction of insulin therapy in 1922 (Bantin*; and Best, 1922), clinical management of dialietes mellitns has been directed almost exclusively toward controlling the wide fluctuations in blood glncose levels. Rigid snpervision of exogenous insulin therapy may result in good control of most hypoglycemic and hyperglyceniic episodes. Despite snch control, liowever, the jnvenile diabetic is still faced with early vascular changes and a consideralily shortened life span. One approach to the problem of stabilizing lx)th blood glncose flnctnations and, possibly, vascular complications wonld be allotranslalantation of a functioning pancreiis to the diabetic recipient. During the past decade, many investigators have performed transplants of whole or partial pancreatic organs in animals {Bergan and Teixeira, 1967; Lillehei et ai, 1970) and, more recently, in man (Connolly et al., 1973; Cliedman et al, 1973). Although the number of ca.ses is small, morbidity and mortality • Present address: University Dept. of Surgery, St. Vincents Hospital, Fitzroy, Victoria.
58
BRUCE N. GRAY AND ELTON WATKINS, JR.
rates are unacceptably liigh; 32 partial or total pancreatic allografts had been recorded by the Organ Transplant Registr>' as of 1973 (ACS/NIH, 1973). Of these, oiily 2 patients had survived with functioning grafts longer than one year. To circumvent the technical difRculties associated with whole organ transplants, several groups ha\e invf.stlgated trau.sphintatioii of isolated pancreatic islets. This paper reports onr investigation into transplanting isolated panereatic islets in tho rut. MATEBIALS AND METHODS. Animals. All donor rats and syngeneic retipients were inbred Wistar/Lewos (LEW/Crl) strain malt' rats weighing between 200 and 300 !^. All diabetic animals receiving allografts were inbred Fischer (F344/Crl) strain male rats weighing between 120 and ISO K- The Fi.scher .strain presents histocompatibility genes, a. h, c, Ag-A^, Ag-B^, Ag-C, and H-V, while the Wistar/Lewis strain presents Ag-A*, Ag-B^, Ag-C, H-P, and H~b^. Fischer and Wi.star/ Lewis strains also differ at the Ag-D locus (Palm, 1971). Weak histocompalibility differences exist between the two .strains. Induction of diabetes. Graft rcoipients and control diabetic animals were rendered diabetic by tl'ie intravenous administration of the antibiotic, strepluzotocin, 6.5 mg/kg-^ of body weight. Transplantation of islets was carried out after the animals had shown sustained signs of diabetes for a period of at least 2 weeks. IsohUon of islets. Islets were obtained by a modification of the method of Ballinger (Ballinger and Lacy, 1972). Techniques of mecfianical dissociation and enzymatic digestion of ths pancreas were similar to those described by those authors. Final separation of i.slets from acinar tissue was performed by centrifugatioo at 700 g for 7 min in a differential density gradient of 31-5% (w/v) Ficoll-suline, 25S P'icoU-saiint', and Hank's solution. The i.slets separated out at the uppcnnost interface and toirld easily be aspirated off and used for transplantation by injection through a 23-gauge hypodermic needle. From each donor pancreas at least 200 intact islets could be obtained. Total isolation time was 60 to 70 min. At least 95^ of the material obtained at the i.ippeniiost interface of the Ficoll density gradient consisted of islet tissue and only 5% consisted of acinar debris. The combined islets from three donor animals were used for each graft. This >'ielded approx. 600 intact islets, and the suitability of different sites was assessed using this same number in each case. The assessment of transplant site suitability was perfonnetl using syngeneic grafts only to obviate rejection. The transplant sites tested included intraperitoneal implantation, subcutaneous implantation, testicular implantation, mesenterie root implantation, sy.steniic venous enibolization, superior mesenteric artery embolization and portal venous enibolization. Methods of immunosuppression. Preliminary investigations suggested that rabbit anti-rat antilymphocyte serum (ALS) was siq>erior to azathioprine (Imuran) in its ability to suppress rejeetion of allografted paiiCI%atic islets. Uiabt'tic Fischer strain anininls receiving allogeneic Wistar/Lewis strain islets recpived ALS (Microbiological Assoc. Lot #13104) in either a high- or low-dose schedule as shown in Table L
ISLET TRANSPLANTATION IN DIABETES
59
Non-fasting senim glucose estimations were performed on blood samples obtained by tail dipping (Glucostratc. Warner-Chilcott). Urine output was measured each day from test aniinul.'i housed in metabolic cages. Test animals were weighed to tlu^ nearest t;rain two or three times a week. Glucose tolerance tests were perfonned on animals fasted for 4 b by administering U-glucose solution. 2-5 g/kg^^ of body weigbt, via an intragastric tube. Serum glucose levels were measured as described above. A successful transplant was defined as one in whicb nonfasting serum glucose levels, 24 h urinary volumes and rate of weight gain returned to nonnai control levels for a miniimun period of 30 days. Animals witb partial improvement only or animals witb initial complete recover>' followed b>' regression to the diabetic state were regarded a.s failures.
TABLE 1. Schedule of immimosiippressian with ALS. High dose 0-1 ml per animal, 6 days/wk for I month 0-1 ml per animal, 3 days/wk for next 2 wks 0-1 ml per animal, 1 day/wk for next 3 months 0- ] mi per animal, I day/month for next 6 months nil thereafter
Low dose 0' 1 ml per animal, 3 days/wk for 1 wk 0-1 ml per animal, I day/wk for next 4 months 0-1 ml per animal, 1 day/month for next 3 monlhs nil thereafter
Skin grafts. Full-tbicltness grafts of approx. 1-5 by 1-5 cm were taken from tbe abdominal skin of male Wistar/Lewis rats and placed with 180" rotation on the central back just below the scapula of Fiscber recipients. Tbe time of rejection was defined as tlie point at wbich the whole graft had been converted to a dry scab and conld be lifted from tbe graft bed. Three groups of Fiscber animals received Wistar/Lewis skin grafts. Cnnip 1 consisted of 10 animals which had received allogeneic Wistar/Lewis islet tran.splantM and had been iiimmnosuppressed with ALS. Group 2 amsisted of 5 normal animals wbicb received ALS in tbe bigh-dose protocol starting at the time of .skin grafting. Group 3 compn.sed nonnai animals without iininnno.suppression. Mixed lymphocyte cultures (MLC). Blood was obtained from animals by cardiac puncture and lympbocytes separated on a FicoU-IIypaque gradient (Thorsby and Bratlie. 1970). Cultures were perfonned in niundbottomed micioculture plates (Linbro Chemical Co.) in 0-22 ml of medium lU'MI 1640 supplemented witb IS- glutamine (Microbiological As.sae.) and 10% lieat-decomplemented nonnai rat serum. Pulse tagging was on days 2 to 6 witb 0-5 microcuries of tritiated tbymidine (New England Nuclear) per well. Each microculture well contained 0-25 by 10" eirector lympbocytes and 0-25 by 10** stinmlator lymphocytes. Polynucleotide synthesis was blocked in stimulator lymphocytes by prior incubation with mitomycin-C (Nutritional Biochemical Inc.), 25 micrograms/IO*' cells/ml for 1 b. Ceils were harvested after 17 h on glass fibre strips using a MASH II cell culture barvesler (Microbiological Assoc), and radioactivity was measured by conventional litpiid scintillation techniques. Activation indexes (AI) for one-way MLC assays were calculated by tbe following formula; (CPM. effector cells + stinuilator cells) - (GPM. blank) AI = (CPM. effector cells) (CPM. stimulator cells) - 2x (CPM. blank) alone alone Significant deviation of the activation index from an iiide.\ value of 1-0. denoting absence of bla.stosis. was detennined by Student's t-test.
60
BRUCE N. GRAY AND ELTON WATKINS, JR.
J^ses.iment of renal and ophthalmic changes. Three groups of Fi.scber animals were used to determine tbe ability of tran.splanted islets to prevent the occurrence of renal and opbthalniic changes, Group I consisted of 10 animals wbich were allowed to remain diabetic. Group 2 consisted of 10 animals wbicb received a corrective portal venous transplant of allogeneic Wistar/Lewis islets 2 to 4 weeks after the induction of diabetes with streptozotocin. In group 2, 6 aniiiiats were immunosuppressed witb tbL' high dose protocol of ALS and 4 animals witb tbe low dose protocol. Group 3 consisted of 10 normal agfMnatcbed animals. Nine months after the induction of diabetes the left eye and kidney were removed from eacb animal in groups 1 and 2. The left eye and kidney were fi.xed in 10^ formalin, sectioned at 2 and 5 nM for the kidney and 5 nM for the retina, and stained witb baeniato.xylin-eosin and periodic acid-Scbiil (PAS) stiiiiis. Retinal sections were prepared from 10 animals and renal sections were prepared from 3 animals in group 3 for comparison witb groups 1 and 2. Codec! bistnlogic sections were randtimized antl scored without knowledge of the code by tbe first author (BNG) on an arbitrary scale of one to four for tbe following changes. In the kidnty, pathologic changes assessed were amount of mesangial deposition of PASpositive material, degree of thickening of the gtomerular afferent arteriole wall, and tbt* degree of thickening nf the purii^tal layer of Bowman's capsule. The greatest diameter of .30 randomly selected glomeruli from each of the test animals in tbe tbree groups was mea.sured, using a microscope reticle in order to assess any cbange in gloinerular size. For tbe assessment of retinal changes, randomized .sections from the three groups were scored on a seal;- of one to four tci .show thif degree (if capillarj- ancf small vessel dilatation of vessels in tbe stratum opticum and outer plexiform layer. For tbe as.se.ssment of retinal new vtssel formation. 10 artas of 0-1 nun- were randomly .selected in each retina and the number of capillaries counted in each of tbe four layers as suggested by Sosula et ai (1972). The four layers were tbe outer plexiform layer, inner nuclear layer, inner plexiform layer and stratum opticum. In scoring renal and retina! changes, a score of one was given wben tbe degree of change being assessed was minimal, whereas a score of four was given to sections showing tbe most advanced cbange. Statistical significance between groups was compared using tbe Student's t-lest. P values of less than 0-05 were considered significant. Test animals were inspected periodically for the presence of cataract formation.
RESULTS. Induction of diabetes. Animals injected with streptozotocin showed diabetic rnanife.stati()ns within hours. Non-fasting serum ghico.sf le\'els increased from a range of 100 to 200 nig per 100 ml of serum to 450 to 900 mg per 100 ml of serum. Uriiu; output iiicreased from a mean of 5 ml per 24 h to between 50 and 130 ml per 24 h. Xjbe rate of weight gain dropped dramaHcally and was invariably less than 30% itrfthe normal rate. In many cases weight loss was observed. Transplantation to various anatomic sites. Of all sites tested, emboUzation into the portal venoxts system proved to be the most efFeetive method (Gray and Watkins, 1973). Althongh implantation into the base of the mesenter)' proved effective in the 1 animal tested, its technical difficulty precluded further evaluation. In 2 of 3 animals in which the islets were ernbolized into the superior niesenteric artery, there was initial complete relief for 7 and H days respectively followed by return to the diabetic state. Postoperative haemorrhage commonly occurred when the superior mesenterie artery
ISLET TRANSPLANTATION IN DIABETES
61
route was used. Only those animals surviving the procedure have been included. In contrast, embolization into the portal venous system was free of operative morbidity and mortality. lntraperitoneal implantation using 600 islets per transplant was met with only partial improvement by the three criteria. Transplantation to other sites failed to alter signiHcuntly the diabetic state of the graft recipients. Table 2 shows the success rate of the various sites tested. Histologic verification of islet survival in host liver is shown in Fig. 1. TABLE 2. Survival of transplants to varhms sites.
Siic or transplant Intraperitoneal Subcutaneous Intratesticular 5ystemic venous Superior mesenterie ariery Mesenterie root Portal venous
Number of transplants
Transplant survival
4
0 0 0 0
5 2 3 1 10
Q
I 6
Fig. 1. Pancreatic islet lodged in portal venous radicle of host liver 1 month after seeding of isolated syiigeneic islets into the bost portal vein (liaematoxyhn-eosin stain), x 437.
m
BRUCE N. GRAY AND ELTON WATKINS, JR.
Transplantation of allogeneic islets. Of 5 Fischer strain animals which received allogeneic Wistar/Lewis islets into the portal venous system, witliout immunosuppression, all rejected the grafts within 8 to IS days with return of diabetes in all in.stances. Survival and function of allogencie islet transplants were prolonged b\- immunosuppression with ALS. Of 5 transplant recipients receiving high dose ALS, all continued to have functioning grafts at 10 months after tran.splantation. Of 5 recipients receiving low dose ALS, 1 animal rejected the transplant after 4 months, and the remaining 4 animals were still free of diabetes 7 months after transplantation. Serial non-fasting scrum glucose levels and 24 h urinary volumes are shown in Figs 2a and 2b and are compared to a group of 5 Wistar/Lewis animals whieh received syngeneic transplants. Recipients of allogeneic portal venons transplants were able to handle a glncose load in a manner similar to normal animals, in contrast to control diabetic anhnals, as shown in Fig. 3. Granulated beta cells have been demonstrated in host liver 12 months or more after the transplant procedure. liy this time, however, the general architectnre of the islets was lost, although densely staining beta cells could readily be identified within the portal triad (Fig. 4).
FIG.
eooi
2A.
ALLOGENEIC W/L-
SYNGENEIC
400
W/L-^-W/L
SERUM 200 GLUCOSE {MO/100 ML)
•«—•-••
ALLOGENEIC W/L-»-F ALS (High Dose)
ALLOGENEIC W / L - » - f ALS (Low Dose)
-10 -5
Ol
7 15 DAYS
10
TIME
See caption on
MONTHS
page.
IS
ISLET TRANSPLANTATION IN DIABETES
63
TABLE 3. Mixed tymphoeyte eutlure responses. Animals (lymphocyte donor)
Exp. No.
Tolerant lo both pancreatic and skin grafts
1 2 3 4 5 6 7 8 9
Day of maximum response
Activation index
Normal
P Value
9.4 13 7 31 3-7 2 9 31 2 5-6
6 5 5 6 6 6 5
ISOLATED ISLET TRANSPLANTATION IN EXPERIMENTAL DIABETES bv BRUCE N. GRAY* AND ELTON WATKINS, JR. (From the Sias Surgical Lalwratory. Lahey Clinic Foundation, Boston, Massachusetts, 02146.) {Accepted for publication January 15,1976.) Suininury. Pancreatif islets have been isolated from the exocrine pancreas of inbred rats hy the toIlaKcnase digestion method. Transplantation of isolated islets into tlw> portal venous .system of streptozottKrin-cIiabetic retipient.s resiiltptl in complete ahrogation of the diabetic .'itate as measured by non-fasting scrum glucose levels, 24 h urinary output, rate of weight gain and glucose tolerance test. Transplantation to other sites resulted in less than optimal survival and function of islets. Allogeneif islets, transplnnted across weak histocompatibility barriers, can survive and function for prolonged periods of time when transplant recipients are imniunosuppressed with antiljniphocyte serum (ALS). Recipients of allogcneic islets, after a period of immunosuppression with ALS, become permanently tolerant to the allografted islets and to subsequent skin grafts from similar allogcnpic donors. Allografted islets arc able to prevent the occurrence of dialx-tic renal and ophthalmic changes that IKCUT in control dialn-tic animals which had not undergone transplantation.
INTRODUCTION. Since the introduction of insulin therapy in 1922 (Bantin*; and Best, 1922), clinical management of dialietes mellitns has been directed almost exclusively toward controlling the wide fluctuations in blood glncose levels. Rigid snpervision of exogenous insulin therapy may result in good control of most hypoglycemic and hyperglyceniic episodes. Despite snch control, liowever, the jnvenile diabetic is still faced with early vascular changes and a consideralily shortened life span. One approach to the problem of stabilizing lx)th blood glncose flnctnations and, possibly, vascular complications wonld be allotranslalantation of a functioning pancreiis to the diabetic recipient. During the past decade, many investigators have performed transplants of whole or partial pancreatic organs in animals {Bergan and Teixeira, 1967; Lillehei et ai, 1970) and, more recently, in man (Connolly et al., 1973; Cliedman et al, 1973). Although the number of ca.ses is small, morbidity and mortality • Present address: University Dept. of Surgery, St. Vincents Hospital, Fitzroy, Victoria.
58
BRUCE N. GRAY AND ELTON WATKINS, JR.
rates are unacceptably liigh; 32 partial or total pancreatic allografts had been recorded by the Organ Transplant Registr>' as of 1973 (ACS/NIH, 1973). Of these, oiily 2 patients had survived with functioning grafts longer than one year. To circumvent the technical difRculties associated with whole organ transplants, several groups ha\e invf.stlgated trau.sphintatioii of isolated pancreatic islets. This paper reports onr investigation into transplanting isolated panereatic islets in tho rut. MATEBIALS AND METHODS. Animals. All donor rats and syngeneic retipients were inbred Wistar/Lewos (LEW/Crl) strain malt' rats weighing between 200 and 300 !^. All diabetic animals receiving allografts were inbred Fischer (F344/Crl) strain male rats weighing between 120 and ISO K- The Fi.scher .strain presents histocompatibility genes, a. h, c, Ag-A^, Ag-B^, Ag-C, and H-V, while the Wistar/Lewis strain presents Ag-A*, Ag-B^, Ag-C, H-P, and H~b^. Fischer and Wi.star/ Lewis strains also differ at the Ag-D locus (Palm, 1971). Weak histocompalibility differences exist between the two .strains. Induction of diabetes. Graft rcoipients and control diabetic animals were rendered diabetic by tl'ie intravenous administration of the antibiotic, strepluzotocin, 6.5 mg/kg-^ of body weight. Transplantation of islets was carried out after the animals had shown sustained signs of diabetes for a period of at least 2 weeks. IsohUon of islets. Islets were obtained by a modification of the method of Ballinger (Ballinger and Lacy, 1972). Techniques of mecfianical dissociation and enzymatic digestion of ths pancreas were similar to those described by those authors. Final separation of i.slets from acinar tissue was performed by centrifugatioo at 700 g for 7 min in a differential density gradient of 31-5% (w/v) Ficoll-suline, 25S P'icoU-saiint', and Hank's solution. The i.slets separated out at the uppcnnost interface and toirld easily be aspirated off and used for transplantation by injection through a 23-gauge hypodermic needle. From each donor pancreas at least 200 intact islets could be obtained. Total isolation time was 60 to 70 min. At least 95^ of the material obtained at the i.ippeniiost interface of the Ficoll density gradient consisted of islet tissue and only 5% consisted of acinar debris. The combined islets from three donor animals were used for each graft. This >'ielded approx. 600 intact islets, and the suitability of different sites was assessed using this same number in each case. The assessment of transplant site suitability was perfonnetl using syngeneic grafts only to obviate rejection. The transplant sites tested included intraperitoneal implantation, subcutaneous implantation, testicular implantation, mesenterie root implantation, sy.steniic venous enibolization, superior mesenteric artery embolization and portal venous enibolization. Methods of immunosuppression. Preliminary investigations suggested that rabbit anti-rat antilymphocyte serum (ALS) was siq>erior to azathioprine (Imuran) in its ability to suppress rejeetion of allografted paiiCI%atic islets. Uiabt'tic Fischer strain anininls receiving allogeneic Wistar/Lewis strain islets recpived ALS (Microbiological Assoc. Lot #13104) in either a high- or low-dose schedule as shown in Table L
ISLET TRANSPLANTATION IN DIABETES
59
Non-fasting senim glucose estimations were performed on blood samples obtained by tail dipping (Glucostratc. Warner-Chilcott). Urine output was measured each day from test aniinul.'i housed in metabolic cages. Test animals were weighed to tlu^ nearest t;rain two or three times a week. Glucose tolerance tests were perfonned on animals fasted for 4 b by administering U-glucose solution. 2-5 g/kg^^ of body weigbt, via an intragastric tube. Serum glucose levels were measured as described above. A successful transplant was defined as one in whicb nonfasting serum glucose levels, 24 h urinary volumes and rate of weight gain returned to nonnai control levels for a miniimun period of 30 days. Animals witb partial improvement only or animals witb initial complete recover>' followed b>' regression to the diabetic state were regarded a.s failures.
TABLE 1. Schedule of immimosiippressian with ALS. High dose 0-1 ml per animal, 6 days/wk for I month 0-1 ml per animal, 3 days/wk for next 2 wks 0-1 ml per animal, 1 day/wk for next 3 months 0- ] mi per animal, I day/month for next 6 months nil thereafter
Low dose 0' 1 ml per animal, 3 days/wk for 1 wk 0-1 ml per animal, I day/wk for next 4 months 0-1 ml per animal, 1 day/month for next 3 monlhs nil thereafter
Skin grafts. Full-tbicltness grafts of approx. 1-5 by 1-5 cm were taken from tbe abdominal skin of male Wistar/Lewis rats and placed with 180" rotation on the central back just below the scapula of Fiscber recipients. Tbe time of rejection was defined as tlie point at wbich the whole graft had been converted to a dry scab and conld be lifted from tbe graft bed. Three groups of Fiscber animals received Wistar/Lewis skin grafts. Cnnip 1 consisted of 10 animals which had received allogeneic Wistar/Lewis islet tran.splantM and had been iiimmnosuppressed with ALS. Group 2 amsisted of 5 normal animals wbicb received ALS in tbe bigh-dose protocol starting at the time of .skin grafting. Group 3 compn.sed nonnai animals without iininnno.suppression. Mixed lymphocyte cultures (MLC). Blood was obtained from animals by cardiac puncture and lympbocytes separated on a FicoU-IIypaque gradient (Thorsby and Bratlie. 1970). Cultures were perfonned in niundbottomed micioculture plates (Linbro Chemical Co.) in 0-22 ml of medium lU'MI 1640 supplemented witb IS- glutamine (Microbiological As.sae.) and 10% lieat-decomplemented nonnai rat serum. Pulse tagging was on days 2 to 6 witb 0-5 microcuries of tritiated tbymidine (New England Nuclear) per well. Each microculture well contained 0-25 by 10" eirector lympbocytes and 0-25 by 10** stinmlator lymphocytes. Polynucleotide synthesis was blocked in stimulator lymphocytes by prior incubation with mitomycin-C (Nutritional Biochemical Inc.), 25 micrograms/IO*' cells/ml for 1 b. Ceils were harvested after 17 h on glass fibre strips using a MASH II cell culture barvesler (Microbiological Assoc), and radioactivity was measured by conventional litpiid scintillation techniques. Activation indexes (AI) for one-way MLC assays were calculated by tbe following formula; (CPM. effector cells + stinuilator cells) - (GPM. blank) AI = (CPM. effector cells) (CPM. stimulator cells) - 2x (CPM. blank) alone alone Significant deviation of the activation index from an iiide.\ value of 1-0. denoting absence of bla.stosis. was detennined by Student's t-test.
60
BRUCE N. GRAY AND ELTON WATKINS, JR.
J^ses.iment of renal and ophthalmic changes. Three groups of Fi.scber animals were used to determine tbe ability of tran.splanted islets to prevent the occurrence of renal and opbthalniic changes, Group I consisted of 10 animals wbich were allowed to remain diabetic. Group 2 consisted of 10 animals wbicb received a corrective portal venous transplant of allogeneic Wistar/Lewis islets 2 to 4 weeks after the induction of diabetes with streptozotocin. In group 2, 6 aniiiiats were immunosuppressed witb tbL' high dose protocol of ALS and 4 animals witb tbe low dose protocol. Group 3 consisted of 10 normal agfMnatcbed animals. Nine months after the induction of diabetes the left eye and kidney were removed from eacb animal in groups 1 and 2. The left eye and kidney were fi.xed in 10^ formalin, sectioned at 2 and 5 nM for the kidney and 5 nM for the retina, and stained witb baeniato.xylin-eosin and periodic acid-Scbiil (PAS) stiiiiis. Retinal sections were prepared from 10 animals and renal sections were prepared from 3 animals in group 3 for comparison witb groups 1 and 2. Codec! bistnlogic sections were randtimized antl scored without knowledge of the code by tbe first author (BNG) on an arbitrary scale of one to four for tbe following changes. In the kidnty, pathologic changes assessed were amount of mesangial deposition of PASpositive material, degree of thickening of the gtomerular afferent arteriole wall, and tbt* degree of thickening nf the purii^tal layer of Bowman's capsule. The greatest diameter of .30 randomly selected glomeruli from each of the test animals in tbe tbree groups was mea.sured, using a microscope reticle in order to assess any cbange in gloinerular size. For tbe assessment of retinal changes, randomized .sections from the three groups were scored on a seal;- of one to four tci .show thif degree (if capillarj- ancf small vessel dilatation of vessels in tbe stratum opticum and outer plexiform layer. For tbe as.se.ssment of retinal new vtssel formation. 10 artas of 0-1 nun- were randomly .selected in each retina and the number of capillaries counted in each of tbe four layers as suggested by Sosula et ai (1972). The four layers were tbe outer plexiform layer, inner nuclear layer, inner plexiform layer and stratum opticum. In scoring renal and retina! changes, a score of one was given wben tbe degree of change being assessed was minimal, whereas a score of four was given to sections showing tbe most advanced cbange. Statistical significance between groups was compared using tbe Student's t-lest. P values of less than 0-05 were considered significant. Test animals were inspected periodically for the presence of cataract formation.
RESULTS. Induction of diabetes. Animals injected with streptozotocin showed diabetic rnanife.stati()ns within hours. Non-fasting serum ghico.sf le\'els increased from a range of 100 to 200 nig per 100 ml of serum to 450 to 900 mg per 100 ml of serum. Uriiu; output iiicreased from a mean of 5 ml per 24 h to between 50 and 130 ml per 24 h. Xjbe rate of weight gain dropped dramaHcally and was invariably less than 30% itrfthe normal rate. In many cases weight loss was observed. Transplantation to various anatomic sites. Of all sites tested, emboUzation into the portal venoxts system proved to be the most efFeetive method (Gray and Watkins, 1973). Althongh implantation into the base of the mesenter)' proved effective in the 1 animal tested, its technical difficulty precluded further evaluation. In 2 of 3 animals in which the islets were ernbolized into the superior niesenteric artery, there was initial complete relief for 7 and H days respectively followed by return to the diabetic state. Postoperative haemorrhage commonly occurred when the superior mesenterie artery
ISLET TRANSPLANTATION IN DIABETES
61
route was used. Only those animals surviving the procedure have been included. In contrast, embolization into the portal venous system was free of operative morbidity and mortality. lntraperitoneal implantation using 600 islets per transplant was met with only partial improvement by the three criteria. Transplantation to other sites failed to alter signiHcuntly the diabetic state of the graft recipients. Table 2 shows the success rate of the various sites tested. Histologic verification of islet survival in host liver is shown in Fig. 1. TABLE 2. Survival of transplants to varhms sites.
Siic or transplant Intraperitoneal Subcutaneous Intratesticular 5ystemic venous Superior mesenterie ariery Mesenterie root Portal venous
Number of transplants
Transplant survival
4
0 0 0 0
5 2 3 1 10
Q
I 6
Fig. 1. Pancreatic islet lodged in portal venous radicle of host liver 1 month after seeding of isolated syiigeneic islets into the bost portal vein (liaematoxyhn-eosin stain), x 437.
m
BRUCE N. GRAY AND ELTON WATKINS, JR.
Transplantation of allogeneic islets. Of 5 Fischer strain animals which received allogeneic Wistar/Lewis islets into the portal venous system, witliout immunosuppression, all rejected the grafts within 8 to IS days with return of diabetes in all in.stances. Survival and function of allogencie islet transplants were prolonged b\- immunosuppression with ALS. Of 5 transplant recipients receiving high dose ALS, all continued to have functioning grafts at 10 months after tran.splantation. Of 5 recipients receiving low dose ALS, 1 animal rejected the transplant after 4 months, and the remaining 4 animals were still free of diabetes 7 months after transplantation. Serial non-fasting scrum glucose levels and 24 h urinary volumes are shown in Figs 2a and 2b and are compared to a group of 5 Wistar/Lewis animals whieh received syngeneic transplants. Recipients of allogeneic portal venons transplants were able to handle a glncose load in a manner similar to normal animals, in contrast to control diabetic anhnals, as shown in Fig. 3. Granulated beta cells have been demonstrated in host liver 12 months or more after the transplant procedure. liy this time, however, the general architectnre of the islets was lost, although densely staining beta cells could readily be identified within the portal triad (Fig. 4).
FIG.
eooi
2A.
ALLOGENEIC W/L-
SYNGENEIC
400
W/L-^-W/L
SERUM 200 GLUCOSE {MO/100 ML)
•«—•-••
ALLOGENEIC W/L-»-F ALS (High Dose)
ALLOGENEIC W / L - » - f ALS (Low Dose)
-10 -5
Ol
7 15 DAYS
10
TIME
See caption on
MONTHS
page.
IS
ISLET TRANSPLANTATION IN DIABETES
63
TABLE 3. Mixed tymphoeyte eutlure responses. Animals (lymphocyte donor)
Exp. No.
Tolerant lo both pancreatic and skin grafts
1 2 3 4 5 6 7 8 9
Day of maximum response
Activation index
Normal
P Value
9.4 13 7 31 3-7 2 9 31 2 5-6
6 5 5 6 6 6 5
