Traumatic injury remains a leading cause of mortality for individuals 45 years and younger and hemorrhage remains the leading cause of potentially preventable death postinjury.1–4 Despite improvements in trauma resuscitation,5–8 the highest rate of mortality from hemorrhage occurs during the first hours after arrival to definitive care,9,10 underscoring the importance of early interventions which provide benefit.11–15

Early platelet transfusion is an essential component of trauma resuscitation and is associated with attenuation of endothelial barrier disruption, improved hemostasis, and attributable survival in severely injured patients.7,16–21 The storage temperature employed for platelets may affect hemostatic function and can impact the platelet supply due to shelf-life constraints.22–24 Cold stored platelets (1–6°C) demonstrate improved hemostatic capabilities, may have a lower risk of bacterial contamination, and can be utilized for urgent transfusion for the acutely bleeding patient.25–30 Recent Food and Drug Administration (FDA) guidance allows for the use of cold stored platelets with 14 days of storage to treat active bleeding when conventional room temperature platelets are not available, or their use is not practical.31 Despite this recent guidance from the FDA, there exists no robust clinical trial evidence demonstrating the safety and efficacy of cold stored platelets (CSPs) in the severely injured patient population at risk of hemorrhage.

The Cold Stored Platelet for Hemorrhagic Shock (CriSP-HS) trial aimed to characterize the safety and efficacy of incorporating early CSP transfusion in injured patients at risk of hemorrhagic shock as compared with standard care trauma resuscitation. We enrolled patients at risk of hemorrhagic shock and allocated patients to early CSP transfusion and concomitant standard care resuscitation or standard care resuscitation alone. We hypothesized that early CSP transfusion would be safe and associated with clinical outcome benefits.

METHODS Trial Design

The CriSP-HS study was a phase 2, multicenter, open label, randomized clinical trial that compared outcomes in patients at risk for hemorrhagic shock receiving CSP transfusion and concomitant standard care resuscitation versus standard care resuscitation alone. A total of 422 patients who arrived at one of 5 participating level 1 trauma centers were screened for inclusion criteria and assessed for eligibility. We enrolled 200 patients from June 23, 2022, through September 7, 2023. The intervention arm received a single apheresis unit of urgent release CSPs which was transfused as soon as feasible, while concomitant standard care resuscitation was occurring. Resuscitation in the standard care arm followed standard procedures at each participating site. We did not alter any other aspects of a care besides the administration of early CSPs. Patients in the standard care arm were not required to receive platelet transfusion unless indicated by standard resuscitation procedures at the enrolling site. The transfusion of CSPs was not standard care for sites preceding or during enrollment of the trial; room temperature platelets were utilized as standard care when required. The trial protocol can be found as a Supplemental Data File (Supplemental Digital Content 1, https://links.lww.com/SLA/F93).

The US FDA (Investigational New Drug 19467), the Office of Human Research Oversight of the US Department of Defense, and the Human Research Protection Office at the University of Pittsburgh approved the clinical trial. An external data and safety monitoring board oversaw the trial. The single institutional review board (IRB) at the University of Pittsburgh, with review and acknowledgment from local site IRBs, approved an exception from informed consent to enroll participants. This approval included community consultation and public disclosure/notification. We notified enrolled participants and/or their legally authorized representatives as soon as feasible and obtained consent for continued participation.32 The study followed the Consolidated Standard of Reporting Trials (CONSORT, Supplemental Digital Content 2, https://links.lww.com/SLA/F94 and Supplemental Digital Content 3, https://links.lww.com/SLA/F95) reporting guideline.

Study Patient Population

We selected inclusion criteria to include a severely injured population of patients at risk of hemorrhagic shock and large volume transfusion. Injured patients arriving at a participating trauma center were eligible for enrollment if they met Assessment of Blood Consumption (ABC) criteria33,34 (2 or more of the following): (1) hypotension (systolic blood pressure≤90 mm Hg); (2) penetrating mechanism of injury; (3) positive Focused Assessment for the Sonography of Trauma (FAST) examination; (4) Heart rate ≥ 120 and who within 60 minutes required a hemorrhage control procedure in the operating room or interventional radiology suite. Patients were eligible if they met ABC criteria qualifying vital signs in the prehospital or in-hospital phase of care. A FAST examination that was deferred due to the expedient transport to the operating room was considered as meeting one of the ABC criteria.33 Exclusion criteria included ages less than 15 years or more than 90 years, fall from standing, penetrating brain injury, >5 minutes of cardiopulmonary resuscitation (CPR), known prisoners or known pregnancy, isolated drowning or hanging victims, burns >20% total body surface area, objection to the study voiced by subject or family or wearing a “No CriSP” opt-out bracelet.

Randomization and Matching

We generated a 1:1 ratio random allocation sequence to either CSP arm or standard care arm with variable block sizes of 4 to 6, stratified by site, using a computer random-number generator. Predefined randomization assignment envelopes were maintained within easy access in the trauma bay or emergency department. Randomization was performed by a member of the study team with randomization cards in sealed envelopes after eligibility was confirmed at the individual patient level and in real-time. Clinical providers were unable to be masked to treatment assignment. Arm assignment was concealed to outcome assessors.

Intervention and Comparison Arms

After meeting all inclusion and no exclusion criteria, patients randomized to the CSP arm received a single apheresis cold stored platelet unit prepared following all FDA approved guidelines as soon as feasible.31 CSP units were refrigerated within 4 hours of donation and were stored in a monitored refrigerator at 1 to 6°C for up to 14 days from the date of donation. The CSP unit was transfused as soon as feasible in the trauma bay, during transport, or in the operating theater or interventional radiology suite. Participating site standard care resuscitation was performed concomitantly.

Patients who met all inclusion criteria and no exclusion criteria and who were randomized to the standard care arm received the site’s standard care resuscitation. Standard care arm patients were not required to received room temperature platelet transfusion.

Outcomes

The primary outcome for the trial was feasibility, including: the proportion of patients able to be randomized, the proportion of eligible patients enrolled, the proportion of patients enrolled with adherence to the study protocol and the proportion of patients for whom the principal clinical outcome of 24-hour mortality was obtained.

Our prespecified principal outcome for assessing safety and efficacy for the trial was 24-hour mortality. Additional secondary outcomes included 3-hour mortality, in-hospital mortality, 30-day mortality, death from adjudicated hemorrhage, acute respiratory distress syndrome, time to hemostasis, incidence of arterial or venous thromboembolic events, laboratory coagulation assessment, and 24-hour blood component transfusion requirements.

Achievement of hemostasis was defined when a patient received a single unit or less of blood transfusion in any 60-minute period in the first 4 hours from enrollment. Those patients who did not reach this blood transfusion nadir were considered to not have achieved hemostasis.

Prespecified subgroup analyzes for the 24-hour mortality outcome included: (1) patients who did or did not require in-hospital blood transfusion; (2) patients who did or did not require massive transfusion (≥10 U blood in first 24 hours); (3) patients with and without significant traumatic brain injury (head Abbreviated Injury Score—AIS >2); (4) patients who arrived from the scene of injury versus from a referring hospital; and (5) patients who received CSPs with an age since donation of ≤7 days as compared with 8 to 14 days.

Statistical Analysis

The feasibility of enrollment was evaluated by determining the proportion of patients in each feasibility category. These proportions were estimated directly as the observed ratio of numbers of patients, with 95% CIs being calculated.

The principal intent-to-treat safety and efficacy analysis compared 24-hour mortality across the CSP and standard care arms using a 2-sided Z-test for proportions. The Kaplan-Meier curves were generated for each treatment group and a log-rank test was used to compare the distribution of the cumulative proportion. A logistic regression model was used to assess the independent impact of the CSP intervention on 24-hour survival after controlling for site of enrollment and the potential effects of baseline characteristics with imbalance between treatment groups.

Analyzes to test for the homogeneity of the treatment effect were carried out for the 4 prespecified randomized subgroups. Regression models appropriate for the outcome variable were used to test for the homogeneity of the treatment effect. For the prespecified CSP age less than or equal to 7 days versus 8 to 14 days old analyzes, outcomes were compared across each age subgroup. Since the CSP age was not randomly assigned, a propensity score was generated as an indicator of the age of the CSP (≤7 vs 8–14 days) and multivariable regression models were used to assess the independent relationship of age of the CSP product on 24-hour mortality. All other secondary outcomes were also compared across the CSP product age groups.

We estimated there would be 80% power to detect a 14.2% difference using a mortality estimate of 23% in the control arm, a type I error rate of 0.05, a 2-sided alternative hypothesis, and 1 df with a sample size of 100 patients per treatment arm.

RESULTS

A total of 422 patients were assessed for trial eligibility, of which 384 met all inclusion and no exclusion criteria, with 200 patients being randomized, constituting the intent-to-treat cohort. There were 184 patients who met all inclusion criteria but were unable to be randomized due to staff or product unavailability (Fig. 1). Of the 200 patients in the intent-to-treat cohort, 102 were randomized to the early CSP arm, and 98 were randomized to the standard care arm. The primary and 24-hour mortality outcomes were available for 102 (100%) patients in the CSP arm and 98 (100%) patients in the standard care arm. Eleven patients in the CSP arm and 16 patients in the standard care arm refused consent for continued participation from secondary outcome comparisons. All events before refusal of consent for continued participation were included in the analyzes. We estimated a 24-month enrollment period for the study. We completed the study after reaching our target enrollment of 200 patients after 15 months.

FIGURE 1:

Flow of patients in the cold stored platelets in hemorrhagic shock (CRISP-HS) trial. Screening, randomization, and follow-up of the study participants. All data and events before refusal of consent for continued participation were included in clinical outcome comparisons.

Enrolled patients had an average age of 34±13 years, were 85.0% male, were injured via a penetrating mechanism 78.5% of the time, had a median Injury Severity Score (ISS) of 17 (interquartile range, 9–28) and had an all-cause mortality of 8.0% within 24 hours postinjury. Surgeons performed operative intervention in 97% of cases, while 3% required interventional radiology hemorrhage control procedures.

The assigned treatment regimen was provided to the enrolled patient 98.0% of the time. For the CSP randomized arm, 98 (96.1%) of the 102 patients received the early CSP transfusion. In the standard care arm, 98 (100%) of the 98 of patients received standard care resuscitation with 47 (48%) patients receiving room temperature platelet transfusion. The CSP and standard care arms were similar in demographics, injury characteristics and shock severity (Table 1). When complete blood count of enrolled patients was compared, the groups were also similar (Supplemental Digital Content 4, Table 1S, https://links.lww.com/SLA/F96).

TABLE 1 - Baseline Characteristics by Treatment Arm* n/N (%) Characteristics Standard care (N=98) Cold stored platelets (N=102) Age [mean (SD)] 35.3 (14.4) 34.4 (13.5) Male sex 78/98 (80) 85/102 (83) Race  White 18/98 (18) 28/102 (28)  Black 51/98 (52) 47/102 (47)  Other 13/98 (13) 13/102 (13)  Hispanic 13/98 (13) 20/102 (20) SBP [median (IQR)] (mm Hg)  Initial 108.5 (84.5, 129.5) 99 (83, 140)  Highest 122 (102, 144) 127 (98, 148)  Lowest 93.5 (78, 117.5) 90 (78, 127) HR [median (IQR)] (beats/min)  Initial 106 (86, 123) 112 (88, 130.5)  Highest 111.5 (93, 130) 116 (96.5, 135)  Lowest 95 (81.5, 117) 101 (75, 123) Transport mode  Ground EMS 66/98 (67) 71/102 (70)  Air EMS 24/98 (24) 20/102 (20)  Does not apply† 5/98 (5) 4/102 (4) Transferred from  Scene of accident/injury 65/98 (66) 73/102 (72)  Home 4/98 (4) 3/102 (3)  Other hospital 26/98 (26) 19/102 (19) Mechanism of injury blunt 23/98 (23) 12/102 (12)  Fall 5/98 (5) 2/102 (2)  MVC occupant ejected 3/98 (3) 1/102 (1)  MVC occupant not ejected 6/98 (6) 3/102 (3)  MVC motorcycle 5/98 (5) 3/102 (3)  MVCpedestrian 1/98 (1) 2/102 (2)  Struck by or against 3/98 (3) 1/102 (1) Mechanism of injury penetrating 76/98 (77) 90/102 (88)  Firearm 53/98 (54) 62 /102 (61)  Impalement 1/98 (1) 1/102 (1)  Stabbing 20/98 (20) 26/102 (25)  Other 2/98 (2) 1/102 (1) AIS [median (IQR)]  Head and Neck AIS 0, (0, 0) 0, (0, 0)  Face AIS 0, (0, 0) 0, (0, 0)  Chest AIS 1.5 (0, 3) 0 (0, 3)  Abdomen AIS 3 (0, 4) 2 (0, 4)  External AIS 1 (0, 1) 1 (0, 1)  Extremity AIS 0 (0, 3) 0 (0, 3) ISS‡ [median (IQR)] 17.5 (10, 29) 17 (9, 27) TBI 9/98 (9) 6/102 (6) Prehospital advanced airway§ 16/98 (16) 12/102 (12)  Supraglottic airway 2/98 (2) 1/102 (1)  Endotracheal intubation 14/98 (14) 11/102 (11) Prehospital crystalloid/colloid fluids§ 42/98 (43) 36/102 (35)  Crystalloid/colloid volume [median (IQR)] (mL) 1000 (300, 1000) 700 (300, 2000) Prehospital blood product§  Whole blood 6/98 (6) 6/102 (6)   Whole blood volume [median (IQR)] (U) 1 (1, 1) 1.5 (1, 2)  Plasma 14/98 (14) 9/102 (9)   Plasma volume [median (IQR)] (U) 2 (1, 2) 2 (2, 2)  Red blood cell 22/98 (22) 14/102 (14)   Red blood cell volume [median (IQR)] (U) 2 (1, 3) 2 (1, 4)  Platelets 1/98 (1) 0/102 (0)

AIS indicates Abbreviated Injury Score; ED, emergency department; EMS, Emergency Medical Services; HR, heart rate; IQR, interquartile range; ISS, Injury Severity Score; MVC, motor vehicle crash; RR, respiratory rate; SBP, systolic blood pressure; TBI, traumatic brain injury.

*No statistically significant differences were observed between baseline characteristics.

†Not delivered to the ED by EMS.

‡The score range was 0 to 75. A score >15 indicates major trauma.

§Provided by EMS during transport or at an outside ED before transfer.

For the primary feasibility outcome, the proportion of eligible patients who could be randomized was 91%, (95% CI, 88.3%–93.7%), with 52.1%, (95% CI, 47.1%–57.1%) being randomized and enrolled in the trial. The proportion of enrolled patients with adherence to study protocol was 98%, (95% CI, 96.1%–99.9%). Finally, the proportion of enrolled patients where the status of the principle clinical outcome of 24-hour mortality was known/available was 100%.

At 24 hours after arrival, 6 deaths occurred in the CSP arm and 10 deaths in the standard care arm. Patients randomized to early CSP compared with standard care did not significantly differ in the rate of 24-hour mortality (5.9% vs 10.2%; difference, -4.3%; 95% CI, −12.8% to 3.5%; P=0.26). Kaplan-Meier survival curves demonstrate early separation in curves which did not reach statistical significance (log-rank χ2=1.19, P=0.27; Fig. 2). In a binary logistic regression model controlling for site, the assignment to the CSP arm did not significantly change the odds of 24-hour mortality (adjusted odds ratio 0.56; 95% CI, 0.18–1.56; P=0.27).

FIGURE 2:

Kaplan-Meier survival analysis at 24-hour postrandomization.

Mortality at 24 hours in the 4 prespecified randomized subgroups is depicted in Figure 3. No heterogeneity of treatment effect across any of the randomized subgroups was found.

FIGURE 3:

Subgroup analysis for 24-hour mortality. A, Differences in 24-hour mortality in the 4 prespecified subgroups. B, Twenty-four hour mortality differences with CIs. The black circles represent the difference in 24-hour mortality and horizontal bars represent the 95% CI. The difference was computed as 24-hour mortality for CSP minus 24-hour mortality for SC. The 95% CIs are exact (Clopper-Pearson) confidence limits. The gray dashed vertical line represents a difference in 24-hour mortality of 0, indicating no difference in 24-hour mortality between CSP and SC treatment arms. P-values are for the interaction term between treatment arms and subgroup in a logistic regression model with 24-hour mortality as the outcome. CSP indicates cold stored platelets; SC, standard care; MT, massive transfusion; TBI, traumatic brain injury.

No group differences were found in 3-hour mortality (difference, −1.3%, 95% CI, −9.1% to 6.3%; P= 0.72), in-hospital mortality (difference, −2.7; 95% CI, −13.8% to 8.2%; P= 0.62) or 30-day mortality (difference, −2.7; 95% CI, −13.8% to 8.2%; P=0.62) (Table 2). No significant differences were found in the incidence of death due to hemorrhage (difference −0.5%; 95% CI, −8.7% to 7.4%; P=0.89), the incidence of ARDS (difference 2.8%; 95% CI, −4.0% to 9.8%; P=0.36) or in the incidence of arterial or venous thromboembolic events (difference, −0.8%; 95% CI, −10.2% to 8.3%; P=0.86). There were no differences in the achievement of hemostasis (difference, 0.3; 95% CI, −6.8 to 7.4; P=0.94) or time to hemostasis (Supplemental Digital Content 4, Figure 1S, https://links.lww.com/SLA/F96) across the groups. There were no significant group differences in coagulation measurements or differences in 24-hour blood transfusion requirements. There were no documented transfusion/allergic reactions in either arm of the trial.

TABLE 2 - Clinical Outcomes by Treatment Group Unadjusted analysis Adjusted analysis Clinical outcome Standard Care (N=98) [n (%)] Cold stored platelets (N=102) [n (%)] Difference* (95% CI) P OR/difference† (95% CI) P Principle clinical outcome  24-h mortality 10 (10.2) 6 (5.9) −4.3% (−12.8, 3.5) 0.26 0.6 (0.2, 1.6) 0.27 Secondary clinical outcome  3-h mortality 7 (7.1) 6 (5.9) −1.3% (−9.1, 6.3) 0.72 0.9 (0.3, 2.9) 0.86  In-hospital mortality‡ 13/82 (15.8) 12/91 (13.2) −2.7% (−13.9, 8.2) 0.62 0.8 (0.3, 1.9) 0.62  30-day mortality ‡§ 13/82 (16) 12/91 (13) −2.7% (−13.9, 8.2) 0.62 0.8 (0.3, 1.9) 0.62  Mortality from hemorrhage 7 (7.1) 7 (6.9) −0.3% (−8.8, 7.4) 0.89 1.0 (0.3, 3.1) 0.99  ARDS 3/93 † (3.2) 6/100†(6.0) 2.8% (−4.0, 9.8) 0.36 2.2 (0.6, 11.0) 0.27  Allergic/transfusion reaction 0 (0) 0 (0) - -  Thromboembolic events 10/93†(10.8) 10/100† (10.0) −0.8% (−10.2, 8.3) 0.86 0.9 (0.4, 2.4) 0.85  Achieve hemostasis 91/98 (92.9) 95/102 (93.1) 0.3 (−6.8, 7.4) 0.94 1.1 (0.4, 3.2) 0.90 Initial rapid thromboelastography measurements∥¶ [median (IQR)]  Activated clotting time (sec) 105 (97, 113) 113 (105, 113) 8.0 (−8.0, 8.0) 0.79 0.5 (−11.2, 5.5) 0.49  K-time (min) 1.9 (1.3, 2.3) 2 (1.4, 2.2) 0.1 (−0.5, 0.5) 0.97 −0.2 (−0.7, 0.3) 0.39  Alpha angle (deg.) 69 (66, 73) 72 (67, 73) 3.0 (−2.0, 5.0) 0.56 1.4 (−2.2, 5.0) 0.43  Maximal amplitude (mm) 56.4 (51.1, 62.2) 57.7 (48.9, 63.6) 1.3 (−2.4, 3.4) 0.55 0.1 (−3.2, 3.5) 0.94  LY30 (%) 0.10 (0, 1.30) 0.35 (0, 1.25) 0.3 (0, 0.2) 0.27 −0.5 (−1.8, 0.9) 0.49 24-h rapid thromboelastography measurements∥# [median (IQR)]  Activated clotting time (sec) 121 (105, 121) 113 (97, 121) −8.0 (−16.0, 8.0) 0.45 −3.7 (−16.9, 9.5) 0.56  K-time (min) 1.6 (1.2, 2.6) 2.0 (1.4. 2.3) 0.4 (−0.6, 0.9) 0.70 0.1 (−0.8, 0.9) 0.90  Alpha angle (deg.) 73 (70, 76) 73 (69, 75) 0.0 (−7.0, 4.0) 0.67 0.3 (−6.0, 6.5) 0.93  Maximal amplitude (mm) 60.2 (56.6, 64.2) 57.4 (52.4, 64.6) −2.8 (−6.0, −0.4) 0.09 −2.5 (−6.2, 1.2) 0.18  LY30 (%) 0.3 (0, 2.1) 0.8 (0, 2.0) 0.6 (−0.2, 0.7) 0.75 −0.1 (−0.8, 0.7) 0.89 Initial coagulation measurements  Prothrombin time (sec)** 12.6 (11.6, 14.2) 12.9 (11.2, 4.3) 0.3 (−1.0, 0.6) 0.58 −0.1 (−0.8, 0.6) 0.74  International normalized ratio†† 1.1 (1.1, 1.2) 1.1 (1.0, 1.2) 0.0 (−0.1, 0.0) 0.46 0.0 (−0.1, 0.1) 0.73 24-h coagulation measurements  Prothrombin time (sec)‡‡ 13.9 (12.2, 15.8) 14.5 (12.6, 7.2) 0.6 (−0.6, 1.8) 0.29 0.4 (−0.6, 1.4) 0.82  International normalized ratio§§ 1.2 (1.1, 1.4) 1.3 (1.1, 1.4) 0.0 (−0.1, 0.1) 0.54 0.0 (−0.1, 0.1) 0.53 24-h transfusion requirements  Red cells (U) 5.5 (3, 9) 6 (2, 10) 0.5 (−1.0, 1.0) 0.94 0.1 (−2.6, 2.8) 0.95  Plasma (U) 5 (3, 9) 4 (2, 9) −1.0 (−1.0, 1.0) 0.75 0.2 (−2.3, 2.7) 0.85  Platelets (U) 1 (1, 3) 1 (1, 3) 0.0 (0.0, 0.0) 0.86 0.1 (−0.9, 1.1) 0.79  Whole cells (U) 1 (1, 2) 1 (1, 2) 0.0 (0.0, 1.0) 0.86 −0.4 (−1.9, 1.1) 0.56  Total (U) 10 (5, 18) 9.5 (5, 19) −0.5 (−3.0, 2.0) 0.92 0.1 (−5.3, 5.5) 0.98  Total (mL) 3089 (1547, 5333) 2881 (1456, 5876) −208 (−754, 754) 0.97 −8.3 (−1619, 1603) 0.99

OR indicates odds ratio.

*Difference in proportions (difference) or medians (Wilcoxon rank-sum test).

†Adjusted for study site.

‡Excluding 27 subjects who refused participation past 24-hour endpoint (16 in standard care arm and 11 in cold stored platelets arm).

§Patients with attempted follow-up without documented mortality at 30-days and with verified in-hospital survival considered alive (standard care, n=10; cold stored platelet, n=14).

∥Thromboelastography measurements provide viscoelastic properties of a blood sample. Activated clotting time is the time in seconds between initiation of the test and the initial fibrin formation and is increased with factor deficiency or severe hemodilution. The α-angle is the slope of the tracing that represents the rate of clot formation, decreasing with hypofibrinogenemia or platelet deficiency. K-time is the time in minutes needed to reach 20 mm clot strength and is generally increased with hypofibrinogenemia or platelet deficiency. The maximal amplitude is the greatest amplitude of the tracing and reflects platelet contribution to clot strength. LY30 is the percent amplitude reduction at 30 minutes after the maximal amplitude and when elevated reflects a state of hyperfibrinolysis.