Cerebral Thromboses during Treatment of Childhood Acute Lymphoblastic Leukaemia.

Thromboembolic events are currently among the most serious complications occurring during treatment of acute lymphoblastic leukaemia (ALL), and account for cases of morbidity and even mortality. The role of L-Asparaginase, particularly that of E. coli L-Asparaginase (L-ASPA) in association or not with corticotherapy has been demonstrated by several authors. Most of the time, these thromboses are located within the central nervous system. Our retrospective study was carried out from December 2000 to December 2004 within 18 French centres. 902 children were prospectively included in the FRALLE 2000 protocol and were assigned to two risk groups: the A group with standard risk ALL (n=502) received dexamethasone during induction and 9 infusions of 6000 U/m² of L-ASPA, while the B group with high risk ALL (n=400) was to receive prednisone. Both groups received a delayed intensification with dex and L-ASPA. 27 thrombotic events were reported, 16 of which were cerebrovascular accidents. The thromboses lay in the superior sagittal sinus (n=14), in the lateral sinus (n=7) and in the sinus rectus (n=2) and were revealed by convulsions in 30,7% of the cases and by diffuse neurologic signs in 68,7% of the cases. The course turned out positively 15 times out of 16, but one child died of haemorrhage. Hereditary thrombophilic factors were evaluable in 15 patients out of 16. We found out a congenital Protein C deficiency in one patient, but no Protein S or anti-thrombin deficiencies. We found out two Factor V G1691A heterozygous mutations and two Factor II G20210A heterozygous mutations. Five patients out of 14 have a MTHFR C677T mutation: one is homozygous in combination with a Factor V G1691A mutation, and 4 are heterozygous, isolated in two cases, in combination with a Factor V G1691A mutation in one case, and in combination with a Factor II G20210A in the other. Hereditary thrombophilic factors (isolated heterozygous MTHFR C677T mutations excepted) were thus reported in 31,2% (n = 5) of the cases.

Conclusion: even if the frequency of thrombophilia may be underestimated in this study due to its retrospective design and non exhaustivity (APL antibodies, lipoprotein A were not investigated), it nevertheless points out CNS thrombosis represents an important cause of morbidity in childhood ALL treated with high dose steroid and L-ASPA. A prospective analysis of all thrombophilic factors seems justified in this setting for prevention possibly based on LMWH.