Campath-1H, an anti-CD52 monoclonal antibody, was used as induction therapy (40 mg i.v. total dose) in 29 primary human renal transplants, and the patients were maintained on rapamycin monotherapy (levels 8–15 ng/mL) post-transplant. Campath-1H profoundly depletes lymphocytes long-term and more transiently depletes B cells and monocytes. All patients are alive and well at 3–29 months of follow up. One graft was lost because of rejection. There have been no systemic infections and no malignancies. Eight of 29 patients have experienced rejection, which was successfully treated in seven of eight patients. Five of these patients had pathological evidence of a humoral component of their rejection. Seven of the 29 patients were converted to standard triple therapy on account of rejection. Rapamycin was generally well tolerated in that there were no significant wound-healing problems; two lymphoceles required surgical drainage; and most patients were treated with a lipid-lowering agent. Flow crossmatch testing post-transplant revealed evidence of alloantibody in two patients tested with previous combined cellular and humoral rejection. Biopsies have shown no chronic allograft nephropathy to date. In view of the relatively high incidence of early humoral rejection, we plan to modify the immunosuppressive regimen in subsequent pilot studies. This clinical trial provides insight into the use of Campath-1H induction in combination with rapamycin maintenance monotherapy.
Experimental as well as clinical efforts to induce transplantation tolerance have relied substantially on lymphocyte depletion at the time of organ transplantation to establish an immunologic milieu free of rejection (1–3). If the number of alloreactive cells is reduced, the potential to avoid graft destruction and to regulate a nondestructive immune response to the donor may be enhanced. In addition to lymphocyte depletion, maintenance pharmacologic immunosuppression has been necessary, at least initially in humans to prevent rejection. Current conventional immunosuppressive therapy in organ transplantation relies largely on a calcineurin inhibitor and steroids used indefinitely, although steroid-sparing regimens are often successful in liver transplantation. Evidence derived from a nonhuman primate model suggested to us that more profound and durable T-cell depletion achieved using an anti-CD3 immunotoxin in nonhuman primate renal allotransplantation would substantially reduce the risk of rejection and the need for maintenance immunosuppression, and in many cases lead to donor-specific tolerance (4,5). Subsequently, Campath-1H, a humanized anti-CD52 monoclonal antibody that depletes T and B lymphocytes, natural killer cells and some monocytes and macrophages, was tested in renal allotransplantation in humans in combination with cyclosporine monotherapy with good results (6,7). In an effort to avoid calcineurin inhibitors and maintenance steroid use, we performed a pilot trial of Campath-1H induction in combination with rapamycin maintenance monotherapy. Rapamycin has the potential advantages of avoiding nephrotoxicity, preventing graft fibrosis (8), and perhaps enhancing the potential for tolerance induction (9). The results of the pilot study of 29 patients undergoing primary renal allotransplantation are summarized. The first 24 patients received only Campath-1H and rapamycin; subsequently, an additional five patients received a supplemental dose of Thymoglobulin and 14 days of steroids. The outcomes of these 29 patients are presented including their renal transplant function, renal biopsy results, measurement of leukocyte depletion, assessment of alloantibody, and cardiovascular profile.
Campath-1H, a humanized anti-CD52 monoclonal antibody, was provided by Millennium Pharmaceuticals (Cambridge, MA), and later Ilex, Inc. (San Antonio, TX). For the first 24 patients, 20 mg was administered intraoperatively on the day of transplant (day 0) and a second dose of 20 mg was given the day following transplantation (day 1). Thirty minutes prior to the Campath-1H infusion, the patients were administered 500 mg of methylprednisolone. Rapamycin (sirolimus, Wyeth, Philadelphia, PA) was administered at a dose of 2 mg orally starting on the day after the transplant. Doses were adjusted to achieve blood levels in the 8–12 ng/mL range. No other immunosuppressive medications were given. Patients #25–29 received a modified protocol consisting of Campath-1H 20 mg i.v. on day −1 and day 0 and one dose of Thymoglobulin (SangStat, Menlo Park, CA) 1.5 mg/kg i.v. on day 1, and a 14-day course of tapering steroids. Steroids were dosed prior to antibody on days −1, 0, and 1 (500 mg × 2 days, 250 × 1, 120 × 3, 90 × 3, 60 × 2, 45 × 2, 30 × 2, discontinued day 15). Rapamycin dosing was the same as for the first 24 patients.
Recipient and donor selection
Patients were enrolled under an IRB-approved protocol at the University of Wisconsin, Madison, following informed consent. The protocol was also conducted under FDA surveillance via an IND to S.J.K. for off-label use of Campath-1H. Patients consented to biopsies of their kidney transplant at 6 months and 12 months, and to monitoring of their blood leukocyte population by flow cytometry. All patients were entered after informed consent regarding the nature of the study. Primary cadaveric and living donor adult renal transplant recipients (ages 18–60 years) were selected based on the following criteria: current PRA < 10%, historical peak PRA < 25%, and body mass index < 32. Patients with a 0 antigen mismatch were excluded, as were patients who were cytomegalovirus (CMV) seronegative but had a CMV seropositive donor. Donor kidneys from nonheart-beating donors were excluded, as were kidneys from donors older than 55 years or kidneys preserved for > 36 h. A negative NIH crossmatch test was required prior to transplantation.
Postoperative monitoring/infection prophylaxis
The patients were monitored for recovery of leukocyte subsets by flow cytometry on days 0, 7, 14, 30, 60, 90 and 180, and graft function was monitored by serum creatinine levels. Renal transplant biopsies were performed at 6 and 12 months and if clinically indicated by graft dysfunction. Infection prophylaxis consisted of one TMP/sulfa double-strength tablet daily for 1 year post-transplant to prevent pneumocystis carinii. CMV prophylaxis consisted of daily intravenous ganciclovir during hospitalization for renal transplantation followed by oral acyclovir for 3 months, which was the routine prophylaxis at this institution at the time of the study. Creatinine values were reported as mean ± standard deviation and compared between rejection and nonrejection using Wilcoxon's rank-sum test.
Renal allograft biopsies were performed in patients with graft dysfunction (elevation of baseline serum creatinine > 20%), and per protocol at 6 and 12 months. Hematoxylin and eosin-stained frozen sections were used for routine staining of all biopsies. Immunostaining for the C4d complement component was performed on six biopsies performed because of graft dysfunction. A 3-step immunofluorescence technique with monoclonal antibody 10–11 (Biogenesis, Brentwood, NH) was used following the protocol by Collins et al. (10). Frozen section slides from two cases were stained using a Benchmark automated stainer (Ventana Medical Systems, Tucson, AZ) applying the same primary antibody at a 1 : 600 dilution and using a horseradish peroxidase detection system. All biopsies were interpreted at the University of Wisconsin (A.F.), and biopsies with evidence of rejection were also reviewed independently by Dr Robert Colvin (Massachusetts General Hospital) as an external study monitor.
T-cell flow cytometric crossmatch
Dual color flow cytometric crossmatch was performed in 19 patients for whom adequate donor lymphocytes were available. This was performed retrospectively using serum collected pretransplant and at 3-month intervals post-transplant. Testing was performed by Dr L. K. Lebeck using a previously published method (11). Briefly, a Becton-Dickinson FACSCalibur (San Jose, CA) flow cytometer was used with a fluorescein-conjugated goat F (ab′)2 fragment to human IgG Fc (Cappel Research Products, Organon Teknika, Durham, NC). The cutoff for interpretation of a positive T-cell crossmatch was a mean channel shift of greater than 15 on a 256 scale. CellQuest software was used for acquisition and analysis.
Twenty-nine renal transplants were performed from six cadaveric donors, 16 living-related donors, and seven living-unrelated donors. The mean age of recipients was 41 years with a range of 18–60 years. There were 19 males and 10 females, with 16 patients on dialysis pretransplant and 13 not yet on dialysis. The causes of renal failure are listed in Table 1. The distribution of HLA matching among the donors and recipients is listed in Table 2.
Table 1. Recipient demographics
Mean 41, range 19–60
Sex (M : F)
19 : 10
Type of donor
Cause of renal failure
Polycystic kidney disease
Table 2. Degree of HLA mismatch
Degree of mismatch
Patients with rejection
All 29 patients undergoing transplantation are alive and well and 28 of the 29 grafts are functioning. One patient underwent transplant nephrectomy at 2 months because of rejection. No malignancies have been encountered and there have been no systemic infections. Ten patients have been treated for infection as listed in Table 3. Adverse events requiring hospitalization are listed in Table 3. Two patients required surgical drainage of lymphoceles. One patient discontinued sirolimus because of hepatotoxicity (AST elevated to 172 U/L and ALT to 205 U/L), and sirolimus was replaced with mycophenolate mofetil (CellCept®, Roche, Nutley, NJ) plus prednisone for 3 months. One month after discontinuing the prednisone while remaining on MMF monotherapy, this patient experienced acute rejection (Banff 1B), which was successfully treated. The patient has subsequently been converted to sirolimus monotherapy without recurrence of hepatotoxicity or rejection. Five of the 29 patients were on a lipid-lowering agent prior to transplantation, and of the 20 patients currently on rapamycin monotherapy, 17 are on lipid-lowering agents. The only side-effects of Campath-1H administration was a mild rash or itching, which was seen with a second dose of Campath-1H in 12 patients. All responded to a single dose of oral Benadryl. The mean serum creatinine of all 29 patients at 3 and 12 months' follow up was 1.5 ± 0.4 and 1.8 ± 0.8, respectively. Patients without rejection had a mean 3-month and 12-month creatinine of 1.5 ± 0.4 and 1.6 ± 0.8, whereas patients with rejection had 1.7 ± 0.5 and 2.0 ± 0.5, respectively. The differences in creatinine at 3 months were not as significant (p = 0.3) as the differences at 12 months (p = 0.06) between rejection and nonrejection. Mean sirolimus levels (and dose) at 1 month and 12 months were 8.8 ± 3.7 ng/mL (4.6 ± 1.7 mg/day) and 7.4 ± 3.2 ng/mL (4.2 ± 1.4 mg/day).
Table 3. Adverse events requiring hospitalization and number of patients affected
Urinary tract infection (3)
Fever of unknown origin (2)
Infected seroma (1)
Incisional hernia (1)
Hematoma following kidney biopsy (1)
CABG + MVR, pleural effusion (1)
Ureteroureterostomy for reflux (1)
Perinephric hematoma + lymphatic leak (1)
Diabetic gastroparesis + atrial fibrillation (1)
Thrombophlebitis + pulmonary embolus (1)
Right common iliac artery stenosis, stented (1)
Hepatotoxicity + anemia (1)
Immune cell reconstitution
T cells, B cells, and monocytes were substantially depleted immediately following Campath-1H induction as shown in Figure 1 (A–G). Although total white blood cell counts did not substantially change compared with baseline (Figure 1A), lymphocyte counts dropped profoundly (Figure 1B) and remain one-third of baseline at 1 year. Monocytes dropped transiently (Figure 1C) with recovery to baseline by 1–2 months. Total B- and T-lymphocyte counts remained suppressed at 12 months (Figure 1D,E). The most profound long-term depletion occurred in the CD4 T-cell population, which at 12 months remained at a median of 135 cells/µL compared with 840 cells/µL baseline (Figure 1G). CD8 cells also remained substantially suppressed (130 cells/µL vs. 430 cells/µL baseline) (Figure 1F). Interestingly, this level of T-cell depletion was well tolerated clinically in that no patients experienced systemic infection or malignancy.
Of the first 24 patients, six experienced rejection episodes. Details of the rejection episodes are summarized in Table 4. Three of the patients had early acute humoral rejection at days 6, 11, and 17, and one patient had late (day 147) rejection, predominantly humoral. One of these four rejections was severe, leading to graft loss. The other three were reversed with a combination of plasmapheresis, polyclonal ATG (Thymoglobulin®, Sangstat, Freemont, CA), prednisone, and rituximab. Because of suspected humoral rejection, almost all (7/8) patients empirically received Thymoglobulin and plasmapheresis to treat rejection, at the discretion of the patient's surgeon and physician. All four humoral rejection episodes were associated with C4d-positive staining by immunocytochemistry (Figure 2C). There were two episodes of acute cellular rejection on days 8 and 120, which reversed with steroid bolus therapy. One of these episodes (day 120) occurred in a patient who had been switched from rapamycin to MMF because of hepatotoxicity (AST increased to 217), and the rejection occurred on MMF monotherapy. This patient has now returned to rapamycin monotherapy. Since rejection, renal function as reflected by serum creatinine returned to baseline for the patients with cellular rejection. Of those with humoral rejection, three of four have a higher creatinine level since rejection, and creatinine values at the latest testing are listed in Table 4. Patients with rejection were switched to maintenance therapy with FK506, MMF and prednisone, although the patient noted earlier has been converted to rapamycin monotherapy.
As a result of the unexpectedly high incidence of early humoral rejection in the first 24 patients, the subsequent five patients were treated with a modified protocol including Campath-1H on day −1 and day 0, and one dose of polyclonal ATG (Thymoglobulin, SangStat) 1.5 mg/kg i.v. on day 1. These five patients also received 14 days of steroids tapering from 500 mg i.v. on days −1 and 0 to 30 mg by day 13. The rationale for the protocol modification was the observation by flow cytometry that the recovery T cells were largely CD52 negative, and that 1–3% of a normal subject's T cells are CD52 negative. Therefore, we sought to target the CD52-negative cells with a single dose of Thymoglobulin.
Of the five patients on the modified protocol, two experienced rejection. The first of these patients was diagnosed with Banff IB acute rejection on day 30 and treated with bolus steroid therapy, Thymoglobulin, plasmapheresis, and IVIG. Biopsy was negative for C4d. Following treatment, serum creatinine returned to 1.7 mg/dL, and maintenance therapy with FK506, MMF, and prednisone was initiated.
The second patient developed rejection 24 days post-transplant and was treated with steroids, Thymoglobulin, and plasmapheresis. Biopsy showed Banff IIA cellular rejection and thrombotic microangiopathy, with C4d-positive staining of peritubular capillaries. Following treatment, serum creatinine fell to 1.8 mg/dL and the patient was converted to FK506, MMF, and prednisone.
Figure 3 shows the relationship between rejection and age of the patient. Only one of 16 patients older than 45 years developed rejection, whereas seven of 13 patients younger than 45 years experienced rejection.
Eight patients experienced rejection episodes with two predominant histologic patterns. Two patients were diagnosed with typical acute cellular rejection (both Banff 97; type 1B). Biopsies from the other six patients demonstrated features of antibody-mediated rejection either alone (two patients) or in combination with cellular rejection (four patients).
Typical histopathologic findings in the biopsy samples diagnosed as acute humoral rejection included interstitial edema and acute tubular injury characterized by cellular and granular casts (Figures 2A,B,D). Fibrin thrombi were often present in peritubular capillaries (Figure 2D) and occasionally in glomerular capillaries (Figure 2D). Acute transplant glomerulopathy (increased number of mononuclear cells in an endocapillary location) was seen in four of the six patients with acute humoral rejection. Immunostaining for C4d complement product was performed on five of six biopsy samples suspicious for acute humoral rejection. Positive linear staining in peritubular capillaries was detected in all five cases (Figure 2C). A C4d stain was also carried out on one of the biopsies showing classical acute cellular rejection. As expected, peritubular capillaries were negative for C4d in this case. To date, 87 biopsies have been carried out in these patients, including 20 protocol biopsies at 6 months and 13 at 1 year. No biopsies to date have demonstrated chronic rejection.
Measurement of T-cell flow cytometric crossmatches
Patients were evaluated for antidonor antibodies by dual color flow cytometric crossmatches using recipient serum collected both pretransplant and every 3 months up to 18 months' follow up. Donor lymphocytes for pre- and post-transplant T-cell crossmatches were available from 19 patients. All patients but one had a negative pretransplant flow cytometric T-cell crossmatch. One patient tested borderline positive both pre- and post-transplant and has had a slight decrease in mean channel shift post-transplant (MCF 18–15). This patient has never had clinical or biopsy evidence of rejection. Three patients developed positive antidonor antibodies, two at 3 months and one at 1 year. The former two patients had experienced humoral rejection histologically. One of these patients was tested at 1 year and remained positive. The third patient, who developed alloantibody at 1 year, has not had clinical evidence of rejection. Two patients who tested negative at all time points for T-cell crossmatches had experienced acute cellular rejection. In other words, they did not develop detectable alloantibody prior to or subsequent to acute cellular rejection. Donor T cells were not available for testing of three other patients who had experienced humoral rejection.
The results of this pilot study are notable in that similar patient and graft survival was achieved compared to conventional therapy, yet maintenance steroids were not used and no calcineurin inhibitors were used. Sirolimus monotherapy was well tolerated and low CD4 lymphocyte counts were well tolerated without systemic infection or malignancy. Five episodes of acute humoral rejection out of 29 (17%) recipients represents a higher incidence than the approximately 10% incidence seen with the cyclosporine/azathioprine/prednisone-based regimen (12,13). All patients had low PRA levels (<10% at transplant) and negative NIH crossmatch results. Two patients who experienced humoral rejection had retrospective flow cytometric T-cell crossmatches using pretransplant serum that tested negative. This immunosuppressive protocol appears not to protect against humoral rejection as well as conventional therapy. Reasons for this may be that immune cells that down-regulate the alloantibody response are impaired by the treatment protocol, or that the immunosuppressive regimen permits up-regulation of the alloantibody response in some patients, particularly young patients. The long-term consequences of humoral rejection in these patients are not known at this time. It will be important to continue to assess the development, persistence, or disappearance of antibody in these patients, and whether or not they develop progressive graft injury including chronic allograft nephropathy. There appears to be a significant association between age (>45 years) and rejection in this pilot trial, consistent with the observation that older patients with organ transplants have less graft loss as a result of rejection (14). We did not see a significant association between age and rate of leukocyte recovery.
Follow-up 6- and 12-month biopsies have not shown evidence of acute rejection for patients who did not experience a clinically diagnosed previous episode of rejection. Two patients with previously diagnosed acute rejection had histologic evidence of mild acute rejection (Banff 1A) on protocol biopsy at 12 months. Neither of these two patients had a clinical scenario consistent with acute rejection at the time of their protocol biopsy, but they were nevertheless treated with pulse steroid therapy without a change in their creatinine values. No biopsy to date has revealed chronic rejection.
Concern about the incidence of acute humoral rejection led us to modify the immunosuppressive protocol to include one dose of polyclonal antithymocyte globulin 1 day following the second dose of Campath and to add a 14-day course of steroids as induction. As two doses of Campath-1H effectively depleted two logs of lymphocytes, it was felt that a single dose of Thymoglobulin might be adequate to control the smaller population of alloreactive lymphocytes not targeted by Campath-1H. The brief course of steroids was added to down-regulate cytokines associated with ischemia/reperfusion injury and early nonspecific immune activation associated with the surgical aspects of renal transplantation. Of five patients treated with this regimen, two experienced rejection (2–4 months' follow up at this time): one humoral and one cellular in nature. This protocol modification did not effectively address the incidence of humoral rejection. Both the Cambridge series (6,7) and the experience of the Northwestern University transplant team (ATC 2002, Abstract #1030, Dr F. Stuart) have been marked by less humoral rejection than we report herein. Calcineurin inhibitors therefore may be valuable adjunctive immunotherapy in combination with Campath-1H. In fact, the ideal maintenance immunosuppressive agent to be used in combination with T-cell depletion is not known. It would appear from these results and those of Dr Allan Kirk (ATC 2002, Abstract #958) that despite profound depletion of T cells from blood, Campath-1H in the doses used does not avoid immune activation.
As more severe acute rejection is generally associated with a higher likelihood of humoral rejection, and because the cellular infiltrates of acute graft rejection in these study patients would be scarce because of lymphocyte depletion, the episodes of humoral rejection encountered may represent aggressive alloreactive T and B cells working in concert but with a relative dominance of the humoral rather than cellular component. Alternatively, the degree of immune cell depletion achieved in these patients by Campath-1H may cause disregulation of the humoral immune response. Finally, humoral rejection was studied more systematically in this study than is generally done in our clinical practice. Namely, all biopsies suspicious for rejection were stained for C4d, and if the necessary substrate was available, flow crossmatch testing was carried out to assess the presence of antidonor alloantibody. The incidence of humoral rejection in renal transplant patients on conventional immunosuppression is not known, although estimates based on C4d staining would put the incidence at approximately 10% (12,13).
The results of this pilot study suggest that profound lymphocyte depletion by Campath-1H is well tolerated in renal transplant recipients on concomitant sirolimus immunosuppression. Longer follow up is needed to assess the duration of immune cell depletion in transplant patients, although 7-year results in rheumatoid arthritis patients have been reported and demonstrate persistence of CD4 lymphopenia (15). There have to date been no adverse consequences of switching patients to FK506, MMF and prednisone if they develop rejection. Aggressive treatment of rejection in this series using Thymoglobulin, rituximab, steroids and plasmapheresis has been associated with salvage of seven of eight grafts without adverse events related to overimmunosuppression. The optimal maintenance immunosuppression to accompany Campath induction remains to be determined. Campath-1H has the advantages of being administered by the peripheral intravenous route, of being associated with very few side-effects of administration, and of requiring a small number of doses to achieve profound lymphocyte depletion. So far, prolonged lymphocyte depletion has been well-tolerated clinically with few infections and no malignancies. The high incidence of humoral injury and thrombotic microangiopathy was not expected, and is the reason that this protocol has subsequently been modified. Possible modifications might include increasing the degree of immune cell depletion by giving more Campath-1H or adding agents shown to augment tolerance induction such as DSG, or relying on at least a limited course of calcineurin-inhibitors with or without steroids. If the goal of Campath-1H induction is not tolerance but rather to decrease the use of steroids or to minimize maintenance immunosuppression, then a practical approach that has proven attractive is low-dose calcineurin-inhibitors with or without rapamycin or mycophenolate mofetil (7; ATC 2002 Abstract #1030, Dr F. Stuart).
The patients in this protocol who have avoided rejection (72%) have done remarkably well on minimal maintenance immunosuppression. This group will be followed with regard to long-term graft function, incidence of late acute or chronic rejection, and immunologic testing to assess tolerance. The patients successfully rescued from rejection with good renal transplant function will similarly be studied and compared with those who remain rejection-free. T-cell depletion with Campath-1H appears to be useful for solid organ transplantation and creates the opportunity to assess different approaches to either prevent rejection or possibly induce tolerance. These two goals may require different strategies. Future studies will be aimed at addressing this question.
Many thanks to Prof. R. Y. Calne and Dr. Allan Kirk for helpful discussions.This work was supported by NIH grant no. R01 AI50938 to S.J.K.