REBLOZYL 75mg 50mg / ml powder for injection medication leaflet

Reblozyl 25 mg powder for solution for injection

Reblozyl 75 mg powder for solution for injection

Reblozyl 25 mg powder for solution for injection

Each vial contains 25 mg of luspatercept. After reconstitution, each mL of solution contains50 mg luspatercept.

Reblozyl 75 mg powder for solution for injection

Each vial contains 75 mg of luspatercept. After reconstitution, each mL of solution contains50 mg luspatercept.

Luspatercept is produced in Chinese Hamster Ovary (CHO) cells by recombinant DNAtechnology.

For the full list of excipients, see section 6.1.

Powder for solution for injection (powder for injection).

White to off-white lyophilised powder.

Reblozyl is indicated in adults for the treatment of transfusion-dependent anaemia due to verylow, low and intermediate-risk myelodysplastic syndromes (MDS) (see section 5.1).

Reblozyl is indicated in adults for the treatment of anaemia associated withtransfusion-dependent and non-transfusion-dependent beta-thalassaemia (see section 5.1).

Treatment should be initiated by a physician experienced in treatment of haematologicaldiseases.

Prior to each Reblozyl administration, the haemoglobin (Hb) level of patients should beassessed. In case of a red blood cell (RBC) transfusion occurring prior to dosing, thepre-transfusion Hb level must be considered for dosing purposes.

The recommended starting dose of Reblozyl is 1 mg/kg administered once every 3 weeks.

* Myelodysplastic syndromes

The recommended desired Hb concentration range is between 10 g/dL and 12 g/dL. Doseincrease for insufficient response is provided below.

Table 1: Dose increase for insufficient response

Dose at 1 mg/kg Dose increase

If after at least 2 consecutive doses at 1 mg/kg, a * Dose should be increased topatient: 1.33 mg/kg

* is not RBC transfusion- free, or

* does not reach Hb concentration of≥ 10 g/dL and the Hb increase is < 1 g/dL

Dose at 1.33 mg/kg Dose increase

If after at least 2 consecutive doses at 1.33 mg/kg, a * Dose should be increased topatient: 1.75 mg/kg

* is not RBC transfusion- free, or

* does not reach Hb concentration of≥ 10 g/dL and the Hb increase is < 1 g/dL

The dose increase should not occur more frequently than every 6 weeks (2 administrations) andshould not exceed the maximum dose of 1.75 mg/kg every 3 weeks. The dose should not beincreased immediately after a dose delay.

For patients with a pre-dose Hb level of > 9 g/dL and who have not yet achieved transfusionindependence, a dose increase may be required at the physician’s discretion; the risk of Hbincreasing above the target threshold with concomitant transfusion cannot be excluded.

If a patient loses response (i.e. transfusion independence), the dose should be increased by onedose level (see Table 2).

* Transfusion-dependent β-thalassaemia

In patients who do not achieve a response, defined as a reduction in RBC transfusion burden ofat least a third after ≥ 2 consecutive doses (6 weeks), at the 1 mg/kg starting dose, the doseshould be increased to 1.25 mg/kg. The dose should not be increased beyond the maximum doseof 1.25 mg/kg every 3 weeks.

If a patient loses response (if the RBC transfusion burden increases again after an initialresponse) the dose should be increased by one dose level (see Table 3).

* Non-transfusion-dependent β-thalassaemia

In patients who do not achieve or maintain a response, defined as an increase from baseline inpre-dose Hb of ≥ 1 g/dL, after ≥ 2 consecutive doses (6 weeks) at the same dose level (inabsence of transfusions, i.e. at least 3 weeks after the last transfusion), the dose should beincreased by one dose level (see Table 3). The dose should not exceed the maximum dose of1.25 mg/kg every 3 weeks.

Increase to next dose level

Increase to next dose level based on current dose is provided below.

Table 2: Increase to next dose level for MDS

Current dose Increased dose0.8 mg/kg 1 mg/kg1 mg/kg 1.33 mg/kg1.33 mg/kg 1.75 mg/kg

Table 3: Increase to next dose level for β-thalassaemia

Current dose Increased dose0.6 mg/kg* 0.8 mg/kg0.8 mg/kg 1 mg/kg1 mg/kg 1.25 mg/kg

* Applicable only to non-transfusion-dependent β-thalassaemia.

Dose reduction and dose delay

In case of Hb increase > 2 g/dL within 3 weeks in absence of transfusion compared with the Hbvalue at previous dose, Reblozyl dose should be reduced by one dose level.

If the Hb is ≥ 12 g/dL in the absence of transfusion for at least 3 weeks, the dose should bedelayed until the Hb is ≤ 11 g/dL. If there is also a concomitant rapid increase in Hb from the

Hb value at previous dose (> 2 g/dL within 3 weeks in absence of transfusion), a dose reductionto one step down should be considered after the dose delay.

Dose should not be reduced below 0.8 mg/kg (for MDS or transfusion-dependentβ-thalassaemia) and below 0.6 mg/kg (for non-transfusion-dependent β-thalassaemia).

Reduced dose during treatment with luspatercept are provided below.

Table 4: Reduced dose for MDS

Current dose Reduced dose1.75 mg/kg 1.33 mg/kg1.33 mg/kg 1 mg/kg1 mg/kg 0.8 mg/kg

Table 5: Reduced dose for β-thalassaemia

Current dose Reduced dose1.25 mg/kg 1 mg/kg1 mg/kg 0.8 mg/kg0.8 mg/kg 0.6 mg/kg*

* Applicable only to non-transfusion-dependent β-thalassaemia.

Dose modification due to adverse reactions

Instructions on dose interruptions or reductions for luspatercept treatment-related adversereactions are outlined in Table 6.

Table 6: Dose modification instructions

Treatment-related adverse reactions* Dose instructions

Grade 2 adverse reactions (see section 4.8), * Interrupt treatmentincluding Grade 2 hypertension (see sections 4.4 * Restart at previous dose when adverseand 4.8) reaction has improved or returned tobaseline

Grade ≥ 3 hypertension (see sections 4.4 and * Interrupt treatment4.8) * Restart at reduced dose once the bloodpressure is controlled as per dosereduction guidance

Treatment-related adverse reactions* Dose instructions

Other persistent Grade ≥ 3 adverse reactions * Interrupt treatment(see section 4.8) * Restart at previous dose or at reduceddose when adverse reaction has improvedor returned to baseline as per dosereduction guidance

Extramedullary haemopoiesis (EMH) masses * Discontinue treatmentcausing serious complications (see sections 4.4and 4.8)

* Grade 1: mild; Grade 2: moderate; Grade 3: severe; and Grade 4: life-threatening.

In case of a missed or delayed scheduled treatment administration, the patient should beadministered Reblozyl as soon as possible and dosing continued as prescribed with at least3 weeks between doses.

Patients experiencing a loss of response

If patients experience a loss of response to Reblozyl, causative factors (e.g. a bleeding event)should be assessed. If typical causes for a loss of haematological response are excluded, doseincrease should be considered as described above for the respective indication being treated (see

Table 2 and Table 3).

Reblozyl should be discontinued if patients do not experience a reduction in transfusion burden(for transfusion-dependent β-thalassaemia patients), or an increase from baseline Hb in theabsence of transfusions (for non-transfusion-dependent β-thalassaemia patients), or a decreasein transfusion burden including no increase from baseline Hb (for MDS patients) after 9 weeksof treatment (3 doses) at the maximum dose level, if no alternative explanations for responsefailure are found (e.g. bleeding, surgery, other concomitant illnesses) or if unacceptable toxicityoccurs at any time.

No starting dose adjustment is required for Reblozyl (see section 5.2). Limited data are availablein β-thalassaemia patients ≥ 60 years of age.

No starting dose adjustment is required for patients with total bilirubin (BIL) > upper limit ofnormal (ULN) and/or alanine aminotransferase (ALT) or aspartate aminotransferase (AST)< 3 x ULN (see section 5.2).

No specific dose recommendation can be made for patients with ALT or AST ≥ 3 x ULN orliver injury CTCAE Grade ≥ 3 due to lack of data (see section 5.2).

No starting dose adjustment is required for patients with mild or moderate renal impairment(individual estimated glomerular filtration rate [eGFR] 30 to 89 mL/min).

No specific dose recommendation can be made for patients with severe renal impairment(individual eGFR < 30 mL/min) due to lack of clinical data (see section 5.2). Patients with renalimpairment at baseline have been observed to have higher exposure (see section 5.2).

Consequently, these patients should be closely monitored for adverse reactions and doseadjustment should be managed accordingly (see Table 6).

There is no relevant use of Reblozyl in the paediatric population for the indication ofmyelodysplastic syndromes, or in paediatric patients less than 6 years of age in β-thalassaemia.

The safety and efficacy of Reblozyl in the paediatric patients aged from 6 years to less than18 years have not yet been established in β-thalassaemia. For non-clinical data, see section 5.3.

For subcutaneous use.

After reconstitution, Reblozyl solution should be injected subcutaneously into the upper arm,thigh or abdomen. The exact total dosing volume of the reconstituted solution required for thepatient should be calculated and slowly withdrawn from the single-dose vial(s) into a syringe.

The recommended maximum volume of medicinal product per injection site is 1.2 mL. If morethan 1.2 mL is required, the total volume should be divided into separate similar volumeinjections and administered across separate sites using the same anatomical location but onopposite sides of the body.

If multiple injections are required, a new syringe and needle must be used for each subcutaneousinjection. No more than one dose from a vial should be administered.

If the Reblozyl solution has been refrigerated after reconstitution, it should be removed from therefrigerator 15-30 minutes prior to injection to allow it to reach room temperature. This willallow for a more comfortable injection.

For instructions on reconstitution of the medicinal product before administration, seesection 6.6.

* Hypersensitivity to the active substance or to any of the excipients listed in section 6.1.

* Pregnancy (see section 4.6).

* Patients requiring treatment to control the growth of EMH masses (see section 4.4).

In order to improve the traceability of biological medicinal products, the name and the batchnumber of the administered product should be clearly recorded.

In β-thalassaemia patients, thromboembolic events (TEEs) were reported in 3.6% (8/223) ofpatients treated with luspatercept in the double-blind phase of the pivotal study intransfusion-dependent patients and in 0.7% (1/134) of patients during the open-label phase ofthe pivotal study in non-transfusion-dependent patients. Reported TEEs included deep veinthrombosis (DVT), portal vein thrombosis, pulmonary emboli, ischaemic stroke and superficialthrombophlebitis (see section 4.8). All patients with TEEs were splenectomised and had at leastone other risk factor for developing TEE (e.g. history of thrombocytosis or concomitant use ofhormone replacement therapy). The occurrence of TEE was not correlated with elevated Hblevels. The potential benefit of treatment with luspatercept should be weighed against thepotential risk of TEEs in β-thalassaemia patients with a splenectomy and other risk factors fordeveloping TEE. Thromboprophylaxis according to current clinical guidelines should beconsidered in patients with β-thalassaemia at higher risk.

In MDS patients, TEEs were reported in 3.9% (13/335) of patients treated with luspatercept.

Reported TEEs included cerebral ischemia and cerebrovascular accident in 1.2% (4/335) ofpatients. All TEEs occurred in patients with significant risk factors (atrial fibrillation, stroke orheart failure and peripheral vascular disease) and were not correlated with elevated Hb, plateletlevels or hypertension.

Extramedullary haemopoiesis masses

In transfusion-dependent β-thalassaemia patients, extramedullary haemopoiesis (EMH) masseswere observed in 3.2% (10/315) of patients treated with luspatercept in the pivotal study and inthe long-term follow-up study. Spinal cord compression symptoms due to EMH massesoccurred in 1.9% (6/315) of patients treated with luspatercept (see section 4.8).

In non-transfusion-dependent β-thalassaemia patients, EMH masses were observed in 6.3%(6/96) of patients treated with luspatercept in the pivotal study. Spinal cord compression due to

EMH masses occurred in 1.0% (1/96) of patients treated with luspatercept. During theopen-label portion of the study, EMH masses were observed in 2 additional patients for a totalof 8/134 (6.0%) of patients (see section 4.8).

Patients with EMH masses may experience worsening of these masses and complications duringtreatment. Signs and symptoms may vary depending on anatomical location. Patients should bemonitored at initiation and during treatment for symptoms and signs or complications resultingfrom the EMH masses, and be treated according to clinical guidelines. Treatment withluspatercept must be discontinued in case of serious complications due to EMH masses.

In MDS and β-thalassaemia pivotal studies, patients treated with luspatercept had an averageincrease in systolic and diastolic blood pressure of up to 5 mmHg from baseline (seesection 4.8).

An increased incidence of hypertension was observed in the first 12 months of treatment innon-transfusion-dependent β-thalassaemia patients treated with luspatercept (see section 4.8).

The treatment must be started only if the blood pressure is adequately controlled. Bloodpressure should be monitored prior to each luspatercept administration. Luspatercept dose mayrequire adjustment or may be delayed, and patients should be treated for hypertension as percurrent clinical guidelines (see Table 6 in section 4.2). The potential benefit of treatment with

Reblozyl should be re-evaluated in case of persistent hypertension or exacerbations ofpre-existing hypertension.

Traumatic fracture

In transfusion-dependent β-thalassaemia patients, traumatic fractures were observed in 0.4%(1/223) of patients treated with luspatercept.

In non-transfusion-dependent β-thalassaemia patients, traumatic fractures were observed in8.3% (8/96) of patients treated with luspatercept. Patients should be informed of the risk oftraumatic fracture.

This medicinal product contains less than 1 mmol sodium (23 mg) per dose, that is to sayessentially ‘sodium-free’.

Polysorbate 80 content

This medicinal product contains 0.1 mg of polysorbate 80 in each 25 mg vial or 0.3 mg ofpolysorbate 80 in each 75 mg vial which is equivalent to 0.2 mg/mL. Polysorbates may causeallergic reactions.

No formal clinical interaction studies have been performed. Concurrent use of iron-chelatingagents had no effect on luspatercept pharmacokinetics.

Women of childbearing potential have to use effective contraception during treatment with

Reblozyl and for at least 3 months after the last dose. Prior to starting treatment with Reblozyl, apregnancy test has to be performed for women of childbearing potential and the patient card hasto be provided.

Treatment with Reblozyl should not be started if the woman is pregnant (see section 4.3).

There are no data from the use of Reblozyl in pregnant women. Studies in animals have shownreproductive toxicity (see section 5.3). Reblozyl is contraindicated during pregnancy (seesection 4.3). If a patient becomes pregnant, Reblozyl should be discontinued.

It is unknown whether luspatercept or its metabolites are excreted in human milk. Luspaterceptwas detected in the milk of lactating rats (see section 5.3). Because of the unknown adverseeffects of luspatercept in newborns/infants, a decision must be made whether to discontinuebreast-feeding during therapy with Reblozyl and for 3 months after the last dose or todiscontinue Reblozyl therapy, taking into account the benefit of breast-feeding for the child andthe benefit of therapy for the woman.

The effect of luspatercept on fertility in humans is unknown. Based on findings in animals,luspatercept may compromise female fertility (see section 5.3).

Reblozyl has minor influence on the ability to drive and use machines. The ability to react whenperforming these tasks may be impaired due to risks of fatigue, vertigo, dizziness or syncope(see section 4.8). Therefore, patients should be advised to exercise caution until they know ofany impact on their ability to drive and use machines.

* Myelodysplastic syndromes

The most frequently reported adverse drug reactions in patients receiving Reblozyl (at least 15%of patients) were fatigue, diarrhoea, nausea, asthenia, dizziness, oedema peripheral and backpain. The most commonly reported Grade ≥ 3 adverse drug reactions (at least 2% of patients)included hypertension events (12.5%), syncope (3.6%), dyspnoea (2.7%), fatigue (2.4%) andthrombocytopenia (2.4%). The most commonly reported serious adverse drug reactions (at least1% of patients) were urinary tract infection (1.8%), dyspnoea (1.5%) and back pain (1.2%).

Asthenia, fatigue, nausea, diarrhoea, hypertension, dyspnoea, dizziness and headache occurredmore frequently during the first 3 months of treatment.

Treatment discontinuation due to an adverse event occurred in 10.1% of patients treated withluspatercept. The most common reason for discontinuation in the luspatercept treatment armwas progression of underlying MDS.

Dose delays due to pre-dose Hb ≥ 12 g/dL occurred in 24.3% of luspatercept treated patients.

* Transfusion-dependent β-thalassaemia

The most frequently reported adverse drug reactions in patients receiving Reblozyl (at least 15%of patients) were headache, bone pain and arthralgia. The most commonly reported Grade ≥ 3adverse drug reaction was hyperuricaemia. The most serious adverse reactions reported includedthromboembolic events of deep vein thrombosis, ischaemic stroke portal vein thrombosis andpulmonary embolism (see section 4.4).

Bone pain, asthenia, fatigue, dizziness and headache occurred more frequently during the first3 months of treatment.

Treatment discontinuation due to an adverse reaction occurred in 2.6% of patients treated withluspatercept. The adverse reactions leading to treatment discontinuation in the luspatercepttreatment arm were arthralgia, back pain, bone pain and headache.

* Non-transfusion-dependent β-thalassaemia

The most frequently reported adverse drug reactions in patients receiving Reblozyl (at least 15%of patients) were bone pain, headache, arthralgia, back pain, prehypertension and hypertension.

The most commonly reported Grade ≥ 3 and most serious adverse reaction (at least 2% ofpatients) reported was traumatic fracture. Spinal cord compression due to EMH masses occurredin 1% of patients.

Bone pain, back pain, upper respiratory tract infection, arthralgia, headache and prehypertensionoccurred more frequently during the first 3 months of treatment.

The majority of adverse drug reactions were non-serious and did not require discontinuation.

Treatment discontinuation due to an adverse reaction occurred in 3.1% of patients treated withluspatercept. Adverse reactions leading to treatment discontinuation were spinal cordcompression, extramedullary haemopoiesis and arthralgia.

The highest frequency for each adverse reaction that was observed and reported in patients inthe pivotal studies in MDS, β-thalassaemia and the long-term follow-up study is shown in

Table 7 below. The adverse reactions are listed below by body system organ class and preferredterm. Frequencies are defined as: very common (≥ 1/10), common (≥ 1/100 to < 1/10),uncommon (≥ 1/1 000 to < 1/100), rare (≥ 1/10 000 to < 1/1 000), very rare (< 1/10 000) and notknown (frequency cannot be estimated from the available data).

Table 7: Adverse drug reactions (ADRs) in patients treated with Reblozyl for MDS and /or β-thalassaemia in the four pivotal studies

System organ class Preferred term Frequency Frequency(all grades) for (all grades) for

MDS β-thalassaemia

Infections and infestations bronchitis Common Commonaurinary tract Very common Commonainfectionrespiratory tract Commoninfectionupper respiratory Common Very commonatract infectioninfluenza Common Very common

Blood and lymphatic system extramedullary Not known VII Commondisorders haemopoiesis VIthrombocytopenia Common

Immune system disorders hypersensitivity I, VI Common Common

Metabolism and nutrition hyperuricaemia Common Commondisorders dehydration Commondecreased appetite Commonelectrolyte Very commonimbalanceIX

Psychiatric disorders insomnia Common Very commonbanxiety Common Commonirritability Commonconfusional state Common

Nervous system disorders dizziness Very common Very commonheadache Very common Very commonmigraine Commonbspinal cord CommoncompressionVIsyncope/presyncope Common Commona

Ear and labyrinth disorders vertigo/vertigo Common Commonapositional

Cardiac disorders atrial fibrillation Commoncardiac failure Common

Vascular disorders prehypertension Very commonbhypertensionII, VI Very common Very commontachycardia Commonthromboembolic Common CommoneventsIV, VI

Respiratory, thoracic and cough Very commonmediastinal disorders epistaxis Common CommonbdyspnoeaVIII Very common Common

System organ class Preferred term Frequency Frequency(all grades) for (all grades) for

MDS β-thalassaemia

Gastrointestinal disorders abdominal pain Common Very commonbabdominal Commondiscomfortdiarrhoea Very common Very commonanausea Very common Very common

Skin and subcutaneous tissue hyperhidrosis Commondisorders

Musculoskeletal and connective back pain Very common Very commontissue disorders arthralgiaVI Common Very commonbone painVI Common Very commonmyalgia Commonmuscular weakness Common

Renal and urinary disorders proteinuria Commonbalbuminuria Commonbkidney injuryX Common

General disorders and non-cardiac chest Commonadministration site conditions paininfluenza-like Commonillnessfatigue Very common Very commonaasthenia Very common Very commoninjection site Common CommonreactionsIII, VIoedema peripheral Very common

Investigations alanine Common CommonVaminotransferaseincreasedaspartate Common Very commonVaminotransferaseincreasedblood bilirubin Common Very commonVincreasedgamma- Commonglutamyltransferaseincreased

Injury, poisoning and procedural traumatic fractureVI Commonbcomplications

The four pivotal studies are ACE-536-MDS-001(ESA-refractory or -intolerant MDS), ACE-536-MDS-002 (MDS),

ACE-536-B-THAL-001 (transfusion-dependent β-thalassaemia) and ACE-536-B-THAL-002(non-transfusion-dependent β-thalassaemia).

I Hypersensitivity includes eyelid oedema, drug hypersensitivity, swelling face, periorbital oedema, face oedema,angioedema, lip swelling, drug eruption.

II Hypertension includes essential hypertension, hypertension and hypertensive crisis.

III Injection site reactions include injection site erythema, injection site pruritus, injection site swelling and injectionsite rash.

IV TEEs include deep vein thrombosis, portal vein thrombosis, ischaemic stroke and pulmonary embolism.

V Frequency is based on laboratory values of any grade.

VI See section 4.8 Description of selected adverse reactions.

VII Reported only in post-marketing.

VIII Dyspnoea includes dyspnoea exertional for ACE-536-MDS-002.

IX Electrolyte imbalance includes bone, calcium, magnesium and phosphorus metabolism disorders and electrolyteand fluid balance conditions.

X ADR includes similar/grouped terms.a ADRs observed in transfusion-dependent β-thalassaemia study ACE-536-B-THAL-001.b ADRs observed in non-transfusion-dependent β-thalassaemia study ACE-536-B-THAL-002.

Bone pain

Bone pain was reported in 2.4% of MDS patients treated with luspatercept with all events being

Grade 1-2.

Bone pain was reported in 19.7% of transfusion-dependent β-thalassaemia patients treated withluspatercept (placebo 8.3%) with most events (41/44) being Grade 1-2, and 3 events Grade 3.

One of the 44 events was serious, and 1 event led to treatment discontinuation. Bone pain wasmost common in the first 3 months (16.6%) compared to months 4-6 (3.7%).

Bone pain was reported in 36.5% of non-transfusion-dependent β-thalassaemia patients treatedwith luspatercept (placebo 6.1%) with most events (32/35) being Grade 1-2, and 3 events

Grade 3. No patient discontinued due to bone pain.

Arthralgia

Arthralgia was reported in 7.2% of MDS patients treated with luspatercept with 0.6% being≥ Grade 3.

Arthralgia was reported in 19.3% of transfusion-dependent β-thalassaemia patients treated withluspatercept (placebo 11.9%) and led to treatment discontinuation in 2 patients (0.9%).

Arthralgia was reported in 29.2% of non-transfusion-dependent β-thalassaemia patients treatedwith luspatercept (placebo 14.3%) with most events (26/28) being Grade 1-2, and 2 events

Grade 3. Arthralgia led to treatment discontinuation in 1 patient (1.0%).

MDS and β-thalassaemia patients treated with luspatercept had an average increase in systolicand diastolic blood pressure of up to 5 mmHg from baseline not observed in patients receivingplacebo.

Hypertension events were reported in 12.5% of MDS patients treated with luspatercept (placebo9.2%). Grade 3 hypertension events were reported in 25/335 patients (7.5%) treated withluspatercept (placebo 3.9%).

Hypertension was reported in 19.8% of non-transfusion-dependent β-thalassaemia patientstreated with luspatercept (placebo 2.0%). Most events (16/19) were Grade 1-2 with 3 events

Grade 3 (3.1%) in patients treated with luspatercept (placebo 0.0%). An increased incidence ofhypertension was observed over time in the first 8-12 months in non-transfusion-dependentβ-thalassaemia patients treated with luspatercept. See section 4.4.

Hypertension was reported in 8.1% of transfusion-dependent β-thalassaemia patients treatedwith luspatercept (placebo 2.8%). See section 4.4. Grade 3 events were reported in 4 patients(1.8%) treated with luspatercept (placebo 0.0%).

Hypersensitivity-type reactions included eyelid oedema, drug hypersensitivity, swelling face,periorbital oedema, face oedema, angioedema, lip swelling, drug eruption.

Hypersensitivity-type reactions were reported in 4.6% of MDS patients (placebo 2.6%) with allevents being Grade 1-2 in patients treated with luspatercept.

Face oedema occurred in 3.1% of non-transfusion-dependent β-thalassaemia patients (placebo0.0%).

Hypersensitivity-type reactions were reported in 4.5% of transfusion-dependent β-thalassaemiapatients treated with luspatercept (placebo 1.8%) with all events being Grade 1-2.

Hypersensitivity led to treatment discontinuation in 1 patient (0.4%).

Injection site reactions included injection site erythema, injection site pruritus, injection siteswelling and injection site rash.

Injection site reactions were reported in 3.6% of MDS patients.

Injection site reactions were reported in 2.2% of transfusion-dependent β-thalassaemia patients(placebo 1.8%) with all events Grade 1 and none leading to discontinuation.

Injection site reactions were reported in 5.2% of non-transfusion-dependent β-thalassaemiapatients (placebo 0.0%) with all events Grade 1 and none leading to discontinuation.

TEEs included deep vein thrombosis, portal vein thrombosis, ischaemic stroke and pulmonaryembolism.

TEEs were reported in 3.9% of MDS patients (placebo 3.9%). Reported TEEs included cerebralischemia and cerebrovascular accident in 1.2% of patients. All TEEs occurred in patients withsignificant risk factors (atrial fibrillation, stroke or heart failure and peripheral vascular disease)and were not correlated with elevated Hb, platelet levels or hypertension. See section 4.4.

TEEs occurred in 3.6% of transfusion-dependent β-thalassaemia patients receiving luspatercept(placebo 0.9%).

TEE (superficial thrombophlebitis) occurred in 0.7% of patients in the open-label phase of thepivotal study in non-transfusion-dependent β-thalassaemia.

All TEEs events were reported in patients who had undergone splenectomy and had at least oneother risk factor. See section 4.4.

Extramedullary haemopoiesis masses

EMH masses occurred in 10/315 (3.2%) transfusion-dependent β-thalassaemia patientsreceiving luspatercept (placebo 0.0%). Five events were Grade 1-2, 4 events were Grade 3, and1 event was Grade 4. Three patients discontinued due to EMH masses. See section 4.4.

EMH masses occurred in 6/96 (6.3%) non-transfusion-dependent β-thalassaemia patientsreceiving luspatercept (placebo 2.0%). Most (5/6) were Grade 2 and 1 was Grade 1. One patientdiscontinued due to EMH masses. During the open-label portion of the study, EMH masseswere observed in 2 additional patients for a total of 8/134 (6.0%) of patients. Most (7/8) were

Grade 1-2 and manageable with standard clinical practice. In 6/8 patients, luspatercept wascontinued after onset of event. See section 4.4.

EMH masses may also occur after extended treatment with luspatercept (i.e. after 96 weeks).

Spinal cord compression

Spinal cord compression or symptoms due to EMH masses occurred in 6/315 (1.9%)transfusion-dependent β-thalassaemia patients receiving luspatercept (placebo 0.0%). Fourpatients discontinued treatment due to Grade ≥ 3 symptoms of spinal cord compression.

Spinal cord compression due to EMH masses occurred in 1/96 (1.0%)non-transfusion-dependent β-thalassaemia patient with a history of EMH masses receivingluspatercept (placebo 0.0%). This patient discontinued treatment due to Grade 4 spinal cordcompression. See section 4.4.

Traumatic fracture

Traumatic fracture occurred in 1 (0.4%) transfusion-dependent β-thalassaemia patient receivingluspatercept (placebo 0.0%).

Traumatic fracture occurred in 8 (8.3%) non-transfusion-dependent β-thalassaemia patientsreceiving luspatercept (placebo 2.0%) with Grade ≥ 3 events reported for 4 patients (4.2%)treated with luspatercept and in 1 patient (2.0%) receiving placebo.

In clinical studies in MDS, an analysis of 395 MDS patients who were treated with luspaterceptand who were evaluable for the presence of anti-luspatercept antibodies showed that 36 (9.1%)patients tested positive for treatment -emergent anti-luspatercept antibodies, including 18 (4.6%)patients who had neutralising antibodies against luspatercept.

In clinical studies in transfusion-dependent and non-transfusion-dependent β-thalassaemia, ananalysis of 380 β-thalassaemia patients who were treated with luspatercept and who wereevaluable for the presence of anti-luspatercept antibodies showed that 7 (1.84%) patients testedpositive for treatment emergent anti-luspatercept antibodies, including 5 (1.3%) patients whohad neutralising antibodies against luspatercept.

Luspatercept serum concentration tended to decrease in the presence of anti-luspaterceptantibodies. There were no severe systemic hypersensitivity reactions reported for patients withanti-luspatercept antibodies. There was no association between hypersensitivity type reactionsor injection site reactions and presence of anti-luspatercept antibodies. Patients withtreatment -emergent anti-luspatercept antibodies were more likely to report a serioustreatment -emergent adverse event (69.4% [25/36] for anti-luspatercept antibodies -positivepatients vs. 45.7% [164/359] for anti-luspatercept antibodies -negative patients) or a Grade 3 or4 treatment -emergent adverse event (77.8% [28/36] for anti-luspatercept antibodies -positivepatients vs. 56.8% [204/359] for anti-luspatercept antibodies -negative patients) compared topatients without anti-luspatercept antibodies in the TD MDS pool.

Other special population

MDS patients without ring sideroblast (RS-)

RS- patients are more likely to experience serious adverse events, Grade 5 treatment -emergentadverse events, adverse events leading to drug discontinuation or dose reduction compared topatients with ring sideroblasts (RS+). In ACE-536-MDS-002 study, RS- patients showed higherincidence of some adverse reactions compared to RS+ patients in both treatment arms. Whencomparing RS subgroups in the luspatercept arm, asthenia, nausea, vomiting, dyspnoea, cough,thromboembolic events, alanine aminotransferase increased, aspartate aminotransferaseincreased, and thrombocytopenia occurred more frequently in the RS- subgroup.

MDS patients with mutational status SF3B1 non-mutated

Patients with mutational status SF3B1 non-mutated are more likely to experience Grade 3 or 4treatment -emergent adverse events, serious adverse events, Grade 5 treatment -emergentadverse events, adverse events leading to drug discontinuation, dose reduction as well as doseinterruption compared to patients with mutational status SF3B1 mutated. Known luspaterceptadverse reactions with a frequency ≥ 3% higher in the non-mutated SF3B1 luspatercept armsubgroup included vomiting, dyspnoea, and hypertension.

Reporting suspected adverse reactions after authorisation of the medicinal product is important.

It allows continued monitoring of the benefit/risk balance of the medicinal product. Healthcareprofessionals are asked to report any suspected adverse reactions via the national reportingsystem listed in Appendix V.

Overdose with luspatercept may cause an increase of Hb values above the desired level. In theevent of an overdose, treatment with luspatercept should be delayed until Hb is ≤ 11 g/dL.

Pharmacotherapeutic group: Antianaemic preparations, other antianaemic preparations, ATCcode: B03XA06.

Luspatercept, an erythroid maturation agent, is a recombinant fusion protein that binds selectedtransforming growth factor-β (TGF-β) superfamily ligands. By binding to specific endogenousligands (e.g. GDF-11, activin B) luspatercept inhibits Smad2/3 signalling, resulting in erythroidmaturation through expansion and differentiation of late-stage erythroid precursors(normoblasts) in the bone marrow, thereby restoring effective erythropoiesis. Smad2/3signalling is abnormally high in disease models characterised by ineffective erythropoiesis, i.e.

MDS and β-thalassaemia, and in the bone marrow of MDS patients.

Somatic mutations in MDS patients

Luspatercept demonstrated clinical benefit and favourability over epoetin alfa across multiplegenomic mutations that are frequently observed in lower-risk MDS with the exception of CBLgene mutations.

* Myelodysplastic syndromes

The efficacy and safety of luspatercept were evaluated in a Phase 3 multicentre, randomised,open-label, active controlled study COMMANDS (ACE-536-MDS-002) comparing luspaterceptversus epoetin alfa in patients with anaemia due to International Prognostic Scoring System-

Revised (IPSS-R) very low-, low- or intermediate-risk MDS or withmyelodysplastic/ myeloproliferative neoplasm with ring sideroblasts and thrombocytosis(MDS/MPN RS-T) in ESA naïve patients (with endogenous sEPO levels of < 500 U/L) whorequire red blood cell transfusions. For eligibility, patients were required to have had 2 to6 RBC units/8 weeks confirmed for a minimum of 8 weeks immediately precedingrandomization. Patients with deletion 5q (del5q) MDS were excluded from the study.

Patients were treated for at least 24 weeks, unless the patient experienced unacceptabletoxicities, withdrew the consent or met any other treatment discontinuation criteria. Thetreatment was continued beyond week 24 in case of clinical benefit (defined as a transfusionreduction of ≥ 2 pRBC units/8 weeks compared with baseline) and absence of diseaseprogression. Based on the outcome of these assessments, patients either were discontinued fromtreatment and entered into the Post-Treatment Follow-up Period or continued open-labeltreatment (with luspatercept or epoetin alfa) as long as the above criteria continued to be met oruntil the patient experienced unacceptable toxicities, withdrew consent, or met any otherdiscontinuation criteria.

A total of 363 patients were randomized to receive subcutaneously luspatercept (N = 182) orepoetin alfa (N = 181) at 1 mg/kg every 3 weeks or at 450 U/kg every week, respectively.

Randomization was stratified by RBC transfusion burden, RS status, and endogenous serumerythropoietin (sEPO) level at baseline. Two dose level increases were allowed for luspatercept(to 1.33 mg/kg and to 1.75 mg/kg). Doses were held and subsequently reduced for adversereactions, reduced if the haemoglobin increased by ≥ 2 g/dL from the prior cycle, and held if thepre-dose haemoglobin was ≥ 12 g/dL. All patients received best supportive care, which included

RBC transfusions, use of antibiotic, antiviral and antifungal therapy, and nutritional support asneeded. BSC for this study excluded the use of ESAs outside of the study treatment. The keybaseline disease characteristics in MDS patients in ACE-536-MDS-002 are shown in Table 8.

Table 8: Baseline demographics and disease characteristics of MDS patients in

ACE-536-MDS-002

Luspatercept Epoetin alfa(N = 182) (N = 181)

Demographics

Agea (years)

Median (min, max) 74 (46, 93) 74 (31, 91)

Age categories, n (%)≤ 64 years 27 (14.8) 25 (13.8)65-74 years 68 (37.4) 66 (36.5)≥ 75 87 (47.8) 90 (49.7)

Sex, n (%)

Male 109 (59.9) 92 (50.8)

Female 73 (40.1) 89 (49.2)

Race, n (%)

Asian 19 (10.4) 25 (13.8)

Black 2 (1.1) 0

White 146 (80.2) 143 (79)

Not collected or reported 15 (8.2) 13 (7.2)

Disease Characteristics

Hb (g/dL), n (%)b

Median (min, max) 7.80 (4.7, 9.2) 7.80 (4.5, 10.2)

Time since original MDS diagnosis (months)c

Median 7.97 5.13

Serum EPO (U/L) categories, n (%)d≤ 200 145 (79.7) 144 (79.6)> 200 37 (20.3) 37 (20.4)

Median serum EPO 77 245 85 370

Serum ferritin (mcg/L) 623.00 650.00

Median (min, max) (12.4, 3170.0) (39.4, 6960.5)

Baseline transfusion burden/8 weekse (pRBCunits), n (%)< 4 units 118 (64.8) 111 (61.3)≥ 4 units 64 (35.2) 70 (38.7)

Luspatercept Epoetin alfa(N = 182) (N = 181)

MDS Classification WHO 2016 at baseline, n (%)

MDS-SLD 1 (0.5) 4 (2.2)

MDS-MLD 50 (27.5) 47 (26.0)

MDS-RS-SLD 2 (1.1) 6 (3.3)

MDS-RS-MLD 127 (69.8) 118 (65.2)

MDS/MPN-RS-T 2 (1.1) 5 (2.8)

Missing 0 1 (0.6)

IPSS-R classification risk category, n (%)

Very low 16 (8.8) 17 (9.4)

Low 130 (71.4) 133 (73.5)

Intermediate 34 (18.7) 29 (16.0)

Other/missing 2 (1.1) 2 (1.1)

Ring sideroblast status (per WHO criteria), n (%)

RS+ 133 (73.1) 130 (71.8)

RS- 49 (26.9) 50 (27.6)

Missing 0 1 (0.6)

SF3B1 mutation status, n (%)

Mutated 114 (62.6) 101 (55.8)

Non-mutated 65 (35.7) 72 (39.8)

Missing 3 (1.6) 8 (4.4)

Hb = haemoglobin; IPSSR = International Prognostic Scoring System-Revised; MDS-SLD = MDS with singlelineage dysplasia; MDS-MLD = MDS with multilineage dysplasia; MDS-RS-SLD = MDS with ring sideroblasts withsingle lineage dysplasia; MDS-RS-MLD = MDS with ring sideroblasts with multilineage dysplasia;

MDS/MPN-RS-T = myelodysplastic/myeloproliferative neoplasms with ring sideroblasts and thrombocytosis;

RS+ = with ring sideroblasts; RS- = without ring sideroblasts; SF3B1 = Splicing Factor 3B Subunit 1A MDSmutationa Age was calculated based on the informed consent signing date.b After applying the 14/3-day rule (only Hb values that are measured at least 14 days after a transfusion may be usedunless there is another transfusion within 3 days after the Hb assessment. If a transfusion within 3 days after the Hbassessment occurs, that Hb value will be used despite being < 14 days after the previous transfusion), the baseline Hbvalue (efficacy) is defined as the lowest Hb value from the central, or local laboratory, or pre-transfusion Hb fromtransfusion records that is within the 35 days prior to the first dose of study drug, if it was available.c The number of months from the date of original diagnosis to the date of informed consent.d Baseline EPO was defined as the highest EPO value within the 35 days preceding the first dose of study drug.e Collected over 8 weeks prior to randomisation.

The efficacy results are summarised below.

Table 9: Efficacy results in MDS patients in ACE-536-MDS-002

Endpoint Luspatercept Epoetin alfa(N = 182) (N = 181)

Primary endpoint

* RBC-TI for 12 weeks with associated concurrent mean Hb increase of ≥ 1.5 g/dL(Weeks 1-24)

Number of responders (response rate %) 110 (60.4) 63 (34.8)(95% CI) (52.9, 67.6) (27.9, 42.2)

Common Risk Difference (95% CI)a 25.4 (15.8, 35.0)p-value < 0.0001

Odds Ratio (95% CI)a 3.1 (2.0, pct. 4.8)

Secondary endpoints

* HI-E per IWG ≥ 8 weeks (Weeks 1-24)b

Number of responders (response rate %) 135 (74.2) 96 (53.0)(95% CI) (67.2, 80.4) (45.5, 60.5)

Common Risk Difference (95% CI)a 21.5 (12.2, 30.7)p-value < 0.0001

Odds Ratio (95% CI)a 2.8 (1.8, 4.5)

* RBC-TI for 24 weeks (Weeks 1-24)

Number of responders (response rate %) 87 (47.8) 56 (30.9)(95% CI) (40.4, 55.3) (24.3, 38.2)

Common Risk Difference (95% CI)a 16.3 (7.1, 25.4)p-value 0.0003

Odds Ratio (95% CI)a 2.3 (1.4, 3.7)

* RBC-TI for ≥ 24 weeks (Weeks 1-48) 163 167

Number of responders (response rate %) 99 (60.7) 66 (39.5)(95% CI) (52.8, 68.3) (32.1, 47.4)

Common Risk Difference (95% CI)a 20.7 (10.8, 30.6)p-value p < 0.0001c

Odds Ratio (95% CI)a 2.6 (1.6, pct. 4.3)

Hb = haemoglobin; RBC = red blood transfusiona Based on CMH test stratified by baseline RBC transfusion burden (< 4, ≥ 4 pRBC units), RS status (RS+, RS-) andsEPO level (≤ 200, > 200 U/L). 1-sided p-value is presented.b HI-E = haematological improvement - erythroid. The proportion of patients meeting the HI-E criteria as per

International Working Group (IWG) 2006 criteria sustained over a consecutive 56-day period during the indicatedtreatment period. For patients with baseline RBC transfusion burden of ≥ 4 units/8 weeks, HI-E was defined as areduction in RBC transfusion of at least 4 units/8 weeks. For patients with baseline RBC transfusion burden of< 4 units/8 weeks, HI-E was defined as a mean increase in Hb of ≥ 1.5 g/dL for 8 weeks in the absence of RBCtransfusions.c Nominal p-value

The treatment effect of luspatercept on RBC-TI ≥ 12 weeks and Hb increase of ≥ 1.5 g/dL washigher than epoetin alfa across all clinically relevant baseline demographic and most diseasecharacteristic subgroups, except in patients without ring sideroblasts, where the treatment effectof luspatercept was comparable to epoetin alfa.

* Myelodysplastic syndromes in ESA-refractory or -intolerant patients

The efficacy and safety of luspatercept were evaluated in a Phase 3 multicentre, randomised,double-blind, placebo-controlled study MEDALIST (ACE-536-MDS-001) in adult patients withanaemia requiring RBC transfusions (≥ 2 units/8 weeks) due to IPSS-R very low-, low- orintermediate-risk MDS who have ring sideroblasts (≥ 15%). Patients with del5q MDS orwithout ring sideroblasts (RS-) were not included in the study. Patients were required to haveeither received prior treatment with an ESA with inadequate response, to be ineligible for ESAs(determined to be unlikely to respond to ESA treatment with serum erythropoietin (EPO)> 200 U/L), or intolerant to ESA treatment.

Patients in both arms were treated for 24 weeks, then continued treatment if they haddemonstrated clinical benefit and absence of disease progression. The study was unblinded foranalyses when all patients had at least received 48 weeks of treatment or discontinued treatment.

A total of 229 patients were randomised to receive luspatercept 1 mg/kg (N = 153) or placebo(N = 76) subcutaneously every 3 weeks. A total of 128 (83.7%) and 68 (89.5%) patientsreceiving luspatercept and placebo respectively completed 24 weeks of treatment. A total of 78(51%) and 12 (15.8%) patients receiving luspatercept and placebo respectively completed48 weeks of treatment. Dose titration up to 1.75 mg/kg was allowed. Dose could be delayed orreduced depending upon Hb level. All patients were eligible to receive best supportive care(BSC), which included RBC transfusions, iron-chelating agents, use of antibiotic, antiviral andantifungal therapy, and nutritional support, as needed. The key baseline disease characteristicsin patients with MDS in study ACE-536-MDS-001 are shown in Table 10.

Table 10: Baseline demographics and disease characteristics of MDS patients with < 5%marrow blasts in study ACE-536-MDS-001

Luspatercept Placebo(N = 153) (N = 76)

Demographics

Agea (years)

Median (min, max) 71 (40, 95) 72 (26, 91)

Age categories, n (%)≤ 64 years 29 (19.0) 16 (21.1)65-74 years 72 (47.1) 29 (38.2)≥ 75 52 (34.0) 31 (40.8)

Sex, n (%)

Male 94 (61.4) 50 (65.8)

Female 59 (38.6) 26 (34.2)

Race, n (%)

Black 1 (0.7) 0 (0.0)

White 107 (69.9) 51 (67.1)

Not collected or reported 44 (28.8) 24 (31.6)

Other 1 (0.7) 1 (1.3)

Luspatercept Placebo(N = 153) (N = 76)

Disease characteristics

Serum EPO (U/L) categoriesb, n (%)< 200 88 (57.5) 50 (65.8)200 to 500 43 (28.1) 15 (19.7)> 500 21 (13.7) 11 (14.5)

Missing 1 (0.7) 0

Serum ferritin (mcg/L)

Median (min, max) 1089.2 1122.1(64, 5968) (165, 5849)

IPSS-R classification risk category, n (%)

Very low 18 (11.8) 6 (7.9)

Low 109 (71.2) 57 (75.0)

Intermediate 25 (16.3) 13 (17.1)

Other 1 (0.7) 0

Baseline RBC transfusion burden/ 8 weeksc, n (%)≥ 6 units 66 (43.1) 33 (43.4)≥ 6 and < 8 units 35 (22.9) 15 (20.2)≥ 8 and < 12 units 24 (15.7) 17 (22.4)≥ 12 units 7 (4.6) 1 (1.3)< 6 units 87 (56.9) 43 (56.6)≥ 4 and < 6 units 41 (26.8) 23 (30.3)< 4 units 46 (30.1) 20 (26.3)

Hbd (g/dL)

Median (min, max) 7.6 (6, 10) 7.6 (5, 9)

SF3B1, n (%)

Mutated 149 (92.2) 65 (85.5)

Unmutated 12 (7.8) 10 (13.2)

Missing 0 1 (1.3)

EPO = erythropoietin; Hb = haemoglobin; IPSS-R = International Prognostic Scoring System-Reviseda Age was calculated based on the informed consent signing date.b Baseline EPO was defined as the highest EPO value within 35 days of the first dose of study drug.c Collected over 16 weeks prior to randomisation.d Baseline Hb was defined as the last value measured on or before the date of the first dose of investigationalproduct (IP). After applying the 14/3-day rule, baseline Hb was defined as the lowest Hb value that was within35 days on or prior to the first dose of IP.

The efficacy results are summarised below.

Table 11: Efficacy results in patients with MDS in study ACE-536-MDS-001

Endpoint Luspatercept Placebo(N = 153) (N = 76)

Primary endpoint

* RBC-TI ≥ 8 weeks (Weeks 1-24)

Number of responders (response rate %) 58 (37.9) 10 (13.2)

* Common risk difference on response rate (95% CI) 24.56 (14.48, 34.64)

Odds ratio (95% CI)a 5.065 (2.278, 11.259)p-valuea < 0.0001

Endpoint Luspatercept Placebo(N = 153) (N = 76)

Secondary endpoints

* RBC-TI ≥ 12 weeks (Weeks 1-24)

Number of responders (response rate %) 43 (28.1) 6 (7.9)

* Common risk difference on response rate (95% CI) 20.00 (10.92, 29.08)

Odds ratio (95% CI)a 5.071 (2.002, 12.844)p-valuea 0.0002

* RBC-TI ≥ 12 weeks (Weeks 1-48)

Number of responders (response rate %)b 51 (33.3) 9 (11.8)

* Common risk difference on response rate (95% CI) 21.37 (11.23, 31.51)

Odds ratio (95% CI)a 4.045 (1.827, 8.956)p-valuea 0.0003

Transfusion event frequencyc

* Weeks 1-24

Interval transfusion rate (95% CI) 6.26 (5.56, 7.05) 9.20 (7.98, 10.60)

Relative risk vs. placebo 0.68 (0.58, 0.80)

* Weeks 25-48

Interval transfusion rate (95% CI) 6.27 (5.47, 7.19) 8.72 (7.40, 10.28)

Relative risk vs. placebo 0.72 (0.60, 0.86)

RBC Transfusion unitsc

* Weeks 1-24

Baseline transfusion burden < 6 units/8 weeks

LS Mean (SE) 7.2 (0.58) 12.8 (0.82)95% CI for LS mean 6.0, 8.3 11.1, 14.4

LS mean difference (SE) (luspatercept vs. placebo) -5.6 (1.01)95% CI for LS mean difference -7.6, -3.6

Baseline transfusion burden ≥ 6 units/8 weeks

LS Mean (SE) 18.9 (0.93) 23.7 (1.32)95% CI for LS mean 17.1, 20.8 21.1, 26.4

LS mean difference (SE) (luspatercept vs. placebo) -4.8 (1.62)95% CI for LS mean difference -8.0, -1.6

* Weeks 25-48

Baseline transfusion burden < 6 units/8 weeks

LS Mean (SE) 7.5 (0.57) 11.8 (0.82)95% CI for LS mean 6.3, 8.6 10.1, 13.4

LS mean difference (SE) (luspatercept vs. placebo) -4.3 (1.00)95% CI for LS mean difference -6.3, -2.3

Baseline transfusion burden ≥ 6 units/8 weeks

LS Mean (SE) 19.6 (1.13) 22.9 (1.60)95% CI for LS mean 17.4, 21.9 19.7, 26.0

LS mean difference (SE) (luspatercept vs. placebo) -3.3 (1.96)95% CI for LS mean difference -7.1, 0.6

RBC-TI: RBC Transfusion Independent; CI: confidence interval; CMH = Cochran-Mantel-Haenszela CMH test stratified for average baseline transfusion burden (≥ 6 units vs. < 6 units per 8 weeks), and baseline

IPSS-R score (very low or low vs. intermediate).b After the Week 25 disease assessment visit, patients who were no longer deriving benefit discontinued therapy;few placebo patients contributed data for evaluation at the later timepoint compared with luspatercept(N = 12 vs. N = 78 respectively).c Post-hoc analysis using baseline imputation.

A treatment effect in favour of luspatercept over placebo was observed in most subgroupsanalysed using transfusion independence ≥ 12 weeks (during week 1 to week 24), includingpatients with high baseline endogenous EPO level (200-500 U/L) (23.3% vs. 0%, explorativeanalysis).

Only limited data are available for the group with transfusion burden of ≥ 8 units/8 weeks.

Safety and efficacy have not been established in patients with a transfusion burden of> 12 units/8 weeks.

Exploratory findings

Table 12: Exploratory efficacy results in patients with MDS in study ACE-536-MDS-001

Endpoint Luspatercept Placebo(N = 153) (N = 76)mHI-Ea

* Weeks 1-24

Number of responders (response rate %) 81 (52.9) 9 (11.8)(95% CI) (44.72, 61.05) (5.56, 21.29)

RBC transfusion reduction of 4 units/8 weeks, n (%) 52/107 (48.6) 8/56 (14.3)

Mean Hb increase of ≥ 1.5 g/dL for 8 weeks, n (%) 29/46 (63.0) 1/20 (5.0)

* Weeks 1-48

Number of responders (response rate %) 90 (58.8) 13 (17.1)(95% CI) (50.59, 66.71) (9.43, 27.47)

RBC transfusion reduction of 4 units/8 weeks, n (%) 58/107 (54.2) 12/56 (21.4)

Mean Hb increase of ≥ 1.5 g/dL for 8 weeks, n (%) 32/46 (69.6) 1/20 (5.0)

Mean change from baseline in mean serum ferritin with imputation by baseline (ITTpopulation)

Mean change from baseline in mean serum ferritin averagedover Weeks 9 through 24 (mcg/L)b

LS Mean (SE) 9.9 (47.09) 190.0 (60.30)95% CI for LS Mean -82.9, 102.7 71.2, 308.8

Treatment comparison (luspatercept vs. placebo)c

LS mean difference (SE) -180.1 (65.81)95% CI for LS mean difference -309.8, -50.4

Hb=haemoglobina mHI-E = modified haematological improvement - erythroid. The proportion of patients meeting the HI-E criteriaas per International Working Group (IWG) 2006 criteria sustained over a consecutive 56-day period during theindicated treatment period. For patients with baseline RBC transfusion burden of ≥ 4 units/8 weeks, mHI-E wasdefined as a reduction in RBC transfusion of at least 4 units/8 weeks. For patients with baseline RBC transfusionburden of < 4 units/8 weeks, mHI-E was defined as a mean increase in Hb of ≥ 1.5 g/dL for 8 weeks in the absenceof RBC transfusions.b If a patient did not have a serum ferritin value within the designated postbaseline interval, the serum ferritin isimputed from the baseline value.c Analysis of covariance was used to compare the treatment difference between groups (including nominal p-value),with the change in serum ferritin as the dependent variable, treatment group (2 levels) as a factor, and baselineserum ferritin value as covariates, stratified by average baseline RBC transfusion requirement (≥ 6 units vs.< 6 units of RBC per 8 weeks), and baseline IPSS-R (very low or low vs. intermediate).

The median duration of the longest RBC Transfusion Independent (RBC-TI) period amongresponders in the luspatercept treatment arm was 30.6 weeks.

A total of 62.1% (36/58) of the luspatercept responders who achieved RBC-TI ≥ 8 weeks from

Weeks 1-24 had 2 or more episodes of RBC-TI at the time of analysis.

* Transfusion-dependent β-thalassaemia

The efficacy and safety of luspatercept were evaluated in a Phase 3 multicentre, randomised,double-blind, placebo-controlled study BELIEVE (ACE-536-B-THAL-001) in adult patientswith transfusion-dependent β-thalassaemia-associated anaemia who require RBC transfusions(6-20 RBC units/24 weeks) with no transfusion-free period > 35 days during that period.

Patients in both the luspatercept and placebo arms were treated for at least 48 and up to96 weeks. After unblinding, placebo patients were able to cross-over to luspatercept.

A total of 336 adult patients were randomised to receive luspatercept 1 mg/kg (N = 224) orplacebo (N = 112) subcutaneously every 3 weeks. Dose titration to 1.25 mg/kg was allowed.

Dose could be delayed or reduced depending upon Hb level. All patients were eligible to receive

BSC, which included RBC transfusions, iron-chelating agents, use of antibiotic, antiviral andantifungal therapy, and nutritional support, as needed. The study excluded patients with Hb

S/β-thalassaemia or alpha (α)-thalassaemia or who had major organ damage (liver disease, heartdisease, lung disease, renal insufficiency). Patients with recent DVT or stroke or recent use of

ESA, immunosuppressant or hydroxyurea therapy were also excluded. The key baseline diseasecharacteristics in patients with β-thalassaemia in study ACE-536-B-THAL-001 are shown in

Table 13.

Table 13: Baseline demographics and disease characteristics of patients withtransfusion-dependent β-thalassaemia in study ACE-536-B-THAL-001

Luspatercept Placebo(N = 224) (N = 112)

Demographics

Age (years)

Median (min, max) 30.0 (18, 66) 30.0 (18, 59)

Age categories, n (%)≤ 32 129 (57.6) 63 (56.3)> 32 to ≤ 50 78 (34.8) 44 (39.3)> 50 17 (7.6) 5 (4.5)

Sex, n (%)

Male 92 (41.1) 49 (43.8)

Female 132 (58.9) 63 (56.3)

Race, n (%)

Asian 81 (36.2) 36 (32.1)

Black 1 (0.4) 0

White 122 (54.5) 60 (53.6)

Not collected or reported 5 (2.2) 5 (4.5)

Other 15 (6.7) 11 (9.8)

Disease characteristics

Pretransfusion Hb thresholda, 12-week run-in (g/dL)

Median (min, max) 9.30 (4.6, 11.4) 9.14 (6.2, 11.5)

Baseline transfusion burden 12 weeks

Median (min, max)(units/12 weeks) (Week -12 to Day 1) 6.12 (3.0, 14.0) 6.27 (3.0, 12.0)β-thalassaemia gene mutation grouping, n (%)β0/β0 68 (30.4) 35 (31.3)

Non-β0/β0 155 (69.2) 77 (68.8)

Missingb 1 (0.4) 0a The 12-week pretransfusion threshold was defined as the mean of all documented pretransfusions Hb values for apatient during the 12 weeks prior to Cycle 1 Day 1.b “Missing” category includes patients in the population who had no result for the parameter listed.

The study was unblinded for analyses when all patients had at least received 48 weeks oftreatment or discontinued treatment.

The efficacy results are summarised below.

Table 14: Efficacy results in patients with transfusion-dependent β-thalassaemia in study

ACE-536-B-THAL-001

Endpoint Luspatercept Placebo(N = 224) (N = 112)

Primary endpoint≥ 33% reduction from baseline in RBC transfusionburden with a reduction of at least 2 units for12 consecutive weeks compared to the 12-weekinterval prior to treatment

Weeks 13-24 47 (21.0) 5 (4.5)

Difference in proportions (95% CI)a 16.5 (10.0, 23.1)p-valueb < 0.0001

Secondary endpoints

Weeks 37-48 44 (19.6) 4 (3.6)

Difference in proportions (95% CI)a 16.1 (9.8, 22.3)p-valueb < 0.0001≥ 50% reduction from baseline in RBC transfusionburden with a reduction of at least 2 units for12 consecutive weeks compared to the 12-weekinterval prior to treatment

Weeks 13-24 16 (7.1) 2 (1.8)

Difference in proportions (95% CI)a 5.4 (1.2, 9.5)p-valueb 0.0402

Weeks 37-48 23 (10.3) 1 (0.9)

Difference in proportions (95% CI)a 9.4 (5.0, 13.7)p-valueb 0.0017

CI: confidence interval.a Difference in proportions (luspatercept + BSC - placebo + BSC) and 95% CIs estimated from the unconditionalexact test.b P-value from the Cochran Mantel-Haenszel test stratified by the geographical region.

Exploratory findings

Table 15: Exploratory efficacy results in patients with transfusion-dependentβ-thalassaemia in study ACE-536-B-THAL-001

Endpoint Luspatercept Placebo(N = 224) (N = 112)≥ 33% reduction from baseline in RBC transfusion burden with a reduction of at least2 units for 12 consecutive weeks compared to the 12-week interval prior to treatment

Any consecutive 12 weeks* 173 (77.2) 39 (34.8)

Difference in proportions (95% CI)a 42.4 (31.5, 52.5)

Any consecutive 24 weeks* 116 (51.8) 3 (2.7)

Difference in proportions (95% CI)a 49.1 (41.3, 56.2)

Endpoint Luspatercept Placebo(N = 224) (N = 112)≥ 50% reduction from baseline in RBC transfusion burden with a reduction of at least2 units for 12 consecutive weeks compared to the 12-week interval prior to treatment

Any consecutive 12 weeks* 112 (50.0) 9 (8.0)

Difference in proportions (95% CI)a 42.0 (32.7, 49.9)

Any consecutive 24 weeks* 53 (23.7) 1 (0.9)

Difference in proportions (95% CI)a 22.8 (16.5, 29.1)

Least square (LS) mean change from baseline in transfusion burden (RBC units/48 weeks)

Weeks 1 to Week 48

LS mean -4.69 +1.17

LS mean of difference (luspatercept-placebo) -5.86(95% CI)b (-7.04, -4.68)

Weeks 49 to Week 96

LS mean -5.43 +1.80

LS mean of difference (luspatercept-placebo) -7.23(95% CI)b (-13.84, -0.62)

ANCOVA = analysis of covariance; CI: confidence interval.a Difference in proportions (luspatercept + BSC - placebo + BSC) and 95% CIs estimated from the unconditionalexact test.b Estimates are based on ANCOVA model with geographical regions and baseline transfusion burden as covariates.

* Placebo patients are assessed up to prior to crossing over to luspatercept. For the rolling analyses at any consecutive12/24 weeks, luspatercept treatment arm does not include placebo patients who crossed over to luspatercept.

A reduction in mean serum ferritin levels was observed from baseline in the luspatercept armcompared to an increase in the placebo arm at Week 48 (-235.56 mcg/L vs. +107.03 mcg/Lwhich resulted in a least square mean treatment difference of -342.59 mcg/L(95% CI: -498.30, -186.87).

A total of 85% (147/173) of luspatercept responders who achieved at least a 33% reduction intransfusion burden during any consecutive 12-week interval achieved 2 or more episodes ofresponse at the time of analysis.

Non-transfusion-dependent β-thalassaemia

The efficacy and safety of luspatercept were evaluated in a Phase 2 multicentre, randomised,double-blind, placebo-controlled study BEYOND (ACE-536-B-THAL-002) in adult patientswith non-transfusion-dependent β-thalassaemia-associated anaemia(Hb concentration ≤ 10 g/dL).

A total of 145 adult patients receiving RBC transfusions (0-5 RBC units in the 24-week periodprior to randomization), with a baseline Hb level ≤ 10 g/dL (defined as average of at least 2 Hbmeasurements ≥ 1 week apart within 4 weeks prior to randomization) were randomized toreceive luspatercept (N = 96) or placebo (N = 49) subcutaneously every 3 weeks. Patients werestratified at randomization based on their baseline Hb level and their non-transfusion-dependentβ-thalassaemia (NTDT) patient-reported outcome (PRO; NTDT-PRO) Tiredness/Weakness(T/W) weekly domain score. Dose titration to 1.25 mg/kg was allowed. Dose could be delayedor reduced depending upon Hb level. Overall, 53% of luspatercept patients (N = 51) and 92% ofpatients on placebo (N = 45) had their dose increased to 1.25 mg/kg within the 48-weektreatment period. Among patients receiving luspatercept, 96% were exposed for 6 months orlonger and 86% were exposed for 12 months or longer. A total of 89 (92.7%) patients receivingluspatercept and 35 (71.4%) patients receiving placebo completed 48 weeks of treatment.

All patients were eligible to receive BSC, which included RBC transfusions, iron-chelatingagents, use of antibiotic, antiviral, and antifungal therapy, and nutritional support, as needed.

Concurrent treatment for anaemia with blood transfusions was allowed, at the discretion of thephysician, for low haemoglobin levels, symptoms associated with anaemia (e.g. haemodynamicor pulmonary compromise requiring treatment) or comorbidities. The study excluded patientswith Hb S/β-thalassaemia or alpha (α)-thalassaemia or who had major organ damage (liverdisease, heart disease, lung disease, renal insufficiency), active hepatitis C or B, or HIV.

Patients with recent DVT or stroke or recent use of ESA, immunosuppressant or hydroxyureatherapy, or on chronic anticoagulant or uncontrolled hypertension were also excluded. Only alimited number of patients with comorbidities associated with underlying anaemia such aspulmonary hypertension, liver and kidney disease and diabetes were included in the study.

The key baseline disease characteristics in the Intention-To-Treat (ITT) population withnon-transfusion-dependent β-thalassaemia in study ACE-536-B-THAL-002 are shown in

Table 16.

Table 16: Baseline demographics and disease characteristics of patients with non-transfusion-dependent β-thalassaemia in study ACE-536-B-THAL-002

ITT population

Luspatercept Placebo(N=96) (N=49)

Demographics

Age (years)

Median (min, max) 39.5 (18, 71) 41 (19, 66)

Sex, n (%)

Male 40 (41.7) 23 (46.9)

Female 56 (58.3) 26 (53.1)

Race, n (%)

Asian 31 (32.3) 13 (26.5)

White 59 (61.5) 28 (57.1)

Other 6 (6.3) 8 (16.3)

Disease characteristicsβ-thalassaemia diagnosis, n (%)β-thalassaemia 63 (65.6) 34 (69.4)

HbE/β-thalassaemia 28 (29.2) 11 (22.4)β-thalassaemia combined with α-thalassaemia 5 (5.2) 4 (8.2)

Baseline Hb levela (g/dL)

Median (min, max) 8.2 (5.3, 10.1) 8.1 (5.7, 10.1)

Patients with mean baseline Hb levela category(g/dL), n (%)< 8.5 55 (57.3) 29 (59.2)

Baseline NTDT-PRO T/W domain scoreb, n (%)

Median (min, max) 4.3 (0, 9.5) 4.1 (0.4, 9.5)

Baseline NTDT-PRO T/W domain scorebcategory, n (%)≥ 3 66 (68.8) 35 (71.4)

Baseline transfusion burden (units/24 weeks)

Median (min, max) 0 (0, 4) 0 (0, 4)

Splenectomy, n (%)

Yes 34 (35.4) 26 (53.1)

MRI LIC (mg/g dw)c, n 95 47

Median (min, max) 3.9 (0.8, 39.9) 4.1 (0.7, 28.7)

MRI spleen volume (cm3), n 60 22

Median (min, max) 879.9 1077.0(276.1, 2419.0) (276.5, 2243.0)

ITT population

Luspatercept Placebo(N=96) (N=49)

Baseline use of ICT, n (%) 28 (29.2) 16 (32.7)

Baseline serum ferritin (mcg/L)d 456.5 (30.0, 360.0 (40.0,

Median (min, max) 3528.0) 2265.0)

Hb = haemoglobin; HbE = haemoglobin E; ICT = Iron Chelation Therapy; LIC = liver iron concentration; max =maximum; min = minimum; MRI = magnetic resonance imaging; NTDT-PRO T/W = non-transfusion-dependentβ-thalassaemia patient-reported outcome tiredness and weakness domain score;a Mean of at least 2 Hb values by the central laboratory during the 28-day screening period.b Baseline defined as the average of non-missing NTDT-PRO T/W domain score over 7 days before Dose 1 Day 1.c The value of LIC was either the value collected from the electronic Case Report Form (eCRF) or the value derivedfrom T2*, R2*, or R2 parameter depending on which techniques and software were used for MRI LIC acquisition.d Baseline mean serum ferritin was calculated during the 24 weeks on or prior to Dose 1 Day 1. Baseline ICT wascalculated during the 24 weeks on or prior to Dose 1 Day 1.

The efficacy results are summarised below.

Table 17: Efficacy results in patients with non-transfusion-dependent β-thalassaemia instudy ACE-536-B-THAL-002

ITT population

Endpoint Luspatercept Placebo(N = 96) (N = 49)

Primary endpoint

Increase from baseline ≥ 1 g/dL in mean Hb over continuous 12-week interval (inabsence of transfusions)

* Weeks 13-24

Response ratea, n 74 0.0[(%) (95% CI)]b [(77.1) (67.4, 85.0)] [(0.0) (0.0, 7.3)]p-valuec < 0.0001

CI = confidence interval; Hb = haemoglobina Defined as number of patients with ≥ 1 g/dL Hb increase in the absence of RBC transfusion compared to baseline(i.e. the average of ≥ 2 Hb measurements at ≥ 1 week apart within 4 weeks before Dose 1 Day 1).b The 95% CI for response rate (%) was estimated from the Clopper-Pearson exact test.c The odds ratio (luspatercept vs. placebo) with 95% CI and p-value were estimated from the CMH test stratified bybaseline Hb category (< 8.5 vs. ≥ 8.5 g/dL) and baseline NTDT-PRO T/W domain score category (≥ 3 vs. < 3)defined at randomization as covariates.

Note: Patients with missing Hb at Weeks 13-24 were classified as non-responders in the analysis.

A total of 77.1% of luspatercept treated patients achieved an increase from baseline ≥ 1 g/dL inmean Hb over continuous 12-week interval (in absence of transfusions) (Weeks 13-24). Thiseffect was maintained in the 57.3% of patients who reached Week 144 of treatment.

* Myelodysplastic syndromes

The European Medicines Agency has waived the obligation to submit the results of studies with

Reblozyl in all subsets of the paediatric population in myelodysplastic syndromes (seesection 4.2 for information on paediatric use).

* β-thalassaemia

The European Medicines Agency has deferred the obligation to submit the results of studieswith Reblozyl in one or more subsets of paediatric population older than 6 years of age inβ-thalassaemia (see section 4.2 for information on paediatric use).

In healthy volunteers and patients, luspatercept is slowly absorbed following subcutaneousadministration, with the Cmax in serum often observed approximately 7 days post-dose across alldose levels. Population pharmacokinetic (PK) analysis suggests that the absorption ofluspatercept into the circulation is linear over the range of studied doses, and the absorption isnot significantly affected by the subcutaneous injection location (upper arm, thigh or abdomen).

Interindividual variability in AUC was approximately 37% in both β-thalassaemia and MDSpatients.

At the recommended doses, the geometric mean apparent volume of distribution was 9.56 L for

MDS patients and 7.26 L for β-thalassaemia patients. The small volume of distribution indicatesthat luspatercept is confined primarily in extracellular fluids, consistent with its large molecularmass.

Luspatercept is expected to be catabolised into amino acids by general protein degradationprocess.

Luspatercept is not expected to be excreted into urine due to its large molecular mass that isabove the glomerular filtration size exclusion threshold. At the recommended doses, thegeometric mean apparent total clearance was 0.47 L/day for MDS patients and 0.44 L/day forβ-thalassaemia. The geometric mean half-lives in serum were approximately 14.1 days for MDSpatients and 11 days for β-thalassaemia patients.

The increase of luspatercept Cmax and AUC in serum is approximately proportional to increasesin dose from 0.125 to 1.75 mg/kg. Luspatercept clearance was independent of dose or time.

When administered every three weeks, luspatercept serum concentration reaches the steady stateafter 3 doses, with an accumulation ratio of approximately 1.5.

Hb response

In patients who received < 4 units of RBC transfusion within 8 weeks prior to the study, Hbincreased within 7 days of treatment initiation and the increase correlated with the time to reachluspatercept Cmax. The greatest mean Hb increase was observed after the first dose, withadditional smaller increases observed after subsequent doses. Hb levels returned to baselinevalue approximately 6 to 8 weeks from the last dose (0.6 to 1.75 mg/kg). Increasing luspaterceptserum exposure (AUC) was associated with a greater Hb increase in patients with ESArefractory or -intolerant MDS or β-thalassaemia.

In non-transfusion-dependent β-thalassaemia patients who had a baseline transfusion burden of0 to 5 units within 24 weeks, increasing luspatercept serum exposure (time-averaged AUC) wasassociated with a greater probability of achieving a Hb increase (≥ 1 g/dL or ≥ 1.5 g/dL) and alonger duration of such Hb increases. The luspatercept serum concentration achieving 50% ofthe maximum stimulatory effect on Hb production was estimated to be 7.6 mcg/mL.

Population PK analysis for luspatercept included patients with ages ranging from 27 to 95 and18 to 71 years old, for MDS and β-thalassaemia patients, respectively, with a median age of72.5 years for MDS patients and of 33 years for β-thalassaemia patients. No clinicallysignificant difference in AUC or clearance was found across age groups in MDS patients (≤ 64,65-74, and ≥ 75 years) or in β-thalassaemia patients (18 to 71 years).

Population PK analysis for luspatercept included patients with normal hepatic function (BIL,

ALT, and AST ≤ ULN; N = 62 for β-thalassaemia patients and N = 311 for MDS patients), mildhepatic impairment (BIL > 1 - 1.5 x ULN, and ALT or AST > ULN; N = 89 for β-thalassaemiapatients and N = 126 for MDS patients), moderate hepatic impairment (BIL > 1.5 - 3 x ULN,any ALT or AST; N = 157 for β-thalassaemia patients and N = 32 for MDS patients), or severehepatic impairment (BIL > 3 x ULN, any ALT or AST; N = 73 for β-thalassaemia patients and

N = 1 for MDS patients) as defined by the National Cancer Institute criteria of hepaticdysfunction. Effects of hepatic function categories, elevated liver enzymes (ALT or AST, up to3 x ULN) and elevated total BIL (4 - 246 μmol/L) on luspatercept clearance were not observed.

No clinically significant difference in mean steady state Cmax and AUC was found across hepaticfunction groups. PK data are insufficient for patients with liver enzymes (ALT or AST)≥ 3 x ULN. No PK data are available for patients with liver cirrhosis (Child-Pugh Classes A, Band C) as no dedicated study was performed.

Population PK analysis for luspatercept included patients with normal renal function (individualeGFR ≥ 90 mL/min N = 302 for β-thalassaemia patients and N = 169 for MDS patients), mildrenal impairment (individual eGFR 60 to 89 mL/min; N = 74 for β-thalassaemia patients and

N = 204 for MDS patients), or moderate renal impairment (individual eGFR 30 to 59 mL/min;

N = 4 for β-thalassaemia patients and N = 88 for MDS patients) as defined by Modification of

Diet in Renal Disease (MDRD) formula. Luspatercept steady-state serum exposure (AUC) was24% to 41% higher in patients with mild to moderate renal impairment than in patients withnormal renal function. PK data are insufficient for patients with severe renal impairment(individual eGFR < 30 mL/min) or end-stage kidney disease.

Other intrinsic factors

The following population characteristics have no clinically significant effect on luspatercept

AUC or clearance: sex and race (Asian vs. White).

The following baseline disease characteristics had no clinically significant effect on luspaterceptclearance: serum erythropoietin level (2.4 - 1680 U/L for β-thalassaemia patients and 7.80 -2920 U/L for MDS patients), RBC transfusion burden (0 - 43.4 units/24 weeks), MDS ringsideroblasts, β-thalassaemia genotype (β0/β0 vs. non-β0/β0) and splenectomy.

The volume of distribution and clearance of luspatercept increased with increase of body weight(33 - 124 kg), supporting the body weight-based dosing regimen.

Single and repeat-dose toxicity

Following repeated administration of luspatercept in rats, toxicities included:membranoproliferative glomerulonephritis; congestion, necrosis and/or mineralisation of theadrenal glands; hepatocellular vacuolation and necrosis; mineralisation of the glandularstomach; and decreased heart and lung weights with no associated histology findings. A clinicalobservation of swollen hindlimbs/feet was noted in several studies in rats and rabbits (includingjuvenile and reproductive toxicity studies). In one juvenile rat, this correlatedhistopathologically with new bone formation, fibrosis, and inflammation.

Membranoproliferative glomerulonephritis was also seen in monkeys. Additional toxicities inmonkeys included: vascular degeneration and inflammatory infiltrates in the choroid plexus.

For the 6-month toxicity study, the longest duration study in monkeys, the no-observed-adverse-effect level (NOAEL) was 0.3 mg/kg (0.3-fold of clinical exposure at 1.75 mg/kg every3 weeks). A NOAEL was not identified in rats and the lowest-observed-adverse-effect-level(LOAEL) in the rat 3-month study was 1 mg/kg (0.9-fold of clinical exposure at 1.75 mg/kgevery 3 weeks).

Carcinogenesis and mutagenesis

Neither carcinogenicity nor mutagenicity studies with luspatercept have been conducted.

Haematological malignancies were observed in 3 out of 44 rats examined in the highest dosegroup (10 mg/kg) in the definitive juvenile toxicity study. The occurrence of these tumours inyoung animals is unusual and the relationship to luspatercept therapy cannot be ruled out. At the10 mg/kg dose, at which tumours were observed, the exposure represents an exposure multipleof approximately 4 times the estimated exposure at a clinical dose of 1.75 mg/kg every threeweeks.

No other proliferative or pre-neoplastic lesions, attributable to luspatercept, have been observedin any species in other non-clinical safety studies conducted with luspatercept, including the6-month study in monkeys.

In a fertility study in rats, administration of luspatercept to females at doses higher than thecurrently recommended highest human dose reduced the average number of corpora lutea,implantations and viable embryos. No such effects were observed when exposure in animalswas at 1.5 times the clinical exposure. Effects on fertility in female rats were reversible after a14-week recovery period.

Administration of luspatercept to male rats at doses higher than the currently recommendedhighest human dose had no adverse effect on male reproductive organs or on their ability tomate and produce viable embryos. The highest dose tested in male rats yielded an exposureapproximately 7 times the clinical exposure.

Embryo-foetal development (EFD)

Embryo-foetal developmental toxicology studies (range-finding and definitive studies) wereconducted in pregnant rats and rabbits. In the definitive studies, doses of up to 30 mg/kg or40 mg/kg every week were administered twice during the period of organogenesis. Luspaterceptwas a selective developmental toxicant (dam not affected; foetus affected) in the rat and amaternal and foetal developmental toxicant (doe and foetus affected) in the rabbit. Embryofoetaleffects were seen in both species and included reductions in numbers of live foetuses and foetalbody weights, increases in resorptions, post-implantation loss and skeletal variations and, inrabbit foetuses, malformations of the ribs and vertebrae. In both species, effects of luspaterceptwere observed in the EFD studies at the lowest dose tested, 5 mg/kg, which corresponds to anestimated exposure in rats and rabbits of approximately 2.7 and 5.5 times greater, respectively,than the estimated clinical exposure.

Pre- and post-natal development

In a pre- and post-natal development study, with dose levels of 3, 10, or 30 mg/kg administeredonce every 2 weeks from gestational day (GD) 6 through post-natal day (PND) 20, adversefindings at all doses consisted of lower F1 pup body weights in both sexes at birth, throughoutlactation, and post weaning (PND 28); lower body weights during the early premating period(Weeks 1 and 2) in the F1 females (adverse only at the 30 mg/kg/dose) and lower body weightsin F1 males during the premating, pairing and post-mating periods; and microscopic kidneyfindings in F1 pups. Additionally, non-adverse findings included delayed male sexualmaturation at the 10 and 30 mg/kg/dose. The delay in growth and the adverse kidney findings,in the F1 generation, precluded the determination of a NOAEL for F1 general and developmentaltoxicity. However, there was no effect on behavioural indices, fertility or reproductiveparameters at any dose level in either sex, therefore the NOAEL for behavioural assessments,fertility and reproductive function in the F1 animals was considered to be the 30 mg/kg/dose.

Luspatercept is transferred through the placenta of pregnant rats and rabbits and is excreted intothe milk of lactating rats.

Juvenile toxicity

In a study in juvenile rats, luspatercept was administered from postnatal day (PND) 7 to PND 91at 0, 1, 3, or 10 mg/kg. Many of the findings seen in repeat-dose toxicity studies in adult ratswere repeated in the juvenile rats. These findings included glomerulonephritis in the kidney,haemorrhage/congestion, necrosis and mineralization of the adrenal gland, mucosalmineralization in the stomach, lower heart weights, and swollen hindlimbs/feet. Luspatercept-related findings unique to juvenile rats included tubular atrophy/hypoplasia of the kidney innermedulla, delays in the mean age of sexual maturation in males, effects on reproductiveperformance (lower mating indices), and non-adverse decreases in bone mineral density in bothmale and female rats. The effects on reproductive performance were observed after a greaterthan 3-month recovery period, suggesting a permanent effect. Although reversibility of thetubular atrophy/hypoplasia was not examined, these effects are also considered to beirreversible. Adverse effects on the kidney and reproductive system were observed at clinicallyrelevant exposure levels and seen at the lowest dose tested and, thus, a NOAEL was notestablished. In addition, haematological malignancies were observed in 3 out of 44 ratsexamined in the highest dose group (10 mg/kg). These findings are all considered potential risksin paediatric patients.

Citric acid monohydrate (E330)

Sodium citrate (E331)

Polysorbate 80 (E433)

Sucrose

Hydrochloric acid (for pH adjustment)

Sodium hydroxide (for pH adjustment)

This medicinal product must not be mixed with other medicinal products except thosementioned in section 6.6.

Unopened vial5 years.

When stored in the original container, chemical and physical in-use stability of the reconstitutedmedicinal product has been demonstrated for up to 8 hours at room temperature (≤ 25 °C) or forup to 24 hours at 2 °C - 8 °C.

From a microbiological point of view, the medicinal product should be used immediately. If notused immediately, in-use storage times and conditions prior to use are the responsibility of theuser and should not be longer than 24 hours at 2 °C - 8 °C.

Do not freeze the reconstituted solution.

Store in a refrigerator (2 °C - 8 °C).

Do not freeze.

Store in the original carton in order to protect from light.

For storage conditions after reconstitution of the medicinal product, see section 6.3.

Reblozyl 25 mg powder for solution for injection3 mL Type I glass vial with a hydrophobic inner coating closed with a bromobutyl rubberstopper and aluminium seal with yellow polypropylene flip-off cap.

Reblozyl 75 mg powder for solution for injection3 mL Type I glass vial with a hydrophobic inner coating closed with a bromobutyl rubberstopper and aluminium seal with orange polypropylene flip-off cap.

Pack size: 1 vial

Reblozyl must be reconstituted gently prior to administration. Aggressive shaking should beavoided.

Reconstitution of the product

Reblozyl is supplied as a lyophilised powder for reconstitution before use. Only water forinjections (WFI) should be used when reconstituting Reblozyl.

The appropriate number of Reblozyl vials should be reconstituted to achieve the desired dose. Asyringe with appropriate graduations must be used for reconstitution to ensure accurate dose.

The following steps should be followed for reconstitution:

1. Remove the coloured cap from the vial and wipe the top with an alcohol wipe.

2. Reblozyl 25 mg powder for solution for injection

Add 0.68 mL WFI into the vial by means of a syringe with appropriate graduations with aneedle directing the flow onto the lyophilised powder. Allow to stand for one minute.

Each 25 mg single-dose vial will deliver at least 0.5 mL of 50 mg/mL luspatercept.

Reblozyl 75 mg powder for solution for injection

Add 1.6 mL WFI into the vial by means of a syringe with appropriate graduations with aneedle directing the flow onto the lyophilised powder. Allow to stand for one minute.

Each 75 mg single-dose vial will deliver at least 1.5 mL of 50 mg/mL luspatercept.

3. Discard the needle and syringe used for reconstitution. Do not use them for subcutaneousinjection.

4. Gently swirl the vial in a circular motion for 30 seconds. Stop swirling and let the vial sitin an upright position for 30 seconds.

5. Inspect the vial for undissolved powder in the solution. If undissolved powder isobserved, repeat step 4 until the powder is completely dissolved.

6. Invert the vial and gently swirl in an inverted position for 30 seconds. Bring the vial backto the upright position and let it sit for 30 seconds.

7. Repeat step 6 seven more times to ensure complete reconstitution of material on the sidesof the vial.

8. Visually inspect the reconstituted solution prior to administration. When properly mixed,

Reblozyl reconstituted solution is a colourless to slightly yellow, clear to slightlyopalescent solution which is free of visible foreign particulate matter. Do not use ifundissolved product or foreign particulate matter is observed.

9. If the reconstituted solution is not used immediately, see section 6.3 for storageconditions.

Any unused medicinal product or waste material should be disposed of in accordance with localrequirements.

Bristol-Myers Squibb Pharma EEIG

Plaza 254

Blanchardstown Corporate Park 2

Dublin 15, D15 T867

Ireland

EU/1/20/1452/001

EU/1/20/1452/002

Date of first authorisation: 25 June 2020

Date of last renewal:

Detailed information on this medicinal product is available on the website of the European

Medicines Agency http://www.ema.europa.eu.