Case report: single-cell transcriptome sequencing reveals the clonal origin of mature plasmacytoid dendritic cell proliferation in early T-cell precursor lymphoblastic leukemia

Plasmacytoid dendritic cells (pDCs) represent a rare, crucial subset of immune cells, accounting for 0.1–2.5% of nucleated cells in healthy individuals [1]. pDCs play a key role in antitumor immunity and innate and adaptive immune response modulation [2]. Furthermore, studies have shown that pDC proliferation within the tumor microenvironment adversely impacts the prognosis of patients with cancer [3].

According to the 5th edition of the World Health Organization Classification of Hematolymphoid Tumors, clonal proliferation of pDCs ultimately culminates in two distinct clinical and pathological tumor forms [4]: (1) blastic plasmacytoid dendritic cell neoplasm, characterized by a highly aggressive clinical course and unfavorable prognosis; [5] and (2) myeloid neoplasm-associated mature plasmacytoid dendritic cell proliferation (MPDCP), which is most frequently encountered in chronic myelomonocytic [6] and acute myeloid leukemia [7]. Herein, we present a case of early T-cell precursor acute lymphoblastic leukemia (ETP-ALL) accompanied by mature plasmacytoid dendritic cell proliferation. Comprehensive data analysis encompassing clinical manifestations, cellular morphology, immunophenotyping, histopathology, cytogenetic analysis, next-generation sequencing (NGS), T-cell gene rearrangement, and single-cell transcriptome sequencing was conducted.

A 69-year-old male patient presented with multiple enlarged preauricular, cervical, inguinal, and axillary lymph nodes to our hospital. A complete blood count revealed a white blood cell count of 4.81 × 10⁹/L, neutrophil count of 1.20 × 10⁹/L, lymphocyte count of 2.79 × 10⁹/L, red blood cell count of 3.68 × 10¹²/L, hemoglobin concentration of 121 g/L, and platelet count of 118 × 10⁹/L. Peripheral blood smear revealed the presence of blasts (12%). The initial suspicion was for a lymphoproliferative disorder based on clinical presentation and peripheral blood findings. Bone marrow smears and flow cytometry were performed to identify the nature of abnormal cells. The bone marrow aspiration smear revealed 43.0% blasts with variable cell sizes, a high nucleocytoplasmic ratio, loose nuclear chromatin, visible nucleoli, irregular nuclei, and scanty cytoplasm (Fig. 1A, red arrow). Notably, there was cytoplasmic tailing, and approximately 32.0% of the cells exhibited pseudopod-like features, suggestive of DCs (Fig. 1A, green arrow). Multiparameter flow cytometry confirmed the blasts in the bone marrow to be early T-cell precursor lymphoblasts, expressing CD34 (partially), CD38, CD7 (bright), CD5, CD2, CD33 (partially), CD45 (dim) on the cell membrane surface, as well as TDT and CD3 in the cytoplasm, and lacking CD3, CD4, CD8, CD117, CD13, CD123 on the cell membrane surface, and MPO in the cytoplasm (Fig. 1B, red). Additionally, the suspected DC population in the bone marrow was confirmed as mature pDC by flow cytometry, expressing CD123, CD7, CD303, HLA-DR, and CD4 on the cell membrane surface and lacking CD56 and CD34 on the cell membrane surface, as well as TDT, CD3, and MPO in the cytoplasm (Fig. 1B, green). A bone marrow biopsy also revealed the presence of these two cell populations (Fig. 1C). Thus, based on the morphological and immunophenotypic characteristics, the patient was diagnosed with ETP-ALL and MPDCP.

Subsequently, chromosomal analyses, NGS, and T-cell gene rearrangement assays were performed. R + G banding chromosomal analysis revealed a normal male karyotype. NGS identified mutations in CSF3R (p.T640N, variant allele frequency (VAF): 4.0%; p.S783Kfs*4, VAF: 4.4%), DNMT3A (c.1543_1554 + 8del20, VAF: 32.2%; p.Q692Hfs*14, VAF: 37.3%), NOTCH1 (p.L1585Q, VAF: 34.9%), IDH1 (p.R132G, VAF: 38.3%), and KRAS (p.G12V, VAF: 36.4%) with an average sequencing depth of 2567.6×. These mutations were detected in the bulk bone marrow sample without lineage-specific cell sorting. With the exception of NOTCH1, all identified mutations are indicative of lymphoid neoplasms with a poor prognosis [8‐12]. Polymerase chain reaction amplification of the target fragments was performed, revealing negative rearrangements in the TCRB, TCRG, and TCRD genes.

To further investigate the relationship between early T-cell precursor lymphocytes and DC differentiation in ETP-ALL with MPDCP, single-cell RNA sequencing (scRNA-seq) was performed on the bone marrow samples, resulting in 13,702 high-quality cells after filtering. The density peak clustering method was used to perform clustering analysis on early T cell precursor lymphocytes and DC populations. Using Discriminative Dimensionality Reduction with Trees method for trajectory analysis to construct unsupervised transcription trajectories. Cluster analysis of early T-cell precursor lymphocytes and DC populations revealed 14 distinct cell clusters (Fig. 2a). Preliminary annotation of these clusters revealed cell types comprising five T_ALL subsets and two pDC subsets (Fig. 2b). Unsupervised transcriptional trajectories were constructed for the five T_ALL subsets and two pDC subsets based on pseudo-time analysis. The T_ALL and pDC subsets were located in distinct trajectory branches, indicating distinct differentiation states. Notably, a portion of T_ALL5 cells was positioned between the T_ALL and pDC branches, suggesting the presence of a transitional differentiation state (Fig. 2c). Furthermore, visualization of the rate projection from the dynamic model into dimensionality reduction (streamline plot) revealed that pDC1 and pDC2 originated from a T_ALL5 subset, illustrating the differentiation process (Fig. 2d). This observation underscores the intricate interplay between early T-cell precursor lymphocytes and DCs in ETP-ALL with MPDCP.

The patient received active and regular chemotherapy. Two months later, abnormal T cells in the bone marrow, detected by flow cytometry for minimal residual disease (MRD), were below the lower limit of quantitative detection. Considering the possibility of relapse driven by pDCs, the patient underwent monthly flow cytometry monitoring for MRD. During the 9-month follow-up, the patient achieved complete remission, and no relapse was detected.

To date, only three cases of MPDCP associated with lymphoid tumors have been reported [13‐15]; therefore, the definition and classification of MPDCP remain poorly understood. Some studies have demonstrated that in MPDCP associated with myeloid tumors, both pDCs and abnormal myeloid cells originate from a clonally related myeloid progenitor [16]. In contrast, other studies suggest that lymphocytes share precursors with pDCs [17, 18]. Using single-cell RNA sequencing, this study shows that T-ALL cells initiate a pDC transcriptional program, leading to pDC differentiation and expansion. This phenomenon suggests a potential plasticity between the T-ALL and pDC lineages, which may be driven by common progenitor cells or shared signaling pathways (such as the NOTCH signaling pathway, whose abnormal activation may lead to abnormal differentiation of T-ALL cells into pDCs [19]). Therefore, our findings support the concept of “MPDCP associated with hematologic malignancies” proposed by Fei et al., [13] suggesting that pathogenic mutations in MPDCP may occur at an early stage of lineage differentiation.

Different MPDCP-associated hematological malignancies exhibit distinct pathogenic mutations. Combining existing cases of lymphoid tumor-related MPDCP, the most frequent mutation was DNMT3a (4/4, 100%), which is also relatively common in AML-related MPDCPs (20%), whereas the most prevalent mutation in AML-related MPDCPs is RUNX1 (> 70%) [7, 16]. In contrast, CMML-related MPDCPs tend to harbor RAS pathway gene mutations [6]. A systematic association between different types of hematological malignancies, different gene mutations, and excessive pDC proliferation may be used to identify potential therapeutic targets for tumors. In this case, the presence of mutations such as DNMT3a and KRAS suggests potential targets for therapy, although their specific roles in MPDCP-associated ETP-ALL remain to be explored. For instance, DNMT3a mutations, which are frequently observed in MPDCP-associated malignancies, may indicate a potential therapeutic benefit from DNA methyltransferase inhibitors [20]. Similarly, KRAS mutations, which are known to be involved in multiple signaling pathways, could be targeted with specific inhibitors that are currently under investigation [21]. Future studies should focus on identifying actionable mutations and developing targeted therapies for patients with MPDCP-associated hematologic malignancies. This approach could lead to more personalized and effective treatment strategies for these patients.

This case highlights the presence of MPDCP in lymphoid tumors through morphological and immunophenotypic characteristic analysis and provides evidence for the shared pDC and lymphocyte precursor. However, the clinical implications of MPDCP in ETP-ALL are not fully understood. The presence of MPDCP may influence prognosis, treatment response, or therapeutic strategies, but further studies are needed to elucidate these aspects. In this case, the patient achieved complete remission after chemotherapy, suggesting that MPDCP did not significantly impact the initial treatment response. However, long-term follow-up reports for similar patients are limited, highlighting the need for further research to better understand the prognosis and optimal monitoring strategies for such cases. Regular monitoring with flow cytometry and molecular diagnostics is recommended to detect any signs of disease recurrence early. This approach can help in the early identification of relapse or clonal evolution, allowing for timely intervention and improved patient outcomes.

It is important to acknowledge the limitations of our study. First, we did not use external datasets for validation, which is recommended for further confirmation of our findings. Second, NGS identified mutations in CSF3R, DNMT3a, NOTCH1, IDH1, and KRAS. It is currently unclear whether these mutations are present in both T-ALL and pDC populations, as cell sorting was not performed to distinguish lineage-specific mutations. Future studies should include cell sorting and targeted sequencing to determine the clonal relationship between T-ALL and pDCs in such cases. Third, while our data suggest a shared precursor, definitive markers like Ly6D⁺ lymphoid progenitors were not assessed. Future studies should integrate surface protein detection with transcriptomic profiling to validate this hypothesis.

In conclusion, our study provides valuable insights into the presence of MPDCP in lymphoid malignancies and the potential link between pDCs and lymphocytes. However, further research is needed to fully understand the clinical implications, genetic landscape, and therapeutic potential of MPDCP in hematologic malignancies. Strengthening both basic and clinical research will be crucial to uncover the nature of these conditions and provide clinicians with more accurate and effective diagnostic and therapeutic options.