Photochemotherapy and Graft-versus-leukemia in Acute-leukemia

This is a 10-year follow-up of patients with Acute-myeloid-leukemia (AML) or acute-lymphatic-leukemia (ALL). AML is diagnosed by the French-American-British criteria (FAB-criteria) and ALL is separated into chief forms by immunohistological methods. All patients underwent myeloablative Hematopoietic Stem Cell Transplantation (HSCT) between 1985 and 2005 at the center for allogeneic stem cell transplantation (CAST) at Karolinska University Hospital. All patient receive GVHD-prophylaxis.

The risk for relapse after HSCT is graded into low-risk if the disease is in first complete remission before HSCT, all other disease states are classified as high-risk.

Eligible patients received photochemotherapy (Ultraviolet radiation type A on skin photosensitized by oral 8-methoxypsoralen) for acute graft-versus-host disease (GVHD within 100-days after HSCT). Photochemotherapy may be given as primary or later aGVHD therapy. Patients with aGVHD after booster doses of stem-cells or donor-lymphocyte-infusions are not included.

Additional treatment are registered where present. Methotrexate is not considered as an additional GVHD treatment as intravenous methotrexate a part of the governing GVHD prophylaxis and as the effects of methotrexate as a secondary aGVHD treatment is weak.

At the start, the end, at maximum and up until two weeks after end of PUVA-therapy the GVHD is diagnosed in accordance with Glucksberg and indexed by CIBMTR.

Relapse is diagnosed when leukemic cells is present extra medullary or with a bone marrow biopsy with ≥ 30% blasts. Early relapse is diagnosed when the medulla contain 5 - 30% blasts

The primary outcome is GVL i.e. abscence of relapse in malignant disease or minimal residual disease (MRD) i.e. threatening relapse in malignant disease demanding donor lymphocyte infusion (DLI).

Primary predictor: Time-to-treatment by photochemotherapy at day 0 - 7 vs. start at day 8 ≤ of aGVHD.

Continuous secondary predictor: Time-to-treatment by photochemotherapy as a continuous variable (days after start of aGVHD).

Binary secondary predictors: Risk (Low/High), Sibling donor-recipient (Yes/No), Mismatched related (Yes/No), Unrelated donor (Yes/No), (Male recipients of female grafts (Yes/No), T-cell depletion or Anti-Thymocyte Globulin (Yes/No).

Categorical secondary predictors: AGVHD organ disease stage and disease grade; Skin (+, ++, +++, ++++), Liver (+, ++, +++, ++++), Gastro-intestinal (+, ++, +++, ++++), Center for International Blood and Marrow Transplant Research CIBMTR index (A, B, C, D) respectively.

Statistical analysis:

Cox proportional Hazards ratio is used to conduct a univariate data analysis of all adequate variables in patient characteristics and disease towards the primary outcome. In the analysis, death, DLI or retransplantation due to graft-failure was treated as a competing event. The primary predictor (binary) and all binary or categorical covariates identified from the patient and disease characteristics are to be included in a multivariate forward regression analysis, controlled for with backward regression based on the log-likelihood method. P=0.05 is considered as significant and p=0.10 as a trend. StatSoft, Inc. (2013). STATISTICA (data analysis software system), version 12. www.statsoft.com. are used for statistical computation.