Prediction of Primary Treatment Outcome Using Gene Expression Profiling of Pre-Treatment Biopsies Obtained from Childhood and Adolescent Hodgkin Lymphoma Patients
Bibliographic record
Abstract
Abstract Introduction: Hodgkin lymphoma (HL) is a common malignancy of children and adolescents and is highly curable with a 5-year overall survival (OS) rate of > 97%, yet dose-intensified chemotherapy regimens in combination with radiation therapy come with a high cost in form of long-term toxicity and morbidity (Castellino et al., Blood 2011). This major clinical challenge has resulted in the evaluation of risk-adapted treatment regimens in clinical trials aiming to achieve the optimal equilibrium between high survival rates and prevention of treatment-related toxicity. However, risk stratification is currently limited to the use of clinical factors as there are no validated integral biomarkers that can be employed to either improve risk stratification or as surrogate markers of treatment outcome in pediatric HL. The aim of our study was to perform gene expression profiling (GEP) to uncover disease biology underlying treatment response and develop a prognostic model to tailor first-line therapy in pediatric HL. Methods: We selected 203 formalin-fixed, paraffin-embedded tissue (FFPET) specimens from patients enrolled in a randomized phase 3 clinical trial (AHOD0031) of the Children's Oncology Group (COG) based on the availability of archived FFPET blocks. That trial was designed to assess the value of early chemotherapy response for tailoring subsequent therapy in intermediate-risk pediatric HL. We performed GEP on RNA extracted from pre-treatment FFPET biopsies using NanoString technology and a customized codeset encompassing probes for 784 genes. These genes were either previously reported to be associated with prognosis and outcome in HL or represent the cellular diversity of the tumor microenvironment. Event free survival (EFS) and OS were estimated using the Kaplan-Meier method. Gene expression data were used to develop a predictive model for EFS using penalized Cox regression with parameters trained using leave-one-out cross-validation. Results: Of the 203 tissue samples obtained from the Biopathology Center at the Cooperative Human Tissue Network, 182 (89.7%) passed quality assurance testing, similar to the pass rate achieved for adult HL samples obtained from the Eastern Cooperative Oncology Group trial E2496 (Scott et al., JCO 2013). We applied our previously published 23-gene predictor for OS (Scott et al., JCO 2013), developed using biopsies from adult HL patients to our pediatric cohort. After calibrating the new codeset, 53 patients were classified as "high-risk" and 129 as "low-risk". Importantly, the model failed to predict inferior outcomes in the "high-risk" group (5-year OS 100% vs 95%, log-rank P = 0.125; 5-year EFS 82% vs 70%, log-rank P = 0.159), with patients in the "high risk" group trending to have superior outcomes than the "low risk" patients. Moreover, only 2 genes from this model, IFNG and CXCL11, were significantly associated with EFS in univariate Cox regression analysis (P = 0.003 and 0.048, respectively) but with inverse hazard ratios in the pediatric group compared to adult patients. Therefore, we sought to develop a novel EFS predictive model for pediatric patients treated in the AHOD0031 trial. Using univariate Cox regression we identified 79 genes significantly associated with EFS (raw P < 0.05). Using the expression of these 79 genes as the input to penalized Cox regression, we developed a 16-gene model to predict EFS in our training cohort. Using an optimized cut-off for the model score, 31% of patients were designated high-risk and had significantly inferior post-treatment outcome (5-year EFS 38% vs 89%, log-rank P < 0.0001). When multivariate analyses were performed including our EFS-model score, disease stage and initial treatment response as variables, only the model score was significantly associated with EFS (P < 0.0001, HR 11.3 (95% CI 5.5-23.4)). Conclusions: Failure of the GEP-based model developed in adult HL suggests distinct biology underlies treatment failure in the pediatric age group. We describe the development of a novel predictive model for EFS in intermediate-risk pediatric HL patients that will be validated in an independent cohort of patients treated in the AHOD0031 trial. Successful validation of the model may provide a clinically relevant biomarker for pediatric and adolescent HL patients allowing refinement of risk stratification and the combination of molecular and clinical risk factors at diagnosis. Disclosures Scott: Celgene: Consultancy, Honoraria; NanoString: Patents & Royalties: Inventor on a patent that NanoString has licensed.
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 | 0.000 | 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".