Duvelisib
| 證據等級: L5 | 預測適應症: 10 個 |
目錄
Using the drug-repurposing report template as instructed. Key data-quality issue up front: predicted_indications[0] (rank 1, “Hodgkins lymphoma”) is the required focus disease per the template rules, but the evidence pack’s own repurposing_rationale for that entry states this is very likely a disease-ontology mapping artifact — none of the 11 trials or 16 papers actually enroll classical Hodgkin lymphoma patients; they’re all CLL/SLL/NHL. I’m reporting this transparently rather than reframing it as a clean positive signal.
Duvelisib: From CLL/SLL to Hodgkin’s Lymphoma
One-Sentence Summary
Duvelisib (DrugBank DB11952) is a dual PI3K-δ/γ inhibitor originally developed for relapsed/refractory Chronic Lymphocytic Leukemia (CLL) / Small Lymphocytic Lymphoma (SLL), later expanded to follicular lymphoma — per literature in this evidence pack, not a Taiwan license (the drug is currently not marketed in Taiwan). The TxGNN model’s top-ranked prediction is Hodgkin’s Lymphoma, supported on paper by 11 clinical trials and 16 publications. On review, however, every one of those trials and papers actually enrolls non-Hodgkin lymphoma (NHL), CLL/SLL, or mixed lymphoid-malignancy populations — none specifically targets classical Hodgkin lymphoma, so this top prediction is best treated as a probable disease-label mapping error rather than a genuine repurposing signal.
Quick Overview
| Item | Content |
|---|---|
| Original Indication | Chronic Lymphocytic Leukemia (CLL) / Small Lymphocytic Lymphoma (SLL) (per literature; not a TFDA-licensed indication — drug not marketed in Taiwan) |
| Predicted New Indication | Hodgkin’s Lymphoma |
| TxGNN Prediction Score | 99.94% |
| Evidence Level | L4 |
| Taiwan Market Status (TFDA) | 未上市 (Not marketed) |
| Number of NDAs/Licenses | 0 |
| Recommended Decision | Hold |
Note on the evidence set: a lower-ranked candidate in this same pack — “B-cell neoplasm” (rank 9, score 99.80%) — carries much stronger evidence (Evidence Level L1, includes the pivotal Phase 3 DUO trial NCT02004522, decision stage S3, recommendation “Proceed with Guardrails”). Per its own rationale, that entry essentially restates duvelisib’s already-known CLL/SLL/follicular lymphoma activity rather than a new indication. It is not a repurposing candidate, but it is worth flagging separately from the Hodgkin lymphoma prediction reviewed below.
Why is This Prediction Reasonable?
Duvelisib is a first-in-class oral dual inhibitor of PI3K-δ and PI3K-γ, blocking B-cell receptor (BCR) downstream signaling — a pathway central to indolent B-cell malignancies such as CLL/SLL and follicular lymphoma (per literature PMID 30430368, PMID 28388280). Detailed structured MOA data was not returned from DrugBank in this evidence pack (Data Gap DG002); the mechanism above is reconstructed from the literature evidence attached to these predictions.
Mechanistically, this pathway dependency does not extend cleanly to classical Hodgkin lymphoma: the malignant Reed-Sternberg cells in classical Hodgkin lymphoma typically lose BCR expression and are instead driven by constitutive JAK/STAT and NF-κB signaling, which is only indirectly related to PI3K-δ/γ inhibition. Consistent with this, a line-by-line review of the 11 trials and 16 papers attached to this prediction found that every one addresses indolent/aggressive non-Hodgkin lymphoma, CLL/SLL, mantle cell lymphoma, or T-cell lymphoma — not classical Hodgkin lymphoma. The most plausible explanation is that TxGNN’s underlying disease ontology grouped a broadly-named “lymphoma” node in a way that pulled in this high similarity score without genuine target-specific evidence.
Because of this, the mechanistic rationale for duvelisib in classical Hodgkin lymphoma specifically should be treated as unconfirmed pending disease-ontology verification, not as a supported hypothesis.
Clinical Trial Evidence
| Trial Number | Phase | Status | Enrollment | Key Findings |
|---|---|---|---|---|
| NCT04038359 | Phase 2 | Completed | 103 | Compared two duvelisib dosing schedules (with dose holidays) in indolent non-Hodgkin lymphoma (iNHL) — not classical Hodgkin lymphoma |
| NCT05065866 | Phase 1 | Completed | 14 | Duvelisib + BMS-986345 in broad lymphoid malignancies; no Hodgkin-specific cohort |
| NCT02640833 | Phase 1b/2 | Withdrawn | 0 | Planned duvelisib + venetoclax in R/R CLL/SLL/NHL; withdrawn, no data generated |
| NCT04836832 | Phase 1 | Withdrawn | 0 | Planned duvelisib + acalabrutinib in R/R indolent NHL (DUAL trial); withdrawn |
| NCT04379167 | Phase 2 | Unknown | 140 | Single-arm study of a different PI3K inhibitor (YY-20394) in R/R follicular NHL; status unknown, not duvelisib-specific |
| NCT01871675 | Phase 1b | Completed | 48 | IPI-145 (duvelisib) + rituximab/bendamustine in lymphoma/CLL — NHL population |
| NCT02576275 | Phase 3 | Withdrawn | 0 | Planned duvelisib + R-bendamustine vs placebo in previously-treated indolent NHL; withdrawn before enrollment |
| NCT05923502 | N/A (real-world) | Not yet recruiting | 200 | Planned real-world observational study of duvelisib in NHL; not yet started |
| NCT05044039 | Phase 1 | Active, not recruiting | 42 | Duvelisib post-CAR-T to improve CAR-T persistence; mechanism study, not Hodgkin-specific |
| NCT04803201 | Phase 2 | Suspended | 170 | Duvelisib-CHOP vs alternative regimens in untreated CD30-negative peripheral T-cell lymphoma; suspended |
None of the above trials enroll a classical Hodgkin lymphoma population; all relevance gradings in the source data are “C” (low relevance to the stated indication).
Literature Evidence
| PMID | Year | Type | Journal | Key Findings |
|---|---|---|---|---|
| 36685572 | 2022 | Systematic review/meta-analysis | Frontiers in Immunology | Meta-analysis of duvelisib safety/efficacy across relapsed/refractory lymphoid neoplasms (CLL/NHL); does not isolate classical Hodgkin lymphoma |
| 31490009 | 2019 | Phase 1b trial | American Journal of Hematology | Duvelisib + rituximab ± bendamustine in NHL/CLL; disease-specific expansion cohorts were NHL/CLL, not Hodgkin lymphoma |
| 29191916 | 2018 | Phase 1 trial | Blood | First-in-human dose-escalation in 210 patients with advanced hematologic malignancies; established MTD 75 mg BID |
| 28017967 | 2017 | Mechanistic/translational | Leukemia | Duvelisib alters apoptotic regulators, sensitizing CLL cells to venetoclax — CLL-focused |
| 36882482 | 2023 | Preclinical | Scientific Reports | PI3Kγ/δ role in mantle cell lymphoma (a non-Hodgkin B-cell lymphoma) proliferation/migration |
| 30799261 | 2019 | Review | The Lancet Oncology | Overview of duvelisib in indolent non-Hodgkin lymphoma |
| 31580408 | 2019 | Review | Am J Health-Syst Pharm | Summary of targeted therapies approved for B- and T-cell lymphomas |
| 33616890 | 2021 | Review | Drugs | Novel therapy approaches in follicular lymphoma (an NHL subtype) |
| 32356174 | 2020 | Review | Curr Treat Options Oncol | PI3K inhibitors as targeted therapy across lymphoma broadly |
| 32658557 | 2020 | Review | Future Oncology | Copanlisib (a related PI3K inhibitor) in non-Hodgkin lymphoma |
As with the trials, none of these publications address classical Hodgkin lymphoma specifically.
Taiwan Market Information (TFDA)
Duvelisib currently has 0 TFDA licenses and is not marketed in Taiwan (market_status: 未上市). No NDA/license records are available to summarize.
Cytotoxicity
Duvelisib is an antineoplastic agent (PI3K-δ/γ inhibitor used to treat hematologic malignancies), so this section applies.
| Item | Content |
|---|---|
| Cytotoxicity Classification | Targeted therapy (dual PI3K-δ/γ inhibitor) — not a conventional cytotoxic chemotherapy |
| Myelosuppression Risk | Not available in this evidence pack — pending TFDA package insert (Data Gap DG001, Blocking severity) |
| Emetogenicity Classification | Not available in this evidence pack — pending TFDA package insert (Data Gap DG001) |
| Monitoring Items | Not available in this evidence pack — pending TFDA package insert (Data Gap DG001) |
| Handling Protection | Not available in this evidence pack — pending TFDA package insert (Data Gap DG001) |
Safety Considerations
Please refer to the package insert for safety information. TFDA warnings/contraindications (Data Gap DG001) and DDI data (query returned “not found”) are not yet available in this evidence pack.
Conclusion and Next Steps
Decision: Hold
Rationale: The top-ranked TxGNN prediction (Hodgkin’s lymphoma, 99.94% score) is not supported by disease-specific evidence — all 11 trials and 16 papers attached to it involve non-Hodgkin lymphoma, CLL/SLL, or mixed lymphoid malignancies, indicating a likely disease-ontology mapping error rather than a genuine signal. Combined with a Blocking-severity data gap on TFDA warnings/contraindications and the drug’s unmarketed status in Taiwan, this candidate does not meet the bar to advance.
To proceed, the following is needed:
- TFDA package insert retrieval and parsing (resolves DG001, currently blocking safety review)
- DrugBank MOA confirmation (resolves DG002)
- Disease-ontology validation to determine whether the TxGNN “Hodgkins lymphoma” node was correctly mapped, or should be re-run against a corrected disease vocabulary
- If repurposing is still of interest, direct clinical or preclinical evidence in a genuine classical Hodgkin lymphoma population (JAK/STAT- or NF-κB-relevant models), since none currently exists in this pack
- Separately, Taiwan market-entry/licensing assessment, since duvelisib has 0 TFDA licenses today
Disclaimer
This content is for research purposes only and does not constitute medical advice. Clinical validation is required before any clinical application.