In 2020, both Li et al. (2020) and Sentís et al. (2020) performed WGS which both focused on the somatic alterations observed during diagnosis and relapsed stage for both B-ALL and T-ALL patients. SNVs, indels, copy number variations (CNVs), and structural variations (SVs) were identified. Li et al. (2020) reported that 12 genes were enriched in relapse-specific alterations, including 11 known relapse-related genes. Seven of the 12 genes showed relapse-specific alterations in both B-ALL and T-ALL, whereas PRPS1, MSH2, FPGS, CREBBP, and WHSC1 alterations were exclusively observed in B-ALL. Sentís et al. (2020) compared the mutational profiles of T-ALL and B-ALL patients of varying ages and found no significant difference. However, slight variations were observed between pediatric and adult malignancies. The study also revealed that NOTCH1 and FBXW7 were overrepresented in both pediatric and adult T-ALL as compared to B-ALL (Sentís et al., 2020).

Waanders studied the genomic patterns and mutational pathways that cause relapse in pediatric patients with relapsed B-ALL and T-ALL (Waanders et al., 2020). Fifty genes were enriched as mutational targets in relapsed diseases. Epigenetic regulators such as PRDM2, PHF19, TET3, and SIN3A, were also enriched in the relapse samples. Mutations observed in different gene regulation pathways showed different frequencies of relapse-enriched genes between B- and T-ALL, implying that distinct biological mechanisms drive the genetic alterations responsible for disease progression (Waanders et al., 2020). Oshima et al. (2020) analyzed paired diagnostic and relapse samples from children and adults with B precursor and T-ALL. They found an average of 27–34 coding mutations in diagnostic and relapse samples, respectively. Mutational signatures associated with microsatellite instability were observed along with significant associations between certain mutations at diagnosis and relapse. In T-ALL, they observed a significant co-occurrence of JAK1 and JAK3 mutations (Oshima et al., 2020), which was similarly reported in a previous study (Sentís et al., 2020), and an association between WT1 and NRAS mutations, and between JAK1 and WHSC1 mutations at relapse (Oshima et al., 2020). In B precursor ALL, a significant association was observed between SETD2 mutations and ETV6 deletions and between NRAS and CREBBP mutations at relapse (Oshima et al., 2020).

Antic et al. performed WGS on 10 samples from 2 pediatric BCP-ALL patients with multiple relapses (Antic et al., 2021). They detected 8,922 and 8,759 single-base substitutions and 686 and 646 indels in patients 1 and 2, respectively. Signatures for single-base substitutions (SBS)2 and SBS13 were established at diagnosis and persisted throughout disease progression (Antić et al., 2022). Sayyab et al. (2021) studied eight subtypes of BCP-ALL, the B-other subgroup, and T-ALL. They reported that somatic mutations differed between subtypes and individual samples. Nine ALL-driver genes had recurrent protein-coding mutations. In addition, copy-number aberrations and deletions of 9p21.3 remained high from diagnosis to relapse (Sayyab et al., 2021).

In summary, several investigations utilized WGS and analyzed samples obtained during diagnosis and relapse stages to identify somatic changes in different subtypes of ALL. The results of these studies varied depending on their objectives, with different somatic mutational patterns observed in pediatric and adult patients with ALL based on the specific subtype of the disease.

In 2017, Vesely et al. (2017) analyzed WES data from 41 patients with leukemic BCP-ALL and found subclonal and unstable JAK/STAT- and RTK/Ras pathway-activating mutations in 76% of cases at diagnosis and almost all relapses. Because the P2RY8–CRLF2 fusion, a JAK/STAT- or RTK/Ras pathway mutation, or all three, was lost at relapse (Vesely et al., 2017). Ding et al. (2017) later performed WES and TS in 11 trios of B-ALL patients with diagnoses, full remissions, and relapses. They found the RTK-Ras signaling pathway, epigenetic regulators, transcriptional factors, and p53/cell cycle pathway were the most significantly mutated genes throughout the entire cohort (Ding et al., 2017).

Kimura et al. examined 30 cases of T-ALL, encompassing both initial diagnosis and relapse, through WES (Kimura et al., 2019). Their findings indicated a higher mutation rate in relapsed samples than in those at the time of diagnosis. At the outset, NOTCH1, FBXW7, DNM2, and PHF6 mutations were common, with NOTCH1 mutations continuing to appear frequently during relapse. PEST-rich alterations were more prevalent in relapsed cases than in non-relapsed cases at diagnosis (Kimura et al., 2019). Similarly, Hoell et al. (2019) studied leukemic blasts from 10 children with post-allogeneic stem cell transplantation (SCT) relapses and discovered that NOTCH1 was the sole recurrent gene among the 50 genes in T-cell leukemia. In their study using targeted exome sequencing, Zhang et al. (2020) observed distinct mutation patterns between B-ALL and T-ALL patients. KRAS was most frequently mutated in B-ALL, whereas T-ALL exhibited enrichment for NOTCH1, FBXW7, PHF6, and PTEN mutations (Zhang et al., 2020), which is consistent with the findings of Ding et al. (2017). Both B-ALL and T-ALL frequently show mutations in the Ras and Notch pathways (Zhang et al., 2020).

In 2020, Yu et al. (2020) employed WES and examined 299 diagnostic and 73 relapse samples from 372 ALL patients to assess the incidence of FPGS mutations, as well as those in two crucial thiopurine pathway genes, NT5C2 and PRPS1 which have been previously demonstrated to be important mechanisms for resistance in leukemia (Schroeder et al., 2019). Three additional FPGS mutations were detected in two patients, NT5C2 mutations in six, PRPS1 mutations in two, and both NT5C2 and PRPS1 mutations in one. According to Yu et al. (2020), three children with relapsed ALL had four acquired relapse-specific mutations in FPGS. Unexpectedly, Schroeder et al. (2019) and Li et al. (2020) identified a new relapse-specific mutation in the FPGS gene that solely affected B-ALL and exhibited relapse-specific lesions and stable losses. In addition, similar investigations concluded that most of these changes were induced by a subclone or were acquired during relapse (Schroeder et al., 2019; Li et al., 2020). Therefore, FPGS relapse-specific mutations constitute a pharmacogenetic pathway associated with pediatric ALL relapse and should be regarded as a factor of relapse in pediatric ALL (Yu et al., 2020).

In a study conducted by Oshima et al. (2020), WES was performed on matched germline diagnosis and relapse DNA samples. This study revealed diagnostic and relapse-specific mutational mechanisms, along with genetic chemoresistance drivers. Somatic copy number variations (CNVs) were identified, with an average of 18 somatic CNVs per sample for 6,475 alterations in the series. Of these, 3,589 CNVs were detected at the time of diagnosis and 2,876 at the time of relapse, with 2,575 variants present in both the diagnostic and relapsed samples (Oshima et al., 2020).

In 2021, deep sequencing validation of all detected mutations was performed by Antic et al. (2021) after conducting genomic analysis of diagnosis–relapse paired samples of patients who relapsed very early. Two patients with TCF3–PBX1-positive leukemia who experienced very early relapse had E1099K WHSC1 mutations at the time of diagnosis. This hotspot mutation has also been frequently detected in other cases of very early TCF3–PBX1-positive leukemia relapses. This finding suggests that minor subclones at diagnosis typically cause early relapses in BCP-ALL (Antic et al., 2021).

Shirai et al. reported that, in this investigation, KMT2D mutations and 6q LOH were found to be recurring changes in seven patients with TCF3–PBX1-positive B-LBL (Shirai et al., 2022). Additional genetic mutations were found in the relapsed tumor that were not present in the primary tumor, including the 6q LOH. Analysis of recurrent cases revealed that the relapsed clone may have originated from a minor BM clone at the time of diagnosis (Shirai et al., 2022). According to Zou et al. (2022)’s analysis of data from patients with BM relapse, 62.5% of patients with BM relapse had abnormalities discovered using NGS, and uncommon or previously unreported fusion genes and/or gene mutations were detected. Patients with adverse molecular genetic mutations, such as TP53, CREBBP, and IKZF1 had their BMs relapsed (Forero-Castro et al., 2017; Montaño et al., 2020).

As an outcome, using all these findings, this pharmacogenetic knowledge and research can be very helpful in situations with limited resources for identifying the requirements of patients based on their genetic risk profiles; this situation is also being resolved. Hence, Figure 1 presents the genes involved in chemotherapy resistance that cause relapse and play various roles.

In 2017, Vesely et al. (2017) performed RNA-seq and discovered that IKZF1 mutations result in distinct transcriptional signatures. They studied the RNA-seq data based on IKZF1 status, as IKZF1 alterations are strongly associated with relapses in P2RY8–CRLF2-positive ALL cases. The 200 genes with the highest differential expression in the IKN and IKD groups were clustered by IKZF1 status using unsupervised hierarchical clustering. This observation was supported by secondary RNA-seq data collection of 20 B-other ALL cases with known IKZF1 status (Vesely et al., 2017).

In 2020, Waanders et al. (2020) performed RNA-seq on 115 samples made up of TRIzol-extracted RNA samples from 66 patients. They discovered that the expected human leukocyte antigen (HLA)--binding mutant peptide count per tumor increased with disease progression due to a higher mutation burden, and hence, notably in hypermutated samples (Waanders et al., 2020). With increased neoepitope burden and possible immunotherapy vulnerability, a subset of leukemia predisposed to recurring relapse exhibits hypermutation caused by at least three different mutational mechanisms (Waanders et al., 2020).

Oshima et al. conducted chimeric gene transcript analyses using RNA-seq data for 85 cases (Oshima et al., 2020). These analyses revealed the existence of oncogenic gene transcripts as well as a possible initiating rearrangement between TBL1XR1 and JAK2. Despite the exception of two patients who tested positive only at relapse, one for PICALM–MLLT10 and the other for NUP214–ABL, fusion oncogene transcripts were typically observed at both diagnosis and relapse (Oshima et al., 2020).

In 2021, Sayyab et al. (2021) performed RNA-seq and identified genome-wide patterns of somatic mutations in pediatric Nordic ALL patients. RNA-seq libraries were constructed using the RNA from 22 ALL patients who were diagnosed, 25 patients who experienced their first relapse, and 7 patients who experienced their second relapse, as well as the control B cells (CD19⁺) and T cells (CD3⁺) from five Swedish healthy individuals. RNA-seq data showed evidence of the involvement of exons in gene fusion (Sayyab et al., 2021).

Zou et al. (2022) studied pediatric B-ALL patients without specific fusion genes for outcomes and risk factors in three Chinese institutes. Analysis of the RNA-seq data revealed molecular abnormalities in six patients and reversed molecular abnormalities in three (CREBBP, TP53, and P2RY-CRLF2), as well as previously unreported genetic mutations in B-ALL (Zou et al., 2022).

In terms of transcriptional characteristics including newly transcribed regions, RNA editing, allele-specific expression, variant splicing, RNA sequencing, or RNA-seq offers a previously unheard-of accuracy. Owing to these advancements, the transcriptome and its implications for fundamental biology and personalized medicine may now be studied using a powerful toolkit.