Safety of Lumbar Puncture (LP) and Instillation of Intrathecal (IT) Chemotherapy during Thrombocytopenia in Adults with Acute Lymphocytic Leukemia (ALL).

Background: The platelet (PLT) count at which LPs can safely be performed in adults has varied in practice. A few pediatric studies have suggested a threshold level of 10 K/μL above which prophylactic PLT transfusion is not required; however, this has not been systematically studied in adults with ALL. We therefore investigated the incidence of traumatic LPs (TLP >10 RBCs/μL), the incidence of complications, and the relationship to thrombocytopenia.

Methods: Adults with ALL undergoing frontline therapy with hyper-CVAD (fractionated cyclophosphamide, vincristine, doxorubicin, dexamethasone alternating with methotrexate and cytarabine) or variations incorporating rituximab and/or imatinib were reviewed. Patients (pts) were stratified to receive standard intrathecal treatment (IT) with methotrexate alternating with cytarabine based on CNS risk assessment (proliferative index ≥ 14% or serum lactate dehydrogenase ≥ 1400) [Cortes et al, Blood 86:2091, 1995]. Low risk pts received 6 IT, high or indeterminate risk 8 IT, and BLL 16 IT. Pts with PLT count ≤ 30 K/μL received PLT transfusions immediately prior to LP. Procedures were performed by trained practitioners.

Results: Outcome of 1998 LPs (282 pts) performed at our institution was reviewed. Median age was 42 years (range, 15–84). PLT count was ≤ 30 K/μL in 292 LPs (15%), 31–50 K/μL in 157 (8%), 51–80 K/μL in 211 (11%) and > 80 K/μL in 1338 (67%). Criteria for TLP was met in 910 (45%) of the LPs. The incidence of TLP was higher in patients with PLT count ≤ 50 K/μL (52% versus 44%, p=0.004). Four pts developed subdural hematomas. Of those, 3 had a history of CNS trauma, with only 1 considered possibly related to the LP. The latter occurred with PLT count of 28 K/μL prior to transfusion. Three pts developed diskitis or epidural abscesses after 3, 6 and 9 LPs, respectively, and were treated successfully with antibiotics.

Conclusions: Although the incidence of TLP appears to be higher with a lower PLT, it appears safe to perform LPs in thrombocytopenic pts with a threshold of 30 K/μL for prophylactic PLT transfusion as per our standard practice. However, this study does not address whether a lower threshold for PLT count (e.g., 20 K/μL) would yield similar findings.