A miRNA Risk Score for the Prediction of Response to Rituximab-CHOP Therapy and Survival of Patients with Diffuse Large B-Cell Lymphoma
Bibliographic record
Abstract
Abstract Abstract 324 Background: Diffuse large B-cell lymphoma (DLBCL) may be curable with current immuno-chemotherapy, however nearly 30% of patients fail to benefit from this therapeutic approach. Gene expression profiling studies identified lymphocyte as well as stroma-based mRNA signatures that are predictive of response to R-CHOP; however the need for optimally cryopreserved samples has limited their clinical applicability. MicroRNAs (miRNA) are highly preserved non-coding RNAs that act post-transcriptionally to regulate gene expression by binding to the 3′UTR of mRNAs. To date multiple studies have reported selective miRNAs expression at different lymphocyte differentiation stages, distinct expression in mRNA-defined DLBCL subgroups and have correlated the expression of “selected” miRNAs with disease outcomes. Herein, we have conducted a comprehensive profiling of miRNA expression in R-CHOP treated DLBCL patients and established a miRNA-based risk score that is predictive of response to therapy. Methods and results: We have postulated that a miRNA signature in DLBCL is predictive of survival post R-CHOP chemotherapy. To test this hypothesis, we analyzed the miRNA and mRNA signatures of R-CHOP sensitive “S” and resistant “R” DLBCLs (n=20). miRNAs were hybridized to the miRNA Affymetrix gene-chip (847 hsa-miRNA probes) and raw miRNA expression values were log2 transformed and normalized (miRNA-QC tool, Affymetrix). Comparison of normalized miRNAs expression in “S” versus “R” patients (Anova testing) identified 59 differentially expressed miRNAs (Fold change < -1.5 or > 1.5 with a p value and FDR <0.05). In order to establish a risk score based on the expression of these 59 miRNAs, column-dendrogram branches were then sorted left to right based on each patient's difference between the average log2-scale expression of the 37 up-regulated and the 12 down-regulated miRNAs: this difference is interpreted as an up-/down-regulated mean ratio (ie, geometric mean) on the log2 scale [Log2 Geometric mean ratio [GMR] up-/down-regulated miRNA = Log2 [(2^ΣupregulatedümiRNA/ni)/(2^ΣdownregulatedümiRNA/nii)] where ni and nii represent the number of up-regulated and down-regulated miRNAs. This univariate summary (ie, GMR) of the 59-miRNA expression profiles for each patient enabled accurate prediction of all S versus R patients. Patients with log2 GMR > 0 had a 6-years PFS rate of 100%, while all patients with log2 GMR < 0 relapsed (HR=0.293 [95% CI 0.132–0.647]). Cox regression was then used to model relapse free survival times from treatment as a function of miRNA expression. Separate regression models were also built looking at fit measures, proportionality assumption, and discrimination ability (Harrell C statistics) and only miRNA with a discrimination ability (Harrell C) of > 85% (n=12 miRNA) were used to calculate the log2 GMR of up-/down-regulated miRNAs. A simplified model based on 12 miRNAs (Sensitive vs Resistant: upregulated: hsa-let-7i; hsa-miR-130a; hsa-miR-199a-3p; hsa-199b-3p; hsa-miR-223; hsa-miR22; hsa-miR-24; hsa-miR-26b; hsa-miR-27a; hsa-miR-331-5p; downregulated: hsa-miR-1288) accurately predicted sensitivity or resistance to R-CHOP. With this 12 miRNAs model, only 1 patient in each group (S and R) was misclassified (Fig 1). In addition we have validated the differential expression of these miRNA by short-stem loop RT-PCR and found a strong correlation between the gene-chip and qRT-PCR results (correlation coefficient 0.717). mRNA profiling (U133A Plus2 array chip) was also performed on the whole lymph node sections; 176 genes were identified as differentially expressed between S and R patients with many of these genes belonging to the “stroma-1 and -2” DLBCL signature. Using the TargetScan miRNA target mRNA prediction tool, combinatory analysis of miRNA and mRNA expression profiles of DLBCL patients identified positive and negative correlations (P <0.05) between differentially expressed miRNA and mRNAs. Lastly in a multivariate Cox regression analysis that included the IPI and the 12 miRNA based GMR risk score, both variables were independent predictors of survival post R-CHOP therapy. Conclusion: We believe that this log2 GMR score based on the 12 identified miRNA provides a robust method of predicting sensitivity to R-CHOP in DLBCL patients and it is currently being tested in a larger independent validation cohort. Disclosures: No relevant conflicts of interest to declare.
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 | 0.001 | 0.001 |
| 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) | 0.001 | 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".