Specific tolerance induction across a xenogeneic barrier: production of mixed rat/mouse lymphohematopoietic chimeras using a nonlethal preparative regimen.

The development of safe methods for inducing donor-specific tolerance across xenogeneic barriers could potentially relieve the critical shortage of allograft donors that currently limits the applicability of organ transplantation. We report here that such tolerance can be induced in a xenogeneic combination (rat----mouse) using a nonmyeloablative and nonlethal preparative regimen. Successful induction of chimerism and donor-specific transplantation tolerance required pretreatment of recipients with monoclonal antibodies (mAbs) against NK1.1, Thy-1.2, CD4 and CD8, followed by administration of 3 Gy whole body radiation (WBI), 7 Gy thymic irradiation, and infusion of T cell-depleted rat bone marrow cells (BMC). Rat cells appeared among peripheral blood lymphocytes (PBL) of such recipients by 2-3 wk, and rat T cells by 2-5 wk following bone marrow transplantation (BMT). Donor-type rat skin grafts placed 4 mo after BMT were accepted, while simultaneously placed non-donor-type rat skin grafts were promptly rejected. In addition to its clinical potential, the ability to induce donor-specific tolerance across xenogeneic barriers using such a nonlethal preparative regimen provides a valuable model for the study of mechanisms of xenogeneic transplantation tolerance.

Quality of life after aortic valve replacement at the age of >80 years.

BACKGROUND: The optimal management of aortic valve disease in patients >80 years old depends on functional outcome as well as operative risks and late survival. METHODS AND RESULTS: We retrospectively identified 133 patients (62 men, 71 women) aged 80 to 91 years (mean 84+/-3 years) who underwent aortic valve replacement alone or in combination with another procedure between January 1, 1993, and April 31, 1998. Demographics included hypertension 68%, diabetes mellitus 17%, and history of stroke 11%. Operative (30 day) mortality rate was 11%. Urgent or emergent surgery, aortic insufficiency, and perioperative stroke or renal dysfunction were risk factors for operative death by multivariable analysis. Intensive care unit and total hospital length of stay were prolonged at 6.2 and 14.7 days, respectively. Late follow-up between July 1, 1998, and November 1, 1999, was 98% complete. Actuarial survival at 1 and 5 years was 80% and 55%, respectively. Predictors of late mortality were preoperative or perioperative stroke, chronic obstructive pulmonary disease, aortic stenosis, and postoperative renal dysfunction. The mean New York Heart Association functional class for 65 long-term survivors improved from 3.1 to 1.7. Quality of life assessed with the Medical Outcomes Study Short Form-36 was comparable to that predicted for the general population >75 years old. CONCLUSIONS: Functional outcome after aortic valve replacement in patients >80 years old is excellent, the operative risk is acceptable, and the late survival rate is good. Surgery should not be withheld from the elderly on the basis of age alone.

Influence of coronary heart disease on morbidity and mortality after carotid endarterectomy: a population-based study in Olmsted County, Minnesota (1970-1988)

To evaluate the prognostic importance of coronary artery disease among patients undergoing carotid endarterectomy, 177 residents of Olmsted County, Minnesota who underwent carotid endarterectomy during the period 1970 through 1988 were followed up to July 1, 1989. Patients were stratified as to the presence (n = 64) or absence (n = 93) of overt coronary artery disease or prior myocardial revascularization (n = 20) at the time of endarterectomy. At 30 days after carotid endarterectomy, there were no significant differences between patients with or without coronary artery disease in the occurrence of death, myocardial infarction or stroke. Kaplan-Meier estimate of 8-year relative survival after carotid endarterectomy (assessed as a percent of survival in age- carotid endarterectomy (assessed as a percent of survival in age- and gender-matched control subjects) was 89% in those without and 75% in those with overt coronary artery disease. Of the 59 total deaths, 29 (49%) had a cardiac cause and 4 (7%) were due to stroke (p less than 0.0001). The cumulative incidence of a cardiac event at 8 years after carotid endarterectomy was greater in those with than in those without overt coronary artery disease (61% vs. 25%, p less than 0.0001). In multivariable analysis, uncorrected coronary artery disease and diabetes were the only independent predictors of subsequent cardiac events, whereas age was the only independent predictor of death. These population-based data suggest that carotid endarterectomy can be safely undertaken in patients with stable coronary artery disease. In long-term follow-up of these patients, coronary rather than cerebral vascular disease is the most frequent cause of morbidity and mortality. Thus, these data lend strong support to the concept of early identification and management of coronary artery disease in patients undergoing carotid endarterectomy.

Is early anticoagulation with warfarin necessary after bioprosthetic aortic valve replacement?

OBJECTIVES: Freedom from anticoagulation is the principal advantage of bioprosthesis; however, the American Heart Association/American College of Cardiology and the American College of Chest Physicians guidelines recommend early anticoagulation with heparin, followed by warfarin for 3 months after bioprosthetic aortic valve replacement. We examined neurologic events within 90 days of bioprosthetic aortic valve replacement at our institution. METHODS: Between 1993 and 2000, 1151 patients underwent bioprosthetic aortic valve replacement with (641) or without (510) associated coronary artery bypass. By surgeon preference, 624 had early postoperative anticoagulation (AC+) and 527 did not (AC-). In the AC- group, 410 patients (78%) received antiplatelet therapy. Groups were similar with respect to gender (female, 36% AC+ vs 40% AC-, P = .21), hypertension (64% AC+ vs 61%, P = .27), and prior stroke (7.6% AC+ vs 8.5% AC-, P = .54). The AC+ group was slightly younger than the AC- group (median, 76 years vs 78 years, P = .006). RESULTS: Operative mortality was 4.1% with 43 (3.7%) cerebrovascular events within 90 days. Excluding 18 deficits apparent upon emergence from anesthesia, we found that postoperative cerebrovascular accident occurred in 2.4% of AC+ and 1.9% AC- patients. By multivariable analysis, the only predictor of operative mortality was hypertension ( P < .0001). Postoperative cerebrovascular accident was unrelated to warfarin use ( P = .32). The incidence of mediastinal bleeding requiring reexploration was similar (5.0% vs 7.4%), as were other bleeding complications in the first 90 days (1.1% vs 0.8%). No variables were predictive of bleeding by multivariate analysis. CONCLUSIONS: Although these data do not address the role of antiplatelet agents, early anticoagulation with warfarin after bioprosthetic aortic valve replacement did not appear to protect against neurologic events.

Brain pH in focal cerebral ischemia and the protective effects of barbiturate anesthesia.

Intracellular brain pH was measured with a lipid-soluble, pH-sensitive fluorescent indicator in 16 squirrel monkeys. Eight of these were under halothane anesthesia and eight under barbiturate anesthesia. Measurements were taken before, during, and after focal incomplete ischemia. Brain pH following 3 h of cerebral focal ischemia changed from a normal value of 7.0 to 6.5 and 6.2 in animals studied under barbiturate and halothane anesthesia, respectively. Brain pH returned toward normal after flow was restored in animals under barbiturate anesthesia but continued to deteriorate in animals under halothane anesthesia.

Regional cerebral blood flow and focal cortical perfusion: a comparative study of 133Xe, 85Kr, and umbelliferone as diffusible indicators.

We report that regional CBF determined by the initial slope technique using 133Xe and 85Kr in cats and rabbits can be significantly influenced by the size of the field of measurement. The clearance curves of umbelliferone, a lipid-soluble intracellular pH fluorescent indicator, from a visually avascular 80-micron field were used to cross-correlate rCBF with focal cortical perfusion. Our findings indicate that in the cat, as the gamma or beta detector's field of volume was reduced, regional CBF (rCBF) measured by intraarterially injected 133Xe or 85Kr decreased in value by 33% and 28%, respectively, and the slope of the rCBF-PaCO2 response curve became less steep by 56% and 45%, respectively. Umbelliferone, measuring a much smaller volume of tissue, showed a lower normocapnic flow and a more oblique PaCO2 response curve. In the rabbits studied, the normocarbic rCBFs and the rCBF-PaCO2 response curves measured with the three techniques corresponded to those measured in the cat. These results suggest that large field/volume measurements assess a measure of flow that is a weighted average of several distinct flow compartments and that these compartments differ in their response to changes in PaCO2.

The novel dihydronaphthyridine Ca2+ channel blocker CI-951 improves CBF, brain pHi, and EEG recovery in focal cerebral ischemia.

The effects of the novel dihydronaphthyridine Ca2+ antagonist CI-951 on focal cerebral ischemia were assessed during MCA occlusion in 30 white New Zealand rabbits under 1.0% halothane anesthesia. In vivo brain pHi and focal CBF were measured with umbelliferone fluorescence. Baseline normocapnic brain pHi and CBF were 7.02 +/- 0.02 and 48.4 +/- 2.9 ml/100 g/min, respectively. In the severe ischemic regions, 15 min postocclusion brain pHi and CBF were 6.62 +/- 0.04 and 14.4 +/- 0.7 ml/100 g/min in controls vs. 6.60 +/- 0.02 and 12.9 +/- 2.3 ml/100 g/min, respectively, in animals destined to receive CI-951. Twenty minutes after MCA occlusion, CI-951 was administered at 0.5 microgram/kg/min and brain pHi and CBF were determined in both regions of severe and moderate ischemia for 4 h postocclusion. Control severe ischemic sites demonstrated no significant improvement in brain pHi and only mild increases in CBF over the next 4 h. CI-951 caused significant improvement in both of these parameters. Postocclusion 4 h brain pHi and CBF measured 6.69 +/- 0.04 and 18.5 +/- 3.2 ml/100 g/min in controls vs. 7.01 +/- 0.04 and 41.7 +/- 5.3 ml/100 g/min, respectively, in CI-951 animals (p less than 0.001). Similar improvements were observed in moderate ischemic sites. In animals that demonstrated postocclusion EEG attenuation, 75% of CI-951 animals had EEG recovery as compared to 18% in controls. CI-951 may be a useful therapeutic agent for focal cerebral ischemia if histological and outcome studies verify these data.

Cerebral blood flow, brain pH, and oxidative metabolism in the cat during severe insulin-induced hypoglycemia.

The effects of severe hypoglycemia on brain pH, cerebral blood flow (CBF), and other physiologic and metabolic parameters were studied in 26 cats subjected to insulin hypoglycemia. Two groups were utilized to compare the effects of anesthesia. The halothane group was composed of 14 animals and the barbiturate group contained 12 animals. Insulin was administered by both the intravenous and intramuscular routes until there was a severe electroencephalographic (EEG) change or until 6 h had elapsed. The cerebral responses to hypoglycemia demonstrated the following: CBF was unaffected by severe hypoglycemia in normotensive animals with or without EEG changes; brain pH was essentially constant in all groups regardless of glucose levels, CBF, or EEG; and the EEG abnormalities corresponded closely to brain glucose levels. Cerebral adenosine triphosphate and phosphocreatine levels were lowest in the animals with isoelectric EEGs. We conclude that CBF and brain pH in the normotensive cat under general anesthesia are relatively unaffected by insulin hypoglycemia despite the presence of severe EEG changes and diminished cerebral energy reserves. The study suggests tha the PaCO2-CBF response curve is ot dependent upon the metabolic integrity of cerebral tissue and is mediated by pathways separate from those of autoregulation.

Intracellular brain pH, indicator tissue perfusion, electroencephalography, and histology in severe and moderate focal cortical ischemia in the rabbit.

Intracellular brain pH and indicator tissue perfusion were measured with a lipid-soluble, pH-sensitive fluorescent indicator in 10 rabbits who had either severe or moderate focal ischemia depending on whether the middle cerebral artery was occluded at its main trunk or bifurcation. Preocclusion tissue indicator perfusion was 50.1 ml/100 g/min and intracellular brain pH was 7.03. In severe focal ischemia, immediate postocclusion tissue perfusion was 12.7 ml/100 g/min and intracellular brain pH was 6.64. Four hours after occlusion, the perfusion was 5.2 ml/100 g/min and intracellular brain pH was 6.08. There was EEG and histological confirmation of infarction. In the moderate focal ischemia group, immediate postocclusion flow was 20.0 ml/100 g/min and intracellular brain pH was 6.92. At 3 h, postocclusion tissue perfusion was 22.6 ml/100 g/min and intracellular brain pH was 6.86. Therefore, for the first 3 h, this ischemic penumbra was stable. At the fourth hour, both cerebral tissue perfusion and intracellular brain pH worsened. This suggests that the ischemic penumbra is a dynamic state. The rabbit is a good experimental model for the production of both severe and moderate focal ischemia.

Spontaneous internal carotid dissection, hemicrania, and Horner's syndrome.

We describe five patients with angiographic features of dissecting aneurysm of the cervical portion of the internal carotid artery who were seen initially with unilateral headaches and ipsilateral oculosympathetic paresis. This combination of symptoms is a rather common mode of manifestation of this entity.

Weakness of the lower extremity in carotid occlusive disease.

Nineteen patients experienced progressive or episodic weakness of one lower extremity caused by severe stenosis or occlusion of the internal carotid artery. The majority of patients (84.2%) had occlusion or severe stenosis at the origin. Based on clinical profiles, angiographic findings, and cerebral blood flow patterns, we concluded that the pathophysiologic mechanism was hypoperfusion in the border zone between the anterior cerebral artery and the middle cerebral artery and that patients with progressive weakness had more extensive compromise in cerebral circulation. Following surgical treatment in 17 patients, progressive and episodic weakness disappeared and the majority of them (76.4%) became asymptomatic. However, the patients with stenosis at the siphon and those with progressive weakness from occlusion at the origin appeared to be at increased risk for cardiac death.

Minimal oculomotor nerve paresis secondary to unruptured intracranial aneurysm.

In 12 patients with minimal oculomotor nerve deficits due to unruptured intracranial aneurysm, the nerve-related findings were incomplete and at least one element (ptosis, mydriasis, or extraocular muscle weakness) was spared in every patient. Although symptomatic, the affected cranial nerve III functions were only partially lost. Six patients had ptosis and mydriasis, three had ptosis and diplopia, two had mydriasis and diplopia, and one had mydriasis alone. Eleven patients had accompanying headaches that were remarkably variable and difficult to categorize. Cerebral angiography showed the aneurysms, but computed tomography missed one third of them. The symptomatic aneurysm arose from the internal carotid artery in six patients, and from the distal basilar artery in six. Fragments of a cranial nerve III palsy associated with recent onset and ipsilateral headache suggest an enlarging internal carotid or distal basilar artery aneurysm.

Cerebral protection by barbiturate anesthesia. Use after middle cerebral artery occlusion in Java monkeys.

Regional cerebral ischemia was produced in 18 Java monkeys by permanent middle cerebral artery (MCA) occlusion. All monkeys were thereafter paralyzed (pancuronium bromide, 0.05 mg/kg/hr) and sedated (diazepam, 0.1 mg/kg/hr) for a 48-hour period. Thirty minutes after MCA occlusion, pentobarbital sodium anesthesia was induced in nine of the monkeys (14 mg/kg) and maintained for 48 hours (7 mg/kg every two hours), with continuous supportive care. After 48 hours, all drugs were discontinued; the monkeys were observed for five days, and then killed. Seven of the control monkeys developed a cerebral infarction, three did not survive past the 48 hours of intensive care, and the other four had a notable neurologic deficit. All pentobarbital monkeys survived the seven days, but four had a cerebral infarction and two of these had a notable neurologic deficit. These differences were statistically significant.

The unchanging pattern of subarachnoid hemorrhage in a community.

The average annual incidence of subarachnoid hemorrhage (SAH) from aneurysm rupture in Rochester, Minnesota, has remained remarkably constant at about 11 per 100,000 population. Age-specific incidence increased with age. Survival after SAH depended on: (1) clinical grade, (2) time after onset of SAH, and (3) presence of intracerebral hematoma. Among those who survived to receive medical attention, 48% were clinical grade 1 or 2, 20% were grade 3, and 32% were grade 4 or 5. Proved rebleeding occurred within 10 days of the first SAH in 20% of patients who survived until hospital admission.

Patterns of T cell-accessory cell interaction in the generation of primary alloresponses in the pig.

Partially inbred, MHC-homozygous miniature swine provide a unique model for the study of organ transplantation and the induction of tolerance in large animals. Models of both vascularized solid organ transplantation and bone marrow transplantation have previously been established. The availability of monoclonal antibodies reactive with porcine leukocyte subset antigens now makes possible studies of the cellular immunology in this species, affording the opportunity to examine mechanisms of transplant tolerance and graft rejection in increasing detail. Using such antibodies and peripheral blood leukocytes from pigs of recombinant MHC haplotypes, we have examined porcine T cell-accessory cell interactions in vitro with attention to T cell subsets and the class of MHC alloantigen stimulation. Primary allospecific MLR and CML cultures were studied after depletion of accessory cells from responder and/or stimulator populations. Although class II MHC antigens were expressed on the majority of porcine T cells before and after depletion, these cells were insufficient for antigen presentation, since there was an absolute requirement for ACs in the generation of primary alloresponses. Proliferative and CTL alloresponses could be generated provided that ACs of either stimulator or responder type were present. Selective depletion of CD4+ T cells from the responder population demonstrated: (a) that the interaction mediated by self ACs was CD4-dependent; (b) that two pathways exist for interaction involving allogeneic ACs; and (c) that the interaction involving allogeneic class II is CD4-dependent, while that with allogeneic class I is not.

In vivo treatment with antiporcine T cell antibodies.

The effects of in vivo administration of monoclonal antibodies directed against the two major mature T cell subsets of miniature swine have been examined. Antibody 76-2-11, specific for the porcine CD8 cell-surface antigen, caused elimination of CD8+ T cells from the peripheral lymphocyte pool by 4-6 days following a single intravenous infusion. This depletion persisted while excess antibody was detectable in the serum and CD8+ cells returned rapidly to normal levels following the disappearance of serum antibody. In vitro functional assays performed on PBL during the period of depletion showed no effect on proliferative responses but a profound inhibition of ability to generate CTL. In contrast, administration of monoclonal antibody 74-12-4, specific for CD4+ porcine T lymphocytes, caused no consistent change in the level of CD4+ cells in the peripheral lymphocyte pool. By flow microfluorometry, CD4+ peripheral lymphocytes were found to be coated with the antibody, and there also was excess antibody in the serum for at least 4 days in two of the three animals treated. No effects on in vitro proliferative responses or generation of CTL in vitro were observed, even during the period of excess circulating antibody. Treatment with both antibodies was tolerated well by all animals. These findings are similar to those that have been found in clinical studies of monoclonal antibody treatment and suggest that the pig will provide a large animal preclinical model in which the effects of in vivo treatment with monoclonal antibodies on parameters of transplantation immunity can be tested.

Induction of kidney transplantation tolerance across major histocompatibility complex barriers by bone marrow transplantation in miniature swine.

Previous studies from our laboratory have shown that permanent lymphohematopoietic chimerism can be induced in MHC-disparate miniature swine by bone marrow transplantation after lethal total-body irradiation. The purpose of the present study was to determine in this large animal model whether such chimerism would lead to permanent tolerance to a vascularized allograft without a requirement for exogenous immunosuppression. Eight miniature swine that had received MHC-mismatched BMT more than five months earlier underwent kidney transplantation (KTx) from a donor MHC matched (n = 5) or MHC mismatched (n = 3) with the BMT donor. All animals had regained in vitro responsiveness to third-party MHC antigens, as measured by mixed lymphocyte reaction (MLR), before KTx but remained nonresponsive to MHC antigens of the BMT donor and self. All three animals that received KTx mismatched for BMT donor MHC rejected promptly (mean survival time 7.0 days). Of the five animals that received KTx matched for BMT donor MHC, four showed no evidence of rejection and have functioning KTx greater than 200 days after KTx. The fifth animal had excellent renal function for 60 days but then developed a slowly rising BUN and serum creatinine, and died 75 days after KTx. The course of this animal's rejection is consistent with that previously described for rejection due to minor antigen disparities. The difference in survival of KTx matched or mismatched for the MHC of the BMT donor was statistically significant (P = 0.0062). The survival of KTx matched for the MHC of the BMT donor was significantly different from that of control animals without BMT receiving KTx mismatched for MHC (P = 0.0018). We therefore conclude that BMT is an effective means for induction of tolerance to an MHC mismatched KTx in this large animal model.

Bone marrow transplantation in miniature swine: IV. Development of myeloablative regimens that allow engraftment across major histocompatibility barriers.

Studies of the myeloablative regimens capable of permitting successful BMT across MHC barriers in miniature swine have been performed. To minimize graft-versus-host disease (GVHD), engraftment was studied in the F1-->P combination (i.e., MHC homozygous ["parental"] swine receiving bone marrow from one-haplotype matched MHC heterozygous ["F1"] donors). Animals given total body irradiation (TBI) up to 1100 cGy, 10 cGy/min, in a single dose failed to engraft. Increasing the dose rate led to unacceptable extramedullary toxicity without improving engraftment. Eleven different fractionated TBI regimens were tested in this F1-->parent model. At all of the dose rates tested, a total dose of less than 1000 cGy was insufficient for engraftment, and a total dose of 1400 cGy led to unacceptable toxicity. Between these extremes, a window was defined in which engraftment could be obtained without unacceptable extramedullary toxicity utilizing 2 equally divided fractions of TBI delivered 24 hr apart. The addition of 50 mg/kg cyclophosphamide i.v. to fractionated TBI (1150 cGy total dose [500 + 650]) also permitted engraftment, with decreased incidence of interstitial pneumonitis as compared to fractionated TBI (1300 cGy total dose [650 x 2]). Both of these regimens were also confirmed to permit engraftment between heterozygous donors and recipients sharing a single common haplotype ("F1-->F1"). The regimen of 1300 cGy (650 x 2) also permitted engraftment in completely MHC mismatched BMT, but with subsequent death from GVHD. These studies of the myeloablative regimens permitting engraftment across defined MHC barriers in miniature swine provide a basis for further studies of allogenic BMT and GVHD in this large animal preclinical model.

Retransplantation in miniature swine. Lack of a requirement for graft adaptation for maintenance of specific renal allograft tolerance.

In miniature swine, one-haplotype class I disparate renal allografts are accepted without exogenous immunosuppression by approximately 35% of recipients. Alternatively, transplants bearing a two-haplotype class I mismatch are always rejected acutely. However, long-term acceptance in the latter animals can be achieved uniformly with a 12-day course of cyclosporine. In vitro studies of recipient cell-mediated lymphocytotoxicity responses have shown donor-specific cytotoxic T lymphocyte clones in tolerant animals, suggesting that tolerance may be a local phenomenon or a central phenomenon activated in the milieu of the graft. Six animals were retransplanted with kidneys MHC-matched to their original allograft to determine whether (1) tolerance is a central phenomenon; (2) host tolerance can be broken with a fresh challenge of donor antigen and antigen-presenting cells; and (3) graft adaptation is required for maintenance of tolerance. Four of the retransplanted animals had been spontaneous acceptors of one-haplotype class I-disparate grafts and two had been rendered tolerant to two-haplotype class I-mismatched kidneys with CsA induction. All six explanted allografts showed no histological evidence of rejection and all six retransplants were accepted without exogenous immunosuppression. These findings suggest that in miniature swine tolerance of class I-disparate kidneys is a stable, centrally mediated phenomenon that cannot be broken with a challenge of fresh donor antigen and donor-type APCs. Furthermore, successful retransplantation without immunosuppression in animals receiving CsA induction therapy for their first transplant suggests that graft adaptation is not necessary for the maintenance of tolerance.

The failure of skin grafting to break tolerance to class I-disparate renal allografts in miniature swine despite inducing marked antidonor cellular immunity.

Long-term specific tolerance to one haplotype class I plus minor antigen disparate renal allografts develops without exogenous immunosuppression in approximately 35% of miniature swine (n = 128). Previous studies have suggested that this phenomenon is related to limited class I-specific helper T cell activity as evidenced by the failure of antibody class switching in vivo and the ability of exogenous interleukin 2 to elicit antidonor responses in vitro. To determine whether tolerance could be broken by inducing antidonor reactivity with donor antigen and a source of T cell help, multiple skin grafts bearing donor class I plus third-party class II antigens were placed on tolerant animals. Skin grafts were placed at least 3 months after the kidney transplant, at which time all recipients had normal renal function as measured by blood urea nitrogen and serum creatinine. First-set rejection of skin grafts by SLAad and SLAdd hosts occurred in 11.8 +/- 1.1 days (mean +/- SEM, n = 6) and in 9.3 +/- 0.9 days (n = 4), respectively. Coincident with skin rejection, most animals developed a transient rise in BUN to 62 +/- 11 mg/dl (n = 10) and a similar rise in Cr to 4.9 +/- 1.2 mg/dl (n = 10), with normal levels returning in all animals within two weeks. Subsequent skin grafts with the same disparity did not undergo second-set rejection and did not induce BUN or Cr elevations. Prior to skin grafting, animals showed no antidonor activity in mixed lymphocyte reaction or cell-mediated lymphocytotoxicity assays. After two skin grafts, all animals developed donor-specific CML and secondary MLR responses, and additional skin grafts amplified this cellular immunity. Development of marked antidonor immunity without a break in tolerance suggested that either graft adaptation or local suppression might be involved in maintaining tolerance to class I MHC antigens. In preliminary studies, an immunized SLAad animal and an immunized SLAdd animal were retransplanted with kidneys MHC matched to their first allografts. In both cases, the second graft was accepted permanently without immunosuppression, suggesting that graft adaptation is not necessary for the maintenance of tolerance to renal allografts in miniature swine.