Emerging Treatments and Technologies BRIEF REPORT

Autologous Umbilical Cord Blood Transfusion in Young Children With Type 1 Fails to Preserve C-Peptide

1 2 MICHAEL J. HALLER, MD THERESA M. SUMRALL, BS basic metabolic panel, HbA ,andflow 2 3 1c CLIVE H. WASSERFALL, MS JOHN R. WINGARD, MD 2 4 cytometry, with staining for CD3, CD4, MAIGAN A. HULME, BS DOUGLAS W. THERIAQUE, MS 1 4 CD8, CD25, CD62L, CD45RA, and MIRIAM CINTRON, BS JONATHAN J. SHUSTER, PHD 2 2 FOXP3 (2–4). UCB was thawed and TODD M. BRUSKO, PHD MARK A. ATKINSON, PHD 2 1 washed per standard operating proce- KIERAN M. MCGRAIL, BS DESMOND A. SCHATZ, MD dures. Subjects received pretreatment with diphenhydramine and acetamino- phen. No other preparative therapy was OBJECTIVEdWe conducted an open-label, phase I study using autologous umbilical cord given. Thawed UCB cells were infused blood (UCB) infusion to ameliorate (T1D). Having previously reported on the – first 15 patients reaching 1 year of follow-up, herein we report on the complete cohort after through a peripheral IV over 10 20 min. 2 years of follow-up. Subjects returned for follow-up testing (identical to baseline) every 3 months RESEARCH DESIGN AND METHODSdA total of 24 T1D patients (median age 5.1 in the 1st postinfusion year and every years) received a single intravenous infusion of autologous UCB cells and underwent metabolic 6 months in the 2nd postinfusion year. and immunologic assessments.

RESULTSdNo infusion-related adverse events were observed. b-Cell function declined after Statistical analysis UCB infusion. Area under the curve C-peptide was 24.3% of baseline 1 year postinfusion (P , To determine changes from baseline, we 0.001) and 2% of baseline 2 years after infusion (P , 0.001). Flow cytometry revealed increased calculated fractional change in area under regulatory T cells (Tregs) (P = 0.04) and naive Tregs (P = 0.001) 6 and 9 months after infusion, the curve (AUC) C-peptide for each sub- respectively. 2 ject as [(Yn/Y0) 1], where the subscript CONCLUSIONSdAutologous UCB infusion in children with T1D is safe and induces n is the month number. Fractional changes changes in Treg frequency but fails to preserve C-peptide. were tested for a target population null hypothesis of a median of 0 by the two- Diabetes Care 34:2567–2569, 2011 sided Wilcoxon rank-sum test, a non- parametric procedure. Slope analysis was utologous umbilical cord blood detailed description of the study, as well performed to calculate the rate of decline (UCB) is an attractive source for po- as results from the first 15 subjects to reach in AUC C-peptide over time. The pilot A nature of this study dictated against con- tential cell therapies in young chil- 1 year of postinfusion follow-up, was pre- dren. On the basis of preclinical efficacy viously published (1). trolling study-wide errors via either a Bon- and safety data, we performed an un- ferroni correction or formal multivariate blinded observational pilot study to deter- Procedures analysis. Because of concern for outliers, mine whether autologous UCB infusion Peripheral blood and an aliquot of UCB descriptive statistics are presented as me- could preserve remaining endogenous in- from potential subjects were shipped to dian (quartiles). sulin production. the University of Florida for infectious disease testing, HLA confirmation, and vi- RESULTSdBetween 15 December 2003 RESEARCH DESIGN AND ability screening. Thereafter, the UCB unit and 21 November 2008, 24 children with METHODSdSubjects .1 year of age, of qualified subjects was shipped to the T1D (10 males, 14 females) underwent a with type 1 diabetes (T1D), and for whom University of Florida and stored until single autologous UCB transfusion. No autologous UCB was stored, were recruited transfused. Subjects underwent a baseline adverse events were observed in asso- for participation (clinical trial reg. no. 2-h mixed meal tolerance test and had ciation with autologous UCB infusion. NCT00305344; FDA IND BB-11918). A blood drawn for complete blood count, All aliquots of UCB had negative Gram stains, and none grew pathogenic organ- ccccccccccccccccccccccccccccccccccccccccccccccccc isms when cultured for virus, bacteria, or From the 1Department of Pediatrics, University of Florida, Gainesville, Florida; the 2Department of Pathology, fungus. University of Florida, Gainesville, Florida; the 3Department of Medicine, University of Florida, Gainesville, 4 Baseline and longitudinal postinfusion Florida; and the Department of Epidemiology and Health Policy Research and the Clinical and Trans- characteristics are provided in Table 1. lational Science Institute, University of Florida, Gainesville, Florida. Corresponding author: Michael J. Haller, [email protected]fl.edu. Median age at infusion was 5.1 years Received 26 July 2011 and accepted 20 September 2011. (3.4–6.9). Median time from diagnosis to DOI: 10.2337/dc11-1406. Clinical trial reg. no. NCT00305344, clinicaltrials.gov. UCB infusion was 0.25 years (0.19–0.52). This article contains Supplementary Data online at http://care.diabetesjournals.org/lookup/suppl/doi:10 The median infused total nucleated cell .2337/dc11-1406/-/DC1. 3 7 © 2011 by the American Diabetes Association. Readers may use this article as long as the work is properly count was 1.88 10 cells/kg. Median cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/ viability was 97% (95–99%). Overall, the licenses/by-nc-nd/3.0/ for details. total nucleated cell count recovered was care.diabetesjournals.org DIABETES CARE, VOLUME 34, DECEMBER 2011 2567 2568 diabetes 1 type for blood cord Autologous

Table 1dBaseline and postinfusion characteristics of autologous UCB recipients D IABETES

Postinfusion (Preinfusion–to–1 (Preinfusion–to–2 C

ARE Preinfusion 3 months 6 months 9 months 1 year 1.5 years 2 years year ratio) 2 1 year ratio) 2 1 , VOLUME HbA1c (%) 7.4 7.5 7.5 7.1 7.1 7.1 7.6 20.2 0.1 (6.5–8.4) (7.1–8.5) (7.1–8.5) (6.7–7.1) (6.7–7.8) (6.6–7.9) (7.2–8.0) (20.6 to 0.7) (20.5 to 0.9)

4 D 34, {24} {24} [NS] {23} [NS] {23} [NS] {24} [NS] {20} [NS} {21} [NS] {24} [NS] {21} [NS] use

ECEMBER (units/kg per day) 0.37 0.46 0.58 0.58 0.69 0.63 0.66 20.22 20.44 (0.22–0.51) (0.28–0.58) (0.44–0.79) (0.44–0.79) (0.55–0.81) (0.57–0.77) (0.61–0.81) (20.53 to 20.085) (20.68 to 20.21) {24} {23} [NS] {23} [NS] {23} [NS] {23} [,0.001] {20} [0.001] {21} [,0.001] {20} [,0.001] {21} [,0.001] 01care.diabetesjournals.org 2011 Peak C-peptide (ng/mL) 1.16 0.83 0.73 0.6 0.28 0.17 0.05 20.53 20.7 (0.7–1.71) (0.24–1.28) (0.47–1.35) (0.08–1.36) (0.16–0.86) (0–0.95) (0–0.58) (21.0 to 20.16) (21.4 to 20.5) {22} {24} [0.006] {23} [0.003] {20} [,0.001] {21} [,0.001] {19} [,0.001] {21} [,0.001] {20} [,0.001] {19} [,0.001] AUC C-peptide (ng/mL) 0.95 0.66 0.64 0.46 0.22 0.15 0.02 20.49 20.6 (0.51–1.4) (0.2–1.0) (0.2–1.1) (0.03–1.1) (0.02–0.63) (0–0.7) (0–0.52) (20.82 to 20.11) (21.0 to 20.37) {22} {24} [0.002] {23} [,0.001] {20} [,0.001] {21} [,0.001] {21} [,0.001] {21} [,0.001] {21} [,0.001] {19} [,0.001] IA-2A 11.0 7.5 4.5 4.1 2.9 2.2 2.3 23.8 27.2 (1.8–23.1) (0.69–20) (0.2–16.3) (0–16.5) (0–14.5) (0–9.9) (0–8.9) (27.5 to 20.35) (212.7 to 22.3) {22} {21} [0.01] {22} [0.01] {20} [NS] {21} [0.001] {18} [,0.001} {18} [,0.001] {20} [0.001] {16} [,0.001] GADA 2.5 1.7 1.95 2.0 2.6 1.9 1.2 21.0 20.8 (0.7–12.3) (0.3–6.4) (0.2–9.2) (0.5–15.7) (0.3–15.8) (0.4–5.3) (0.2–8.9) (23.6 to 0.6) (28.4 to 0) {23} {21} [NS] {22} [NS] {20} [NS] {22} [NS] {18} [0.01] {19} [NS] {22} [NS] {18} [NS] WBC (cell 3 109/L) 5.6 5.75 6.1 5.8 5.5 5.8 5.5 20.4 20.5 (5.0–7.3) (4.7–7.85) (4.7–6.9) (4.6–6.9) (4.4–6.7) (3.9–7.6) (4.25–7.35) (20.7 to 0.2) (21.8 to 0.6) {23} {24} [NS] {23} [NS] {20} [NS] {22} [NS] {18} [NS] {20} [NS] {22} [NS] {19} [NS] CD4-to-CD8 ratio 1.89 1.89 1.89 1.96 1.97 1.90 2.15 0.1 0.15 (1.5–2.4) (1.6–2.2) (1.5–1.9) (1.6–2.2) (1.7–2.2) (1.4–2.1) (1.7–2.4) (20.28 to 20.24) (20.2 to 20.4) {22} {23} [NS] {23} [NS] {20} [NS] {22} [NS] {18} [NS] {20} [NS] {21} [NS] {18} [NS] Peripheral blood Treg (%) 3.1 4.1 4.4 3.6 4.4 3.0 3.3 0.13 0.52 (0.8–5.4) (1.0–5.5) (2.0–7.5) (1.9–5.1) (2.0–7.6) (1.8–5.1) (1.9–6.5) (21.3 to 1.0) (21.8 to 2.9) {22} {23} [NS] {23} [0.04] {20} [NS] {22} [NS] {20} [NS] {20} [NS] {21} [NS] {18} [NS] CD45RA Treg (%) 39.0 42.6 40.5 43.5 40.9 40.2 42.8 2.24 1.82 (25.7–45.9) (27.8–49.8) (34.1–51.3) (37.9–54.3) (31.9–50.4) (29.9–46.2) (28.6–48.8) (25.53 to 11.0) (23.8 to 1.8) {22} {23} [NS] {23} [NS] {20} [0.001] {22} [NS] {18} [NS] {20} [NS] {21} [NS] {18} [NS] Data are median (interquartile range) and {n}[P value vs. baseline]; N = 24 (10 males, 14 females), median age at infusion 5.1 years (3.4–6.9). IA-2A, insulinoma-associated 2 antibody; GADA, GAD antibody; WBC, white blood cell count. Haller and Associates commonly 1–2 log fold less than that typ- regenerative or immunoregulatory capacity analysis, interpretation of data, or writing of ically observed in samples obtained from may have been transferred to patients with the manuscript. public banks (5). T1D. In addition, the ongoing autoim- No potential conflicts of interest relevant to Median AUC C-peptide at the time of mune response in new-onset T1D subjects this article were reported. autologous UCB infusion was 0.95 ng/mL may contain memory T cells, refractive to M.J.H. researched data and wrote the manu- – script. C.H.W. researched data and contributed (0.5 1.4). Median AUC C-peptide de- regulation by Tregs (6), that facilitate the to discussion. M.A.H. and M.C. researched data. clined at all subsequent study visits com- ongoing autoimmune destruction of en- P , b T.M.B. and D.A.S. researched data, contributed pared with baseline ( 0.01 at all time dogenous or de novo -cells. to discussion, and wrote the manuscript. K.M.M., points) (Table 1). At post-UCB infusion To address the first issue, efforts are T.M.S., J.R.W., and D.W.T. researched data. visits at 3, 6, and 9 months and 1 and 2 underway to isolate and expand specific J.J.S. and M.A.A. researched data and edited the years, median AUC C-peptide was 0.66 cell populations within UCB to augment manuscript. ng/mL (0.2–1), 0.64 ng/mL (0.2–1.1), their therapeutic potential. As proof of The authors acknowledge ongoing cord 0.46 ng/mL (0.03–1.1), 0.22 ng/mL concept, a phase I clinical trial is currently blood collaborations with the teams of Dr. (0.02–0.63), and 0.02 ng/mL (0–0.52), under way in adult patients with recent- Annette Zeigler (University of Munich, Munich, respectively (Supplementary Fig. 1). onset T1D using autologous expanded Germany) and Dr. Olli Simell (University of Turku, Turku, Finland). The authors thank the AUC C-peptide was 24.3% of baseline at Tregs isolated from peripheral blood following for their assistance: Hilla-Lee Viener 1 year after infusion and 2% of baseline (clinical trial reg. no. NCT01210664). In (laboratory technician, University of Florida), 2 years after infusion. Slope analysis of terms of the second limitation, studies the University of Florida Stem Cell Laboratory C-peptide demonstrated AUC C-peptide from our laboratory suggest that a com- staff and nurses, the General Clinical Research change of 22.4% per month (23.09 to bination therapeutic approach involving Center staff and nurses, and most importantly, 21.43) (P , 0.001). transient immune depletion and subse- the children and families who participated in Baseline and 1- and 2-year median quent induction of immune regulation this phase I trial. total peripheral white blood counts were is optimal (7). As such, we believe that 5.6 3 109,5.53 109,and5.53 109 cells/L, therapies combining transient immune fi respectively (not signi cant vs. base- depletion and subsequent infusion of ex- References line). Total T-cell numbers, CD4-to-CD8 panded UCB Tregs may more effectively 1. Haller MJ, Viener HL, Wasserfall C, Brusko + ratio, CD62L regulatory T cells (Tregs), reset the balance of Tregs and effector T, Atkinson MA, Schatz DA. Autologous CD62L+ T effectors, CD45RA+ T conven- T cells in T1D. umbilical cord blood infusion for type 1 tional cells, and CD45RO+ Tregs were Although this effort failed to demon- diabetes. Exp Hematol 2008;36:710–715 not different at all study visits compared strate benefit, the potential of UCB to 2. Brusko T, Atkinson M. Treg in type 1 di- abetes. Cell Biochem Biophys 2007;48: with baseline (data not shown). However, participate in future T1D interventional – an increase in total Treg (CD4+CD25+ therapies remains. Efforts to use autolo- 165 175 + 3. Brusko T, Wasserfall C, McGrail K, et al. No FOXP3 ) was observed at the 6-month gous UCB in the treatment of T1D will alterations in the frequency of FOXP3+ visit (median 4.4%, P = 0.04), as was an continue with emphasis on improved + regulatory T-cells in type 1 diabetes. Diabetes increase in CD45RA Treg at the 9-month understanding of UCB Treg function, 2007;56:604–612 visit (median 43.5%, P = 0.001) (Table 1). the addition of generally regarded as safe 4. Brusko TM, Wasserfall CH, Clare-Salzler therapies (i.e., vitamin D and n-3 fatty MJ, Schatz DA, Atkinson MA. Functional fl CONCLUSIONSdA single infusion acids) to UCB infusion (clinical trial reg. defects and the in uence of age on the no. NCT00873925), and perhaps most frequency of CD4+ CD25+ T-cells in type 1 of minimally manipulated autologous – UCBinyoungchildrenwithT1Dis important, the potential use of expanded diabetes. Diabetes 2005;54:1407 1414 5. Smith L, Haller MJ, Staba-Kelly S. Char- feasible and safe but fails to preserve autologous UCB Tregs either alone or in combination with other immunomodu- acteristics and cell composition of pri- C-peptide. Lack of control subjects (in vately banked autologus cord blood (UCB) this case, attributable to internal review latory agents. units utilized for autologous infusion in board and U.S. Food and Drug Adminis- children with type 1 diabetes (Abstract). tration restrictions) makes it difficult to Biol Blood Marrow Transplant 2008;14 form definitive conclusions regarding ef- AcknowledgmentsdThis study was funded (Suppl.):45 ficacy. The observation that total Treg by Juvenile Diabetes Research Foundation 6. Schneider A, Rieck M, Sanda S, Pihoker C, frequency was increased up to 6 months (JDRF) Innovative Grant 1-2005-362, JDRF Greenbaum C, Buckner JH. The effector after infusion suggests that autologous UCB Center Grant 4-2007-1065, a gift from the Arie T cells of diabetic subjects are resistant to Kurtzig Memorial Fund, National Institutes of infusion may favorably alter the T-cell regulation via CD4+ FOXP3+ regulatory Health (NIH) Grant 1R21-DK-077580-01, T cells. J Immunol 2008;181:7350–7355 repertoire in children with T1D. and an NIH/National Center for Research 7. Parker MJ, Xue S, Alexander JJ, et al. Im- The reasons for an inability of autolo- Resources Clinical and Translational Science mune depletion with cellular mobilization gous UCB to effectively halt autoimmune Award to the University of Florida (UL1- imparts immunoregulation and reverses progression are at least twofold. First, an RR029890). The sponsors of the study had no autoimmune diabetes in nonobese diabetic insufficient number of cells carrying role in the study design, data collection, data mice. Diabetes 2009;58:2277–2284

care.diabetesjournals.org DIABETES CARE, VOLUME 34, DECEMBER 2011 2569