¢

American Journal of Patbolol!.l'. vol. 140, No. 5. •\I~ 1992 C~ngbt C American AsWcialion of PatboiogislS

Islet Cell Allotransplantation in Diabetic Patients

Histologic Findings in Four Adults Simultaneously Receiving Kidney or Transplants

C. E. Sever,· A. J. Demetris,* Y. Zeng,t techniques and better control of rejection have led to a A. Tzakis,t J. J. Fung,t T. E. Starzl,t number of successful transplants in type I insulin depen­ and C. Ricordit dent diabetics with some patients achieving insulin inde­ 7 10 From the Departmf!P/I of PlIIboloPJ·.· and the PittsbUTRb pendence. - At our own institution, treatment with the • Transpla1Jt In.~titute, t ('nil'f!l"Sity of PittsbllrR'J. PittshurRb. powerful immunosuppressive agent FK506 has contrib­ Pennsylt,(lnia uted to successful islet cell transplantation in patients re­ r ceiving multiorgan transplants after upper abdominal ex­ Re/ined methods 0/ islet cell PUrification bave led to enteration,11-13 During the past year an additional num­ unprecedented success 0/ islet cell allotransplanta­ ber of patients with type I insulin-dependent diabetes I tion via portal vein infusion in diabetic patients, mellitus received islet cell transplants simultaneously with resulting in marked reduction 0/ exogenous insulin either liver or kidney allografts. survival in these pa­ requirements and recently even insulin indepen­ tients is further complicated by potential recurrence of the dence. The authors report tbe bistologic findings of primary disease similar to that reported in patients receiv­ islet cell allogra/ts in tbe liver of/our patients who ing pancreas allografts. 14 To date. histologic examination bad undergone combined kidney-islet or liver· islet of successful islet cell transplants has been largely con­ transplantation. Islet cell clusters U!ere detected in fined to a"nimal studies. 11>-17 Little is known about mor­ Subcapsular location at tbe edge ofportal triads. Tbe phologic findings in human islet cell transplantation via early post· transplant period was cbaracterized by portal vein infusion. which could potentially add valuable /latchy mixed portal infiltrates. Only minimal in­ information in addition to functional assays. In four of our flammation but decreased islet cell granulation UJaS patients. liver biopsy specimens or autopsy tissues con­ ,/bserved in one patient 6 montbs alter transplanta­ taining islet cells were available and the histologic find­ tion. As bistologic detection 0/ transplanted islet cells ings are presented in this study. becomes available. additional parameters for evalu­ ation 0/ graft suroival migbt be defined by morpho­ Patients and Methods II/glc assessment. (Am J Patbol 1992. 140:1255- IlGO) Tissues Liver tissue was examined from one needle biopsy taken Despite successful treatment with exogenous insulin di­ at 14 days post-transplant. one wedge biopsy taken 2 abetIC patients suffer from a myriad of long·term compli­ days post-transplant and two entire organs obtained at ':Jet cell transplantation IS an attractive altemative and ~been Conducted experimentally for years. Until re­ Accepted tor pubIicabcn December 26. 1991. been: the ~n ~bstacles to long-term success have Address reprinlrequaIII to Or. Carillo Aicon:Ii. ~ 01 &6. _InSufficient Yield of functional islets and allograft re­ gtIl'I. UnivenIiIY 01 PiIIstu\1I. SchociI 01 MediCine. 3601 FtIIh AY8flJ8 • .--..n 3-7 \Ar.ithi • Ht n the last 2~ years. refined purification PilIsbugh. PA 15213.

1255 1256 Sever et al AlP May 1992. Vol. 140. No.5

biotin complex (ABC) method as described by Hsu. 18 A amount of islet cells he had received initially was cons.o panel of antibodies was used W1th antibodies directed ered suboptimal. an addrtional Isolate from a third donor against Insulin. glucagon. chromogranln (all Blogenex, was infused 5 days later. To date. he stili requires, 6 UI"oII San Ramon. CAl. UCHL-1 (DAKO. Santa Barbara. CAl. of insulin per day: however. this represents an ~ r S and LN3IHLA-DR (Biotest Diagnostics, DenVIlle. NJ) as duction compared with pretransplant reqUirements e summarized in Table 1.

Case 3 Patients A 46-year-old man was admitted with end-stage renal disease secondary to diabetes mellrtus. He unde!wenl All patients underwent islet cejl transplantation via portal cadaveric renal transplant and Simultaneous islet trans vein infusion as previously described. 12 Each patient re­ plant with islets prepared from one donor pancreas ceived purified isolates prepared from one or two donors Three days after transplant he suffered an epiSOde of as previously described. 19.20 In addition. two of the pa­ aspiration with prolonged cardiac arrest. He was trans tients received a cadaveric renal transplant and the other ferred to the intensive care unit, and two received liver transplants. Immunosuppression was was stopped. He did not recover and died 5 days alter achieved With FK506 and variable amounts of pred­ transplant. An autopsy was performed, which showed nisone In all patients. Details of the clinical course and diffuse alveolar damage and ear1y focal bronchopneu. Islet cell function of all four patients are published as part monia but no cardiac findings. specifically. no myocardi­ of a larger series of 22 patients receiVing Islet cell allot­ al infarction could be found. Other findings are deSCribed ransplantatlon. 21 under Results.

Case 1 Case 4 A 48-year-old man was admitted with end-stage liver A 31-year-Old woman was admitted with end-stage disease due to alcohol abuse and a history of upper GI liver disease secondary to cystic fibrosis. In addition she bleeding. In addition he had been an insulin-dependent had a history of diabetes mellitus. diagnosed at age' 5. diabetic requiring four dally In)8Ctlons ranging from 40 to She underwent orthotopic liver transplant and simu~a­ 80 units of insulin. He underwent orthotopIC liver trans­ neous islet-cell transplant with islets prepared from pan­ plant and Simultaneous Islet-cell transplant With Islets pre­ creas of two donors. Blood obtained dunng the operation pared from one donor pancreas Fourteen days after the before Infusion of islets revealed high levels of C-peptide. operation. a liver needle biopsy was done to evaluate the indicating endogenous Insulin production and therefore liver allograft: 6 months after the combined transplant he an Insulin resistant form of diabetes mellitus. There was died of hepatitis B and sepsIs. No postmortem tissue was no Significant reduction of insulin requirements. and she available for additional studies. had multiple complications In the months after transplan­ tatIOn. Including renal failure. marginal respiratory func­ lion. and multiple infections. She died With clinical signs Case 2 of sepsIs 9 months after transplant at an outside hospital. A 34-year old man was admitted W1th end-stage renal where permiSSion was granted for an autopsy. Blocks disease secondary to diabetes mellitus. He underwent from formalin-fixed. paraffin-embedded liver tissue ob­ cadaveriC renal transplant and Simultaneous Islet-cell tained at autopsy were sent to our institution for analysis. transplant with islets prepared from pancreas of two do­ nors. Postoperatively. he was taken to the operating room again for evacuatIOn of a pennephric hematoma. A liver Results wedge biopsy was done at the same time. Since the In all patients. islet cells were found predominantly in a

Table 1. Specificity. WOTkin!? Dilution. and Sou.rce of the sutx:apsular location. In the first two patients that under­ Antibodies Used in the Study went liver biOPSies. only Insulin and glucagon stains were Antibody Dilution Source pertormed. In patient 1. an islet cluster was Situated at the edge of a portal triad (Figure 1). The Inflammatory infil­ anti-Insulin 11067 Biogenex trate conSiSted predominantly of neutrophlls and did not anti-glucagon predlluted Biogenex antl-chromogranin 1.320 Biogenex seem to be particularly concentrated around this cluster UCHl-1 (anti-CD45R) 110 DAKO but was scattered diffusely throughout the triads that lN3 (anli-HLA-DR) 15 Biotest Diagnostics were visualized during the biopsy. r ,I II !.

Histology of Islet Cell Transplantation 1257 I AlP Mav 1992, Vol 140, No 5 I'

• ,"~ .• /~ ... " ...... I' immunoperoxidase staining was performed demon­ .. " ...... \ "". . " ,. ~ "'., .... ~ ,.' ., " • .;~:.:..~:,..... , 1l~ strated islet cells containing insulin granules, mostly lo­ 3 °r.. :_ ~:.~. t,,-.~ ...... , .. ,., !ft,' j' " 0 "" .• ,.' ...... :.....;~:~.~. 1'.·/·;. cated within the portal triads (Figure 3a). ApprOXimately • lJ , i ' _ ~··-"i . . ,. ".... ' ... J \" , ~ .. ... e- ,,~~'.::t.- •• ,...... ,,". \ ,," ~ 20% of cells Within the same cluster stained positive for • . .. t' .\ - ...'... r...... ,.." "...... ". ~ ,_t).: & .•, .... , '.,.,-:' .. \ ...... h~ .. 1 : .. ~/ .. "! J ".!'~" cr. ,.~: ':-. "" ~'.:. ; .• ' ,~ '" glucagon (not shown) An occasional single Insulin­ ~. . -J L,<\" •• ' ."t' ~ ... ,,, • '~, •. ~/" .... , ,"~'. ". ', ... -:;: ...... \ ,.J~.\~...... ,f: containing cell was detected WIthin the lobule along the 0·.· . I

Discussion

~.. 0' Islet-cell transplantation traditionally has been monitored .~ -;,.; ..-~., ~ . -I '> ,: .' ~ •• by production of C-peplide and blood glucose levels. As . -.. ...)\ -~ _

.',-

Figure 3, A-C: SectIOn of iiI 'I!/" fro", {>llllew 3, obtained at at4l0PSI' '; dm's after combmed IZldm .... ··islet tramplalll, A: Islet cell cluster u'llb tlPlcal granular C','loplasmlc Slamm,l:!. for I'NIIiIl ('lIIti·i'L~ulin mlnl/moper-oxulase, X3'iO) B: Identical area stailled for actila/ed T-cell marker. .Ilost of the mU11I round ll'mplJo<1'les are pos//I/'e 71Jf! cell cllLwers lluilcated VI' arrowheads could represent foreign lxxii, gialll cells cleanng up cellular deiJrlS from a01UIT cells or oliJer noncellular C01UamllumL~ generated durinR tbe isolation procedure ({'CHL·! immu, noperaxuiase, x3'iO) C: Idellllcal ar!'a stailled for class II mlligl!11 SelY?ral strongly POSitllV! blstioCl'IlC cells u'ith di!1ulrilic C','tOPIasm,C prOJectrOIlS could reprl!Sl!1l1 allll,1.(I'l1 preseWlIlg cells ,I/am' ,~lIl1l.,.,npb()cl'les also stam (JOS1I11V!, lluiicatm,,! r -{.'ell actimtlon rather than B-cell Imeag!' (a1ll1H/~'[)R L \' ~ I11lmWlOpe1'O,\'ldlL,e, x; ';0) 0: SeCI/OII of allograft kui1U")'from same {>lilli'll I, sholl'll/g a palcJn', mlid 10 moderate Il'mpboc\'llc mfiltrate, COllSlSI£'111 u'llb earl\' {Kille cellulw' reJeCtIon (II{:-E, X NO).

preferential location for Islet cell engraftlng. The same ob· domlnal exenteration and liver replacement '2,13 and servatlons are being reported In two other patients who leave little doubt as to where sampling should be per­ had undergone Islet cell transplantation after upper ab- formed, These findings are Similar to Intraportal islet cell transplants in rats and dogs, which also resulted in pre­ dOminantly subcapsular location of endocrine cells,lS17 In our matenal islet cell clusters were mostly confined to locations Within and at the edge of portal triads, whereas only rare Single subSlnuSOldal Insulin poSitive cells were observed Within the lobule, In canine autografts, most of the reported microscopic findings are similar with the ex­ ception of slightly more numerous endocrine cells wrthin the lobule,'7 This could reflect subtle differences in the disperSion of islets dunng the pUrification procedure and could result in altered function and survival of a small subset of grafted cells, The fact that Islet cells could even be found on a liver needle biOpsy from one of our patients suggests that large numbers of Islets do survive the transplantation procedure, Even In the Immediate post­ Figure 4. Sectron of 1i1Y?r from paiumt 4, oblained at auropsl' 9 monlbs after comhined iiI Y?r'L~Iel transplalll, Setwal clusters ofL~/et transplant period inflammation around the islet cell cells. all/lined 171' arrou beads, stam 011/)' u'f'Oklv lI~tb Insulin No thrombi can be minimal as in our second patient and 'lRnificant l1t/1ammaror)' I7ljiltrate is I1sihle (anit·insulin immlma­ peroxidase, X350) Insulin stains are helpful for detection of small islet clus- • r I

Histology of Islet Cell Transplantation 1259 AlP Mtz:I, 1992, Vol. 140, No.5

ters. In our hands the stain for insulin was consistently nents and cannot be reliably distinguished from early re­ more Intense than for chromogranin and IS preferred as a jection at this point. Findings 6 months after transplanta­ survey stain. Glucagon-secreting cells could be demon­ tion included large Islet clusters at the edge of portal tri­ strated in larger Islet clusters in the third and fourth pa­ ads with no or minimal inflammation and only weak tient. Small clusters were usually lost on deeper cuts. The granulation. As more material becomes available for hiS­ intense portal InflammatIOn which was seen 2 and 5 days tologic examination. morphologic findings might prOVide post-transplant In some triads of the second and third valuable informalion for prognoSIs and management of patient IS likely to be a reaction to nonlslet cell contami­ Isiet transplant recipients. nants such as aCinar cells. soft tissue and hematolym­ phold elements. Our own studies on the compOSition of islet-cell isolates have revealed that even the most highly References pUrified preparations contain approximately 10-20% other cell elements and can be as high as 60%.22 Islet Gray DWR MOrriS PJ Development In Isolated pancreatic cells have been shown to engraft In a great variety of sites Islet transplantation. Transplantation 1987,43.311-331 2 Sutherland DE Pancreas and Islet Transplantation II. Clln­ such as forearm muscle. omentum, pentoneum, spleen, leal trials. Diabetologla 1981.20435-450 kidney. ' .3.46 but the liver has emerged as an especially 3 Scharp D. lacy P Ricordl C. Boyle P Santiago J. Cryer p. favorable location once the Infusate was reduced to such GlngenCK R. Jaffe A. Anderson C Human Islet transplanta­ a low volume as to prevent portal hypertenSion. A slgnlf­ tion In palients With type I diabetes. Transplant Proc 1989. I~ant change of portal vein pressure was not observed In 21.2744-2745 any of the four patients. although islet-cell thrombi COuld 4 Tuch BE. Sheil AR. Ng AB. Turtle JR. Expenence With hu­ be demonstrated histologically. The portal inflammation man fetal pancreatic allografts over a three-year period. as seen In the second (not Illustrated) and third patient, 2 Transplant Prce 1987. 19.2357-2358 and 5 days. respectively. after islet-cell infusion. could 5. Farkas G, KaracsonYI S. Szabo M. Kaiser G: Results of CUl­ represent cleanng up of degenerating aCinar and other tured fetal pancreatic Islet transplantation In Juvenile diabetiC transplanted non-Islet cells. Nevertheless. an additional patients Transplant Prce 1987. 192352-2353 6. Robertson RP. Lafferty KJ. Haug CEo Weil III R. Effect of component of early cellular rejection cannot be ruled out human fetal on secretion of C-pep­ In the third patient. especially since Immunosuppression tide and glucose tolerance In type I diabetiCS Transplant had been Withdrawn 2 days before the patient's death Proc 1987. 192354-2356 and changes conSistent With early cellular rejection were 7. Scharp DW, Lacy PE, Santiago JV. McCullough CS, Weide found In his allograft kidney (Figure 3d) Despite the pro­ LG, Boyle PJ. Pal qui L Marchetti P. Ricordi C. Gingerich RL, nounced Inflammation In many triads, Islet cells appear Jaffe AS. Cryer PE. Hanto DW. Anderson CB, Flye MW well granulated and Clinically no exogenous Insulin was Results of our first nine Intraportal Islet allografts In type 1, reqUired to maintain normal blood glucose levels on the Insuhndependent diabetiC patients Transplantation 1991, day of the patient s death. Mononuclear Infiltrates Within 51.76-85 ,slet cell clusters such as described In rejection of whole 8. Warnock GL, Kneteman NM. Ryan EA. Evans MG. Seelis pancreas transplants (Isletitis) were not seen.'4 RE. Halloran PF, Rabinovltch A. RajOtte RV Continued func­ tion of pancreatic Islets after transplantation In type I diabe­ More matenal needs to be studied to determine morpho­ tes. lancet 1989. 2570-572 logiC correlates of relection more accurately. In the fourth 9. Scharp OW. Lacy PE, Santiago JV. McCullough CS, Weide patient, the weak staining of islet cell clusters for insulin lG. Pal qui L, Marchetti P, Gingerich RL. Jaffe AS. Cryer PE, (Figure 4) IS Interpreted as excessive degranulation Anderson CB, Flye MW: Insulin Independence after islet whICh correlates With the clinical Situation of insulin resIs­ transplantation Into type 1 diabetiC patient. Diabetes 1990. tant diabetes and parenteral hyperalimentatlon. both of 39:515-518 MilCh Chronically stimulate sustained insulin release. De­ 10 Warnock GL Kneteman NM. Ryan EA. Evans MG, Seeils generative changes such as descnbed In failed canine RE. Rablnovltch A. RajOlte RV Normoglycaemla after trans­ ISlet grafts were not observed. 17 Although the patient did plantatIOn of freShly isotated and cryopreserved pancreatic not benefit from the islet transplant, the findings dOCu­ Islets In type 1 (Insuhn-dependent) diabetes mellitus. Dlabe­ ~t the long-term survival of morphologically Intact al­ tologlca 1991. 34.55-58 'Clgraft Islets In sufficient numbers to be detected on hls­ , 1 Tzakls AG. Rlcerdl C. Alejandro R. Zeng Y Fung JJ. Todo S. DemetriS AJ. Mintz DH. Starzl TE. Pancreatic Islet cell trans­ l0loglc examination. In summary, the early posttransplant plantabOn after upper abOomlnal exenteration and liver re­ Period In diabetIC patients receiving a combined kldney­ placement. Lancet 1990.336.402-405 'Slet or liver-Islet aJlograft IS charactenzed by focal portal 12. Ricordi C. Tzakls A. Alejandro R, Zeng y, Demetns AJ. Car­ InflammatIOn ConSiSting of lymphocytes, macrophages, roll P. Mintz DH Starzl TE Detection of pancreatic Islet tiS­ ~ OCcasional eo5Inop/111s. which probably represents sue following islet allotransplantation in man, Transplanta­ l ...... 0 01 decaying nonendocnne aJk>g ... compo- !Jon 1991, 52107S-1 1260 Sever et 81 AjP May /992. VoJ. /40. No 5

13 Rlcordl C. Tzak!s A. Alejandro R. Zeng Y. Demetns N. Car­ peroxidase complex (ABC) in Immunoperoxldase leC/' roll P Fung JJ. Mintz DH. Slarzi TE Detection of intrahepatic nlQues. A comparison between ABC and unlabeled ar' human Islets follOWIng combined liver-Islet allotransplanta- body (PAP) procedures J Hlstochem CytOChem '98' tion Pancreas (in press) _ 29577-580 14 Sibley RK. Sutherland DE: Pancreas TransplantatIOn. An Im­ 19. Ricordi C. Lacy PE. Finke EH. alack BJ. Scharp OW Al.10 I' munohistologic and histopathologic ExamlnallOn of 100 mated method for isolation of human pancreatic Islets D.t . Grafts. Am J Pathol1987. 128:151-17015 betes 1988,37413-420 15 Griffith RC. Scharp OW. Ballinger WF. Lacy PE. A morpho­ 20. Ricordi C. Alejandro R, Zeng K, Tzakis A. CasaVi11a A Jall4> logic study of intrahepatic portal vein islet isografts. Diabe­ R. Mintz DH. Starzl TE: Human Islet isolation and Punflea! .... tes 1975.24419 from pediatric-age donors. Transplant Proc 1991. 23783 16. Millard PR. Reece-Smith H, Clark A. Jeffery EL. McShane P. 784 Morris PR The long-term allografted rat Islet. A Histologic and immunohistologic study. Am J Pathol 1983. 111.16&- 21. Ricordi C. Tzakis AG. Carroll PB. Zeng Y, Rita HLR, AIEl!iIr' 173 dro R, Shapiro R, Fung JJ, Demetns AJ, Mintz DH. Starzl T[ 17. Alejandro R. Cutfield RG. Shienvold FL. Polonsky KS, Noel J. Human islet Isolation and allotransplantation in 22 COOsec Olson L. Dillberger J. Miller J. Mintz 0 Natural history of utive cases. Transplantation 1992.53:407-414 Intrahepatic canine Islet cell autografts. J Clin Invest 1986, 22. Sever CE, Demetris AJ. Zeng Y, Carroll P. Tzakts A. Star, 78.1339-1348 TE, Ricordl C Cellular composition of Islet cell SU5penSI()r\<; 18 Hsu SM, Raine L. Fanger H The use of aVldin-blotin- for transplantation. Transplant Proc (In press)

------..~