Texture Analysis on Computed Tomography Predicts the Outcome of Patients with Hodgkin Lymphoma
Bibliographic record
Abstract
Abstract Background Texture analysis is a novel approach, which analyses the distribution of pixel on computed tomography (CT) within a region of interest. CT-based texture analysis (CTTA) can be applied to images that are acquired in routine clinical practice and provide prognostic information in patients with non-small cell lung cancer or colorectal adenocarcinoma. We aimed to assess whether CTTA could predict the outcome in patients with Hodgkin lymphoma (HL). Methods All consecutive patients who received chemotherapy for HL at our center between 2007 and 2015 were considered for analysis. Patients without available CT at baseline and/or at interim analysis in the picture archiving and communication system (PACS) of our center were excluded. Overall, 56 patients (15 early and 41 advanced stages) were included in this retrospective study. Median age was 34.6 years (range, 17-78) and 64% patients were male. First line chemotherapy were ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine; n=42), escalated-dose BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone; n=12) and ABVD without bleomycin (n=2). Median follow-up was 37.6 months (5-103). CTTA was performed using TexRAD® software on single-slices on baseline CT and interim CT (after 2 or 4 cycles of chemotherapy). CTTA parameters and their changes were compared with disease response, relapse and disease-free survival (DFS). Receiver operating characteristic (ROC) curve analyses were performed to identify the strongest parameter and the optimal cutoff value for predicting relapse. Results On baseline CT, the strongest parameter predictive of relapse was MPP (mean positive pixels). With a spacial scale filter (ssf) of 2, CTTA < 27.5 (cut off value) were defined as positive. Cumulative incidence of relapse (IR) and 5-year DFS were 36.4% and 54.5% in patients with positive baseline CTTA compared to 11.1% and 82.2% in patients with negative baseline CTTA (p=0.053 and p=0.077, respectively). Baseline CTTA significantly predicted relapse and DFS in early stages patients (p=0.015 and p=0.015, respectively) and in advanced stages patients treated with ABVD (p=0.042 and p=0.025, respectively). However, outcomes were not significantly different according to baseline CTTA in patients treated with escalated-dose BEACOPP. Indeed, DFS tented to be improved with escalated-dose BEACOPP (87.5%) compared to ABVD (67.5%). On interim CT, the strongest parameter predictive of relapse was the entropy (ssf=4). CTTA > 5 (cut off value) were defined as positive. Patients with positive interim CTTA had a cumulative IR and a 5-year DFS of 50% and 37.5% compared to 10.4% and 83.3% in patients with negative baseline CTTA (p=0.002 and p=0.004, respectively). Interim CTTA could significantly predict the outcome both in early and advanced stages patients. Interestingly, only one patient treated with BEACOPP had a positive interim CTTA. Finally, positive interim CTTA was significantly associated with DFS in patients with negative (p=0.015) or positive (p=0.049) interim 18F-Fluorodeoxyglucose (FDG) positron emission tomography (PET). Conclusion Texture analysis on baseline and interim CT can provide prognostic information in patients with HL. It may be used at diagnosis to identify high-risk patients and guide the initial treatment choice. At interim analysis, the combination of CTTA and PET may help clinicians to better adapt therapy. Disclosures Garderet: Takeda: Consultancy; Amgen: Consultancy; Novartis: Consultancy; Bristol-Myers Squibb: Consultancy.
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 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 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".