T Cell Depleted Allogeneic Stem Cell Transplants for Hematological Malignancies: A 20 Year Experience Shows No Relationship Between Choice of Transplanted T Cell Dose or Delayed T Cell Add-Back on Major Outcomes

Introduction: T cell depletion of the stem cell allotransplant (SCT) has the advantage of reducing incidence and severity of GVHD but can be complicated by relapse and infection. An optimum residual T cell dose in T depleted SCT is not known.  To optimize T cell depletion we  delivered a series of fixed T cell transplant doses with scheduled T cell add back post-SCT in a series of protocols to determine the T cell doses that secured immune reconstitution, and minimized relapse and infection. Here we retrospectively examined large series of patients (pts) transplanted in a single institution where T cell dose at SCT and add back were quantitated. Unique to these protocols was the precise measurement of CD3 dose/kg at transplant for every patient which enabled us to relate transplant T cell doses to outcome.

Patients and methods: 205 pts (106 males, 99 females) underwent HLA identical sibling allogeneic T cell depleted SCT between 1994 and 2014. Diagnosis at SCT included AML (48%), MDS (17%), ALL (27%), CLL (5%), MM (3%). Disease risk at SCT was classified as high (57%) or standard (43%).  The median age at SCT was 37 (range: 10-75) yrs. TBI based myeloablative conditioning was used. The graft source was marrow in 20 pts and peripheral blood in 195 pts. GVHD prophylaxis was low dose cyclosporine. Different T cell depleted methods were used consecutively: elutriation, Isolex ®, Cellpro ® CD34, and Miltenyl ^© CD34 selection.  A defined T cell dose was allocated at SCT by protocol ranging from 2 – 50 × 10⁴ CD3+ cells/kg.  Various schedules were used to add back T lymphocytes between day 30 to 90 with doses ranging 5 – 60 ×10⁶ CD3+ cells/kg by protocol and no T cell add-back was given in 28 pts in recent protocols.  Overall survival (OS) was estimated by the Kaplan-Meier method, and cumulative incidence of relapse and nonrelapse mortality (NRM) was estimated by Gray’s method to account for competing risks. Cox proportional hazard regression models were used to assess the association of factors at baseline, day 100 and GVHD with the post-SCT outcomes.

Results:  At a median follow-up of 8.6 yrs (range: 0.7- 19.8) for surviving pts, 112 pts died (52 from NRM) and 68 pts relapsed.  OS was 47%, 43% and 41%, NRM was 24%, 27% and 27%, relapse was 32%, 34% and 38% at 5 yr, 10 yr and 15 yr post SCT, respectively.  Grade II-IV and III-IV acute GVHD were 41% and 13%, and chronic GVHD was 42% (25% limited, 17% extensive).  In the multivariate models of baseline risk factors that adjusted for age at SCT and disease risk, T-cell doses at SCT did not affect OS, NRM or relapse. A higher dose of CD34+ cells at SCT was significantly associated with better OS and lower NRM. Disease risk was an independent predictor, with high-risk pts having more relapse and worse OS, compared to pts with standard risk.  A landmark analysis of 156 pts surviving and relapse-free beyond day 100 was carried out to examine the effects of add back T cell schedules by day 100 and aGVHD.  The total T-cell dose at add back by day 100  and different add-back T-cell schedules from day 30-90 had no impact on any outcome, controlling for T-cell dose at SCT.   In the models controlling for age, risk, CD34+ and T-cell dose at SCT,  pts with grade III-IV aGVHD by day 100 had an increased risk for overall mortality and NRM beyond day 100 (HR= 3 and 3.6, both P<0.001), but did not affect relapse.  An analysis of pts surviving and relapse-free beyond 1 yr  showed pts with extensive cGVHD had higher NRM rate (39%) and lower relapse rate (0%)  after 1 yr post-HSCT compared to pts with no cGVHD ( NRM, 9%, P= 0.020; relapse 24%, P=0.026).

 Conclusion: These findings indicate that for myeloablative matched sibling SCT there is no ideal prescription of T cell dose at transplant within a range of 10⁴ – 10⁵ or for scheduled add back of lymphocytes within a range of T depletion 5 - 60 × 10⁶. Instead, factors other than T cell dose either at transplant or when add-back was delayed determined GVHD incidence, relapse, and survival.   These findings set a limit on the efficacy of any T cell depletion procedure to optimize transplant outcome.