Kidney International, Vol. 21(1982), pp. 323—329

Prolongation of canine allograft survivalwith donor pretreatment

HENRI L. F. BROM, PETER J. C. VAN BREDA VRIESMAN, and JOHAN L. TERPSTRA

Departments of Surgery and Pathology, Leiden University Hospital, and the Department of Immunopathology,University of Limburg at Maastricht, The Netherlands

Prolongation of canine allograft survival with donor pretreatment.Donor pretreatment of 100 mg/kg each of cyclophosphamide (CY) andmethylprednisolone (P) infused 5 hours before nephrectomy invariablyprolongs the survival of DLA mismatched, MLC incompatible non-littermate Beagle renal allografts as well as the survival of mongrel renalallografts. The effect of donor pretreatment appears to be mediated bycyclophosphamide and its metabolites because methylprednisolonepretreatment does not significantly prolong survival. Methyipredniso-lone is needed, however, because it abolishes cyclophosphamide pre-treatment mediated early but transient postoperative renal (allograft)insufficiency. The effect of donor pretreatment appears to be mediatedby drugs residing in the graft; mannitol infusions given 1 hour prior todonor nephrectomy or peroperatively into the recipient decrease therenal cortical content of carbon 14 cyclophosphamide and its metabo-lites and abolishes the prolonged survival. Because donor pretreatedkidneys contain less than 0.5% of the infused dose of carbon 14cyclophosphamide, the drugs appear to exert their effect locally in thetransplanted kidney. Donor pretreatment mediating prolonged caninerenal allograft survival appears to be an example of influencing abiological process by a localized drug delivery by virtue of uniqueproperties of the drug and because early postoperatively host sensitiza-tion occurs mainly at the site of the graft.

Prolongation de l'allogreffe survie chez le chien au moyen du pre-traitement du donneur. Le pré-traitement du donneur (des 100 mg/kg dechaque drogue) au moyen de cyclophosphamide (CY) et de methyl-prednisolone (P) perfusés en 5 heures avant Ia néphrectomie, prolongedans tous les cas Ia survie d'allogreffes rénales chez des chiens Beaglen'appartenant pas a Ia méme portée, MLC incompatibles, DLA non-identiques. De mCme Ia survie de l'allogreffe rénale est allongée chez lechien bâtard. L'effet du pré-traitement du donneur parait avoir pourmédiateurs le cyclophosphamide et ses métabolites parce que le pré-traitement par Ia methyiprednisolone ne modifie pas significativementIa survie de Ia greffe. La methyiprednisolone est cependant nécessaireparce qu'elle abolit l'insuffisance rénale transitoire précoce de Ia greffe,elle-méme dépendante du cyclophosphamide. L'effet du pré-traitementdu donneur paralt avoir pour médiateurs des drogues presentes dans Iagreffe; Ia perfusion de mannitol, administrée une heure avant Ianephrectomie ou au receveur, diminue le contenu renal cortical encarbone 14 cyclophosphamide et en ses métabolites et abolit l'allonge-ment de Ia survie. Dans Ia mesure oi les reins pré-traités contiennentmoms de 5% de la dose de carbone 14 cyclophosphamide perfuse, lesdrogues paraissent exercer leur effet localement dans le rein trans-plante. L'allongement de Ia survie de l'allogreffe rénale de chien aumoyen du pré-traitement du donneur parait être un exemple de Iamodification d'un processus biologique par Ia délivrance locale d'unedrogue en fonction des propriétés remarquables de Ia drogue et du faitque Ia sensibilisation de l'hôte survient dans Ia phase post-opératoireprécoce essentiellement au lieu de Ia greffe.

Clinically prolonged allograft survival is brought about bymodulating the immune response of the host together withattempts to reduce immunogenicity by means of tissue typing or

donor pretreatment using large dosages of cyclophosphamideand methylprednisolone [1]. The rationale for donor pretreat-ment derives from studies in rats which suggested that leuko-cytes were immunogenic to allogeneic hosts [21!. The observa-tion that depletion of peripheral blood leukocytes of the donor(by lethal x-irradiation with 1200 rad administered 2 to 4 daysprior to donor nephrectomy) enhanced the survival of renalallografts in the rat and dog by means of a systemic effect on thedonor, and not by a local effect on the graft [3], confirmed theimmunogenecity of passenger leukocytes. Precisely how pas-senger leukocytes sensitize the host is unclear, but lymphocytesresiding in the interstitial tissue [4] of the kidney have beenshown to be capable of migrating out of a renal allograft within24 hours postoperatively in order to relocate in the centrallymphoid tissue of the host [51; there they may sensitize thehost because they share class I alloantigens [61 and Ia-likealloantigens [7] with the endothelial cells of the renal graft or,alternatively, they may amplify the host versus graft responseby a proliferative reaction directed against the host [8].

Since reduction of renal graft immunogenecity by means oflethal x-irradiation of the cadaver donor is not feasible clinicallybecause of the 48-hour interval needed to bring about leukocytedepletion and because lethal x-irradiation may damage thekidney irreversibly, [9] pharmacological reduction of graftimmunogenecity by pretreating cadaveric donors with massivedosages of cyclophosphamide and methylprednisolone wasintroduced in 1973 [10]. This strategy was derived from studiesin rats which showed that dosages of cyclophosphamide suffi-cient to cause leukocytopenia and to abolish the proliferativecapacity of lymphocytes in vitro [111 enhanced the survival ofkidneys [12] and hearts [13, 14] transplanted over a stronghistocompatibility locus. In man the effectiveness of donorpretreatment with massive dosages of cyclophosphamide andmethylprednisolone is a controversial issue for the prolongedsurvival observed in two uncontrolled series consisting of alarge numbers of patients [15, 16] was not confirmed in anotherstudy consisting of a small number of patients [17] nor observedin a controlled study using randomization of patients [18].

323

Received for publication March 2, 1981and in revised form June 19, 1981

0085-2538/82/0021-0323 $01.40© 1982 by the International Society of Nephrology

324 Brom et a!

Because the rat differs from man in terms of cyclophosphamidemetabolism [19], the lack of a consistent effect of donorpretreatment on cadaveric renal allograft survival need not onlybe due to differences in protocol, but may also pertain tospecific differences in the way each species handles cyclophos-phamide.

Our study addresses itself to the reduction of the first set ofrenal allograft immunogenecity in dogs by means of donorpretreatment with cyclophosphamide and methyiprednisolonebecause dogs resemble man both in terms of cyclophosphamidemetabolism [19] and relative resistance to the effect of cortico-steroids [201. In the dog a 5- to 6-hour period of donorpretreatment with dosages of cyclophosphamide and methyl-prednisolone sufficient to cause clearcut leukocytopenia [20]has been reported to prolong the survival of pretreated kidneysinserted into unmodified [21] and immunosuppressed [201 hosts.Others, however, failed to observe an effect of donor pretreat-ment with cyclophosphamide and methylprednisolone abovethat of methylprednisolone alone [221, and some observed anadverse effect of donor pretreatment on the first set of caninerenal allograft survival [23]. In all of the canine studies mongrelswere used, and consequently, the histocompatibility barriercrossed was not known nor could individual specific differencesin hepatic activation of cyclophosphamide [241 be taken intoaccount. Therefore, the effect of donor pretreatment withdosages of cyclophosphamide and methylprednisolone knownto abolish a proliferative capacity of human lymphocytes invitro [25] and reported to be effective clinically [15] wasassessed in a rigidly standardized protocol using tissue typemismatched inbred Beagles. Because donor pretreatment pro-longed the survival of these renal allografts, we next evaluatedthe effect of donor pretreatment on mongrel renal allografts andaddressed ourselves to the question of whether or not donorpretreatment mediated prolonged canine renal allograft survivalwas brought about by means of an effect on passenger leuko-cyte or alternatively was mediated by passive transfer of drugs,specifically cyclophosphamide metabolites, residing in the kid-ney at the time of transplantation.

Methods

Animals. Donor kidneys were harvested from mongrel dogsor from inbred, tissue typed and for the recipient mismatchednon-littermate Beagles depending on the nature of the experi-ment. When non-littermate Beagle combinations were used,they were tissue typed serologically for DLA antigens [26] andalso in a unilaterally mixed leukocyte culture (MLC) [27] withstandard techniques by A. B. Bijnen, M.D., Department ofSurgery, Erasmus University, Rotterdam, The Netherlands.Donor and recipient Beagle combinations were chosen in such afashion that they differed from each other by at least oneserologically defined DLA haplotype and also stimulated inunilateral MLC [281. The donor provided the stimulator cells (inthe unilateral MLC) and the recipient the responder cells.Unmodified non-littermate Beagle renal allografts mismatchedin this fashion are rejected within 11 days [28]. Mongrels werenot tissue typed for the effect of matching for serologicallydefined DLA-A, B, and C antigens; MLC has far less influenceon mongrel renal allograft survival compared to littermate aswell as non-littermate Beagle renal allograft survival [28, 29].Mongrel dogs were selected using the criteria of good health,

age [30], and normal renal function as shown by preoperativeserum creatinine concentrations below 100 j.moles/liter. Bothmale and female dogs were used between 1 and 4 years of age.All recipients were inbred Beagles varying in weight between 9and 14 kg. None of the recipients had been sensitized by bloodor pregnancy.

Surgery. After removal of the donor kidneys by medianlaparotomy, kidneys were perfused ex vivo with ISO ml of aslightly modified sterile Collins C2 solution at 4° C using awaterhead of 100 cm; perfusion duration varied from 8 to 10mm. Kidneys which perfused poorly—duration of perfusionover 15 mm—were discarded. After perfusion the kidneys werestored in the Collins solution on melting ice up to 1 hour, afterwhich they were grafted into the recipient. Kidneys weregrafted into the right iliac fossa of bilaterally nephrectomizedrecipients. Warm ischemia time varied from 17 to 25 mm with amean of 22 mm. Transplantations were considered successfulwhen the serum creatinine concentration was below 200.u.moles/liter on the second day postoperatively. Grafts wereconsidered to be rejected when the serum creatinine concentra-tion increased progressively to 1000 i.moles/liter or when therecipient died because of uremia. All recipients were autopsiedand rejection was confirmed macroscopically and microscopi-cally.

Donor pretreatment. After the induction of fluothan anesthe-sia, 100 mg/kg each of cyclophosphamide (CY) and methyl-prednisolone (P) were infused separately or together as a bolusi.v. Dogs were kept under light anesthesia up to the moment ofnephrectomy. The dosages of drugs chosen were identical tothose used clinically [15]. The duration of donor pretreatmentwith CY and P necessary to bring about enhanced renalallograft survival was determined in preliminary experiments. A3-hour period of donor pretreatment with CY and P failed toprolong the survival of three DLA mismatched, MLC incom-patible inbred Beagle renal allografts; the survival times of 9,11, and 11 days did not differ significantly P < 0.5 from themean survival of 9 1.8 days of unmodified controls, but a 5-hour period of donor pretreatment with CY and P causedinvariably enhanced renal allograft survival (see Results). A 5-hour period of donor pretreatment with CY and P decreased theperipheral blood lymphocytes to 25% of the original value andabolished proliferative capacity in unilaterally mixed leukocytecultures performed as described elsewhere [31].

Mannitol. Mannitol, 5 mg/kg of a 20% solution, was infusedi.v. over a 20-mm period into the donor 60 mm prior tonephrectomy or preoperatively into the recipient after revascu-larization of the transplant. In the donor diuresis was monitoredby an indwelling catheter; diuresis increased usually 10 to 20times after the mannitol infusion. An amount of 5% glucoseequal to the mannitol-induced osmotic diuresis was infused i.v.

Carbon 14 cyclophosphamide. Radioactivity cyclophospha-mide monohydrate (ring-4-'4C) with a specific activity of 250mCi per 6.2 mg of cyclophosphamide was purchased from NewEngland Nuclear Co., Boston Massachusetts. Mongrel dogs,weighing 9 to 10 kg, were infused with radioactive cyclophos-phamide, 10 to 25 p.Ci/kg, together with cold CY and P, 100 mgIkg, each 5 hours prior to sacrifice as a bofus i.v. At the time ofsacrifice one or both kidneys were perfused as described above.Depending on the protocol, the left kidney was transplanted.The perfused right kidney was sliced into 5 mm pieces perpen-

Donor pretreatment 325

Table 1. The effect of donor pretreatment on the survival of inbredand mismatched Beagle renal allografts'

Beagle renal allograft survivat in days

No. experiment

postoperatively

Unmodified Pretreated

1 10 182 8 153 9 164 11 205 6 166 10 18

Mean SD 9 1.8 17 1.8

a Non-littermate Beagles were chosen on the basis of tissue typing insuch a fashion that one donor differed at least one DLA haplotype fromeach of two recipients and also stimulated in MLC. The donor wasinfused with 100 mg/kg of cyclophosphamide and methylprednisoloneafter the removal of one kidney; the remaining kidney was harvested 5hrs later. Both the unmodified and "pretreated kidney" were trans-planted in each of two recipients. Renal allografts were consideredirreversibly rejected when the recipient died because of uremia or theserum creatinine concentration exceed 1000 l.Lmoles/liter.

dicular to the long axis; next the slices were divided intosubcapsular renal cortes with a capsule, mid and inner renalcortex, and medulla and papilla. Multiple 0.5 g tissue pieceswere taken from the inner and mid cortex, medulla and papilla,sealed in plastic, and frozen at —20° C. To determine total renalradioactivity, we stripped perfused kidneys of the capsule,hilus, and ureter which were individually weighed and sealed asdescribed above. Next the decapsulated kidneys were slicedand divided into compartments; 0.5 g tissue aliquots from theouter cortex, mid and inner cortex, papilla, and medulla weretaken and stored at —20° C. Aliquots of serum, urine, and non-perfused extrarenal organs were also sealed and frozen. Inorder to avoid color quench, aliquots of tissue or 0.1 to 0.25 mlof urine, bile or serum were absorbed on paper and burned in asample oxidizer oven, model 306 (Packard International Co.,Brussels, Belgium). The '4C02 was collected into CarbosorbTM and Perma Iluor V (Packard International Co., Brussels,Belgium). Samples were counted in a liquid scintillation counterusing the external standard for counting efficiency.

Lymph node cells. Lymph node lymphocytes were preparedby mincing lymph nodes—removed surgically prior to theinfusion of CY and P into the donor—and pressing the mincethrough five layers of nylon into a minimally essential mediumcontaining 5% heat inactivated calf serum. The cells were nextcentrifuged on Ficoll/Hypaque, density 1077, and those cells inthe interphase were harvested. These lymphocytes were over90% viable by trypan blue exclusion.

Statistics. The survivors of the renal allografts were com-pared by the Wilcoxon mid rank sum test [32].

Results

The effect of donor pretreatment on canine renal a/b graftsurvival. The effect of donor pretreatment with 100 mg/kg eachof cyclophosphamide (CY) and methylprednisolone (P) wasfirst assessed using tissue-typed Beagles in sets of three. OneBeagle donated one unmodified and one pretreated kidney toeach one of two nonlittermate Beagle recipients who differedfrom the donor by at least one DLA haplotype and also had

proliferated in a mixed leukocyte culture. After the removal ofone kidney, the donor received CY and P i.v. The "pretreated"kidney was harvested 5 hours later. Six experiments wereperformed (Table 1): Donor pretreated renal allografts survived17 1.8 days which is about twice the survival of unmodifiedrenal allografts and highly significant (P < 0.0011). Whenmongrel dogs were used in lieu of Beagles as donors, thepretreated kidneys also enjoyed a prolonged survival of 173.2 days (Table 2) which was twice the survival rate ofunmodified mongrel renal allografts (8.3 2.0 days) and alsohighly significant (P < 0.0002). In order to decide whether ornot the prolonged survival was brought about by CY and P orboth, mongrel dogs were pretreated with 100 mg/kg of CY or P.After 5 hours of pretreatment, these kidneys were transplantedinto Beagle recipients. Methylprednisolone pretreated mongrelrenal allografts survived for 10.6 4.3 days which is notsignificantly (P < 0.2) different from the survival of unmodifiedmongrel renal allografts. Cyclophosphamide pretreated mon-grel renal allografts on the other hand showed a significantly (P<0.005) prolonged survival of 15.4 2.3 days. All cyclophos-phamide pretreated renal allografts (Fig. 1), however, experi-enced an episode of early postoperative renal insufficiency asshown by a serum creatinine concentration above 200 p.moles/liter on the second day postoperatively in every instance. Theserum creatinine concentrations decreased in all recipientssubsequently and were below the 150 moles/liter in four out offive recipients within 7 days postoperatively. Operationallytherefore, all cyclophosphamide pretreated renal allografts hadto be considered as 'technical failures". Nevertheless, theeffect of donor pretreatment appeared to be mediated bycyclophosphamide and its metabolites.

On the mechanism of donor pretreatment. If the effects ofdonor pretreatment were to be mediated by passenger leuko-cytes, the passive transfer of viable lymphocytes from thedonor into the recipients of donor pretreated renal allograftsmight abolish the prolonged survival. Therefore, l0 viablelymph node lymphocytes harvested from the donor beforepretreatment with CY and P were infused into Beagle recipientsof pretreated DLA mismatched and MLC incompatible non-littermate Beagle kidneys. The number of lymph node cellsused were derived from rat experiments [5] which showed a l-grat kidney to contain about 106 lymphocytes of which 5% were"mobile" and capable of migrating out of the renal transplant inorder to relocate into the spleen of the allogeneic host within 24hours postoperatively. A 30-g dog kidney was therefore consid-ered to have no more than 1.5 x 106 "mobile" passengerlymphocytes capable of relocating into an allogeneic spleen.The passive transfer of l0 lymph node cells (30% of which maybe taken to relocate into an allogeneic spleen [5]) intravenouslyinto the recipient of a pretreated renal allograft did not abolishthe enhanced survival in two experiments as shown by survivaltimes of 20 and 19 days; also infusion of l0 autologous lymphnode cells into the renal artery of the pretreated transplantimmediately after revascularization failed to influence the effectof donor pretreatment as shown by survival times of 20 and 16days in two experiments.

These experiments suggested that the effect of donor pre-treatment might not be caused by an effect on passengerleukocytes but rather by drugs residing in the donor pretreatedkidney. Because the urinary excretion of cyclophosphamide

326 Bro,n et al

Table 2. Effect of mannitol infusions on the survival of donor pretreated canine renal ailografts

No. exp.

Donor pretreatmentRecipient

pretreatment

M

.RAS in days

postoperatively5CY P M

6 — — — — 8.3+2.111 + + — — 17.0±3.16 + + + — 7.2±1.55 + + — + 8.2±2.2

Mongrel dogs were given 100 mg/kg iv. of both cyclophosphamide (CY) and methyiprednisolone (P); their kidneys were harvested 5 hrs later,perfused with 150 ml cold Collin's solution and transplanted into Beagle recipients. Mannitol (M), 5 mi/kg of a 20% solution, was infused into thedonors 1 hr prior to nephrectomy and into the recipients preoperatively after revascularization of the renal allograft.

b Mean SDAbbreviation used is defined: RAS, renal allograft survival.

600

500

400U)C)C00• 300C(UU)0E 200U)C')

100

I

1? ?E;:

I

F

3Days postoperatively

mongrel kidneys immediately after revascularization of thetransplant. Mannitol infusions given to the recipient also abol-ished the effect of donor pretreatment on renal allograft survival(Table 2).

The cyclophosphamide content of pretreated kidneys. Inorder to decide whether or not cyclophosphamide and itsmetabolites transplanted within the donor-pretreated kidneyexerted a systemic effect on the host or a local effect in the renalaliograft, the dose of cyclophosphamide and metabolites trans-planted was measured by infusing 10 to 25 p.Ci/kg of carbon 14cyclophosphamide together with 100 mg/kg each of cold CYand P into the donor, Well perfused donor-pretreated kidneyscontained about 0.5% of the infused dose of cyclophosphamide.The concentration of radioactive carbon in the renal cortexexceeded that of the medulla. Mannitol infusions given 60 mmprior to donor nephrectomy decreased the renal cortical con-centration of carbon 14 and reduced the transplanted dose byover 50% (Table 3).

F Additional evidence that mannitol decreased the renal con-tent of cyclophosphamide was obtained by transplanting two

• donor-pretreated kidneys from mongrels with comparable7 amounts of radioactive carbon into two recipients, one of which

received mannitol peroperatively (Table 4). Both donor-pre-treated kidneys enjoyed a good and comparable renal functionas shown by an identical diuresis in the first hour postoperative-ly. At the time of sacrifice at two hours postoperatively, therenal graft residing in the mannitol infused recipient containedless than half of the radioactivity that the control renal allografthad. In the mannitol-infused recipient 63% of the transplantedrenal radioactivity was recovered in liver and bile; on the otherhand, in the control renal allograft 76% of the transplanted dosewas recovered from a urine sample and a renal transplant.

5

Fig. 1. The effect of donor pretreatment on early renal allograftfunction. Unmodified (0) and donor-pretreated mongrel kidneys weretransplanted into Beagle recipients. Donor pretreatment consisted ofcyclophosphamide (•) or cyciophosphamide and methylprednisolone(E) 100 mg/kg each infused 5 hrs prior to donor nephrectomy. Thesymbols designate the following: 0: N = 6 up to the sixth daypostoperatively and N = Son the seventh day; N = 5; ON = 11. Themeansof the serum creatinine concentration and one standard deviationare given.

metabolites is influenced by diuresis [19], attempts were madeto flush putative cyclophosphamide metabolites out of CY- andP-pretreated mongrel kidneys by infusing mannitol, 5 mI/kg of a20% solution i.v., into the donor 1 hour prior to nephrectomy.Mannitol was chosen because of its diuretic properties andbecause it may ameliorate ischemia mediated postoperativeperturbation of renal function [33]. Mannitol infusion abolishedthe effect of donor pretreatment on renal allograft survival(Table 2). Additional support for the hypothesis that the effectof donor pretreatment was mediated by drugs residing in thetransplant was obtained by infusing mannitol, 5 mg/kg of a 20%solution i.v., into Beagle recipients of CY- and P-pretreated

Discussion

Donor pretreatment with cyclophosphamide and methyl-prednisolone caused an almost two-fold prolonged survival ofDLA mismatched MLC incompatible, non-littermate Beaglerenal allografts in a controlled series of experiments in everyinstance. The effect of donor pretreatment on renal allograftsurvival was not restricted to the Beagle strain because donorpretreatment caused enhanced survival of mongrel renal allo-grafts equal to that observed in non-littermate and mismatchedBeagles. It is unlikely that this degree of prolonged survival isbrought about exclusively by an effect on passenger leukocytesbecause destruction of passenger leukocytes by lethal total

1R2R3R

SR

Donor pretreatment 327

0.210.210.210.110.07

0.190.090.110.090.05

ND."4.254.96

ND."2.09

Table 4. Effect of preoperatively administered mannitol on the carbon 14 cyclophosphamide residing in donor-pretreated canine renal allografts"

Diuresis jbmi/hr

Radioactivity in % of transplanted dose

Renal medulla WholeKidney Renal cortex" and papilla renal Liver and

no. Mannitol 1st 2nd / g I g transplant Urine" bile2B — 17.5 14 1.19 0.44 25.59 51.12 13.063B + 17.0 49 0.41 0.21 10.05 17.11 63.73

Two donor pretreated and well perfused kidneys containing 0.425% (2B) and 0.496% (3B)of the dose of carbon 14 cyclophosphamide weretransplanted. Mongrel dogs and recipients weighed between 9 and 10kg. After revascularization mannitol, 20% 5 mI/kg, was infused i.v. into the re-cipient of one donor-pretreated kidney. A urine sample was collected by means of an indwelling catheter. Recipients were sacrificed 2 hrs aftertransplantation when the bile was harvested. All radioactive values given represent the mean of two different pieces of tissue or aliquots of urineand bile.

b Diuresis in the first and second hour postoperatively."Mid cortex.d Radioactivity of the two-hour urine volume.

Radioactivity of the entire liver and bile.

body X-irradiation of donor dogs 96 hours before nephrectomyprolongs the survival of these grafts by three days only [31. Theobservation that passive transfer of iO viable donor lymphnode cells i.v. into recipients of donor pretreated Beagle renalallografts or into recipients of donor-pretreated allografts afterrevascularization failed to influence the effect of donor pretreat-ment may be taken as another argument against the effect ofdonor pretreatment being mediated by mobile passenger leuko-cytes. There are good reasons to assume that the effect of donorpretreatment is mediated by drugs residing in the graft at thetime of surgery, particularly cyclophosphamide and its metabo-lites. First methyiprednisolone pretreatment enhanced the sur-vival of only one out of five mongrel renal allografts which wasnot statistically significant. Donor pretreatment with cyclo-phosphamide on the other hand enhanced the survival ofmongrel renal allografts significantly, but all renal allograftswent through an episode of early and transient postoperativeuremia.

Because chronic uremia in men may be associated with amarkedly prolonged survival of cadaveric renal allografts in oneof the earlier reported series [34J and because chronic uremiainterferes with an acute rejection of cardiac allografts in the rat[35], we cannot rule out that the episode of transient uremia

contributed to the enhanced survival of cyclophosphamidepretreated renal allografts. Also, because of cyclophosphamide-mediated nephrotoxicity, the effect of cyclophosphamide pre-treatment in renal allograft survival can only be accuratelyassessed in conjunction with methylprednisolone pretreatment,for methylprednisolone protects the kidney both against ische-mia [36] and cyclophosphamide-mediated injury.

Second, the fact that mannitol infused 1 hour prior to donornephrectomy or peroperatively into the recipient abolished theeffect of donor pretreatment on renal allograft survival anddecreased the dose of cyclophosphamide and/or its metabolitesin the graft, particularly in the renal cortex, indicates that theeffect of donor pretreatment is mediated by drugs residing in therenal allograft. The failure of others to observe an effect ofpretreatment with cyclophosphamide on canine renal allograftsurvival [221 and of pretreatment with cyclophosphamide andmethylprednisolone on cadaveric renal allograft in men [181may be due to the fact that in the first study mannitol wasinfused into the recipient and in the second study into thepretreated donor prior to nephrectomy. Cyclophosphamide isconverted by the liver and much less so by the kidney [37J intowatersoluble active and inactive metabolites which are excretedinto the bile and by means of glomerular filtration into the urine

Table 3. Effect of mannitol infusion on the carbon 14-cyclophosphamide content of donor pretreated canine renal allografts"

Exp.no. Mannitol

Radioactivity perin '/oo of infused

I g of kidneydose

Medulla and papilla TRR"Cortex"

++

Five dogs were given 100 mg/kg i.v. of both Cy and P in a bolus containing 10 to 20 C C/kg of carbon 14 cyclophosphamide 5 hrs prior tonephrectomy. When desired mannitol, 20% 5 mI/kg, was infused 60 mm prior to sacrifice. Kidneys were perfused with 150 ml of a cold Collin's so-lution prior to processing. From dogs 2 and 3 the left kidney was transplanted; the right kidneys (R) were used for measuring renal radioactivity.

b The mean radioactivity expressed per gram of tissue of two 0.5 g pieces of mid and inner cortex, and medulla and papilla is given. Variationfrom the mean was less than 15% for the renal cortex and less than 20% for the medulla/papilla.

"Total renal radioactivity (TRR): The radioactivity of I g of renal outer and inner cortex, medulla and papilla, renal capsule, perirenal fat, andureter was multiplied with the total weight of each tissue. Next all radioactivity was added.

"N.D. means "not done."

328 Brom Ct al

[381. The urinary excretion of metabolites is influenced bydiuresis [191; the renal tubu]es have been shown to be capableof reabsorbtion [39}. Because the donor pretreated kidneyswere perfused with 150 ml of Coltin's solution, it is reasonableto assume that most of the intratublar load of water solublemetabolites was removed and that mainly cyclophosphamideand metabolites residing in renal parenchymal cells and intersti-tial tissue were transplanted.

How mannitol flushes cyclophosphamide metabolites out ofthe kidney we do not know; this need not be by diuresis becauseof two observations. In preliminary experiments we haveobserved in donor-pretreated mongrel dogs that a furosemide-induced diuresis initiated 60 mm prior to nephrectomy does notabolish the effect of donor pretreatment on renal allograftsurvival. Next, a mannitol infusion administered peroperativelyinto the recipient of a donor pretreated allograft (Table 4)increased diuresis and decreased the renal cortical content ofradioactive carbon, but most of the radioactivity flushed out ofthe graft was removed from the liver and bile and not from theurine. Possibly mannitol mediates this effect by enhancing therenal flow of lymph [401 which might be related to a decrease invascular resistance [41] and thus renal blood flow. Probably thedrugs transplanted within the donor pretreated renal allograftexert their effects locally because only 0.5% of the cyclophos-phamide infused resides in the graft. Our donor and recipientdogs weighed usually between 9 and 10kg which meant that lessthan 5 mg of cyclophosphamide and metabolites was given to a10-kg recipient of which half may be lost into the urine in animmediate postoperative period. A single dose of 0.25 mg/kg ofmicrosomally activated cyclophosphamide is unlikely to alterthe immune response by means of a systemic effect, forexample, on T lymphocytes [421. How a local effect is exertedwe do not know, but since carbon 14 cyclophosphamide and itsmetabolites appear to diffuse readily through and out of therenal cortex in order to be secreted into urine and bile in thefirst few hours postoperatively, a working hypothesis may bepresented.

We conclude that host lymphocytes entering the cortex of therenal allograft in the first few hours postoperatively [43] getsensitized [44] and next relocate into the central lymphoidorgans of the host where they proliferate, as well as hostlymphocytes that enter the renal cortex in the early postopera-tive period in order to proliferate in the interstitial tissue of thegraft [45] may be subjected to a concentration of cyclophospha-mide metabolites sufficient to cause an extent of DNA damagewhich, in the absence of sufficient time for repair, wouldinevitably lead to cell death upon mitosis [46]. The in vitroobservation that a 15-mm period of interaction between lym-phocyte and allogeneic skin target cell suffices for sensitizationand subsequent proliferation [47], together with the finding thata 1-hour ex vivo perfusion of host lymphocytes through anallogeneic kidney suffices for host sensitization [44], supportssuch a hypothesis. Donor pretreatment mediating prolongedcanine renal allograft survival appears to be an example ofinfluencing a biological process by local drug delivery by virtueof unique properties of the drug and because host sensitizationoccurs mainly at the site of the graft.

AcknowledgmentsWe thank the nurses Mr. P. G. J. Wijnen, Miss K. H. van der Mey,

Mrs. A. A. Rasser-van der Mey, and M. J. Stol, veterinarian surgeon,for their enthusiastic and competent assistance. Dr. 0. Driessen

allowed us to use the facilities of the Department of Pharmacology ofthe Leiden University, which enabled us to carry out the radioactivestudies. His advice and cooperation is gratefully acknowledged. MissL. F. M. Viek gave expert technical assistance. Mrs. M. G. P. Peters-Vanderschuren typed the manuscript. Dr. A. Muytjens gave statisticaladvice. The study was also supported by grant Cl37 of the DutchKidney Foundation.

Reprint requests go to Dr. H. L. F. Brom, Department of Surgery,Elisabeth Hospital, Haariem, The Netherlands

References

1. GUTTMANN RD: Renal Transplantation. New Engi J Med 301:975—982, 1979

2. GUTTMANN RD, LINDQUIST RR, OCKNER SA: Renal transplanta-tion in the inbred rat IV Hematopoietic origin of an immunogenicstimulus of rejection. Transplantation 8:472—484, 1969

3. STUART PF, GARRICK T, HOLTON A, LYNCH A. BASTIEN E:Delayed rejection of renal allografts in the rat and the dog byreduction of passenger leukocytes. Surgery 70:128—134, 1971

4. BALCH CM, WILSON CB, LEE S. FELDMAN JD: Thymus dependentlymphocytes in tissue sections of rejecting rat renal allografts. JExp Med 138:1548—1590, 1973

5. VAN SCHILFGAARDE RS, HERMANS P, VAN BREDA VRIESMAN PJC:Influence of recirculating lymphocytes of the rejection of first setrat organ allograft. Transpl Proc 11:1485—1488, 1979

6. VAN BREDA VRIESMAN PJC, FELDMAN JD: 1gM alloantibodyelicited by first set renal allografts: In vivo and in vitro studies. JImmunol 108:1188—1189, 1972

7. PAUL LC, MILFORT EL, PARADYSZ JM, CARPENTER CB: Alloanti-gens on the endothelium of rat kidneys. Transpi Proc. in press

8. VAN SCHILFGAARDE RS, HERMANS P, TERPSTRA JL, VAN BREDAVRIESMAN PJC: Role of mobile passenger lymphocytes in therejection of renal and cardiac allografts in the rat. Transplantation29:209—213, 1980

9. MADRAzO JA, SCHWARZ C, CHURG J: Radiation nephritis a review,J Urol 114:822—827, 1975

10. GUTTMAN RD, BEAUDOIN JD, MOREHOUSE DD: Reduction of graftimmunogenecity of human cadaver renal allografts by donor pre-treatment. Transpl Prac 5:663—665, 1973

11. GUTTMANN RD: Membrane properties and functional activity fromcyclophosphamide pretreated rats. Jlmmunol 112:1594—1601. 1974

12. GUTTMANN RD, LINDQUIST RR: Renal transplantation in theinbred rat XI Reduction of graft immunogenecity by cytotoxic drugpretreatment of donors. Transplantation 8:490—495, 1969

13. SooTs A, HAYRY P: Prolongation of rat cardiac allograft survival bydonor pretreatment. Transplantation 25:259—264, 1978

14. GUTTMANN RD, BEAUDOIN JD, MOREHOUSE DD, KLASSEN J,KNAACK J, JEFFREY J, CHASSONT PG, ABBOU CC: Donor pretreat-ment as an adjunct to cadaver renal transplantation. Transpi Proc7:117—121. 1975

15. GUTTMAN RD, MOREHOUSE DD, MEAKINS JL. KLASSEN J,KNAACK J, BEAUDOIN JG: Donor pretreatment in an unselectedseries of cadaver renal allografts. Kidney mt l3:S99—Sl02, 1978

16. ZtNCKE H, WooDs JE: Donor pretreatment in cadaver renaltransplantation. Surg Gynecol Obstet 145:183—188, 1975

17. BARRY JM, BENNETT WM: Primary cadaver kidney transplantsurvival after donor pretreatment with cyclophosphamide andmethylprednisolone. Transplantation 26:202—203, 1978

18. JEFFERY JR, DOWNS A, GRAHAME JW, LYE C, RAMSEY E,THOMSON AE: A randomized prospective study of cadaver donorpretreatment in renal transplantation. Transplantation 25:287—289,1978

19. BROCK N, GROSS R, HOHORST HJ: Activation of cyclophosphamidein man and animals. Cancer 27:1512—1529, 1971

20. DtENST SG, MACDONALD R, TOLEDO-PEREIRA LH, Fox B: Modi-fication of donor pretreatment prolongs canine renal atlograftsurvival. Transpi Proc 10:1465—1466, 1979

21. CHASSOT P, BEAUDOIN JG, GUTFMANN RD: Prolongation ofkidney allograft survival in dogs with donor pretreatment .SurgForum 25:314—3 16, 1974

22. ZINCKE H, WooDs JE: Attempted immunological alteration ofcanine renal allograft donors. Transplantation 18:480—486. 1974

23. BURLESON RL, JONES DB, JENIKONISHIAN C, DE VOL C, DE RITOJ: Renal donor pretreatment with cytotoxic agents: Immunological-

Donor pretreatment 329

ly ineffective and damaging to the graft. Transpl Proc 9:1591—1596,1977

24. MOURIDSEN HT, FABER 0, SKOVSTED L: The biotransformation ofcyclophosphamide in man: Analysis in the variation in normalsubjects. Acta Pharmacol Toxicol 35:98—106, 1974

25. JEFFREY JR, GUTTMANN RD. CHARPENTIER B: In vitro monitoringof cadaver kidney donor pretreatment by lymphocyte culture. ClinExp Immunol 25:437—441. 1976

26. VRIEZENDORP HM, GROSSE-WILDE H, DORF ME: The majorhistocompatibility system of the dog, in The Major Hiswcompati-bility System in Man and Animals, edited by GOTZE, D. New York,Springer-Verlag, 1977, pp. 129—163

27. BIJNEN AB, VRIEZENDORP HM, GROSSE-WILDE H, WESTBROEKDL: Polygenic control of mixed lymphocyte reactions in dogs.Tissue Antigens 9:187—194, 1977

28. BIJNEN AB, WESTBROEK DL, OBERTOP H, VAN URK H,VRIEZENDORP HM: Genetics of kidney allograft survival in dogs II.Relevance of minor histocompatibility systems in recipients with-out immunosuppressive therapy. Transplantation 28:191—195, 1979

29. BIJNEN AB, WESTBROEK DL, OBERTOP H, VRIEZENDORP HM:Genetics of kidney allograft survival in dogs I. Relevance ofsubregions of the major histocompatibility complex in recipientswithout immunosuppressive therapy. Transplantation 28:186—190,1979

30. HABERMEHL KH: Die Alterbestimmung bei Haus- und Labortieren(2nd ed), Berlin, Paul Parey Verlag, 1974

31. TWEEL JG, VAN DEN VRIEZENDORP HM, TERMIJTELEN A,WESTBROEK DL, BACK ML, ROOD JJ VAN: Genetic aspects ofcanine mixed leukocyte cultures. J Exp Med 140:825—836, 1974

32. LEHMANN EL: Non parametrics in statistical methods based onranks, San Francisco, Holden Day Inc., 1975

33. BARRY KG, COHEN A, KNOCKEL JP: Mannitol infusion 11. Theprevention of acute functional renal failure during resection of ananeurysm of the abdominal aorta. New Engl J Med 264:967—971.1961

34. HUME DM, MERRILL JP, MILLER BF, THORN GW: Experienceswith renal homotranspiantation in the human. J C/in Invest 34:327—

382, 1955

35. Souii..rn RL, SMITH J, BRADFIELD JWB: The effect of uremia onorgan graft survival in the rat. Brit J Exp Pathol 54: 183—191, 1973

36. MILLER HC, ALEXANDER JW: Protective effect of methyl predniso-lone against ischemic injury in the kidney. Transplantation 10:57—69, 1970

37. BROCK N, HOHORST Hi: Ueber die Aktivierung von cyclophospha-mide in vivo und in vitro. ArzneimForsc/i 13:1021—1031, 1963

38. BROCK N, HOHORST HJ: Metabolism of cyclophosphamide. Cancer20:900—904, 1967

39. JOA JY, JusKo WJ, COHEN JL: Phenobarbital effect on cyclophos-phamide pharmacokinetics in man. Cancer Res 32:2761—2764, 1972

40. SZWED ii, MAXWELL DR: Diuretics, hepatic, and thoracic ductlymph flows in the dog. C/in Sci Mo! Med 56:211—214, 1979

41. HARVEY RB: Vascular resistance changes produced by hyperos-motic diuresis. Am J Physiol 199:31—34, 1960

42. SHANDS FL, LIEw FY: Differential sensitivity to cyclophospha-mide of helper T cells for humoral responses and suppressor T cellsfor delayed type hypersensitivity. EurJI,nmunol 10:480—483, 1980

43. PORTER KA: Morphological aspects of renal homograft reaction. BrMed Bull 21:171—175, 1965

44. STROBER S, GOWANS JL: The role of lymphocytes in the sensitiza-tion of rats to renal homografts. J Exp Med 122:347—360, 1965

45. FELDMAN JD, LEE S: Renal homotransplantation in rats, I. Alloge-neic recipients. J Exp Med 126:783—794, 1967

46. PUTTEN LM VAN, LELIEVELD P: Factors determining cell killing bychemotherapeutic agents in vivo. I. Cyclophosphamide. Eur JCancer 6:313—321, 1979

47. TANAKA S, SAKAL A: Minimum length of time required for alloanti-gen recognition. Transplantation 27:260—264, 1979