Low Pre-DLI Lymphocytes Counts Are Predictive of Good Disease Responses in Patients with Mixed Chimerism Following Campath-Containing Reduced Intensity Transplants.

Following reduced intensity conditioned (RIC) transplants, donor leukocyte infusions (DLI) are frequently used either to convert mixed chimerism (MC) to full donor chimerism (FDC) or for residual or relapsed disease. Unfortunately, DLI are not universally successful and few factors are known (e. g disease type and level of pre-DLI chimerism) which predict for good responses. We analysed the impact of the chimerism pattern in 125 recipients of (CAMPATH containing) RIC transplants for malignant diseases. Of these, 68 (55%) had FDC (group A), 49 (39%) developed MC and 8 (6%) lost DC and had autologous reconstitution (group D). The patients who developed MC could be further subdivided into those with persisting MC post transplant (27, 55%; group B: non-responders) and MC post transplant with subsequently development of FDC (22, 45%; group C: responders). These two groups were analysed further. The median patient age was 55 (range: 19–71). The donors were siblings (22) or unrelated (27). The diseases were as follows: AML/MDS 14, CML 4, Myeloma, 4, lymphoma/CLL 26, MF 1. Stem cell source was PBSC (38) and bone marrow (11). Conditioning consisted of fludarabine, melphalan and campath (fmc) in 24 patients; fludarabine, busulphan and campath (fbc) in 5; BEAM, campath +/− fludarabine in 18 and FLAG in 2. There were no significant differences in any of these features between groups B and C. 25/49 patients received DLI. This was for disease relapse in10 patients, residual disease in 6 and MC alone in 8 (Unknown in 1). A complete disease response (CDR) was seen in 9/14 (64%) evaluable cases. There was a highly significant difference in CDR between the two groups (group B: 0/4, group C: 9/10, p=0.005). The reason for the difference in response rate was investigated. Median time to DLI was 196 days (range: 57–2123), not significantly different between the groups (p=0.561). The indication for and total number of DLI, the underlying disease and the degree of pre-DLI donor chimerism were not significantly different. In addition there was no significant difference in the incidence of post DLI GvHD (p=0.137), although this was 10/13, 77% in those who responded and 2/5, 40% in those who did not. Conversely, there was a significant difference in the pre-DLI lymphocyte counts (p=0.036). The median count was 2.24 × 109/l. In group B 3/11 (27%) were below this while 10/14 (71%) in group C were below this. The pre-DLI lymphocyte count was not significantly correlated with the time post transplant at which DLI was given, the type of donor, the indication for DLI or the disease, conditioning or post transplant immunosuppression regimen. The predicted overall survival at 2 years was significantly better in group C than in group B (95% versus 57%, p=0.002). This was largely due to the higher relapse risk in group B (77%) compared to group C (32%) (p=0.043). In conclusion, in patients with MC, the development of FDC was significantly associated with a superior OS. In those receiving DLI, the factor most significantly predictive for a ‘responsive’ (C) versus ‘non-responsive’ (B) pattern was the presence of a low pre-DLI lymphocyte count, suggesting that a lack of ‘space’ for expansion or increased suppressor cells in the lymphoid compartment mediate DLI resistance. We postulate that DLI ‘non-responders’ (those with higher lymphocyte counts) may be converted to ‘responders’ by the addition of pre-DLI lymphoreduction, thus reducing relapse and improving outcome.