Combined Islet and Hematopoietic Stem Cell Allotransplantation: A Clinical Pilot Trial to Induce Chimerism and Graft Tolerance
Article first published online: 29 APR 2008
© 2008 The Authors Journal compilation © 2008 The American Society of Transplantation and the American Society of Transplant Surgeons
American Journal of Transplantation
Volume 8, Issue 6, pages 1262–1274, June 2008
How to Cite
Mineo, D., Ricordi, C., Xu, X., Pileggi, A., Garcia-Morales, R., Khan, A., Baidal, D. A., Han, D., Monroy, K., Miller, J., Pugliese, A., Froud, T., Inverardi, L., Kenyon, N. S. and Alejandro, R. (2008), Combined Islet and Hematopoietic Stem Cell Allotransplantation: A Clinical Pilot Trial to Induce Chimerism and Graft Tolerance. American Journal of Transplantation, 8: 1262–1274. doi: 10.1111/j.1600-6143.2008.02230.x
- Issue published online: 30 MAY 2008
- Article first published online: 29 APR 2008
- Received 31 August 2007, revised 25 February 2008 and accepted for publication 26 February 2008
- graft tolerance;
- hematopoietic stem cells;
- islet transplantation;
- type 1 diabetes
To prevent graft rejection and avoid immunosuppression-related side-effects, we attempted to induce recipient chimerism and graft tolerance in islet transplantation by donor CD34+hematopoietic stem cell (HSC) infusion. Six patients with brittle type 1 Diabetes Mellitus received a single-donor allogeneic islet transplant (8611 ± 2113 IEQ/kg) followed by high doses of donor HSC (4.3 ± 1.9 × 106 HSC/kg), at days 5 and 11 posttransplant, without ablative conditioning. An ‘Edmonton-like’ immunosuppression was administered, with a single dose of anti-TNFα antibody (Infliximab) added to induction. Immunosuppression was weaned per protocol starting 12 months posttransplant. After transplantation, glucose control significantly improved, with 3 recipients achieving insulin-independence for a short time (24 ± 23 days). No severe hypoglycemia or protocol-related adverse events occurred. Graft function was maximal at 3 months then declined. Two recipients rejected within 6 months due to low immunosuppressive trough levels, whereas 4 completed 1-year follow-up with functioning grafts. Graft failure occurred within 4 months from weaning (478 ± 25 days posttransplant). Peripheral chimerism, as donor leukocytes, was maximal at 1-month (5.92 ± 0.48%), highly reduced at 1-year (0.20 ± 0.08%), and was undetectable at graft failure. CD25+T-lymphocytes significantly decreased at 3 months, but partially recovered thereafter. Combined islet and HSC allotransplantation using an ‘Edmonton-like’ immunosuppression, without ablative conditioning, did not lead to stable chimerism and graft tolerance.
Recent progress in islet transplantation for brittle Type 1 Diabetes Mellitus (T1DM) has made this procedure a feasible, minimally invasive approach to avoid severe hypoglycemia and secondary complications, improving glucose control with variable insulin-independence and improved quality of life (1–4). Long-term results have shown immunosuppression-related side-effects and progressive decline in islet graft function, most of the patients requiring reintroduction of various levels of insulin therapy within 4–5 years posttransplant (1,5–7).
Donor bone marrow cell (BMC) transplantation following myelo- or lympho-ablative conditioning has been shown to induce recipient chimerism and graft tolerance in solid organ transplantation, with reduction or discontinuation of immunosuppression (8). Similar results were seen in experimental islet transplant models, after minimal or nonablative regimens (9). In animal and clinical studies, infusion of high doses of donor CD34+hematopoietic stem cells (HSC) using minimal or nonablative conditioning resulted in successful engraftment, reduced adverse events, immunomodulation and increased allograft survival (10–15).
In this study, we attempted to induce recipient chimerism and graft tolerance in islet transplant recipients with brittle T1DM by infusing high doses of donor HSC without ablative conditioning to prevent graft rejection and eliminate life-long immunosuppression. An ‘Edmonton-like’ immunosuppression was given, with a single-dose of tumor necrosis factor-α (TNFα) monoclonal antibody (Infliximab) at induction to limit early posttransplant inflammation and maximize single-donor islet engraftment (1,4,16,17). Immunosuppression was weaned per protocol starting 12 months posttransplant to evaluate islet survival and chimerism (18–21). Clinical status, islet function, glucose control, immune reactivity and chimerism were monitored to evaluate efficacy and safety of the protocol.
Materials and Methods
Study design and population
This prospective, nonrandomized pilot study (NCT00315614) was approved by the University of Miami Institutional Review Board (IRB2000/0024) and written informed consent was obtained. The protocol included patients with long-term T1DM receiving a single-donor, freshly isolated allogeneic islet transplant followed by two infusions of high doses of cryopreserved donor HSC at days 5 and 11 posttransplant. No ablative conditioning for HSC engraftment was given to avoid related side-effects and risks. High doses of donor HSC (or BMC) have been shown to overcome the host HLA barrier, allowing for successful bone marrow (BM) engraftment, even without ablative regimens. Delayed timing of HSC (or BMC) infusions has resulted in greater engraftment and improved allograft survival (11–15).
An ‘Edmonton-like’ steroid-free immunosuppression was administered, adding an antiinflammatory agent (Infliximab) at induction (1,4,16). Starting 12 months posttransplant, immunosuppression was weaned. This time-frame was considered adequate to allow for stable donor HSC engraftment (21) and sustained islet function (4). Detectable chimerism was not a prerequisite for weaning of immunosuppression, since occurrence of graft tolerance has been reported even in absence of measurable donor cells (18–20).
Recipient selection and organ procurement
Inclusion and exclusion criteria for islet transplantation were previously reported (1,4,16). Negative cross-match, recipient panel reactive antibodies (PRA) ≤20%, ABO/Rh and CMV serology but not HLA compatibility, were required (22).
Islets and HSC were obtained from 15- to 45-year-old, heart-beating deceased donors. Both donor pancreata and vertebral bodies were preserved in specific cold-storage solutions and procurement media, and then shipped for processing and transplant (16,23,24).
Bone marrow cell extraction and infusion
Ten to twelve vertebral bodies were processed from each donor using a semi-automated system releasing around 3–5 × 1010BMC (25–27). To prevent graft-versus-host disease (GvHD), T-cell burden was indirectly reduced through CD34+ positive selection (Isolex 300i Magnetic Cell Separator; Baxter, Deerfield, IL) with a final cell preparation of >80% purity and >80% viability. Percentage of CD34+ and CD3+ cells was assessed by multicolor flow-cytometry (EPIX XL-MCL, Beckman Coulter, Fullerton, CA). Each BMC preparation was divided in two aliquots and stored in liquid nitrogen until infusion (28). Patients received a minimum of 2 × 106HSC/kg of recipient body weight (at least 10 × 108nucleated cells/kg), with a minimum of 2-logs T-cell reduction. All donors HSC obtained were infused (no maximum) (11–15).
Islet cells isolation and transplantation
Pancreata were processed using a modified automated method (29) and purified on density gradients (30), with a final islet preparation of purity >30% and viability >70% (16). Islets were infused by gravity into the portal vein (16,31,32) within 4 h from isolation. Patients received a minimum of 5000 IEQ/kg of recipient body weight (at least 300 000 IEQ total) in ≤5 mL of packed tissue.
Induction included Daclizumab (Zenapax®, Roche-Pharma, 1 mg/kg intravenously for five doses, starting the day of transplant, then every 14 days) plus Infliximab (Remicade®, Centocor, 5 mg/kg intravenously 2 h prior to islet infusion) (1). Maintenance consisted of sirolimus (Rapamune®, Wyeth-Ayerst, 0.2 mg/kg orally pretransplant then 0.15 mg/kg once daily, to attain trough levels of 12–15 ng/mL the first 3 months and 8–10 ng/mL thereafter) and low-dose tacrolimus (Prograf®, Fujisawa-Astellas, 1 mg orally pretransplant then 1 mg twice daily, to maintain trough levels of 3–6 ng/mL) (1,4,16). Immunosuppression weaning began by reducing tacrolimus (0.5 mg/week), then sirolimus (1 mg/week).
Metabolic assessments included fasting plasma glucose and C-peptide, daily insulin requirement, HbA1c, C-peptide-to-glucose ratio (CPGR) and indexes from the mixed-meal tolerance test (MMTT), namely the 90 minute-glucose (90 min-glc) and the mixed-meal stimulation index (MMSI), as ratio of C-peptide and glucose areas under-the-curve (33, 34).
Occurrences of hypoglycemic unawareness and coma were monitored. Renal function was evaluated by 24-h albumin urine excretion and estimated glomerular filtration rate (eGFR), using the modification of diet in renal disease equation (35), and the National Kidney Foundation stages for chronic kidney diseases (36). Immunosuppression trough levels were measured and drug-related adverse events evaluated.
Insulin-independence: C-peptide positive recipients maintaining, without insulin therapy, an HbA1c <6.5% and a fasting and/or 2-h postprandial finger-stick (capillary) blood glucose (FBG) <140 mg/dL and <180 mg/dL, respectively (34). Islet graft dysfunction: C-peptide positive recipients having, in the same week, three or more fasting and/or 2-h postprandial FBG >140 mg/dL and/or >180 mg/dL, respectively, and/or an HbA1c >6.5% in two consecutive measurements (33, 34). Islet graft failure: recipients with two or more consecutive fasting C-peptide <0.15 ng/mL, in the absence of hypoglycemia, and/or a stimulated peak of C-peptide <0.3 ng/mL during a MMTT within a month (34).
Chimerism was assessed monthly as percentage of donor cells circulating in the recipient's peripheral blood, including leukocytes, CD3+ and CD34+ cells. At 1-year posttransplant, a BM iliac aspirate was obtained and chimerism similarly assessed. A combined PCR and flow-cytometry analysis was used for detection of HLA class II DRβ gene and specific CD cell-surface epitopes on single fixed cells (37).
White blood cell count was routinely measured and immunophenotyping was performed by multicolor flow-cytometry, using fluorochrome labeled antibodies and isotype matched control (38). Cell-surface markers included: total T-lymphocytes (CD45+/3+) and relative subpopulations (CD45+/3+/4+, CD45+/3+/8+, CD3+/4+/25+, CD3+/8+/25+), B-lymphocytes (CD20+/40+/19+) and natural killer (NK) cells (CD56+/16+/3−).
Presence of PRA positivity and HLA class I and II alloantibodies was detected using a complement-dependent microlymphocytotoxic (CDC) technique (LCT assay, One Lambda). Recently, stored sera were retested with more modern and sensitive ELISA and flow-cytometry techniques (LAT assay and LABScreen assay plus LABScan flow analyzer, One Lambda and Luminex) (22).
Recipient ability to respond to donor antigens was evaluated in a one-way mixed lymphocyte reaction (MLR). Recipient peripheral blood mononuclear cells (PBMC) were challenged with γ-irradiated donor splenocytes (stored frozen cells), third-party cells, and self-PBMC (negative control). Recipients PBMC were also cocultured with phytohemoagglutinin (PHA) (positive control). Data were calculated as mean counts per minute (cpm) of quadruplicate cultures. Results were expressed as ratio (Stimulation Index) of proliferation (cpm) observed for donor, third-party, or PHA reactions divided by recipient versus self-cpm (39).
The cytotoxic T-lymphocyte genes (CLG) Granzyme B (GB), perforin (P) and Fas-Ligand (FasL) were evaluated by mRNA expression levels using real-time RT-PCR techniques (40). Results were expressed as the percentage ratio of copy number of the target genes to copy number of the control gene β-actin (40).
Autoantibody levels for GAD65, IA2 and insulin were evaluated for recurrence of autoimmunity, using standard radioimmunoassay (41). Results were expressed as the ratio of values of graft recipients to values of age- and sex-matched healthy control subjects.
Demographics and descriptive statistics were expressed as mean ± standard deviation (SD). Due to the nonnormal distribution of some variables and the small sample size, nonparametric tests for paired (intra-group) comparisons were used (Wilcoxon Rank-Sum test, SPSS 14.0, Chicago, IL). A p-value <0.05 was considered statistically significant.
From April through September 2000, 6 patients with T1DM were transplanted (2 males, 4 females, age 39 ± 6.9 years, disease duration 27 ± 11.8 years, weight 67.5 ± 16.9 kg, body mass index 23.8 ± 3.9 kg/m2), receiving a total of 574 475 ± 141 619 IEQ (8611 ± 2113 IEQ/kg) and 283.3 ± 126 × 106 HSC (4.3 ± 1.9 × 106 HSC/kg). Due to technical problems, patient 1 received a whole BMC infusion, a total of 3.5 × 1010 nucleated cells, resulting in a total of 227.0 × 106 HSC (2.3 × 106 HSC/kg) at flow-cytometry (Table 1). No procedure-related adverse events occurred.
|Pt||Sex||Weight (kg)||HLA mismatch (A, B, DR)||IEQ/kg (Purity%)||HSCs (×106/kg) (Purity%)||Days of insulin independence||Tacrolimus weaning (Day)||Sirolimus weaning (Day)||Graft failure (Day)|
|1||m||98.0||1, 1, 2||7046 (45)||2.3 (2.2)#||–||377||440||478°||503°||451∞||74∞|
|2||f||54.5||2, 1, 2||12 485 (37)||3.7 (87)||50||387||415||438||480||480||93|
|3||f||60.6||2, 2, 2||8012 (45)||2.7 (95)||5||–||–||–||–||178§||–|
|4||f||56.0||1, 2, 2||7039 (50)||6.7 (91)||17||388||430||438||465||471||83|
|5||f||59.5||1, 2, 2||7534 (30)||3.9 (95)||–||–||–||–||–||158§||–|
|6||m||76.5||1, 1, 2||9547 (40)||6.4 (86)||–||395||422||434||463||510||115|
|M||-||67.5||5||8611 (41)||4.3 (91)||24||387||427||437||469||478||91|
|SD||–||16.9||1||2113 (7)||1.9 (4)||23||7||11||2||9||25||18|
Islet graft failure occurred in recipients 3 and 5 within 6 months posttransplant, while having low immunosuppressive drugs trough levels due to poor therapeutic adherence and severe drug-related psychosis requiring reduction of immunosuppression, respectively, and follow-up was accordingly interrupted. The remaining four recipients maintained graft function up to 1 year (Table 1).
All recipients, except for those with early graft failure, had the greatest islet function and the best improvement in glycemic control in the first 3 months posttransplant but slowly deteriorated thereafter. Daily insulin requirements significantly decreased within this period (−65% of pretransplant value), with recipients 2, 3 and 4 achieving insulin-independence for a short time (50, 5 and 17 days, respectively). CPGR, 90 min-glc and MMSI significantly improved at 3 months posttransplant but worsened thereafter. By 12 months posttransplant, the four recipients with preserved grafts showed a relevant reduction in islet function and glycemic control, with increased daily insulin requirements (−40% of pretransplant value) (Tables 1 and 2, Figure 1).
|Pt||Fasting C-peptide (ng/mL)||HbA1c (%)||Daily insulin requirements (UI/kg/day)|
|Pretransplant||3 months||12 months||Graft failure||Pretransplant||3 months||12 months||Graft failure||Pretransplant||3 months||12 months||Graft failure|
|Pt||Albuminuria (mg/24 hr); (eGFR mL/min/1.73 m2)||Tacrolimus (ng/mL)||Sirolimus (ng/mL)|
|Pretransplant||12 months||Graft failure°||1 month||3 months||12 months||Graft failure||1 month||3 months||12 months||Graft failure|
|1||7 (62)||13 (78)||0 (59)||3.41||3.81||3.40||0.30∞||10.08||11.38||8.81||9.85∞|
|2||21(68)||29 (74)||0 (65)||4.10||5.40||5.25||0.00||12.24||12.29||10.80||6.63|
|3||58 (78)||–||65 (86)†||1.62§||3.58§||–||1.50†||6.93§||12.01§||–||7.40†|
|4||26 (46)||80 (43)||19 (47)||3.40||3.53||4.45||0.00||12.59||12.00||10.70||7.55|
|5||0 (66)||–||0 (64)‡||3.19§||1.55§||–||1.90‡||11.91§||8.19§||–||7.44‡|
|6||5 (98)||9 (99)||10 (95)||4.14||3.95||6.10||0.00||14.91||13.05||9.50||7.00|
|M||19.7 (69.5)||32.9 (73.6)||15.6 (69.3)||3.76||4.17||4.80||0.62||12.46||12.18||9.95||7.64|
|SD||21.3 (17.3)||32.6 (26.9)||25.2 (17.8)||0.41||0.84||1.15||0.86||1.98||0.69||0.96||1.13|
Following the 12-month follow-up visit, weaning of immunosuppression began for the four recipients with preserved graft function. Islet graft failure subsequently occurred within 4 months (478 ± 25 days posttransplant). A mild transient amelioration in fasting C-peptide and HbA1c was seen during weaning, with slight reduction of daily insulin requirements and improvement of CPGR, but not of 90 min-glc and MMSI. At islet graft failure, the metabolic parameters approximated pretransplant levels in all recipients, although lower daily insulin requirements were maintained (−35% of pretransplant value) (Tables 1 and 2, Figure 1).
Severe hypoglycemic episodes or coma did not recur posttransplant, with all recipients claiming improved hypoglycemia awareness. Renal function was normal pre-transplant, except for recipient 3 with micro-albuminuria, and recipient 4 with reduced eGFR. At 1-year posttransplant, while on immunosuppression, the remaining four recipients remained stable, except for recipient 4 who developed micro-albuminuria, without changes in eGFR. Recipient 3 showed persistent micro-albuminuria even after the end of the study. After islet graft failure and immunosuppression weaning, no changes were seen in renal parameters except for recipient 4 who reversed to normo-albuminuria with unmodified eGFR (Table 2).
Immunosuppression levels were consistently in the therapeutic range in the 4 recipients who completed 1 year of follow-up, except for recipient 1, whose sirolimus levels were below targeted level in the first 3 months posttransplant (Table 2). During follow-up, all recipients had multiple immunosuppression-related side-effects, with similar frequency and severity previously reported using ‘Edmonton-like’ protocols (5,42,43). The most important were: mucocutaneous (oral ulcers, n = 4 recipients), hematological (leucopenia, n = 5; anemia, n = 3), infective (respiratory tract infections, n = 5), metabolic (hyperlipidemia, n = 3) gynecological (dysfunctional menstrual bleeding, requiring endometrial ablation, n = 1), and neuropsychiatric (depression and anxiety, requiring reduction of immunosuppression, n = 1). All resolved satisfactorily after specific treatments without any sequelae.
Peripheral chimerism, measured as donor leukocytes, was maximal in the first 3 months posttransplant, then progressively decreased, highly reduced at 12 months and was undetectable at islet failure. At 12 months, donors HSC were slightly higher in the circulation than in the BM, while leukocytes and CD3+T-cells showed an opposite trend. Recipients 3 and 5 presented higher levels of chimerism at the time of islet graft failure when compared to the remaining four recipients at 1-year posttransplant, before starting immunosuppression weaning (Table 3, Figure 2).
|Pt||Circulating donor cells%/Bone marrow donor cells%∞|
|1 month||3 months||12 months||1 month||3 months||12 months||1 month||3 months||12 months|
|1||5.10||3.30||0.32 (0.50)||82.6||72.6||11.1 (1.08)||4.33||2.07||0.19 (0.56)|
|2||5.92||4.10||0.18 (0.51)||86.0||78.7||10.1 (1.35)||3.64||2.46||0.11 (0.60)|
|4||5.80||4.56||0.18 (0.32)||80.3||79.4||11.0 (0.87)||4.90||3.52||0.11 (0.44)|
|6||6.10||4.00||0.12 (0.38)||85.3||80.1||5.3 (0.79)||4.11||2.57||0.08 (0.40)|
|M||5.92||4.18||0.20 (0.43)||84.0||78.6||9.4 (1.02)||4.16||2.81||0.12 (0.50)|
|SD||0.48||0.51||0.08 (0.09)||2.2||3.2||2.8 (0.25)||0.44||0.59||0.05 (0.07)|
White blood cell counts presented a significant reduction of leukocytes, mainly neutrophils, within the first trimester posttransplant, with limited recovery at graft failure. Immunophenotyping showed a similar trend for T-lymphocytes, mainly CD4+cells, in the same period, with increasing levels up to graft failure. The CD25+ subpopulations, mainly CD4+cells, significantly decreased in the first 3 months posttransplant, but partially recovered thereafter. B-lymphocytes slightly increased at 1-month post-transplant, decreased during follow-up, and normalized at graft failure. NK-cells were reduced early posttransplant, but tended to increase at immunosuppression weaning (Figure 3).
PRA was ≤20% pretransplant in all recipients by CDC assay. With ELISA and flow re-testing, recipient 1 was slightly positive pretransplant and converted to negative while on immunosuppression, while recipients 3, 4 and 5 were well above the inclusion criteria pretransplant and remained persistently positive during the follow-up. All patients developed high PRA positivity after islet graft failure and immunosuppression weaning, with donor-specific alloantibodies (Table 4).
|Pt||Pretransplant PRA%||On immunosuppression PRA%||Off immunosuppression PRA%|
|HLA I (A,B,C)||HLA II (DR, DQ)||HLA I (A,B,C)||HLA II (DR, DQ)||HLA I (A,B,C)||HLA II (DR, DQ)|
|1||0%||All 0%||All 0%||All 0%||43%/A23; B44, 45, 49||–|
|^11%/A23,24||^87%/A11,24,25,32,66,80; B8,13,26,27,37,44,45,47, 49,51,52,53,57,58,59, 60,61,63,64||^71%/DR1,8,11,12, 13,15,16,51|
|2*||All 0%||All 0%||All 0%||All 0%||50%/A24,23,25; B49,51,57,58||–|
|§82%/A3,23,24; Bw4; Cw8||§66%/DR1,51; DQ8,11|
|^67%/A23,24,25,32; B13,27,37,41,44,47,49, 51,52,53575858,59,63||^43%/DR1,11,12,13,15, 16,51; DQ6|
|^36%/A1,23,32,36,80; B7,13,18,27,37,42,47, 49,50,60; Cw5||^29%/DR1,10,15,16; DQ5,6,8||^31%/A32; B7,13,41,47,49,50, 60,61,67; Cw18||§46%/DR1,4,6, 10,15,16; DQ5,9||^84%/A1,23,24,25,30,31, 32,36,80; B8,13,27,35,37, 44,47,49,51,52,53,57,58, 59,63,64,65; Cw12,17||^97%/DR1,4,8,10,11,12, 13,15,16,51,52,53; DQ4,9|
|§80%/A24,32,66; Bw4||§41%/DR1,10;||§79%/A24; Bw4,||§56%/DR51;DQ5||§79%/A1,24;Bw4,||§72%/DR9; DQ5,|
|^73%/A1,23,24,25,32,80; B8,13,27,37,44,45,47,49, 51,52,53,57,58,59,63,64,65; Cw12||^40%/DR1,7,9,10, 15,16,51,53; DQ6||^84%/A1,11,23,24,25,32,80; B13,27,37,41,44,45, 47,49,51,52,53,57,58,59,60 61,63||^77%/DR1,4,10,15, 16,51; DQ5,6,9||^80%/A1,11,23,24,25,26, 32,34,80; B13,27,33,37,44,45,47, 49,51,52,53,57,58, 59,60,61,63||^86%/DR1,4,7,9,10, 15,16,51,53; DQ5,6,9|
|5||18%/A25; B72; Cw1||All 0%||46%/A23,31; B37,41; Cw7||All 0%||0%||–|
|^56%/A2,11,24,25,26,30,34, 66;B8,38,39,48,51,52,54, 55,56,59,60,64,65,67,81; Cw8||^44%/A3,11,24,26, 32,34,36,66,68; B7,44,45,47,53,54,59,60,81||^84%/A1,3,11,23,24,25, 32,36,80 B8,13,27,35,37 ,44,45,47,51,52,53,57,58, 59,62,63,64, 65,75,76||^43%/DR8,11,12,15,16,51; DQ6|
|6||All 0%||All 0%||All 0%||All 0%||21%/A34; B41,46,53,72||–|
|§73%/A24,66; B35; Bw4||§84%/-|
|^44%/A1,23,24,25,32; B49,50,51,52,57,58,63||^77%/DR1,4,8,10,11, 12,13,51,53;DQ6,7,8,9|
MLR variably decreased in all recipients posttransplant, retaining the anti-third-party response, but increased at the time of reduced graft function and low immunosuppression levels. Recipient 6, with the highest response during the follow-up, exhibited a prolonged graft survival, similar to 2, the lowest response of all (Figure 4).
Similarly, in all recipients, analysis of CLG expression revealed multiple episodes of elevated and sustained GB levels, closely associated with periods of declining islet function and reduced immunosuppression levels, mainly sirolimus, without concomitant infections. Elevations of FasL were seen in recipients 1 and 5 related to graft failure and positivity of anti-IA2 and anti-insulin antibodies, respectively (Figure 4).
Auto-antibodies were positive pretransplant in most of the recipients (anti-GAD65 in recipients 2 and 3; anti-GAD65 and anti-IA2 in recipient 4; and anti-IA2 in recipient 6) and persisted with fluctuating levels during follow-up. Antiinsulin antibodies were negative pretransplant in all recipients. Recipients 1 and 5, autoantibodies negative pretransplant, developed anti-IA2 antibodies just before immunosuppressive weaning and antiinsulin antibodies immediately posttransplant, respectively.
Full or mixed chimerism has been reported in patients after partially matched BMC transplants for lympho-hematological malignancies using myelo-ablative conditioning allowing for the development of tolerance to a solid organ transplant from the same donor without chronic immunosuppression (8). The ability of microchimerism to induce graft tolerance remains unproven, despite detection of donor leukocytes many years after a solid organ transplant in patients with long-term graft function able to reduce or discontinue immunosuppression (18). Previous trials using lympho-ablative or no conditioning did not show a clear reduction in graft rejection nor immunosuppression dosage, despite persistence of <1% of donor cells for over 5 years (8,44–46).
Mismatched BMC transplants in nonhuman primates and rodents, after nonmyelo-ablative or lympho-depleting regimens, even nonobese autoimmune diabetic (NOD) mice, induced transient or stable chimerism, allowing for islet graft tolerance and discontinuation of immunosuppression (8,9,47–49).
High doses of HSC successfully engraft in animal models, even without conditioning, allowing for tolerance to donor islets as well as for diabetes reversal or prevention in NOD mice (8,10–14,50,51). Similar observations were reported in patients with recent T1DM onset after allogeneic BMC transplant for lympho-hematological malignancies (52) or autologous HSC infusion for the autoimmune disease itself (53).
In the clinical setting, infusion of high doses of donor HSC for lympho-hematological malignancies demonstrated similar efficacy in inducing chimerism, when compared to whole BMC, with fewer complications due to the lower antigenicity and T-cell content (10,15,54,55). Nevertheless, GvHD incidence remains significant, especially in mismatched transplants, when highly ablative conditioning is used (56–59).
The immunomodulatory properties of the HSC remain controversial. Persistent donor cells such as CD34+HSC and CD3+T-cells seem capable of inducing recipient immune unresponsiveness (60). A functional immune homeostasis, or operational tolerance, deriving from the recipient chimeric BM, appears able to inhibit the antidonor response by putative donor suppressor T-cells as well as to restore self-tolerance, preventing recurrence of autoimmunity (8,18,46,61–63). Conversely, the upregulation of costimulatory molecules on the HSC surface by the early posttransplant inflammatory response, or the differentiation of HSC into antigen-presenting cells by interaction with host T-cells, may be able to trigger alloreactivity, thereby limiting the safety and efficacy of their infusion (64,65).
This is the first study reporting the attempt to induce recipient hematopoietic chimerism and graft tolerance in clinical islet transplantation by infusing high doses of donor HSC under a modified ‘Edmonton-like’ immunosuppression without ablative conditioning. Recipients with 1-year follow-up, despite insulin dependence, showed C-peptide secretion with significant reduction of insulin requirements, normalization of HbA1c, and resolution of hypoglycemia. Nonetheless, islet graft function started to decline 3 months posttransplant, concomitant to completing Daclizumab course and reducing sirolimus levels.
No GvHD, malignancy or permanent complications were recorded, highlighting the relative safety of this protocol. Renal function did not deteriorate posttransplant, except transiently in a recipient with a preexisting alteration.
Immunosuppressive levels, mainly of sirolimus, seemed to be critical to preserve islet survival. Graft failure occurred in periods of low trough levels or even while in the targeted ranges, leading us to aim at sirolimus trough levels of 10–12 ng/mL after the third month in subsequent trials (66). Recipients 3 and 5 experiencing early graft failure despite high C-peptide at 1-month posttransplant, presented islet dysfunction within 3 months, probably due to low immunosuppressive drug levels, with subsequent graft failure. Receiving lower IEQ/kg and HSC/kg might have also conditioned their outcome. Indeed, patients 2 and 6 receiving the highest IEQ/kg had the longest islet graft survival. Furthermore, patients 4 and 6 receiving higher HSC/kg, despite elevated immune reactivity, had similar or longer islet survival when compared to recipient 2 of lower HSC/kg but optimal IEQ/kg and mild immune reactivity. Conversely, patient 1 receiving the lowest HSC/kg and IEQ/kg had a shorter graft survival, despite moderate immune reactivity. Recurrence of autoimmunity at weaning or islet loss due to low sirolimus levels in the first trimester may have also influenced this outcome.
Peripheral chimerism started to decrease 3 months posttransplant, concomitantly with lowering of immunosuppression. Donor HSC progressively disappeared from the circulation showing poor ability to home to the host BM, as evidenced by the 12-month iliac aspirate.
Immunophenotyping showed persistent and significant reduction of T-lymphocytes while on immunosuppression. The CD25+ subsets were almost undetectable in the first 3 months posttransplant related to Daclizumab infusion, but promptly increased thereafter.
Repetitive episodes of immune reactivity occurred during follow-up, evidenced by increased CLG levels and MLR responses, related to periods of low immunosuppression levels and declined islet function, probably reducing the islet graft mass before weaning.
Most recipients with negative or borderline-positive pre-transplant PRA at CDC assay were found to be highly positive when retested by ELISA or flow-cytometry assays (higher values in females with a previous pregnancy). Confirmed negative or low-positive pretransplant PRA remained negative while under immuno-suppression as previously described (22). Allosensitization occurred in all recipients after weaning (22). Positive PRA has persisted in the only two recipients (2 and 6) with long-term follow-up (80% and 40% HLA-I and 10% and 35% HLA-II, 5 and 6 years posttransplant, respectively). The clinical relevance of these phenomena needs further evaluation.
We hypothesize that infusing highly T-cell-depleted donor HSC without ablative conditioning could have preserved recipient immune-competence while avoiding GvHD, but might have prevented HSC engraftment in the host BM, limiting achievement of stable chimerism. Adding a single dose of TNFα-blocker to a relatively short nonlympo-depleting induction could have reduced early HSC activation, limiting their antigenicity and allorecognition, but might not have avoided a late donor-specific immune response with loss of cell grafts. Chronic immunosuppression could have interfered with HSC engraftment and proliferation, and might have impeded the generation of regulatory T cells, preventing the induction of tolerance. Transplanting a marginal islet mass from a single-donor without adequate lympho-depletion at induction may explain the low rate of insulin-independence in our study (67–68).
A limitation of our study is the lack of a control group receiving an equal islet mass without HSC infusion, to better evaluate the effect of the HSC on islet survival and glucose control. Analyses of chimerism and immunophenotype, which could have assisted in defining a possible immunomodulatory role for both HSC and immunosuppressive protocol, were limited. Other T-cell subsets, such as CD4+/25+/Foxp3+ regulatory T-lymphocytes were not studied, and their influence on graft tolerance could not be defined.
In conclusion, combined islet and HSC allotransplantation under an ‘Edmonton-like’ immunosuppression without ablative conditioning did not lead to stable chimerism and graft tolerance, as evidenced by islet dysfunction during follow-up and graft failure after immunosuppression weaning. Further studies need to address the possible immunomodulatory effects of donor HSC, which may occur only at high doses (i.e. ≥6 × 106). A more powerful conditioning will most likely be necessary to improve the tolerogenic potential of donor HSC together with immune strategies that may favor a regulatory cell reconstitution after induction.
Recently, different approaches have been successfully used in both experimental and clinical settings to increase the likelihood to induce chimerism and operational tolerance (69,70). Pan- or selective lympho-depleting agents, alone or in combination with nonmyeloablative pharmacological or limited irradiative regimens, could be used at induction to improve HSC engraftment and reduce host immune alloreactivity (71). Prolonged anti-inflammatory and costimulatory blockade treatments could be used at induction and maintenance to induce more complete and tolerogenic immunosuppression (72,73). Other types of donor cells with immunomodulatory properties (i.e. lymphocytes or mesenchymal stem cells) might represent possible alternative strategies to increase recipient immunomodulation and graft tolerance while avoiding GvHD (74,75). A careful evaluation of the possible side-effects and risks associated with these regimens should be considered in designing future protocols and clearly explained to patients at recruitment.
This study was supported by: National Institutes of Health/National Center for Research Resources (Islet Cell Resources: U423RR016603; General Clinical Research Center: MO1RR016587); National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases (R01DK056953, DK2580218); and the Diabetes Research Institute Foundation (DRIF) (http://www.diabetesresearch.org).
DM is supported by a Post-Doctoral Research Fellowship in ‘Advanced Technologies and Therapies in Surgery’, Department of Surgery, University of Rome ‘Tor Vergata’.
We express gratitude to the cGMP Cell Transplant Center and the Tissue Typing Laboratory, in particular Mr. Joel Schultz and Mrs. Carmen Gomez, for their support and collaboration. We also deeply thank Mr. John Wilkes (regulatory officer) and Ms. Elizabeth Meyer (CITP manager) for reviewing the manuscript.
- 16Twenty years of clinical islet transplantation at the Diabetes Research Institute–University of Miami. Clin Transpl 2004: 177–204., , et al.
- 47Islet cell transplantation tolerance. Transplantation 2001; 72(8 Suppl): 43–46., , , .
- 57CD34 cell dose in granulocyte colony-stimulating factor-mobilized peripheral blood mononuclear cell grafts affects engraftment kinetics and development of extensive chronic graft-versus-host disease after human leukocyte antigen-identical sibling transplantation. Blood 2001; 98: 3221–3227., , et al.