Background and objetives: Acute graft versus host disease (aGVHD) represents one of the main causes of death in patients who receive allogeneic stem cells transplantation (SCT). Quantification of hematopoietic chimerism by real-time PCR allows the detection of residual receptor cells with a high sensitivity (0.01%). Since aGVHD is supposed to be produced by reactivity of donor cells against host cells, this study was designed to relate the appearance of aGVHD with the amount of receptor cells remaining early after SCT.

Patients and methods: Since January 2003, twenty seven patients who underwent myeloablative allogeneic SCT for acute leukemia (48% myeloid) could be included in the study (a minimum of 100 days follow-up was required). A sibling donor was used in 18 cases (67%) and bone marrow was the source of stem cells in 15 transplants (only in 2 cases the source was umbilical cord blood whereas the remaining were collected from peripheral blood). Conditioning with total body irradiation (TBI) was performed in 9 cases (33%). All the patients received the same GVHD prophylaxis with cyclosporine and methotrexate. Chimerism was determined by quantitative real-time PCR amplification of null alleles or insertion/deletion polymorphisms as previously described (

Leukemia
2005
;
19
:
336
–43
). Granulocytic engraftment, defined by the presence of more than 0.5 ×10e9/L granulocytes in two consecutive determinations, appeared at a median of 14 days (11–27) in our series. At least one chimerism sample had to be collected in the next week after granulocytic engraftment. Risk factors related to aGVHD were recorded and analyzed using Chi square for qualitative variables and non parametric tests for quantitative variables. Multivariate analysis was performed using Cox regression.

Results: Eleven patients (40%) presented aGVHD (grade II–IV) in our series. Median receptor chimerism at the time of engraftment was 0.27% (0.01–7), showing significantly lower levels in patients who developed aGVHD (0.10 vs 0.57%, p=0.001 Mann-Whitney test). A chimerism value of 0.1% at the time of granulocytic engraftment was used to classify patients. COR curves were plotted previously for a better selection of the cut-off point. At 0.1% or lower, 8 of 9 patients (89%) developed aGVHD, whereas only 3 of 18 patients above this value presented aGVHD signs (p<0.001, Chi-square test). In those patients who implanted before aGVHD onset (7 of 11, 64%), median time between chimerism determination and clinical appearance of aGVHD was 8 days (1–31). Univariant analysis showed only 3 significant variables in relation to the development of aGVHD: TBI conditioning (p=0.029, log rank test), unrelated donor (p=0.013) and low level of receptor chimerism (p=0.0002). In the multivariate study, only chimerism remained as a significant variable (p<0.001, Cox regression).

Interpretation and conclusions: Our data show that early monitoring of chimerism may be useful for identification of patients at risk of developing aGVHD after SCT.

With the quantitative method used, a low chimerism receptor level at the time of engraftment (0.1% or lower) predicts the onset of aGVHD and can anticipate it in more than one week. This could allow the use of measures destined to reduce aGVHD in higher risk patients.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution