Page 45 - 2021_06-Haematologica-web

Monocytic switch in pediatric BCP-ALL subtypes
types, although KMT2Ar leukemias can present with a mixed phenotype. However, due to the heterogeneity of KMT2Ar leukemias and the limitations of our cohort, the correlation of monocytic switch with specific KMT2Ar sub- types should be verified in larger cohorts.
The results from the multidimensional analyses of the gene expression profiles (GEP) and immunophenotypes showed that the patients with monocytic switch coclus- tered based on genotype. Similarly, Alexander et al.42 recent- ly showed that gene expression-based clustering primarily distinguished genetic subtypes irrespective of MPAL (mixed phenotype acute leukemia) status. Together, these findings show that monocytic switch is a behavior with varying propensity across BCP-ALL genotypes rather than being exclusive to a distinct genetic subtype of leukemia.
Similarly, switching cannot be predicted using a diagnos- tic immunophenotype according to our data. Nevertheless, immunophenotype can help to predict genotype, as was shown previously and extended in this study.39,40,43,44
Monocytic switch may lead to uncertainty about the continuation of ALL-type therapy. In some patients the phenomenon was very discreet and might be overlooked using routine examination. On the contrary, we observed monocytosis at d+8 of as high as 20,804/mL, 3,969/mL and 15,544/mL for the DUX4r, PAX5-P80R and ZNF384r sub- types, respectively. Although such findings may trigger thoughts of changing to the AML type of treatment, all but one patient with detected monocytic switch achieved complete remission on an ALL type of treatment. One patient died during induction therapy. The EFS for an ALL type of therapy was identical among patients with and without monocytic switch. However, we did observe two patients (one with the DUX4r subtype and one with the KMT2Ar subtype) who relapsed with monocytic AML, showing that the optimal treatment of such rare patients has yet to be determined. Of note, the majority of the published cases of monocytic relapse after primary BCP- ALL had the KMT2Ar genotype.45 In addition to these AML relapses, we discovered that six patients with detectable monocytic switch relapsed later with BCP-ALL. According to published data, the prognosis of patients with the DUX4r or the PAX5-P80R subtype does not seem to be unfavorable.3,38
The clinical significance of switched monocytoids, per se, remains unknown. To date, we have limited evidence about their potential to initiate relapse. Functional tests of
Figure 4. Event-free survival of B-cell precursor acute lymphoblastic leukemia patients with and without switch. The prospective cohort (n=725) is shown. One patient was lost to follow-up.
those switched cells so far have not been performed. Interestingly, we observed a rapidly enlarging spleen in one patient and progressive liver failure in another patient, most likely caused by infiltrating macrophages, which had iden- tical IG/TR rearrangements as the original malignant B pre- cursors.
Despite several observed features of switched monocy- toids in DUX4r- and PAX5-P80R-mutated patients (e.g., CD45RA and CD2 positivity in DUX4r, CD13 negativity and CD66c positivity in PAX5-P80R), their interpretation (especially at time points with myeloid regeneration, including time point d+33) is challenging.
Monocytic switch not only creates discordance between MRD levels determined by FC and MRD levels determined by PCR but also affects the availability of CD19 as a thera- peutic target. In some current pediatric treatment protocols, these patients can be stratified for anti-CD19 treatment regardless of their CD19 expression levels. Thus far, data are limited regarding the efficacy of such treatment in patients with monocytic switch. Recently, the first case report of myeloid relapse in BCP-ALL patients with a ZNF384r subtype after CAR-T therapy was published.46 However, myeloid relapse in ZNF384r patients without tar- geted therapy has also been described.47
In conclusion, we report the frequency of monocytic switch in novel genetic subtypes of BCP-ALL and highlight the discordance between MRD levels determined by FC and PCR during the switch. New markers for discriminat- ing switched monocytoid blasts from nonmalignant mono- cytes are needed to overcome FC underestimation of MRD levels, which is becoming more relevant with the use of tar- geted anti-CD19 therapy.
Disclosures
No conflicts of interest to disclose
Contributions
MN analyzed immunophenotypic data, evaluated FC MRD and wrote the manuscript; MZ was responsible for whole-exome and RNA-seq data and wrote the manuscript; KF analyzed RNA-seq data and wrote the manuscript; BV analyzed immunophenotypic data and evaluated FC MRD; LS performed PCR MRD on sorted samples, evaluated PCR MRD levels and analyzed immunophenotypic data; AM analyzed RNA-seq and immunophenotypic data; MB and MR performed the investiga- tion in German patients; EF was responsible for PCR MRD
haematologica | 2021; 106(8)
2073