Carfilzomib

The epoxyketone warhead reacts with the N-terminal threonine of the beta-5 subunit to form a stable morpholino adduct, which is essentially irreversible — recovery requires synthesis of new proteasomes rather than dissociation of drug. That gives more complete and prolonged inhibition than bortezomib and explains why activity persists despite a plasma half-life under an hour. The epoxyketone chemistry is also far more selective for the proteasome than boronates, which cross-react with serine proteases including those implicated in bortezomib's neuropathy — hence the dramatically lower rate of peripheral neuropathy. The trade-off is a class-recognised cardiovascular toxicity, thought to involve proteasome-dependent protein quality control in cardiomyocytes and endothelium, producing hypertension, heart failure and, rarely, thrombotic microangiopathy.

Targets: 26S proteasome beta-5 subunit (irreversible), Immunoproteasome LMP7 subunit, Unfolded protein response / ER stress

• · 20 mg/m2 on days 1 and 2 of cycle 1, then escalated to 27 mg/m2 (with lenalidomide-dexamethasone) or 56 mg/m2 over 30 minutes (with dexamethasone alone). Dexamethasone premedication and 250 to 500 mL of IV fluid before and after each dose in cycle 1.
• · 20 mg/m2 on day 1 of cycle 1, then 70 mg/m2 weekly. Fewer hospital visits with at least equivalent efficacy to the twice-weekly 27 mg/m2 schedule.

Work out your exact syringe units →

• very commonAnaemia and thrombocytopenia
• very commonFatigue
• very commonHypertension— Frequently needs new or intensified antihypertensive therapy; hypertensive crisis has been reported.
• commonInfusion reactions— Fever, chills, rigors, dyspnoea, arthralgia — usually within 24 hours, which is why dexamethasone premedication is built into cycle 1.
• commonCardiac failure and reduced ejection fraction— The signature carfilzomib toxicity, around 5 to 8 percent grade 3 or higher. Volume loading interacts with it, so hydration is deliberately balanced.
• commonDyspnoea— Must be worked up rather than assumed benign — it can herald cardiac or pulmonary toxicity.
• commonAcute renal failure— More frequent in heavily pretreated patients with high tumour burden.
• commonVenous thromboembolism— Thromboprophylaxis is recommended, particularly in immunomodulator-containing regimens.
• uncommonPeripheral neuropathy— Strikingly lower than with bortezomib — the main reason to choose it.
• rareThrombotic microangiopathy— Including TTP and HUS; stop the drug immediately if suspected.

• Pre-existing significant heart failure or recent myocardial infarction — the cardiac signal is real and dose-related.
• Pregnancy — embryo-fetal toxicity.
• Uncontrolled hypertension until it is controlled.

• redundantbortezomib — Same target. Carfilzomib retains activity in some bortezomib-refractory disease, but they are sequential options rather than a stack.
• synergyLenalidomide and dexamethasone — The ASPIRE triplet is one of the standard relapsed-myeloma backbones.
• cautionAggressive IV hydration — Needed to prevent tumour lysis and renal injury, yet excess volume worsens the cardiac toxicity. This balance is why it is an inpatient-grade decision.

• · Blood pressure at every visit and between visits at home.
• · Baseline echocardiogram in anyone with cardiac risk, plus prompt reassessment for new dyspnoea or oedema.
• · Full blood count before each cycle.
• · Renal function, uric acid and electrolytes, especially during cycle 1.
• · Paraprotein and free light chains each cycle.

FDA and EMA approved for relapsed or refractory multiple myeloma. Hospital or infusion-centre administration only.

• Stewart et al. 2015, ASPIRE phase 3 trial of carfilzomib-lenalidomide-dexamethasone, New England Journal of Medicine (trial)
• Dimopoulos et al. 2016, ENDEAVOR phase 3 trial of carfilzomib versus bortezomib, Lancet Oncology (trial)
• Moreau et al. 2018, ARROW once-weekly versus twice-weekly carfilzomib, Lancet Oncology (trial)
• FDA prescribing information for Kyprolis (label)

The epoxyketone warhead does something chemically elegant that a boronate cannot. It reacts with the beta-5 subunit's N-terminal threonine in two steps: the ketone forms a hemiketal with the threonine hydroxyl, and then the free alpha-amino group of that same threonine opens the epoxide, closing a six-membered morpholino ring. Two bonds, one residue, and the product is essentially irreversible. That double-bond requirement is also what makes it selective: only the proteasome's catalytic threonine presents a nucleophilic hydroxyl and a free alpha-amine in exactly that geometry, so the serine protease off-target reactivity that plagues boronates largely disappears. That is the molecular reason peripheral neuropathy drops from around 40 percent to single digits. The consequence of irreversibility is that recovery of proteasome activity requires synthesis of new proteasome, which takes roughly 24 hours in most tissues but longer in some, so the pharmacodynamic effect long outlives a plasma half-life under one hour. Duration of effect here is set by target turnover, not drug levels, which is why the extremely short half-life is not a problem. Carfilzomib also inhibits the immunoproteasome subunit LMP7 more completely than bortezomib does. Downstream biology is the same story as bortezomib — terminal unfolded protein response, NF-kB shutdown, NOXA stabilisation — but the inhibition is deeper and more sustained, and it retains activity in some bortezomib-refractory disease because incomplete or recoverable inhibition is one mechanism of bortezomib resistance. The cardiovascular toxicity is the price of that depth. Cardiomyocytes and endothelium depend on proteasome-mediated protein quality control to clear damaged contractile and mitochondrial proteins; complete and prolonged inhibition impairs that, and there is additional evidence for reduced nitric oxide bioavailability and endothelial dysfunction driving the hypertension. This is on-target toxicity in a tissue you did not intend to treat.

Cardiac events are the defining problem and the pattern is usually the same — an older patient with pre-existing hypertension, cardiac risk factors and amyloid that was never looked for, given generous fluids, who develops dyspnoea in cycle 1 or 2 that gets attributed to anaemia. By the time an echocardiogram is done, the ejection fraction has fallen. The prevention is unglamorous: control the blood pressure before starting, get a baseline echocardiogram in anyone with risk, screen for cardiac amyloid in patients with unexplained cardiac findings, keep the fluid restrained, and treat new dyspnoea as cardiac until proven otherwise. The second failure is skipping the cycle-1 20 mg/m2 step-up dose to save a visit, which raises tumour lysis and infusion reaction risk for no benefit. Third is thrombotic microangiopathy being missed because falling platelets in a myeloma patient on chemotherapy look ordinary — the discriminator is the combination with haemolysis, so LDH, haptoglobin and a film are the tests that matter. Fourth is reconstitution: adding sterile water too fast produces foam, foam makes the volume unmeasurable, and the classic technique error is shaking rather than directing the stream down the vial wall and letting it stand. Fifth, and easy to forget, is that this drug is intravenous only and needs a hospital visit, so a once-weekly schedule is not merely convenience — for many patients it is the difference between completing therapy and abandoning it.

1. —Cycle 1, days 1 and 2 — 20 mg/m2 over 10 to 30 minutes. This step is mandatory in every regimen and it is not a cautious clinician's choice — it exists to limit tumour lysis and infusion reactions, and skipping it is a documented way to cause harm. Dexamethasone 4 to 8 mg premedication and 250 to 500 mL of intravenous fluid before and after each dose in cycle 1.
2. —Cycle 1 day 8 onward, twice-weekly triplet — Escalate to 27 mg/m2 over 10 to 30 minutes, days 1, 2, 8, 9, 15 and 16 of a 28-day cycle, alongside lenalidomide and dexamethasone. This is the ASPIRE schedule.
3. —Cycle 1 day 8 onward, twice-weekly doublet — Escalate to 56 mg/m2 over 30 minutes with dexamethasone alone. This is the ENDEAVOR schedule and the reason the 30-minute infusion exists — the same AUC with a two to three-fold lower peak.
4. —Cycle 1 day 8 onward, once-weekly — Escalate to 70 mg/m2 over 30 minutes, days 1, 8 and 15 of a 28-day cycle. This is the ARROW schedule: fewer hospital visits and at least equivalent efficacy to twice-weekly 27 mg/m2.
5. —On grade 3 or 4 cardiac, renal, pulmonary or hepatic toxicity — Hold until resolution, then restart one dose level down — 70 to 56 to 45 to 36 mg/m2, or 56 to 45 to 36, or 27 to 20 to 15. A second occurrence at the reduced dose usually means stopping.
6. —From cycle 13 in the twice-weekly triplet — Days 8 and 9 are dropped, and in the ASPIRE protocol carfilzomib stops after cycle 18 while lenalidomide and dexamethasone continue. This de-intensification is written into the regimen, not an improvisation.

• 26S proteasome beta-5 subunit (PSMB5, chymotrypsin-like activity) — Low nanomolar IC50, commonly reported around 5 to 6 nM, but the relevant parameter is irreversibility rather than affinity — the covalent adduct does not dissociate

  Sustained, near-complete chymotrypsin-like inhibition until new proteasome is synthesised. Deeper and longer than bortezomib achieves.

• Immunoproteasome beta-5i subunit (LMP7 / PSMB8) — Also inhibited irreversibly

  Contributes to activity and to the immune effects. Selective LMP7 inhibitors are being developed precisely to separate these.

• Cardiomyocyte and endothelial proteasome — Same target, unintended tissue

  Impaired protein quality control and reduced nitric oxide bioavailability, producing hypertension, reduced ejection fraction and heart failure. On-target toxicity, which is why it cannot be engineered away without losing potency.

• Serine proteases — Minimal cross-reactivity, unlike boronates

  The absence of this off-target activity is the reason neuropathy is so much lower than with bortezomib.

• ASPIRE Phase 3 · n=792 · 2015

  Progression-free survival with carfilzomib, lenalidomide and dexamethasone versus lenalidomide and dexamethasone in relapsed myeloma. Median PFS 26.3 versus 17.6 months, hazard ratio 0.69.

• ENDEAVOR Phase 3 · n=929 · 2016

  Progression-free survival with carfilzomib 56 mg/m2 plus dexamethasone versus bortezomib plus dexamethasone. Median PFS 18.7 versus 9.4 months, hazard ratio 0.53, with grade 2 or higher peripheral neuropathy in 6 percent versus 32 percent. The definitive head-to-head between the two proteasome inhibitors.

• A.R.R.O.W. Phase 3 · n=478 · 2018

  Progression-free survival with once-weekly carfilzomib 70 mg/m2 plus dexamethasone versus twice-weekly 27 mg/m2 plus dexamethasone. Once-weekly was superior on progression-free survival with a comparable safety profile, which established the convenient schedule.

Proteasome inhibition is near-complete within minutes of the infusion and persists until new proteasome is made, roughly 24 hours. Infusion reactions — fever, chills, rigors, dyspnoea, arthralgia — appear within 24 hours of a dose and are concentrated in cycle 1, which is why dexamethasone premedication is built into that cycle. Hypertension often appears within the first two cycles. Cardiac failure clusters early too, with most grade 3 or higher events in the first three cycles rather than late. Paraprotein responses usually appear within one to two cycles. Peripheral neuropathy, if it comes at all, is late and mild.

The two established partners are lenalidomide plus dexamethasone, which is ASPIRE, and dexamethasone alone at the higher 56 mg/m2 dose, which is ENDEAVOR. Daratumumab-containing quadruplets are now common. The interaction that requires actual thought is hydration. Carfilzomib needs volume loading in cycle 1 to prevent tumour lysis and renal injury, and that same volume worsens the cardiac toxicity — so the protocol specifies 250 to 500 mL before and after each cycle-1 dose rather than aggressive open-ended fluids, and it is deliberately restrained. Getting that balance wrong in either direction causes harm. Thromboprophylaxis is recommended, particularly in immunomodulator-containing regimens, because venous thromboembolism rates are meaningful. Antiviral prophylaxis for zoster applies to this drug exactly as it does to bortezomib. Unlike bortezomib, the CYP interaction burden is light because metabolism is peptidase and epoxide-hydrolase driven, so the strong-inducer problem largely disappears. What does not combine: carfilzomib and bortezomib together. They hit the same subunit and they are sequential options, not partners.

Against bortezomib, the head-to-head answer from ENDEAVOR is unambiguous on efficacy — roughly double the progression-free survival — and equally unambiguous on neuropathy, 6 percent versus 32 percent at grade 2 or higher. The reason bortezomib is still the front-line workhorse is everything else: subcutaneous administration, no renal dose adjustment, generic pricing and no cardiac signal. Carfilzomib is the drug you reach for when neuropathy is the constraint or when bortezomib has failed, and the drug you avoid when the heart is the constraint. Against ixazomib, the oral boronate, carfilzomib is substantially more potent and substantially less convenient. Against selinexor, belantamab and the bispecific antibodies now crowding the relapsed setting, carfilzomib is the older mechanism with the longer track record. Against the once-weekly versus twice-weekly question within its own label, ARROW settled it in favour of once weekly at 70 mg/m2 for the doublet.

Not characterised here. Carfilzomib is a brand-only intravenous drug billed per vial by body surface area, delivered in an infusion centre, and almost always combined with lenalidomide or daratumumab whose costs are of the same order. Any single monthly figure would be a guess dressed up as data.

• ?The 5 to 6 nM IC50 for the beta-5 subunit is the commonly quoted figure but I did not resolve a primary source in this session, and for an irreversible inhibitor IC50 is time-dependent and somewhat uninformative anyway.
• ?Time to steady state is left null because the drug does not accumulate and the pharmacodynamic effect is governed by proteasome resynthesis rather than plasma concentration.
• ?The endothelial nitric oxide mechanism for carfilzomib hypertension is one hypothesis among several and is not settled.
• ?ASPIRE, ENDEAVOR and ARROW enrolment figures are as commonly cited; I resolved all three publication records but not participant-level detail from the full texts.
• ?The reported rate of grade 3 or higher cardiac failure varies between 5 and 8 percent across series and depends heavily on how systematically ejection fraction was assessed.
• ?Whether routine cardiac amyloid screening before carfilzomib changes outcomes has not been tested prospectively, though the pathophysiological argument is strong.

• Carfilzomib, lenalidomide, and dexamethasone for relapsed multiple myeloma Stewart AK, Rajkumar SV, Dimopoulos MA, et al., New England Journal of Medicine, 2015 · PMID 25482145

  ASPIRE. The triplet that converted accelerated approval to full approval and is still a standard relapsed-myeloma backbone.

• Carfilzomib and dexamethasone versus bortezomib and dexamethasone for patients with relapsed or refractory multiple myeloma (ENDEAVOR): a randomised, phase 3, open-label, multicentre study Dimopoulos MA, Moreau P, Palumbo A, et al., The Lancet Oncology, 2016 · PMID 26671818

  The direct comparison against bortezomib. Roughly double the progression-free survival and roughly a fifth of the neuropathy, which is the trade this whole class is defined by.

• Once weekly versus twice weekly carfilzomib dosing in patients with relapsed and refractory multiple myeloma (A.R.R.O.W.): interim analysis results of a randomised, phase 3 study Moreau P, Mateos MV, Berenson JR, et al., The Lancet Oncology, 2018 · PMID 29866475

  Established once-weekly 70 mg/m2 dosing, which halves the number of infusion visits.

• KYPROLIS (carfilzomib) for injection — US prescribing information Onyx Pharmaceuticals / Amgen, FDA prescribing information

  Source of the 28 L volume of distribution, 97 percent protein binding, under one hour half-life, 151 to 263 L/h clearance, peptidase and epoxide hydrolase metabolism, and the 25 percent urinary metabolite excretion with 0.3 percent parent drug.