Acute promyelocytic leukemia with TTMV::RARA fusion potentially responds to all-trans retinoic acid/arsenic trioxide treatment

Over 95% of patients diagnosed with acute promyelocytic leukemia (APL) exhibit the fusion of the promyelocytic leukemia (PML) gene with the retinoic acid receptor a (RARA) gene, and significantly respond to all-trans retinoic acid (ATRA) combined with arsenic trioxide (ATO).^1,2 Conversely, less than 5% of morphologically classified APL cases harbor alternative partner genes besides PML, which impacts the efficacy of ATRA+ATO, thereby defining variant APL.³

Torque teno mini virus (TTMV) is prevalent and symbiotically exists with humans. Recent research indicates that TTMV could integrate with RARA, potentially resulting in the development of a distinct type of variant APL.⁴ To date, 11 cases have been reported, including nine children and two adults.⁴-¹⁰ However, the pathogenetic mechanisms of TTM-V::RARA in acute myeloid leukemia (AML) remain uncertain, as does its response to ATRA+ATO. We firstly showed an APL case with TTMV::RARA achieving a long-term complete remission (CR) induced by ATRA+ATO, offering an insight into managing this disease and distinguishing it from other variant APL.

The patient, a 15-year-old boy, initially presented to a local hospital with a persistent headache and weakness. Laboratory findings revealed leukocytosis (white blood cell count: 40.85×10⁹/L), anemia (hemoglobin: 109 g/L), and thrombocytopenia (platelet count: 47×10⁹/L). Coagulation tests demonstrated prolonged prothrombin time (19.6 seconds) and activated partial thromboplastin time (32.6 seconds), as well as elevated D-dimer (38.44 mg/L) and hypofibrinogenemia (0.52 g/L). The bone marrow (BM) morphology showed a predominant population consisting of classical hyper-granular promyelocytes (81%) and exhibiting intense reactivity to myeloperoxidase staining (+++) (Figure 1A). According to the AI-PAL model utilizing age and blood parameters (https://alcazerv.shinyapps.io/ AIPAL/), our case was predicted to be APL with high probability (93.7%; Figure 1B). Flow cytometry identified 93% of granulocytes with abnormal immunophenotype (CD13⁺, CD33⁺, CD123⁺, CD34-, HLA-DR^low/-; Figure 1C). Karyotype was 46,XY[20] (Figure 1D). Real-time polymerase chain reaction (PCR) showed negative expression of PML::RARA, while the targeted next-generation sequencing (including 72 myeloid-associated genes) revealed mutations in KRAS (p.G12D, variant allele frequency [VAF] 1.06%), NRAS (p.G12D, VAF 37%), BRAF (p.G469A, VAF 1.15%) and ARID1B (p.E2051K, VAF 47%) genes. The Principal Component Analysis (PCA) of gene expression confirmed that our case was more similar to APL than to non-M3 AML (Figure 1E).

Given the clinical suspicion of hyper-granular (or typical) APL, the initial leukoreduction treatment consisted of hydroxyurea and a low-dose cytotoxicity drug, followed by four courses of ATRA (20 mg, twice daily, days 1-14) + ATO (10 mg, daily, days 1-28). The patient achieved the first CR (CR1) after the first cycle and maintained the CR status for 6 months. Subsequently, the patient presented with fever and hip pain, along with leukocytosis (14.73×10⁹/L) and central nervous system (CNS) involvement. Flow cytometry revealed 78.7% abnormal promyelocytes in BM, showing a changed karyotype (46,XY,t(3;20)(q26.2;q11)[5]/46,idem,del(9) (p21)[13]/46,XY[2]). Due to relapse, the patient underwent the salvage induction therapy consisting of ATRA (20 mg, twice daily, days 1-14), idarubicin (10 mg, daily, days 1, 3, 7), and cytarabine (200 mg, daily, days 1-7), resulting in secondary CR. Following two additional cycles of venetoclax, homoharringtonine, cytarabine, and granulocyte-colony stimulating factor, the disease relapsed once more and the patient succumbed to progressive disease within 2 months before undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT).

After the relapse following four courses of ATRA+ATO, whole-transcriptome sequencing (WTS) of BM was carried out, revealing no PML::RARA. Samples were collected with informed consent in accordance with the Declaration of Helsinki. The study was approved by the Ethics Committee of Beijing Chao-Yang Hospital (2024-KE-515). Based on the recent literature reporting TTMV::RARA,^4,5,7,11 it was postulated that TTMV genetic material had integrated at this locus. A total of 33 reads from RNA sequencing successfully aligned to RARA exon 3 and TTMV genome, exhibiting a mapping length ≥50 base pairs (bp) and an identity ≥90%, indicating the presence of TTMV genetic material within the genome of this case. To further investigate this phenomenon, transcripts in close proximity to RARA gene were assembled. An assembled transcript sequence was found to partially align with RARA exon 3-9 (NC_000017.11:40,348,316-40,356,914) (Figure 2A). Additionally, a 568 bp sequence located at the 5’ end, which could not be mapped to the human hg38 reference genome, was determined as part of the TTMV genome (Figure 2B). Substantial alignment of the integrated TTMV sequence was noted with the complete genome of TTMV isolate SAfiA-453-9 (accession: MN770239).¹² Two mapping blocks demonstrated 82% and 67% identity across the entire 568 bp sequence (Figure 2B). The insertion of the TTMV sequence was located at chr17:40,347,785 (hg38), with the integrated sequence spanning 568 nucleotides. Notably, the final 226 nucleotides correspond to the coding region that aligns in-frame with RARA exon 3 (Figure 2C). The complete TTMV::RARA transcript spans 2,354 bp, with an estimated 1,437 bp coding sequence. The coding sequence and the TTMV insertion breakpoint were further confirmed by PCR, Sanger sequencing, WTS, and whole-genome sequencing (WGS) (Figure 3). While MN770239 lacks detailed annotation of TTMV open reading frame (ORF) 2, the overlap with ORF1 is consistent with earlier findings.¹³ The conserved TTMV ORF2 5′ untranslated region (UTR) sequence “CGAATGGCTGAGTTT” was identified at nucleotides 267-281 in our assembled TTMV transcript sequence, followed by a 226 nucleotides segment initiating with the start codon (ATG). Additionally, the conserved motif “WX7HX3CXCX5H” of TTMV ORF2 was also found in the TTMV::RARA transcript (Figure 2C). According to recent reports, the TTMV sequence consistently appears in ORF2. This study contributes to the existing literature by presenting an additional documented APL case with TTMV::RARA. Most reported cases were characterized by hyper-granular promyelocytes, displaying morphological and immunophenotypic similarities to typical APL. CNS involvement, which is uncommon in APL with PML::RARA, was present in four of six cases with TTMV::RARA. Specifically, three cases demonstrated CNS involvement at diagnosis,^9,10 while our case displayed it following relapse. Among the four cases undergoing ATRA + chemotherapy (idarubicin or “7+3” regimen), two demonstrated sustained CR durations lasting 8-15 months.^4,6,10 Particularly, our case demonstrated a 6-month CR duration during the four courses of ATRA+ATO. TTMV::RARA was not initially considered in our case, yet classic APL characteristics were observed. An empirical ATRA+ATO treatment was thereby implemented, achieving unexpected CR1, possibly due to the anti-RARA properties of ATRA, with ATO synergistically enhancing its anti-leukemic effects. ATO exhibits anti-proliferative and pro-apoptotic effects in leukemia via cell cycle arrest, apoptosis and autophagy induction, as well as reactive oxygen species generation and caspase activation.¹⁴ ATRA+ATO also inhibits the Nrf2 antioxidant pathway, thereby increasing cytotoxicity in AML.¹⁵ Ongoing trials involving this combination are currently underway in non-M3 AML (clinicaltrials gov. Identifier: NCT03031249, NCT05297123). Given the anti-leukemia effects of ATO, further research on ATRA+ATO in APL with TTMV::RARA is warranted.

Figure 1.Morphologic features, immunophenotypes, karyotype and expression profile of the variant acute promyelocytic leukemia case with TTMV::RARA. (A) Giemsa staining of the bone marrow (BM) aspiration specimen at first diagnosis. The promyelocytes had a round or irregular shape with off-center ovoid nuclei, sometimes binucleated. Their chromatin was loosely dispersed, with inconspicuous nucleoli. The cytoplasm, variable in amount and bluish-gray, contained coarse, dense granules, prominent both inside and outside the cytoplasm. (B) The AI-PAL model included large training (N=477), internal testing (N=202) and external validation (N=731) sets, with 10 clinical parameters. According to the AI-PAL model, our case was predicted to be acute promyelocytic leukemia (APL) with high probability (93.7%). (C) Immunophenotype profiling of BM was characterized by CD13⁺, CD33⁺, CD123⁺, CD34^- and HLA-DR^low/-. (D) Karyotype showing a normal chromosomal constitution. (E) The gene expression data from 12 APL cases and 45 cases with non-M3 acute myeloid leukemia (AML) in the Beat AML1.0 dataset (clinicaltrials gov. Identifier: NCT03013998) were randomly selected and integrated with our case with TTMV::RARA for the Principal Component Analysis (PCA). The resulting plots of the first 2 principal components (PC1 and PC2) revealed that our case was positioned closer to the APL cluster along the PC1 axis, with the high variance explained by PC1 (66.4%) suggesting a similar gene expression profile between our case and APL, rather than non-M3 AML (PC2). ALL: acute lymphocytic leukemia.

Figure 2.Sequence features and integration schematic of TTMV::RARA fusion transcript. (A) The coverage details of the assembled transcript over the RARA gene on the hg38 genome reference. (B) Coverage of TTMV sequence in this case over the TTMV isolate SAfiA-453-9 complete genome was assessed. (C) Schematic representation of integration. The integration schematic depicts TTMV open reading frame 2 (ORF2) and an upstream untranslated region (UTR) integrated into RARA. The chimeric transcript includes a TTMV sequence joined to RARA exon 3-9. The conserved UTR sequence “CGAATGGCTGAGTTT” was observed at the end of UTR sequence. A start codon (ATG) at the 5’ end of TTMV ORF2 may predominate the initial translation phase.

Figure 3.Confirmation of TTMV::RARA fusion by polymerase chain reaction, Sanger sequencing, whole-transcriptome sequencing, and whole-genome sequencing data. (A) Detection of the TTMV::RARA chimeric fusion in cDNA. (B) Sanger sequencing electropherograms of the downstream and upstream breakpoints of TTMV integration on the cDNA. (C) Detection of the TTMV integration site at the RNA level. The Integrative Genomics Viewer (IGV) screenshot of TTMV::RARA reads indicated that a total of 33 reads from RNA sequencing successfully aligned to RARA exon 3 and TTMV genome, exhibiting a mapping length ≥50 bp and an identity ≥ 90% at the RNA level, after the removal of polymerase chain reaction duplicates. The starting base of RARA exon 3 is marked by a dashed black line. (D) Detection of the TTMV integration site at the DNA level was performed using IGV. The integration site at chr17:40,347,785 was marked by a dashed black line. The read coverage on either side of this breakpoint (chr17:40,347,785) showed a notable shift, indicating the occurrence of a chimeric event. bp: base pairs.

The case from Wang,⁹ with high-risk factors like complex karyotype and extramedullary involvement, failed to achieve remission with ATRA+ATO. In contrast, our case lacked such factors before treatment, suggesting that ATRA+ATO could be considered as a first-line option for treating APL with TTMV::RARA without high-risk features, rather than as a salvage therapy.

The efficacy of allo-HSCT for APL with TTMV::RARA is noteworthy. Of eight patients receiving allo-HSCT, four underwent allo-HSCT in CR1, with three maintaining CR for 0.63 to 8 years¹⁰ and one recurring after 13 months.⁹ The other four underwent allo-HSCT post-relapse, with two surviving for 4 years and 20 months,^7,10 respectively, but two relapsing after 7 months and 50 days.^4,6 Allo-HSCT in CR1 showed low relapse and prolonged remission. While post-relapse patients had lower survival, some still achieved CR. Further research with larger cases and longer follow-up is still needed.

Prior studies showed that the retained lengths of RARA intron 2 in fusion events range from 13 to 38 nucleotides. However, RARA intron 2 was absent in the fusion transcript of our case (Figure 2C). It indicates a higher level of complexity in the TTMV::RARA fusion event and suggests that the RARA intron 2 may not be an essential causative factor. Additionally, each case exhibits distinct partner genes integrated with RARA, potentially resulting from differential splicing patterns, necessitating further investigation.

Due to the widespread presence of TTMV, there is a potential for additional cases with TTMV::RARA to remain undetected. The development of a specialized bioinformatics approach designed for the diverse TTMV genome sequence could facilitate the identification of TTMV::RARA through the WTS analysis. Additionally, the establishment of a reliable PCR protocol would be essential for detecting TTMV::RARA.

Furthermore, additional research is necessary to elucidate the oncogenic mechanisms associated with TTMV::RARA fusion protein and identify other instances of viral genome integration with human genes that contribute to hematological malignancies. Given the high prevalence of TTMV in population, it is imperative to ascertain whether the fusion event between TTMV and RARA occurs randomly or is influenced by genetic susceptibility factors. Comprehensive understanding of these potential susceptibility factors is essential for preventing virus-induced malignant transformations. It is noteworthy that, unlike previous APL subtypes which were solely of human origin, APL with TTMV::RARA is characterized by a viral component. This warrants a deeper investigation into the mechanisms by which TTMV integrates into human genome.

Conclusively, APL with TTMV::RARA represents a unique variant APL with a novel oncogenic mechanism and distinct clinical and molecular features, offering new insights into APL.

Footnotes

• Received April 18, 2024
• Accepted January 14, 2025

Correspondence

*QX, YP, SS, YZ and TS contributed equally as first authors

H-H. ZHU - zhuhhdoc@163.com

Disclosures

No conflicts of interest to disclose.

Contributions

Funding

This work was supported by grants from the National Natural Science Foundation of China (82450101, 82370169, 82400147), Postdoctoral Fellowship Program of CPSF (GZC20231748), Beijing Postdoctoral Research Funding (2024-ZZ-185) and Beijing Chao-Yang Hospital Golden Seeds Foundation (CYJZ202303, CYJZ202317). QX is the Outstanding Young Talent of the Beijing Overseas Talent Aggregation Project and the “Phoenix” Project in Chaoyang District, Beijing (Y2024070019). HHZ is supported by the Beijing Hospitals Authority “Dengfeng” talent training plan (DFL20240301). These study funders did neither participate in the data collection, analysis, interpretation, nor any aspect pertinent to the study.