Combined DNA and Transcriptome Sequencing Reveals Discrete Subtypes of Myelodysplasia
Bibliographic record
Abstract
Abstract Introduction Although gene expression profile of myelodysplastic syndromes (MDS) had been widely studied, gene expression-based disease classification was yet to be established. We performed combined DNA and transcriptome sequencing to assess the relationship between genomic lesions, transcriptomic data, hematologic phenotype, and clinical outcome were analyzed. Methods We enrolled a total of 214 patients with myeloid neoplasms with myelodysplasia, for whom complete clinical and pathological data were available. Oncogenic variants and copy number alterations were identified by targeted-capture sequencing using RNA baits designed for 89 known or putative driver genes in myeloid neoplasms and 1,674 single nucleotide polymorphisms. RNA sequencing was performed on both bone marrow mononuclear cells (BMMNCs) and CD34+ cells (n = 51), only CD34+ cells (n = 49), or BMMNCs alone (n = 114). Consensus clustering was performed to identify robust and stable molecular subgroups. Survival analyses were performed with the Kaplan-Meier method. Survival curves were compared using the log-rank test. Multivariate survival analyses were performed by means of Cox proportional hazards regression. Included variables were age, sex, %marrow blast, cytogenetic abnormalities, hemoglobin, absolute neutrophil count, and platelet levels. Results Unsupervised clustering of gene expression data of CD34+ cells from 100 cases identified two stable subgroups. The first subgroup was characterized by a lower blast count, and the up-regulation of genes specifically detected in erythroid lineages. By contrast, the second subgroup was significantly associated with an increased blast count, and expression of the genes related to stem/progenitor cells. These differences became more conspicuous when the comparison was made with healthy adults. Up-regulated expression of many signaling pathway genes, including MAPK, PI3K, and JAK/STAT signaling, was also a conspicuous feature of the second subgroup. To investigate the genetic basis of these unique expression profiles, we compared frequencies of genetic lesions between the two subgroups. The patients in the second subgroup had a higher number of mutations (median 2 [range 0-6] vs. 4 [0-10], P = 0.016) and copy number alterations (median 0 [0-6] vs. 0 [0-9], P = 0.0053) than those in the first subgroup. Among those lesions observed in >10% in either subgroup, SF3B1 and TET2 mutations were significantly enriched in the first subgroup (q-value < 0.1). Del(7)/del(7q), NRAS, and TP53 mutations were also more frequent in the second subgroup (q-value < 0.1). Clinical outcomes also differed substantially between both subgroups. Compared to the first subgroup, the second subgroup was significantly associated with a combined endpoint of death or leukemic transformation in either univariate (hazard ratio 20.3 [95% confidence interval (CI), 4.59-89.6], P < 0.001) or multivariate analysis (hazard ratio 15.5 [95% CI, 3.05-79.2], P < 0.001) at a median follow-up of 8.5 months (range, 0-103 months). Especially no leukemic transformation occurred in the first subgroup, which was in contrast to the very high leukemic transformation rate in the second subgroup (38%). These subgroups were based on the gene expression profile of bone marrow CD34+ cells purified from BMMNCs. To enhance clinical utility, we sought to construct a classifier of the molecular subgroups using gene expression of unfractionated BMMNCs. Among the 100 patients with CD34+ cells, 51 were also analyzed by RNA sequencing for BMMNCs, which were used as a training cohort. Ten-fold cross-validation on the training set identified a logistic regression model with 25 genes, which as applied to the remaining 114 cases with only BMMNC samples. Again, SF3B1 mutations were most significantly enriched in the predicted low-risk subgroup. Predicted high-risk subgroup was significantly associated with poor prognosis in either univariate or multivariate analysis. This gene expression-based classification enabled better stratification of patients at risk of leukemic transformation by combining information on bone marrow blast count. Conclusion We showed that myeloid neoplasms with myelodysplasia can be subgrouped into two major classes with erythroid and stem/progenitor cell signature. This newly developed molecular classification might improve risk prediction and treatment stratification of MDS. Disclosures Kataoka: Yakult: Honoraria; Kyowa Hakko Kirin: Honoraria; Boehringer Ingelheim: Honoraria. Makishima:The Yasuda Medical Foundation: Research Funding. Ogawa:Sumitomo Dainippon Pharma: Research Funding; Takeda Pharmaceuticals: Consultancy, Research Funding; Kan research institute: Consultancy, Research Funding.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".