Cryopreserved vs Liquid-Stored Platelets for the Treatment of Surgical Bleeding: The CLIP-II Randomized Noninferiority Clinical Trial

Reade. JAMA, 2025. doi: 10.1001/jama.2025.23355

Clinical Question

• In adult cardiac surgery patients at high risk of bleeding, is the transfusion of cryopreserved platelets, compared with conventional liquid-stored platelets, non-inferior in reducing postoperative bleeding within the first 24 hours after ICU admission?

Background

• Platelet transfusion is a key component of managing major bleeding, with early transfusion recommended to reduce mortality
  • Conventional liquid-stored platelets are stored at 22°C, with a short shelf-life (5–7 days), leading to limited availability and periodic shortages
  • Cryopreserved platelets use dimethyl sulfoxide (DMSO) and are stored at -80°C, then resuspended in plasma prior to transfusion, extending the shelf life to up to 2 years
  • If cryopreserved platelets are safe and efficacious, this could reduce wastage and improve of platelet availability in settings where access in currently limited currently including regional, remote and military centres
• Existing Evidence
  • Safety studies and pilot trials (Khuri 1999; Slichter 2018; McGuinness 2021) have demonstrated no increase in adverse events with cryopreserved platelets
  • Comparative and observational data (Bohonek 2019) suggest similar survival and transfusion requirements compared with liquid-stored platelets
  • The CLIP-I randomized controlled pilot study (Reade et al 2019) identified chest drain blood loss within the first 24 hours as an appropriate endpoint for assessing non-inferiority

Design

• Multicentre randomised, double blinded, parallel group, non-inferiority Phase III trial
  • Trial protocol and investigator brochure approved by the Austin Health and Australian Red Cross Lifeblood Human Research Committees
  • Written consent obtained at least 1 day prior to planned surgery
  • Trial protocol and statistical analysis plan published prior to interim analysis (50% recruitment)
• Randomised
  • 1:1 randomisation prior to surgery using computer-generated permuted blocks of size of 2 or 4, with allocation concealment
  • Only patients who received platelet transfusions (clinical indication) included in study analysis
  • Stratified by trial site
• Double Blinded
  • Two unblinded nurses at each hospital (not involved in care) conducted routine pretransfusion safety checks
  • Study platelets issued from blood bank with opaque cover
  • Participants, treating clinicians and outcome assessors all remained blinding to group allocation throughout the trial
• Statistical Analysis
  • Prespecified non-inferiority margin of 20%, chosen to approximate the mean volume of a unit of red blood cells (258 mL)
  • Transfusion of 202 patients would have 80% power to detect above non-inferiority limit at a one sided p-value of 0.025
    • Based on local pilot data
  • Primary outcome log-transformed and results presented as geometric means
  • Fine and Grey competing risk model (accounting for competing risk of death) used to assess time to event outcomes

Setting

• 11 Australian Hospitals
• 25^th August 2021 – 16^th April 2024

Population

• Inclusion:
  • Adults ≥18 years
  • Undergoing cardiac surgery
  • High risk of requiring platelet transfusion (ACSePT score ≥1 [equates to a 30% probability of needing platelets]) or clinician judgement
• Exclusion:
  • Prior enrolment in bleeding medication trials
  • History of Deep Vein Thrombosis or Pulmonary Embolism
  • Known coagulopathy or current-anticoagulant related bleeding disorder
  • Females aged 18–55 years who were RhD-negative or RhD-unknown
• Participant numbers
  • 879 screened –> 388 randomised
    • 196 to cryopreserved and 192 to liquid stored
  • 202 received study platelets and were analysed
    • Cryopreserved platelets: 104
    • Liquid-stored platelets: 98
• Baseline Characteristics similar between groups (Cryopreserved vs Liquid-Stored)
  • Demographics
    • Age: 65.5 vs 63.2
    • Female Sex: 22.1% vs 26.5%
    • BMI: 28.5 vs 27.4
    • EuroScore II risk score: 3.5 vs 3.4
  • Nature of Surgery
    • Elective: 70.2% vs 69.4%
    • Deep Hypothermic Arrest: 16.3% vs 17.3%
    • Minimally invasive operation: 1% vs 4.1%
    • Total Cardiopulmonary Bypass Time (mins): 195 vs 180
  • Lab Test Results
    • Creatinine Clearance >85mL/min: 38.5% vs 41.8%
    • Preop Hb <100g/dL: 11.5% vs 12.2%
    • Preop Platelets <100 x 10³/μL: 1% vs 2%
    • Preop aPTT >40s: 7.9% vs 12.6%
    • Preop fibrinogen <200mg/dL: 0% vs 1.1%
    • Preop INR >1.5: 3% vs 2.1%
  • Pre-operative Treatments
    • Tranexamic Acid 24h prior to surgery: 7.7% vs 7.1%
    • Anticoagulant within 7 days: 59.6% vs 59.2%
    • Aspirin within 7 days: 32.7% vs 36.7%
    • Prophylactic enoxaparin within 7 days: 16.3% vs 27.6%
    • Prophylactic heparin within 7 days: 16.3% vs 10.2%

Intervention

• Cryopreserved platelets
  • Thawed cryopreserved group O apheresis platelets
  • Reconstituted in ABO group matched plasma
  • Median of 2 units received
  • 25% received open label liquid platelets prior to 3rd study unit being complete
  • 16% received open label liquid platelets after to 3rd study unit being complete

Control

• Liquid Platelets
  • In a mixture of plasma and platelet additive solution
  • Either derived from whole blood or donated by apheresis with ABO group matching according to hospital practice
  • Median of 1 unit received
  • 3.1% received open label liquid platelets prior to 3rd study unit being complete
  • 9% received open label liquid platelets after to 3rd study unit being complete

Management common to both groups

• Study platelet transfusion at clinician discretion as part of standard perioperative bleeding management
• Open-label liquid-stored platelets permitted after 3 study units if required
• First study platelet administered either intraoperatively or within first 24 hours following ICU Admission –  subsequent units could be administered at any time up until ICU discharge
• Time between ordering and receipt of products did not differ between groups
  • ~85% received first platelets in OR

Outcome

• Primary outcome: Post Surgical Chest Drain Bleeding in first 24 hours
  • Cryopreserved: Geometric mean 605mls (95% CI 532 – 688)
  • Liquid stored: Geometric mean 535mls (95% CI 480 – 596)
    • Ratio of geometric means: 1.13 [95% CI, 0.96 – 1.34]; P= .07
    • Non-inferiority was not established, as the upper confidence limit exceeded the pre-specified 20% margin
    • Remained non-inferior on adjustment (site and EUROSCORE II)
    • All subgroups consistent (platelet compatibility, surgery complexity, aspirin use and location of first transfusion)
• Selected Secondary and Tertiary outcomes:
  • No significant difference:
    • Adverse events (low incidence in both groups)
    • Thrombotic complications
    • 24-hour haemoglobin concentration nadir
    • Inotrope and vasopressor requirements
    • Time to recommence aspirin and heparin (numerically shorter in liquid stored group)
    • Renal replacement therapy
    • Survival
  • Significantly greater in cryopreserved Platelet group
    • Chest Drain Volume at ICU admission
    • Total Postoperative Blood loss
      • Ratio of geometric means 1.31 (95% CI 1.07 – 1.60)
    • BARC-4 composite bleeding outcome
      • 31.7 vs 12.3%
    • Red blood cell transfusions, fresh frozen plasma and cryoprecipitate transfusions
    • Time to chest drain removal
    • Time to extubation
    • ICU and Hospital length of stay
  • Significantly less in Cryopreserved Platelet Group
    • Platelet count at 24 hours lower in cryopreserved group (103 vs 140)

Authors’ Conclusions

• Cryopreserved platelets could not be shown to be non-inferior to liquid stored platelets for post operative bleeding within 24 hours of cardiac surgery

Strengths

• Multicentre, multidisciplinary collaboration involving cardiac surgeons, anaesthetists, intensivists and haematologists
• An important question given the supply constraints in some settings
• Double blinded with robust measures to maintain blinding
• Randomised with allocation concealment
• Pragmatic approach
  • Clinicians titrated platelet transfusion, reflective of real world practice, rather than protocol based
• Objective Primary Outcome
  • Chest drain output reliably measured in post cardiac surgery patients
• Prespecified and appropriate Statistical Analysis Plan
  • Non-inferiority margin defined prior and clinically relevant; based around mean volume of red blood cell transfusion unit
  • Per Protocol Sensitivity Analysis completed whereby those that received open label platelets before 3 study units were excluded – primary outcome consistent with main results
• Minimal Missing Data
  • Only one patient had missing data, limiting risk of attrition bias
• Baseline Demographics well balanced
  • Inclusive of 30% non elective surgery, 65% complex surgery, 60% abnormal renal function

Weaknesses

• Dose Exposure
  • Maximum of 3 study platelets limits assessment of dose response and safety at higher exposures
• Open-Label platelet use higher in cryopreserved group
  • Higher proportion of patients in cryopreserved group received open label platelets prior to administration 3 study platelets, possible confounding treatment effects
  • Is this due to not being as effective or is it due to potential unblinding – DMSO has a distinctive smell
• Potential Inequivalence of “units”
  • Cryopreserved and liquid stored platelets differ substantially in phenotype and post transfusion recovery, raising possibility that unit for unit comparison is not equivocal
• Timing of subsequent study platelets variable and unclear, may not have preceded measurement of primary outcome
• Limited to Cardiac Surgery
  • Patient’s restricted to those undergoing high risk cardiac surgery, limiting translation to trauma, gastrointestinal bleeding and obstetric haemorrhage.
• Limited to Australian Sites
  • Limiting translation across sites internationally, particularly given potential difference in clinician directed transfusion practices

The Bottom Line

• The CLIP-II Trial demonstrated that cryopreserved platelets were not non-inferior to liquid stored platelets for control of bleeding in cardiac surgery patients. Cryopreserved platelets were associated with greater total blood loss, higher transfusion requirements, and longer ICU stays
• In centres where liquid stored platelets are readily available, these should continue to be used for first line therapy for cardiac surgical bleeding
• Cryopreserved platelets may still have utility in settings where platelet availability is limited, however this should be further evaluated

External Links

• article Cryopreserved vs Liquid-Stored Platelets for the Treatment of Surgical Bleeding: The CLIP-II Randomized Noninferiority Clinical Trial
• editorial An Oasis in the Platelet Desert?
• further reading Cost-Effectiveness of Cryopreserved vs Liquid-Stored Platelets for Managing Surgical Bleeding
• Presentation of CLIP-II trial and results at CCR Down Under 2025

Metadata

Summary author: Samuel Finlayson
Summary date: 18th December 2025
Peer-review editor: George Walker

Picture by: AhmadArdity/Pixabay