Methods
Study design
This study targeted patients who underwent induction chemotherapy for ALL at the Department of Hematology/Oncology for Children and Adolescents, Sapporo Hokuyu Hospital, between April 2007 and December 2019. Children and AYAs were defined as < 15 years of age and ≥ 15 years of age, respectively. Analyses were conducted based on all data obtained up to February 31, 2020.
A total of 115 patients were treated at our institution according to the Japanese protocols for ALL, which are based on the Berlin-Frankfurt-Münster (BFM) regimen [19]. Two protocols were adopted to treat the patient with precursor B-cell ALL (BCP-ALL), and all of the enrolled patients underwent 7 days of prephase with steroid and one intrathecal dose of methotrexate, followed by the induction therapy. The patients then received 4 chemotherapeutic agents consisting of vincristine, daunorubicin, L-asparaginase, prednisolone, and intrathecal chemotherapy during the induction. In the protocols used until December 2012, the patients stratified into the intermediate-risk (IR) or high-risk (HR) groups were given an additional dose of cyclophosphamide (1,200 mg/m2/dose). Similarly, those who were classified into IR or HR after 2013 received two additional doses of daunorubicin (25 mg/m2/dose). The treatment regimens for IR and HR were the same in each protocol. As for the risk-stratifying method during the induction therapy, the two protocols in this analysis applied the same approach conventionally used in the BFM-typed regimen, in which patients with ALL were stratified according to National Cancer Institute criteria [20], cytogenetic abnormalities, response to the prephase and blast counts of bone marrow at day 15. In addition, patients with T-cell ALL and mixed-phenotype ALL were treated with the same regimen as the IR/HR protocol for BCP-ALL during the induction therapy. Details of the protocols are shown in Table I.
Of the enrolled patients, 5 infant patients with ALL, one patient with mature B-cell ALL, and one patient with Philadelphia chromosome-positive ALL who received different types of chemotherapy were excluded from this study. Furthermore, one patient with trisomy 21, one patient who started tyrosine kinase inhibitor (TKI) during induction therapy, 3 patients who had already been administered prednisolone at the time of hospitalization, and 4 patients who had not completed all scheduled anti-cancer agents for the induction were excluded. Additionally, two patients who died during induction chemotherapy and two patients who did not achieve complete remission before starting an early intensification therapy were also excluded. Therefore, a total of 95 patients were defined as eligible for the present analyses. Of these, 81 were children and 14 were AYAs. The median age of the children at diagnosis was 5.7 years old (1.3−14.7 years old), and that of the AYAs was 16.9 years old (15.1−24.0 years old). Further details regarding the characteristics of the patients are shown in Table II. No significant differences were found in the patients’ gender, immunophenotypes, or cytogenetic abnormalities except for WBC counts at diagnosis and M3 marrow at day 15.
The neutropenic episodes (neutrophil counts < 0.5 × 109/L) of the analyzed patients were retrospectively investigated, and the D-index was calculated from the duration and depth of the neutropenia. Then, the D-index and the duration of neutropenia in the induction course were then statistically compared between children and AYAs.
Finally, the incidence rates of FN, bacteremia, and IFI were surveyed in the eligible patients and were compared between children and AYAs. The incidence rate was defined as the number of newly identified infectious complications per whole number of accomplished induction chemotherapies. Blood cultures were always conducted whenever patients experienced FN at the institution, and any positive result from the cultures was considered to be bacteremia. IFI was defined according to the standardized definitions of the European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Disease Mycosis Study Groupconsensus group [21]. Antibacterial agents were not administered until the patients exhibited fever or some sign of infection, and granulocyte colony-stimulating factor was not used in all analyzed patients. In terms of the method to prevent infections, trimethoprim-sulfamethoxazole for Pneumocystis jiroveciipneumonia, and oral fluconazole at 10 mg/kg/day was used for all enrolled patients.
The present study was approved by the Institutional Review Board of Sapporo Hokuyu Hospital.
Definition and calculation of the D-index
The D-index is derived by plotting the absolute neutrophil count during neutropenia (< 0.5 × 109/L) and calculating the area over the curve, which is the difference of the expected neutrophil area (Ae) minus the area under the curve or A0. Ae is calculated as the product of 500 and the number of days during which neutrophil counts are at or below the level of neutropenia. For instance, if a patient experienced 6 days of neutropenia, Ae is 3,000 (6×500) days ・ neutrophils/μL. A0 is calculated using the trapezium rule, as follows:
\begin{equation} A_{0}=\sum_{\mathbf{i}=2}^{n}\begin{matrix}{(t}_{i}-t_{i-1})\frac{\begin{matrix}N_{i-1}+N_{i}\\ \end{matrix}}{2}\\ \end{matrix}\nonumber \\ \end{equation}
Here, (ti–ti-1) is the time interval (days) between two consecutive neutrophil counts, and Niand Ni-1 are the respective neutrophil counts (per microliter) at times ti and ti-1. All available neutrophil counts are included in the calculation of the D-index, where the total number of neutrophil counts is n[17].
Statistical analyses
Fisher’s exact test was used to compare categorical variables of the patients’ baseline characteristics, and the incidence of infectious complications. The Mann-Whitney U test was used to compare continuous variables of the D-index and the period of neutropenia. A p- value of 0.05 or less was considered statistically significant. All statistical analyses were performed with EZR, which is a graphical user interface for R (The R Foundation for Statistical Computing). More precisely, it is a modified version of the T commander designed to add statistical functions frequently used in biostatistics [22].