Discussion
Pediatric AML patients with hyperleukocytosis, usually refers to WBC counts greater than 100×109/L, have been reported to have a more dismal prognosis than patients without hyperleukocytosis.6-8 However, WBC counts below 100×109/L can also cause leukocytosis-related complications and affect prognosis. Hyperleukocytosis has been defined as more than 50×109/L in some studies in adults with AML.10-12 Hyperleukocytosis has had no definitive criteria heretofore. To the best of our knowledge, this study was the first report the characteristics and prognosis of hyperleukocytosis (described as WBC count ≥50×109/L) in pediatric AML patients. Moreover, this study was performed to further compare the clinical characteristics and prognoses of childhood AML patients with hyperleukocytosis with WBC count 50-100×109/L and those with WBC count ≥100×109/L in a single institution.
In this study, we defined a WBC count greater than 50×109/L as the cut point for hyperleukocytosis. The incidence of hyperleukocytosis in childhood AML patients at diagnosis was 26.38% in our center, which is similar to published data .10-12 In this study, patients with and without hyperleukocytosis had similar CR rates, and yet patients with hyperleukocytosis had a higher mortality rate than those without hyperleukocytosis. We further confirmed that the 10-year PFS and OS rates (P =.041, P =.051) between groups separated with this cutoff were significantly different at our institution. PFS and OS were similar within the subgroups. This fact indicates that a high leukocyte count (WBC count ≥ 50×109/L) is a significant marker of poor prognosis. Tien et al. observed that WBC counts significantly affected OS and pointed out that hyperleukocytosis was an independent poor prognostic factor.11 Genetic characteristics and clinical treatment response as risk stratifications changed the clinical prognosis. Hence, WBC count ≥ 50×109/L should be classified as one of the risk stratification factors for pediatric AML.
Patients with AML-M5 often present with hyperleukocytosis, and patients with AML-M5 may experience serious symptoms of hyperleukocytosis.17In this study, half of patients had the FAB M5 subtypes in patients with AML and hyperleukocytosis. In all patients with hyperleukocytosis, FAB M5 subtype patients had a significantly inferior survival. Monocytic leukemia cells have more large-volume active lysosomes.18During chemotherapy, leukemia cells destroy and release a large number of lysosomes, which caused coagulopathy, metabolic disturbance, and even DIC and tumor lysis syndrome. It may be the main reason for the high mortality of M5 with hyperleukocytosis.19This may be one of the reasons for the increased mortality of children with AML and hyperleukocytosis.
Moreover, Male predominance was noted in a multicentric study in Brazil,20 which was consistent with our data. In accord with our findings, pediatric AML and hyperleukocytosis were more males, who had significantly poorer 10-year PFS and OS rates than females. More than half of the AML-M5 patients experienced testicular relapse without previous systemic relapse.21 Poor prognosis may be related to predominant acute monoblastic leukemia, extramedullary infiltration, and the blood-testis barrier as a refuge for leukemia cells. All of the above factors may lead to unsatisfactory chemotherapy effects, relapse and high mortality in male pediatric AML patients with hyperleukocytosis.
AML withRUNX1-RUNX1T1((AML1/ETO)or core-binding factor subunit beta- myosin heavy chain 11(CBFβ-MYH11) has been considered a unique entity, which is collectively referred to as CBF-AML.22Most treatment protocols classify all children with CBF-AML as having a low-risk disease.5 However, CBF-AML has a very high degree of clinical and biological heterogeneity.23 WBC count, platelet count and cytogenetic were significant prognostic variables in CBF-AML.24 In our study, CBF-AML had a low proportion of hyperleukocytosis in childhood AML patients. Moreover, hyperleukocytosis with CBF-AML had a significantly higher PFS and OS rate than hyperleukocytosis without CBF-AML. This statistical difference may also be due to relatively small sample size. More sample or experiments are needed to further confirm these findings.
The incidence of FLT3-ITD was 8.14% (25 out of 307 patients) in non-M3 pediatric AML in our study, which was consistent with previous reports.25 Our data suggested that the incidence ofFLT3-ITDmutations (n=16, 19.75%) was higher in AML patients with hyperleukocytosis than without hyperleukocytosis, and FLT3-ITD mutations had a high probability of developing hyperleukocytosis in childhood AML patients. These features were unanimous in WBC count above 100×109/L AML patients in the previous study.8 The conformational changes in the juxtamembrane domain of the FLT3 receptor caused by ITD mutations activateFLT3-ITDreceptor tyrosine kinases, which causes proliferation, inhibits apoptosis and suppresses differentiation.26,27 This may be the cause of leukocytosis in FLT3-ITD patients. In addition, FLT3-ITD mutation was an independent risk factor for poor outcome in pediatric acute myeloid leukemia.28Therefore, It was expected to improve the long-term survival of these patients by promising FLT inhibitors, Midostaurin and Sorafenib, combined with sequential chemotherapy or as maintenance after HSCT.29
In conclusion, more than 20% of AML patients have a WBC count greater than 50×109/L, which is defined as hyperleukocytosis in our study. Poor prognosis in terms of 10-year PFS and OS rates indicated that hyperleukocytosis was a critical predictive adverse factor in pediatric AML. In all patients with hyperleukocytosis, male and FAB M5 subtype patients had a significantly inferior survival, and the prognosis of CBF-AML with hyperleukocytosis was good. To exploit more accurate treatment strategies, a more extensive range of WBC counts is needed to define hyperleukocytosis for stratify risk in children with AML.