Catherine H. Roberts

Fractal Organization of the Human T Cell Repertoire in Health and after Stem Cell Transplantation

T cell repertoire diversity is generated in part by recombination of variable (V), diversity (D), and joining (J) segments in the T cell receptor β (TCR) locus. T cell clonal frequency distribution determined by high-throughput sequencing of TCR β in 10 stem cell transplantation (SCT) donors revealed a fractal, self-similar frequency distribution of unique TCR bearing clones with respect to V, D, and J segment usage in the T cell repertoire of these individuals. Further, ranking of T cell clones by frequency of gene segment usage in the observed sequences revealed an ordered distribution of dominant clones conforming to a power law, with a fractal dimension of 1.6 and 1.8 in TCR β DJ and VDJ containing clones in healthy stem cell donors. This self-similar distribution was perturbed in the recipients after SCT, with patients demonstrating a lower level of complexity in their TCR repertoire at day 100 followed by a modest improvement by 1 year post-SCT. A large shift was observed in the frequency distribution of the dominant T cell clones compared to the donor, with fewer than one third of the VDJ-containing clones shared in the top 4 ranks. In conclusion, the normal T cell repertoire is highly ordered with a TCR gene segment usage that results in a fractal self-similar motif of pattern repetition across levels of organization. Fractal analysis of high-throughput TCR β sequencing data provides a comprehensive measure of immune reconstitution after SCT.

Epigenetic induction of adaptive immune response in multiple myeloma: sequential azacitidine and lenalidomide generate cancer testis antigen‐specific cellular immunity

Patients with multiple myeloma (MM) undergoing high dose therapy and autologous stem cell transplantation (SCT) remain at risk for disease progression. Induction of the expression of highly immunogenic cancer testis antigens (CTA) in malignant plasma cells in MM patients may trigger a protective immune response following SCT. We initiated a phase II clinical trial of the DNA hypomethylating agent, azacitidine (Aza) administered sequentially with lenalidomide (Rev) in patients with MM. Three cycles of Aza and Rev were administered and autologous lymphocytes were collected following the 2nd and 3rd cycles of Aza‐Rev and cryopreserved. Subsequent stem cell mobilization was followed by high‐dose melphalan and SCT. Autologous lymphocyte infusion (ALI) was performed in the second month following transplantation. Fourteen patients have completed the investigational therapy; autologous lymphocytes were collected from all of the patients. Thirteen patients have successfully completed SCT and 11 have undergone ALI. Six patients tested have demonstrated CTA up‐regulation in either unfractionated bone marrow (n = 4) or CD138+ cells (n = 2). CTA (CTAG1B)‐specific T cell response has been observed in all three patients tested and persists following SCT. Epigenetic induction of an adaptive immune response to cancer testis antigens is safe and feasible in MM patients undergoing SCT.

Whole exome sequencing to estimate alloreactivity potential between donors and recipients in stem cell transplantation

Whole exome sequencing (WES) was performed on stem cell transplant donor‐recipient (D‐R) pairs to determine the extent of potential antigenic variation at a molecular level. In a small cohort of D‐R pairs, a high frequency of sequence variation was observed between the donor and recipient exomes independent of human leucocyte antigen (HLA) matching. Nonsynonymous, nonconservative single nucleotide polymorphisms were approximately twice as frequent in HLA‐matched unrelated, compared with related D‐R pairs. When mapped to individual chromosomes, these polymorphic nucleotides were uniformly distributed across the entire exome. In conclusion, WES reveals extensive nucleotide sequence variation in the exomes of HLA‐matched donors and recipients.

In silico Derivation of HLA-Specific Alloreactivity Potential from Whole Exome Sequencing of Stem-Cell Transplant Donors and Recipients: Understanding the Quantitative Immunobiology of Allogeneic Transplantation

Donor T-cell mediated graft versus host (GVH) effects may result from the aggregate alloreactivity to minor histocompatibility antigens (mHA) presented by the human leukocyte antigen (HLA) molecules in each donor–recipient pair undergoing stem-cell transplantation (SCT). Whole exome sequencing has previously demonstrated a large number of non-synonymous single nucleotide polymorphisms (SNP) present in HLA-matched recipients of SCT donors (GVH direction). The nucleotide sequence flanking each of these SNPs was obtained and the amino acid sequence determined. All the possible nonameric peptides incorporating the variant amino acid resulting from these SNPs were interrogated in silico for their likelihood to be presented by the HLA class I molecules using the Immune Epitope Database stabilized matrix method (SMM) and NetMHCpan algorithms. The SMM algorithm predicted that a median of 18,396 peptides weakly bound HLA class I molecules in individual SCT recipients, and 2,254 peptides displayed strong binding. A similar library of presented peptides was identified when the data were interrogated using the NetMHCpan algorithm. The bioinformatic algorithm presented here demonstrates that there may be a high level of mHA variation in HLA-matched individuals, constituting a HLA-specific alloreactivity potential.

Distinct oligoclonal T cells are associated with graft versus host disease after stem-cell transplantation.

Background In patients with hematologic malignancies who receive stem-cell transplantation, donors’ T cells can recognize minor histocompatibility antigens on recipient cells and generate an objective response against the tumor. However, a major side effect of such therapy is graft-versus-host disease (GVHD). The purpose of this study was to characterize distinct T-cell clones that were frequently and exclusively involved in GVHD or graft-versus-tumor (GVT) effects. Methods We hypothesized that distinct GVHD-associated T-cell clones can be identified during the disease progression. To test this, we conducted comparative analysis of T-cell receptor (TCR) V&bgr;s in donor-recipient pairs of patients with GVHD versus those with GVHD-free and relapse-free survival using quantitative reverse-transcriptase polymerase chain reaction and spectratyping analyses. Results We identified three sets of T-cell clones that were either frequently involved in GVHD (TCR V&bgr;4, 11, and 23) or GVT effect (TCR V&;9, 16, and 20), or were increased at the time of GVHD and GVT effects in a patient-specific manner (TCR V&bgr;2, 3, 7, 12, 15, and 17). Spectratyping analysis showed restricted clonality of the identified TCR V&bgr;s. Polymerase chain reaction analysis also confirmed the presence of GVHD-associated T-cell clones at the site of the disease. Conclusions These data suggest that GVHD- and GVT-associated clones can be distinguished by molecular analysis of TCR V&bgr; to develop targeted therapy for GVHD.

Favorable Outcomes in Patients with High Donor-Derived T Cell Count after In Vivo T Cell–Depleted Reduced-Intensity Allogeneic Stem Cell Transplantation

Patients with hematologic malignancies were conditioned using a rabbit antithymocyte globulin-based reduced-intensity conditioning regimen for allogeneic stem cell transplantation. Donor-derived CD3(+) cell count (ddCD3), a product of CD3(+) cell chimerism and absolute CD3(+) cell count, when <110/μL at 8 weeks post-stem cell transplantation predicted a high risk of sustained mixed chimerism and relapse. Alternatively, patients with a higher ddCD3 developed graft-versus-host disease more frequently, and when partially chimeric, had higher rates of conversion to full donor chimerism after withdrawal of immunosuppression. Early data from our small cohort of patients indicate that ddCD3 at 8 weeks may be used to guide decisions regarding withdrawal of immunosuppression and administration of donor lymphocyte infusion in partially T cell-depleted reduced-intensity regimens.

Stem cell transplantation as a dynamical system: are clinical outcomes deterministic?

Outcomes in stem cell transplantation (SCT) are modeled using probability theory. However, the clinical course following SCT appears to demonstrate many characteristics of dynamical systems, especially when outcomes are considered in the context of immune reconstitution. Dynamical systems tend to evolve over time according to mathematically determined rules. Characteristically, the future states of the system are predicated on the states preceding them, and there is sensitivity to initial conditions. In SCT, the interaction between donor T cells and the recipient may be considered as such a system in which, graft source, conditioning, and early immunosuppression profoundly influence immune reconstitution over time. This eventually determines clinical outcomes, either the emergence of tolerance or the development of graft versus host disease. In this paper, parallels between SCT and dynamical systems are explored and a conceptual framework for developing mathematical models to understand disparate transplant outcomes is proposed.

Dynamical System Modeling to Simulate Donor T Cell Response to Whole Exome Sequencing-Derived Recipient Peptides Demonstrates Different Alloreactivity Potential in HLA-Matched and -Mismatched Donor–Recipient Pairs

Immune reconstitution kinetics and subsequent clinical outcomes in HLA-matched recipients of allogeneic stem cell transplantation (SCT) are variable and difficult to predict. Considering SCT as a dynamical system may allow sequence differences across the exomes of the transplant donors and recipients to be used to simulate an alloreactive T cell response, which may allow better clinical outcome prediction. To accomplish this, whole exome sequencing was performed on 34 HLA-matched SCT donor-recipient pairs (DRPs) and the nucleotide sequence differences translated to peptides. The binding affinity of the peptides to the relevant HLA in each DRP was determined. The resulting array of peptide-HLA binding affinity values in each patient was considered as an operator modifying a hypothetical T cell repertoire vector, in which each T cell clone proliferates in accordance with the logistic equation of growth. Using an iterating system of matrices, each simulated T cell clones growth was calculated with the steady-state population being proportional to the magnitude of the binding affinity of the driving HLA-peptide complex. Incorporating competition between T cell clones responding to different HLA-peptide complexes reproduces a number of features of clinically observed T cell clonal repertoire in the simulated repertoire, including sigmoidal growth kinetics of individual T cell clones and overall repertoire, Power Law clonal frequency distribution, increase in repertoire complexity over time with increasing clonal diversity, and alteration of clonal dominance when a different antigen array is encountered, such as in SCT. The simulated, alloreactive T cell repertoire was markedly different in HLA-matched DRPs. The patterns were differentiated by rate of growth and steady-state magnitude of the simulated T cell repertoire and demonstrate a possible correlation with survival. In conclusion, exome wide sequence differences in DRPs may allow simulation of donor alloreactive T cell response to recipient antigens and may provide a quantitative basis for refining donor selection and titration of immunosuppression after SCT.

Dynamical system modeling to simulate donor T cell response to whole exome sequencing-derived recipient peptides: Understanding randomness in alloreactivity incidence following stem cell transplantation

Quantitative relationship between the magnitude of variation in minor histocompatibility antigens (mHA) and graft versus host disease (GVHD) pathophysiology in stem cell transplant (SCT) donor-recipient pairs (DRP) is not established. In order to elucidate this relationship, whole exome sequencing (WES) was performed on 27 HLA matched related (MRD), & 50 unrelated donors (URD), to identify nonsynonymous single nucleotide polymorphisms (SNPs). An average 2,463 SNPs were identified in MRD, and 4,287 in URD DRP (p<0.01); resulting peptide antigens that may be presented on HLA class I molecules in each DRP were derived in silico (NetMHCpan ver2.0) and the tissue expression of proteins these were derived from determined (GTex). MRD DRP had an average 3,670 HLA-binding-alloreactive peptides, putative mHA (pmHA) with an IC50 of <500 nM, and URD, had 5,386 (p<0.01). To simulate an alloreactive donor cytotoxic T cell response, the array of pmHA in each patient was considered as an operator matrix modifying a hypothetical cytotoxic T cell clonal vector matrix; each responding T cell clone’s proliferation was determined by the logistic equation of growth, accounting for HLA binding affinity and tissue expression of each alloreactive peptide. The resulting simulated organ-specific alloreactive T cell clonal growth revealed marked variability, with the T cell count differences spanning orders of magnitude between different DRP. Despite an estimated, uniform set of constants used in the model for all DRP, and a heterogeneously treated group of patients, higher total and organ-specific T cell counts were associated with cumulative incidence of moderate to severe GVHD in recipients. In conclusion, exome wide sequence differences and the variable alloreactive peptide binding to HLA in each DRP yields a large range of possible alloreactive donor T cell responses. Our findings also help understand the apparent randomness observed in the development of alloimmune responses.

Low-dose splenic irradiation prior to hematopoietic cell transplantation in hypersplenic patients with myelofibrosis

For patients with myelofibrosis undergoing hematopoietic cell transplantation (HCT), splenomegaly can lead to a protracted post-transplant course and is associated with lower survival in some series [1–3]. Splenectomy has been used in some patients prior to HCT, and studies have demonstrated that splenectomy prior to HCT reduced engraftment time (13 versus 20 d [2]). However, at many centers, splenectomy has fallen out of favor due to high complication rates (5% peri-operative mortality and 30–45% peri-operative morbidity [4]) and acceptable engraftment rates with HCT even in patients with significant splenomegaly [5]. Splenic irradiation has a long history in treating myelofibrosis, with low-dose treatment effective in alleviating painful splenomegaly [6,7]. In patients undergoing HCT for chronic myeloid leukemia, splenic irradiation has been used prior to transplant to reduce tumor burden and improve post-transplant kinetics [8], with a randomized study showing a survival benefit to splenic irradiation in a subset of patients [9]. Immunomodulatory drugs, JAK inhibitors, and mTOR inhibitors have also demonstrated efficacy in improving splenomegaly but have not been evaluated in the pre-transplant setting [10]. Starting in 2011, it has been standard practice at our institution to refer myelofibrosis patients with splenomegaly for low-dose splenic irradiation (LDSI) prior to HCT to attempt to reduce spleen size and improve posttransplant count recovery without the morbidity of splenectomy. Herein we report the outcomes of these patients. Medical records were examined after approval was obtained from our institution’s institutional review board. Between 2011 and 2015, eight patients received LDSI prior to HCT for myelofibrosis. Radiographic documentation of spleen size preand post-transplantation was available for seven patients. Spleen length is reported as the maximum cranio-caudal dimension on abdominal ultrasound. Regarding HCT, five patients had HLA matched related donors. All patients underwent reduced intensity conditioning regimens with busulfan plus fludarabine or melphalan. Anti-thymocyte globulin was given to one of five related donor patients and to all three unrelated donor patients. No patient received total body irradiation. Graft versus host disease (GVHD) prophylaxis was evenly split between cyclosporine and tacrolimus: one patient, whose donor was an identical twin, did not require GVHD prophylaxis. Median LDSI prescription dose was 4.5 Gray (range 3.0–6.0 Gray), delivered in 3 fractions (range 3–6 fractions) over 6 d (range 3–9 d). Patients finished treatment a median of 14 d prior to HCT (range 7–24 d). The date of graft infusion was defined as day 0 of transplant. Patients achieved neutrophil and platelet engraftment on the first of three consecutive days with an absolute neutrophil count of 0.5 10/L and an absolute platelet count of 50,000 10/L, respectively. At our institution, patients received red blood cell and platelet transfusions if their hemoglobin level fell below 8 g/dL and their platelet level fell below 30,000 10/L, respectively. Change in spleen length was evaluated by the Wilcoxon rank sum test, with continuity correction for continuous variables. Median age and Karnofsky performance score were 60 years and 80, respectively (Table 1). Median pre-transplant spleen length was 27 cm. All patients had a spleen size greater than 20 cm. Median time after HCT to repeat splenic imaging was 5 months (range 2–8 months). The patient cohort experienced a significant reduction in spleen size (median decrease 10cm (36%), p< .01). Acute LDSI treatment toxicity was mild (two patients with grade 1 and two patients with grade 2 nausea); no grade 3 toxicities were observed. Median times to neutrophil and platelet engraftment were 13 and 26 d, respectively. At 2 years post-HCT, 75%