Understanding Telomere Biology in Hematopoietic Cell Transplantation: A Dynamical Systems Perspective

T cell proliferation and repertoire reconstitution is a hallmark of successful hematopoietic cell transplantation (HCT). This process may be modeled as a dynamical system and in such a system, precise telomere length (TL) measurement may reflect the proliferative capacity of donor T cells. TL for different chromosomes span a few orders of magnitude, and different T cell clones will display variable TL; these differences across the population are not represented when examining average TL. This study aims to develop a method that integrates the entire spectrum of TL observed within a sample to better understand the influence on clinical outcomes following HCT.

To better reflect the entire span of TL, we used data generated using the TeLSA PCR technique that provided discrete measurments of individual telomeres for each DNA sample for 72 stem cell transplant (SCT) donor-recipient pairs. Donor and recipient TeSLA TL measurements was performed on samples taken before and 90 days post HCT, respectively. Set correspondence mathematical techniques and area under the curve (AUC) calculations were used to measured donor-recipient TL differences (delta-TL) incorporating the full distribution of measured TL from each sample.

Telomere band lengths ranged from 350 basepairs (BP) to 16.7 kilobases with a logarithmically declining distribution in all samples when arrayed in a descending order. Set correspondence methods yielded TL averages which were highly correlated with AUC calculations (r >0.9, p<0.001 for all). The AUC delta-TL method predicted patient overall survival (P-log rank <0.0001). HCT recipients with intermediate degrees of telomere attrition (25^th-75^th percentile) post-HCT experienced the best outcomes (2 years overall survival; OS=92%), whilst donors with the least (<25^th percentile; 2 years OS=33%; adjusted HR vs. intermediate shortening=9.3, p=0.001) and the greatest (>75^th percentile; 2 years OS=59%; adjusted HR=6.0, p=0.01) shortening had worse outcomes. By contrast, using the traditional method based on donor-recipient difference in TeSLA mean telomere length did not demonstrate survival association in this small sample set (p log-rank=0.95).

The findings described herein suggest that the degree of donor telomere attrition may reflect T cell proliferation and alloreactivity following transplant. Accounting for the entire span of telomere lengths, could better identify post-transplant risk groups.