Dear Editor,

Targeted RNA-sequencing (RNA-seq) using next-generation sequencing (NGS) technology is a highly accurate method for selecting and sequencing specific transcripts of interest [1]. We routinely applied a customized targeted RNA-seq system during the diagnostic phase of hematologic malignancies. Our system detected the first Korean case of NUP98-NSD1 and a novel SNRK-ETV6 fusion with therapy-related acute myeloid leukemia (t-AML) showing a dismal clinical course. NUP98-NSD1 accounts for approximately 4% of pediatric AML cases and shows a poor prognosis [2, 3]. It could be created by a cryptic t(5;11)(q35;p15.5) and exerts a leukemogenic function by binding near the HOX locus and MEIS1 to increase expression via histone modifications [4]. The Institutional Review Board of Chonnam National University Hwasun Hospital (CNUHH), Hwasun, Korea (CNUHH-2020-091) approved this study and granted a waiver of consent due to its retrospective nature. This report highlights the role of high-throughput parallel targeted RNA-seq in enhancing the diagnostic yield of hematologic malignancies.

In April 2020, a 14-year-old girl visited the outpatient clinic of CNUHH 1.5 years and 1.9 years after a matched unrelated peripheral blood stem cell transplantation and initial diagnosis of AML, respectively, for a follow-up bone marrow (BM) examination At initial diagnosis, the Korean AML 2012 regimen (double-induction strategy with idarubicin or mitoxantrone plus cytarabine, followed by consolidation therapy with cytarabine and etoposide) was administered and complete remission was achieved 28 days after the second induction. The laboratory findings showed a leukocyte count of 3.1×10⁹/L, absolute neutrophil count of 0.58×10⁹/L, hemoglobin of 114 g/L, and platelet count of 37×10⁹/L. BM aspirates revealed 28% leukemic blasts corresponding to French-American-British (FAB) type M2. The BM karyotype was 45,XX,add(3)(p25),del(5)(q?),-12,add(12)(p13)[8]//46,XY[12], and the multiplex reverse transcription (RT)-PCR (HemaVision kit; DNA Technology, Aarhus, Denmark) finding was negative.

Targeted RNA-seq (HEMEaccuTest RNA; NGeneBio, Seoul, Korea) of the BM sample using STAR-Fusion (ver 1.8.1) and FusionCatcher (ver 1.20) revealed NUP98-NSD1 and a novel SNRK-ETV6 fusion, which were confirmed by direct sequencing (Fig. 1). DESeq2 (ver 1.18.1) analysis showed that WT1, ERG, and BAALC expression increased 7.1, 5.6, and 4.1-log₂-fold, respectively, compared with 14 normal controls (Table 1). An additional tier II variant of WT1, NM_024426.3:c.1142C>A (p.Ser381*), and three tier III variants were detected by FreeBayes (ver 1.3.1) [5]. Further targeted DNA NGS (HEMEaccuTest DNA) confirmed the variants in targeted RNA-seq and additionally detected a tier II variant of KRAS, NM_033360.4:c.38G>A (p.Gly13Asp), and five tier III variants. However, no significant variant of FLT3, including FLT3-ITD, was detected. Donor lymphocyte infusion (DLI) was conducted on day 7 after the diagnosis; however, the BM blasts increased to 88% on day 29. The combination of fludarabine, cytarabine, idarubicin, and granulocyte colony-stimulating factor chemotherapy was started on day 35 and the BM blasts decreased to <5% on day 71 with sustained thrombocytopenia; however, the condition repeatedly relapsed on day 134 and the patient expired on day 223.

Table 1 . Morphological, phenotypic, cytogenetic, and molecular characteristics of the initial and present phase of the case

	Initial phase				Present phase‡
	Finding	VAF (%)	Method		Finding	VAF (%)	Method
WHO classification	AML, NOS				t-AML
FAB, type	M2		Microscopic observation		M2		Microscopic observation
Expressed marker	CD33, CD34, CD117, HLA-DR, and MPO		Immunophenotyping or cytochemical stain		CD33, CD34, CD117, and MPO		Immunophenotyping or cytochemical stain
Karyotype	46,XX[20]		Karyotyping		45,XX,add(3)(p25),del(5)(q?),-12,add(12)(p13)[8]//46,XY[12]		Karyotyping
Gene fusion	Negative		Multiplex RT-PCR		NUP98-NSD1		Targeted RNA-seq
					SNRK-ETV6
Upregulated gene	WT1		Real-time PCR*		WT1		Targeted RNA-seq
					ERG
					BAALC
					TP63
					FGFR3
					CCND1
					CRLF2
Variants	NM_024426.3(WT1):c.1142C>A (p.Ser381*)†	40.97	DNA NGS		NM_024426.3(WT1):c.1142C>A (p.Ser381*)†	12.61	Targeted RNA-seq and DNA NGS
	NM_016320.4(NUP98):c.3557T>G (p.Leu1186Trp)	47.15			NM_001166693.2(AFF1):c.1021A>G (p.Lys341Glu)§	40.58
	NM_006197.3(PCM1):c.4148A>G (p.Asp1383Gly)	47.50			NM_016320.4(NUP98):c.3557T>G (p.Leu1186Trp)	10.45
	NM_033360.4(KRAS):c.38G>A (p.Gly13Asp)†	17.39			NM_006197.3(PCM1):c.4148A>G (p.Asp1383Gly)	9.47
	NM_002834.3(PTPN11):c.227A>C (p.Glu76Ala)	3.89			NM_033360.4(KRAS):c.38G>A (p.Gly13Asp)†	10.79	DNA NGS
					NM_018036.6(ATG2B):c.1586C>T (p.Thr529Met)§	38.94
					NM_000051.3(ATM):c.2117C>T (p.Ser706Leu)	10.85
					NM_022552.4(DNMT3A):c.920C>T (p.Pro307Leu)	3.11
					NM_000215.3(JAK3):c.2062A>T (p.Ile688Phe)§	34.82
					NM_017617.5(NOTCH1):c.4627G>A, (p.Gly1543Arg)§	35.32

^*Using the WT1 ProfileQuant kit (Ipsogen, Marseille, France). ^†Tier II variants classified by the grading system according to the levels of evidence required to determine significance [⁵]. ^‡Increased gene expression was defined as a >2-log₂-fold increase compared with 14 normal controls. HOXA/B expression could not be determined in the present case owing to the lack of a target RNA-seq panel. ^§Variants with a VAF of approximately 35% to 40% that could possibly be donor-derived germline variants rather than clonal evolution.

Abbreviations: FAB, French-American-British; AML, acute myeloid leukemia; NOS, not otherwise specified; VAF, variant allele frequency; RT-PCR, reverse transcription-PCR; NGS, next-generation sequencing; t-AML, therapy-related AML; RNA-seq, RNA-sequencing.

Figure 1. Schematic representation of the NUP98-NSD1 (A–C) and novel SNRK-ETV6 (D–F) gene fusions and proteins. (A) Integrative genomics viewer (IGV) image showing the NUP98-NSD1 breakpoints with 171 supporting junction read counts. (B) Direct sequencing confirmed the identical breakpoint causing an in-frame fusion of NUP98-NSD1. (C) The predicted fusion protein translated from the NUP98-NSD1 transcript based on a merged sequence produced by STAR-Fusion (ver 1.8.1), which contains domains similar to a previous report [²] but is shorter. (D) IGV image showing the novel SNRK-ETV6 fusion breakpoints with 484 supporting junction read counts. (E) Direct sequencing confirmed the identical breakpoint causing a novel in-frame fusion of SNRK-ETV6. (F) The predicted fusion protein translated from the SNRK-ETV6 transcript based on the merged sequence produced by STAR-Fusion (ver 1.8.1).

NUP98-NSD1⁺ AML is characterized by frequent FAB-type M4/M5, a normal karyotype, and HOXA/B upregulation [2]. Further, NUP98-NSD1 is mutually exclusive with other type II variants, but often co-occurs with type I variants such as FLT3-ITD or WT1 variants [2, 3]. FLT3-ITD is the most common variant in NUP98-NSD1⁺ AML (unlike our case), and its prognosis is dismal. Recent studies showed the promising therapeutic effects of dasatinib and navitoclax combination therapy and preemptive DLI based on minimal residual disease for NUP98-NSD1⁺/FLT3-ITD⁺ AML [6, 7]. Regarding the novel SNRK-ETV6 fusion, the defect in ETV6 is pathogenic in hematologic malignancies caused by rearrangement or deletions [8]. However, the partner SNRK gene defect at 3p22.1 has rarely been studied in hematologic malignancies but reportedly impacts hematopoietic cell proliferation and differentiation [9]. Further studies are needed to clarify the role of this novel fusion. This case also meets the criteria of t-AML, representing del(5q) with a complex karyotype and prior cytotoxic chemotherapy; both NUP98- and ETV6- rearrangements were reported in t-AMLs [10]. Additionally, the patient has a variant in TP53 (rs1042522), known to increase the risk of developing therapy-related myeloid neoplasms. Owing to the retrospective nature of this study, the NUP98-NSD1 and SNRK-ETV6 status at the initial diagnostic phase could not be ascertained.

Compared with previous studies using multiple diagnostic methods to characterize NUP98-NSD1⁺ AML [2], the advantage of the present case was the use of RNA-seq, representing a simplified diagnostic step for gene fusion, expression, and gene variant analyses. Additionally, this system might help uncover novel genetic characteristics in leukemias in future larger-scale studies.

Lim HJ and Lee JH conceived and designed the study and collected and analyzed the data; Baek HJ and Kook H contributed to the data; Lim HJ and Shin MG wrote the final manuscript; Lee YE, Park JH, Lee SY, Choi HW, Choi HJ, Kee SJ, and Shin JH participated in coordination and discussion. All authors have accepted their responsibility for the entire content of this manuscript and approved the submission.

No potential conflicts of interest relevant to this article were reported.

This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Ministry of Science and ICT (MSIT) and the Ministry of Health and Welfare (MOHW) (NRF-2019M3E5D1A02067952).