226 Chapter 7 Patient data and outcomes GR and MR expression levels of 278 ALL patients were available and plotted based on ALL subtype. For these samples, mRNA sequencing (RNA-seq) was performed as previously described.13 In brief, total RNA was isolated using the AllPrep DNA/RNA/Protein Mini Kit (QIAGEN) according to standard protocol on the QiaCube (Qiagen). RNA-seq libraries were generated with 300ng RNA using the KAPA RNA HyperPrep Kit with RiboErase (Roche) and subsequently sequenced on a NovaSeq 6000 system (2x150 bp) (Illumina). The raw sequencing reads were aligned using STAR (version 2.7.0f) to GRCh38 and gencode version 29.13 Finally, expression counts were determined at gene level using Subread Counts.14 Outcome data regarding survival and relapse was available on 131 of the 278 patients. This comprised a subgroup of patients treated according to the Dutch Childhood Oncology Group ALL-11 protocol. Patients and/or patients’ parents or legal guardians provided informed consent to use clinical data and leftover diagnostic material for research, compliant with the biobanking procedure of the Princess Máxima Center (MEC-2016-739, Netherlands Trial Register (NTR) NL7744) and the Declaration of Helsinki. Statistical analyses The association between MR and GR expression levels and prednisone response on day 8 was estimated with a multivariable logistic regression model. MR and GR expression levels were categorized as low or high, with the median as cut-off value, as described before.15 We included National Cancer Institute (NCI) risk category (standard risk: <10 years and white blood cell count at diagnosis <50x109/L, high risk: other)16 and sex as covariates. Event free survival (EFS), defined as time since diagnosis to the occurrence of induction failure, relapse, secondary malignancy or death in complete remission as event of interest, was estimated with Kaplan Meier’s methodology.17 Median follow up was estimated by using reverse Kaplan Meiers.18 To study the effect of prognostic factors on EFS, Cox proportional hazard models were estimated. First, univariable analyses were performed which included patient, disease and treatment characteristics. Statistically significant variables were entered with MR or GR expression levels in a multivariable Cox model. Landmark analysis (with land mark points day 33 or 79) was employed to study the effect of minimal residual disease (MRD) on survival outcomes.19 Differences with a p-value <0.05 were considered statistically significant. All analyses were performed with IBM SPSS Statistics version 26.0.
RkJQdWJsaXNoZXIy MTk4NDMw