Melphalan Flufenamide

Melflufen is far more lipophilic than melphalan itself, so it crosses membranes passively and quickly rather than depending on amino acid transporters. Once inside, peptidases — especially aminopeptidase N (CD13) and leucine aminopeptidase, which multiple myeloma cells overexpress — hydrolyse the conjugate to release melphalan and its metabolites. Because the released payload is hydrophilic, it becomes trapped intracellularly, producing a reported 50-fold-plus enrichment of alkylator inside myeloma cells relative to plasma. Melphalan then cross-links DNA, triggers a DNA damage response and drives apoptosis. The mechanism is elegant; the clinical problem was that the peptidase-mediated enrichment did not translate into a survival advantage in transplant-exposed patients.

Targets: Aminopeptidase N (CD13) and leucine aminopeptidase (activating enzymes), DNA (interstrand cross-linking by released melphalan)

• · 40 mg IV on day 1, with dexamethasone 40 mg (20 mg if over 75) on days 1, 8, 15 and 22. Central venous access is preferred because of local irritation.

Work out your exact syringe units →

• very commonThrombocytopenia— Grade 3 or 4 in the majority of patients — the dominant dose-limiting toxicity.
• very commonNeutropenia— Frequently requires growth factor support and cycle delays.
• very commonAnaemia
• very commonFatigue, nausea and diarrhoea
• commonInfections including pneumonia— A direct consequence of the depth of myelosuppression.
• commonBleeding events— Driven by the thrombocytopenia; platelet transfusion support is common.
• rareSecondary malignancy— An expected long-term risk of any alkylator.

• Prior autologous stem cell transplant with progression within 36 months — this is precisely the subgroup in which the OCEAN trial showed shorter overall survival.
• Pregnancy and breastfeeding — it is genotoxic and embryo-fetal toxic.
• Severe pre-existing marrow failure.

• cautionOther myelosuppressive agents — Additive cytopenias in an already fragile marrow.
• conflictLive vaccines — Avoid during and for some period after treatment.
• cautionbortezomib — Sequenced rather than combined outside of trials; both add to cytopenias.

• · Full blood count at least weekly through the first cycles and before each dose.
• · Renal and hepatic function each cycle.
• · Active surveillance for infection and bleeding — patients need clear instructions on when to present.
• · Paraprotein and free light chains each cycle.

Withdrawn from the US market — the FDA's final decision to withdraw accelerated approval took effect in February 2024. Still authorised in the EU and EEA as Pepaxti for patients with at least three prior lines who are refractory to a proteasome inhibitor, an immunomodulator and an anti-CD38 antibody.

• Richardson et al. 2021, HORIZON phase 2 trial of melflufen plus dexamethasone, Journal of Clinical Oncology (trial)
• Schjesvold et al. 2022, OCEAN phase 3 trial of melflufen versus pomalidomide, Lancet Haematology (trial)
• FDA final decision withdrawing approval of Pepaxto, February 2024 (other)

The design idea is worth understanding because it is genuinely clever and it still failed. Melphalan is a phenylalanine mustard: it is charged, hydrophilic, and it enters cells through the LAT1 amino acid transporter, which is a saturable and rate-limiting step. Esterifying it to a fluorinated phenylalanine ethyl ester strips the charge and adds lipophilicity, so melflufen crosses membranes by passive diffusion instead — fast, unsaturable, and independent of transporter expression. Inside the cell, aminopeptidases hydrolyse the conjugate. Aminopeptidase N, also called CD13, and leucine aminopeptidase are both overexpressed in myeloma plasma cells, so the hydrolysis rate is highest in exactly the cells you want to hit. And crucially, the product is hydrophilic melphalan, which cannot diffuse back out. That combination — fast passive entry, enzymatic uncapping, hydrophilic trapping — produces the reported enrichment of intracellular alkylator relative to plasma. Melphalan then does what melphalan does: forms N7-guanine adducts and interstrand DNA cross-links, which stall replication forks, trigger the ATM and ATR damage response, and drive p53-dependent and independent apoptosis. There is also a reported effect on angiogenesis through the same peptidase-dependent enrichment in endothelial cells. So why did OCEAN show an overall survival detriment in transplant-exposed patients? The honest answer is that nobody fully knows, but the leading explanation is that the same marrow that has already been ablated once cannot tolerate a highly efficient alkylator delivery system, and the profound thrombocytopenia and neutropenia — grade 3 or 4 in the majority — translated into deaths that offset the progression-free survival gain. It is a good lesson in how efficient targeting can amplify a toxicity as reliably as it amplifies an effect.

The regulatory story is the main thing that goes wrong, and it is instructive rather than embarrassing. A drug with an elegant mechanism produced a 29 percent response rate in a desperate population, earned accelerated approval, and then the confirmatory trial showed people living less long on it than on the comparator — with the harm concentrated in exactly the patients whose marrow had already been ablated by transplant. The US approval was withdrawn in February 2024. The EU regulator reached a different conclusion for a narrower population defined by not having had a transplant, or by a long interval since one. Both regulators were looking at the same data. Clinically, the thing that goes wrong is the marrow: grade 3 or 4 thrombocytopenia in the majority, bleeding events that require transfusion support, neutropenic infection including pneumonia and sepsis. The practical error is treating a transplant-exposed patient who progressed within 36 months of transplant, which is the subgroup where OCEAN showed shorter survival. Secondary malignancy is the expected long-term alkylator risk. And the preparation error is time — a solution left sitting degrades, and a delayed chair is a wasted vial.

1. —Every 28 days — 40 mg intravenously over 30 minutes on day 1, with dexamethasone 40 mg on days 1, 8, 15 and 22 — reduced to 20 mg dexamethasone in patients over 75. Central venous access preferred for local irritation.
2. —Baseline, in moderate renal impairment — 30 mg for an eGFR of 30 to 45 mL/min/1.73 m2, because melphalan exposure rises about 32 percent and the platelet effect is larger.
3. —On recurrent grade 4 neutropenia or thrombocytopenia — Reduce to 30 mg. Cycles are routinely delayed for count recovery before any reduction is made.
4. —On further recurrence — Reduce to 20 mg. Below this there is no established dose.

• Aminopeptidase N (CD13 / ANPEP) — Not a binding target in the receptor sense — it is the activating enzyme

  Hydrolyses the conjugate intracellularly to release melphalan. Overexpression in myeloma plasma cells is what gives the selectivity.

• Leucine aminopeptidase (LAP3) — Activating enzyme

  Second hydrolytic route to the active payload.

• Genomic DNA, at N7 of guanine

  Interstrand and intrastrand cross-linking by the released bis-chloroethylamine mustard, stalling replication and triggering the DNA damage response.

• LAT1 amino acid transporter — Deliberately bypassed rather than targeted

  Free melphalan depends on this transporter to enter cells; melflufen does not, which is the pharmacological advance.

• HORIZON Phase 2, single-arm · n=157 · 2021

  Overall response rate in heavily pretreated relapsed and refractory myeloma including triple-class-refractory patients. Response rate around 29 percent overall, which supported US accelerated approval in 2021.

• OCEAN (OP-103) Phase 3, head-to-head · n=495 · 2022

  Progression-free survival versus pomalidomide, both with dexamethasone, in lenalidomide-refractory myeloma. Melflufen met the progression-free survival endpoint but overall survival favoured pomalidomide, with the detriment concentrated in patients who had undergone prior autologous stem cell transplant. This is the readout that ended the US approval.

The parent conjugate is gone from plasma within minutes, with an elimination half-life of 2.1 minutes; the released melphalan follows with a half-life of about 70 minutes. Counts fall over the first two to three weeks of each cycle and are the reason cycles get delayed rather than given on time. Paraprotein responses generally emerge over one to three cycles. There is no fast clinical signal here — the first thing most patients notice is the fatigue and the transfusion requirement, not a response.

The only partner in the approved regimen is dexamethasone, 40 mg weekly or 20 mg in patients over 75. Growth factor support with filgrastim is effectively part of the regimen in practice because the neutropenia is so predictable, and platelet transfusion support is routine. Anything else myelosuppressive stacks badly — this drug is already spending most of the marrow reserve available, and adding a second cytopenia-inducing agent outside a trial is how patients end up transfusion-dependent. Live vaccines are contraindicated during and for a period after treatment. Central venous access is strongly preferred, not for pharmacokinetic reasons — peripheral and central infusion are bioequivalent — but because the solution is locally irritating. The chemical constraint is time: melflufen is unstable in solution and the infusion must be completed within about 60 minutes of reconstitution, which means the pharmacy and the chair have to be coordinated. There is no meaningful CYP interaction because the molecule is consumed by peptidases and spontaneous hydrolysis rather than metabolised hepatically.

Against plain melphalan, the pharmacological advance is real — transporter-independent entry, peptidase-mediated intracellular release, hydrophilic trapping — and it does produce higher intracellular alkylator concentration. Against pomalidomide plus dexamethasone in the OCEAN comparison, it produced longer progression-free survival and shorter overall survival, which is the outcome that should make anyone sceptical of progression-free survival as a standalone endpoint. Against the modern relapsed-myeloma landscape of anti-CD38 antibodies, bispecifics and CAR-T, this is an older mechanism competing in a field that has moved decisively towards immunotherapy, and that context matters more than any head-to-head. Against zoptarelin doxorubicin, the other peptide-drug conjugate in this class, the pattern rhymes: an ingenious targeting mechanism that worked exactly as designed at the molecular level and did not deliver a survival benefit. Two entries in this catalogue, the same lesson.

Not characterised. The drug is withdrawn in the US and authorised in the EU under national pricing arrangements that are not public in any comparable form. Even if a figure existed it would not transfer across markets.

• ?Protein binding is not reported in the EMA product information and is left null.
• ?Blood-brain barrier penetration is unknown. The molecule is lipophilic by design, which would in principle favour penetration, but there is no human data and the plasma half-life of two minutes makes it largely moot.
• ?The reported 50-fold-plus intracellular enrichment of alkylator comes from preclinical work; I did not resolve a primary source for that specific multiple in this session.
• ?HORIZON enrolment of 157 and OCEAN enrolment of 495 are as commonly cited; I resolved both publication records but not participant-level detail.
• ?Why the overall survival detriment occurred in transplant-exposed patients is not established. Cumulative marrow injury is the leading explanation but it remains a hypothesis.
• ?The FDA and EMA reached opposite conclusions on the same dataset. Which regulator is right is genuinely unresolved, and this record does not take a position.

• Melflufen and Dexamethasone in Heavily Pretreated Relapsed and Refractory Multiple Myeloma Richardson PG, Oriol A, Larocca A, et al., Journal of Clinical Oncology, 2021 · PMID 33296242

  HORIZON. The single-arm phase 2 that produced the accelerated approval.

• Melflufen or pomalidomide plus dexamethasone for patients with multiple myeloma refractory to lenalidomide (OCEAN): a randomised, head-to-head, open-label, phase 3 study Schjesvold FH, Dimopoulos MA, Delimpasi S, et al., The Lancet Haematology, 2022 · PMID 35032434

  The confirmatory trial. Met its progression-free survival endpoint and lost on overall survival, which is a combination that should make anyone cautious about progression-free survival as a surrogate.

• Benefit Versus Risk Assessment of Melflufen and Dexamethasone in Relapsed/Refractory Multiple Myeloma: Analyses From Longer Follow-up of the OCEAN and HORIZON Studies Sonneveld P, Richardson PG, Ludwig H, et al., Clinical Lymphoma, Myeloma & Leukemia, 2023 · PMID 37355418

  The longer-follow-up reanalysis that underpins the argument for the narrower transplant-naive population the EU authorisation covers.

• Pepaxti (melphalan flufenamide) — EMA summary of product characteristics, section 5.2 Oncopeptides AB / European Medicines Agency, EMA product information

  Source of every pharmacokinetic number here: 2.1-minute parent half-life, 70-minute melphalan half-life, 35 L and 76 L volumes, 692 and 23 L/h clearances, and the renal impairment exposure percentages.