Molecular or cytogenetic monitoring and preemptive therapy for central nervous system relapse of acute promyelocytic leukemia

We read with great interest the article by Montesinos et al.1 concerning central nervous system (CNS) relapse of acute promyelocytic leukemia (APL). They reported a low incidence of CNS involvement at first relapse in APL patients following therapy without CNS prophylaxis.1 The optimal management of APL relapse in CNS has taken on increasing significance.2 Here we report our experience concerning CNS relapse of APL and introduce a new approach with molecular or cytogenetic monitoring in cerebrospinal fluid (CSF) as contrasted with the observations by Montesinos et al.1

Since 2005, we experience of 6 patients with first relapse of APL at The University of Tokyo Hospital. These patients received different first-line therapies with or without prophylactic intrathecal chemotherapy (IT) and high-dose cytarabine (HDAraC), which are effective for CNS leukemia. All these first-line therapies included all-trans retinoic acid (ATRA) and anthracycline. Patients’ characteristics are shown in Table 1. The patients with relapsed APL received arsenic trioxide and prophylactic IT except case 1, who received ATRA, HDAraC, and prophylactic IT. Autologous hematopoietic stem cell transplantation was performed after achieving molecular complete remission.

Table 1.Characteristics of patients with first relapse of acute promyelocytic leukemia at our institute.

Early detection of CNS invasion of malignant cells by polymerase chain reaction (PCR) or fluorescence in situ hybridization (FISH) has been adopted in some cases with CNS lymphoma or acute lymphoblastic leukemia.3–5 Although an article reported that the introduction of routine lumbar puncture detected subclinical CNS leukemia in some newly diagnosed APL cases,6 monitoring by PCR or FISH in CSF of APL patients has never been described. We evaluated the PCR and/or FISH monitoring in cases 3–6 (Table 1). None of the 6 cases showed CNS involvement at the moment of first relapse and all the cases initially had isolated hematologic or molecular bone marrow relapse. However, 4 out of 6 cases subsequently showed overt or subclinical CNS relapse during salvage therapy for first relapse of APL. All the cases with CNS relapse were with intermediate- or high-risk score,7 which was consistent with the observations by Montesinos et al.1 Case 5 showed simultaneous appearance of CNS involvement of APL and subarachnoid hemorrhage, as we described previously.8 He received therapeutic IT repeatedly until all markers of CNS relapse, including cytology, PCR, and FISH, turned negative. Surprisingly, in the other 3 cases, we detected subclinical CNS relapse at the moment of routine prophylactic IT. For these patients, we provided preemptive IT or HDAraC repeatedly until the markers of CNS relapse turned negative. Although cytological examination detected subclinical CNS relapse in case 1, cytology remained negative in cases 4 and 6 when PCR and/or FISH markers turned positive. In addition, because cytology turned negative earlier than PCR and FISH markers in case 5, these novel monitoring methods were also useful for determining when to finish therapy for overt CNS relapse. Therefore, PCR and/or FISH were more sensitive than conventional cytological monitoring for detecting CNS relapse and residual disease during the preemptive therapy. On the other hand, it has been mentioned that the PCR approach in CSF has some pitfalls.3,9 It is sometimes difficult to obtain sufficient DNA for PCR when cell counts are low, and false-positive results may be obtained owing to contamination with blood in the lumbar puncture procedure.3,9 In fact, we could not obtain sufficient DNA for PCR in some CSF samples (Table 1). It is noteworthy that FISH analysis was positive in spite of negative PCR result in some samples of cases 4 and 5. This indicates that PCR results were false-negative because of insufficient material, suggesting superiority of FISH to PCR in monitoring CNS relapse in CSF samples when cell counts are low. False-positive results owing to contamination with blood were not suggested in our PCR and FISH analyses.

Montesinos et al.1 did not mention whether their patients with isolated first bone marrow relapse experienced subsequent CNS relapse.

We consider this information to be indispensable because our observations indicated that subsequent CNS relapses during salvage therapy of relapsed APL are highly common, even if an incidence of CNS involvement at the moment of first relapse is low. We prevented and managed overt CNS relapses by adopting molecular monitoring and prophylactic or preemptive IT. Because overt CNS relapse can cause neurological symptoms and is difficult to cure, early detection is important. Therefore, we believe that these strategies should be considered for patients with relapsed APL, as contrasted with the undetermined role of CNS prophylaxis for patients with newly diagnosed APL.