Allo-HSCT can be burdened by life-threatening complications, being GvHD the major cause of morbility and mortality. Clinical and physio-pathological evidences showed that vascular endothelium could be a target of GvHD in very early phase; therefore markers of endothelial damage are warranted as valuable support in GvHD diagnosis. We conducted a study with primary endpoint to identify and count circulating endothelial cells (CEC) in peripheral blood of patients undergoing allo-HSCT as a function of endothelial damage.

The CellSearch System® is used to capture and enumerate CEC. Enriched and stained cells are dispensed into a MagNest® cartridge that is scanned and individual images of cells are scored as CEC, based on CD146+, CD105+, DAPI+ and CD45- phenotype. Patients undergoing allo-HSCT were tested before (T1), after the conditioning regimen (T2), at engraftment (T3), at GvHD onset (T4) and at 1 weeks after steroids treatment (T5). Ten healthy subjects served as controls.

We enrolled 40 patients with hematologic neoplastic diseases (7 HD, 13 AML, 5 ALL, 8 MM, 3 CLL, 1 NHL, 1 CML, 2 SAA) undergoing allo-HSCT from either HLA-matched familial (n=12) or unrelated donor (n=28). GvHD (grade I-IV) manifested in 19/39 patients. No clinical and transplant characteristics differences were present between patients with and without GvHD. The median CEC/ml pre allo-HSCT was 20 (n=40, range 4-718), in comparison to a value of 2 (range 1-14) in the 10 healthy subjects. At time of engraftment CEC/ml were 47 (range 16-148) in patients with GvHD and 92 (range 23-276) in patients without GvHD (P=0.006). This difference remained significant in multivariate analysis by logistic regression model (OR 0.97, 95% C.I. 0.96-0.99; P=0.02). At GvHD onset, the relative increase of CEC counts (T4 vs T3) was 44% (range, -43 - 569%) in GvHD patients versus 0% (range, -49 – 2%) in patients without GvHD (P=0.003), being confirmed in multivariate analysis (OR 1.04, 95% C.I. 1.0-1.08; P=0.04).

Circulating endothelial cells can represent a promising marker to monitor endothelial damage in patients undergoing allo-HSCT. The confirmation of the clinical utility of CEC counts in a larger series of patients, together with the use of a semi-automatic, standardized and reproducible technology, will allow a valuable help in the diagnostic definition of GvHD in early phase, and moreover could be a valid complement in the prognostic stratification of patients candidates to allo-HSCT.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.