Transfusion Trigger After Operations in High Cardiac Risk Patients

Background: Despite the need for clinically appropriate use of blood products in the postoperative setting, blood transfusion practices are empiric and variable. In the absence of a physiologic test that can effectively guide transfusion-related decisions after an operative intervention, hemoglobin-based transfusion triggers have been suggested as clinical tools. Traditionally, clinicians have transfused patients to maintain hemoglobin (Hb) above a minimum (typically 10 gm/dl) level, in order to prevent adverse cardiac events and death. Recent randomized trials, however, have shown that restrictive transfusion policies (transfusion when Hb falls below 8 gm/dl or even lower) are well tolerated by specific patient populations. Furthermore, these trials have demonstrated that in some patient subsets reduced transfusion is associated with reduction in postoperative complications and death. Thus, guidelines have been developed emphasizing the need for a restrictive transfusion strategy in most stable hospitalized patients.

Despite the emerging enthusiasm with respect to the safety of restrictive transfusion strategies, high quality evidence on the value of such an approach in patients at high risk for postoperative adverse cardiac events remains scarce. This is a serious limitation of the current literature, as ischemic heart disease (IHD) is highly prevalent and represents the leading cause of mortality in this country, accounting for the death of one American every minute. Furthermore, there is evidence from small trials and secondary analyses that in this subset of patients withholding transfusion when Hb falls below 10 gm/dl increases the risk of death or myocardial infarction, suggesting that widespread use of restrictive transfusion policies may actually result in patient harm. This uncertainty on transfusion thresholds in high cardiac risk patients has created a knowledge gap that requires urgent attention. IHD is highly prevalent in patients with peripheral arterial disease (PAD), and myocardial infarction represents the leading cause of postoperative mortality in patients undergoing PAD-related surgical interventions. Furthermore, a substantial proportion of patients undergoing Vascular and General Surgery operations have history of prior IHD, making this patient population an ideal high cardiac risk group in which to analyze the effect of transfusion strategies. In order to address the knowledge gap of postoperative transfusion thresholds in patients at high risk for postoperative adverse cardiac events, the investigators propose the current study under the hypothesis that transfusion strategy will affect important postoperative outcomes after major surgical interventions in high cardiac risk patients.

Objectives: The goal of the proposed study is to determine whether a liberal postoperative transfusion strategy (transfusion trigger at Hb<10gm/dl) in patients at high risk for postoperative adverse cardiac events will reduce the risk of adverse postoperative outcomes after major vascular and general surgery operative interventions. The primary end point is the composite rate of all-cause mortality, acute myocardial infarction (MI), coronary revascularization, stroke, or acute renal failure within 90 days from the time of randomization. The secondary end points are rates of postoperative infectious complications (wound infections, pneumonia, and sepsis), and postoperative cardiac complications (new cardiac arrhythmias, congestive heart failure exacerbation, and cardiac arrest) at 90 days post-randomization; the composite rate of all-cause mortality, MI, coronary revascularization, stroke, or acute renal failure, within 30 days from the time of randomization; the length of hospital stay; and all-cause mortality up to one year after randomization.

Design: CSP #599 - TOP study is a randomized, intent-to-treat, two-arm, parallel design, single blind, multicenter trial. Vascular and General Surgery programs at Veterans Affairs Medical Centers with expertise in performing the operations of interest will be invited to participate and participants will be screened for enrollment using established inclusion/exclusion criteria. Enrolled participants will be randomized to one of the two arms; liberal (transfusion trigger at Hb < 10gm/dl) or restrictive (transfusion trigger at Hb < 7gm/dl). Consent for the study will be obtained prior to the index surgical intervention. Randomization will be performed postoperatively after the patient has a confirmed Hb < 10gm/dl. Assessments will be collected pre/post-operatively and at discharge, or at 30 days after randomization if the patient is still hospitalized. Follow up forms will be filled out during a clinic visit after the 30th and 90th post-randomization day. Patients who cannot present to the clinic will have a phone call for follow up. One year post-randomization mortality will be ascertained using electronic medical records, phone follow-up, and search of the national death index.

Sample Size and Study Duration: This study will randomize 1520 participants undergoing major vascular and general operations at 15 VA medical centers with expertise in Vascular and General surgical procedures. Assuming a recruitment rate of 3 participants per site per month, total recruitment will take approximately four years to complete. The duration of the study will be five years with four year recruitment, three months active follow up, and nine months passive follow up which will be performed by the Chairman's office in order to collect 1-year post-randomization data on all-cause mortality.

Study Population: The study will include a) patients who undergo open (non-endovascular) PAD - related operations, and b) patients who undergo selected major Vascular and General Surgery operations and have prior history of PAD or IHD. Vascular Surgery operations examined will include, but not be limited to, PAD-related: aortobifemoral or aortobiiliac bypass, open abdominal aortic aneurysm repair with simultaneous repair of aortoiliac occlusive disease, visceral bypass, iliofemoral bypass, femoral bypass or endarterectomy, infrainguinal bypass, supra-aortic trunk bypass or endarterectomy, carotid endarterectomy, and major lower extremity amputations (transfemoral, through the knee, or transtibial); Other vascular surgeries: open aneurysm repair (including but not limited to carotid, subclavian, abdominal aortic, iliac, femoral, or popliteal aneurysms); and complex endovascular aneurysm repair (defined as fenestrated endograft, or endograft with need for iliac conduit, or endovascular aneurysm repair with simultaneous femoral artery reconstruction or bypass). General Surgery operations examined will include open cholecystectomy or other complex biliary reconstruction (such as open common bile duct exploration for stones, reconstruction as part of oncologic operations such as palliative pancreatic cancer procedures) , small bowel resection, pancreatectomy, colon resection, rectal resection, splenectomy, transhiatal esophagectomy, liver resection, and open ventral hernia repair. Patients will be included in the study if their postoperative Hb falls below 10gm/dl within 15 days after surgery. Intervention: Liberal transfusion strategy is the conventional transfusion method for the surgical procedures the investigators will include in the study, and is defined as transfusion when the postoperative Hb drops below 10gm/dl, with a goal to maintain Hb above 10 gm/dl. Restrictive transfusion strategy will be defined as transfusion when the postoperative Hb drops below 7gm/dl with goal to maintain Hb above 7 gm/dl.

Significance: The proposed CSP# 599 study is uniquely positioned to address the knowledge gap of postoperative transfusion thresholds in high cardiac risk patients undergoing major surgical interventions, who are currently transfused based on data from patient populations with different risk profile. The study will include patients with high IHD burden, a population that should be the most likely to benefit from a liberal transfusion strategy and CSP#599 will examine this hypothesis directly. Conversely, if the proposed trial demonstrates that this high cardiac risk patient population can tolerate restrictive transfusion well, then the results will be easy to generalize to other patient populations and the question of transfusion thresholds will be definitively addressed. Thus, the investigators believe that regardless of outcome this trial will have significant clinical and policy implications, and will substantially impact the VA and national transfusion-related guidelines.