Stimulating erythropoiesis before hip fracture repair for reducing blood transfusion: should we change the hemoglobin cutoff level for defining anemia in females?
2016; Wiley; Volume: 56; Issue: 9 Linguagem: Inglês
10.1111/trf.13750
ISSN1537-2995
AutoresManuel Múñoz, Susana Gómez‐Ramírez, Michael Auerbach,
Tópico(s)Cardiac Arrest and Resuscitation
ResumoHip fractures (HFs) are a common cause of morbidity and mortality among the elderly population. The incidence of HF is increasing as the mean age of populations rises in western countries. It has been estimated that the number of HFs worldwide will increase from 1.7 million in 1990 to 6.3 million in 2050.1 Blood loss at the fracture site and during surgical repair, iatrogenic hemodilution, and inhibition of erythropoiesis due to trauma and surgery-induced inflammation may lead to acute perioperative anemia requiring allogeneic blood transfusion (ABT). In addition, on admission to hospital, up to 40% to 50% may present with low preoperative hemoglobin (Hb),2 which is one of the risk factors for ABT.3 As a result, 30% to 70% are transfused perioperatively to rapidly and effectively, but transiently, restore Hb levels, although it is a suboptimal means of treating anemia.4 Evidence of clinical and economic disadvantages of ABT for perioperative anemia has prompted a growing interest in multidisciplinary, multimodal, individualized strategies, collectively termed "patient blood management" (PBM), aimed to minimize ABT and improve patient outcomes.5 PBM relies on the perioperative implementation of three pillars of care: stimulation of erythropoiesis, reduction of blood loss, and tolerance of normovolemic anemia (restrictive transfusion threshold). This new standard is now being implemented in many countries, especially for elective cardiac and orthopedic surgery.6 The integration of PBM strategies into routine care for patients undergoing nonelective surgery, such as HF repair, may present challenging limitations.7 After HF, the preponderance of published evidence supports the use of restrictive transfusion thresholds using lower Hb concentrations or symptoms of anemia in deference to liberal transfusion thresholds based on a 100 g/L Hb trigger.8 Although a restrictive transfusion protocol should be the cornerstone of any PBM program, other strategies to reduce both the frequency and the volume of ABT should be implemented. Reduction of perioperative blood loss with administration of tranexamic acid is being evaluated in a number of ongoing trials (NCT02664909, NCT02580227, NCT02736073, NCT01714336, NCT0155781, NCT10940536). At present, data are inconclusive, especially in terms of safety.9, 10 One option, low vacuum drains, has only been reported in observational studies on subcapital HF.11 In contrast, very short-term stimulation of erythropoiesis with intravenous (IV) iron, with or without recombinant human erythropoietin (EPO), has been shown to reduce ABT requirements in two randomized trials12, 13 and a large observational study.14 In this issue of TRANSFUSION, Bernabeu-Wittel and colleagues15 reported on the results of a multicenter, randomized, double-blinded, clinical trial, assessing efficacy of ferric carboxymaltose (FCM) with or without EPO in reducing ABT perioperatively after HF repair. A total of 306 patients with anemia (Hb 90-120 g/L) with pertrochanteric or subcapital HF who received preoperative IV FCM (FE group), IV FCM plus EPO (EPOFE group), or placebo were evaluated. The observed ABT rate was approximately 50% with no significant differences in the number of ABTs per patient, survival, health-related quality of life (HRQoL), and adverse events among treatment groups, although early postoperative complications (hypotension, major bleeding, other anesthetic or surgical complications) were decreased in EPOFE (23%) and FE (30%), compared to placebo (41%). A significant increase in Hb levels was achieved at discharge and 60 days after discharge in EPOFE compared to the placebo arm. There was a higher rate of anemia recovery in EPOFE compared with placebo 60 days after discharge. The authors concluded that preoperative treatment with FCM alone or in combination with EPO did not reduce ABT in patients with HF. There are issues of concern regarding this study. We believe that there are two major problems with the inclusion criteria. First, virtually all HF patients with preoperative Hb level of less than 100 g/L will be transfused perioperatively due to a further Hb decrease induced by perioperative blood loss.14 Therefore, they should not have been included in a trial evaluating the effect of IV iron with or without EPO on transfusion rate as the primary outcome. Second, observational and randomized studies in HF repair showed a benefit of IV iron administration for those with a Hb level of at least 120 g/L13, 16, 17 (Table 1), and in Europe EPO administration is approved for orthopedic surgical procedures with baseline Hb level of 100 to 130 g/L.18 Why were patients with Hb levels of 120 to 130 g/dL excluded? The authors stated that this decision was based on the sex of patients with HF (90% of women), to avoid treating nonanemic patients in whom the need for ABT is significantly reduced. Although low preoperative Hb is a major risk factor for perioperative ABT, blood circulating volume and total RBC mass may also play a role. Female patients have a lower blood circulating volume and lower RBC mass for the same Hb level than males, but similar volume of blood loss when undergoing similar surgical procedures. In the Austrian benchmark study,19 the mean relative perioperative RBC loss in knee arthroplasty, hip arthroplasty, and coronary artery bypass grafting was significantly higher in women than in men, which was reflected by a significantly higher proportion of females receiving ABT. Similarly, the OSTHEO study20 reported that for a given preoperative Hb level (e.g., 120 g/L), ABT rates were higher in females (42%) than in males (30%) who underwent lower limb arthroplasty. Additionally, for women undergoing major, noncardiac surgery, the prevalence of iron deficiency (as defined by ferritin level < 30 µg/L) for Hb 120 to 129 g/L is similar to that for Hb < 120 g/L, and significantly higher than for Hb ≥130 g/L (42%, 51%, and 24%, respectively; p < 0.05). This observation translates to inadequate iron availability for erythropoiesis, as defined by transferrin saturation <20% (58%, 69%, and 34%, respectively; p < 0.05).21 Thus, although considered nonanemic according to World Health Organization definitions,22 most women with Hb 120-129 g/dL would benefit from iron replacement. We reanalyzed our previously published ABT data on HF patients who received IV iron sucrose (400-600 mg), with or without EPO (40,000 IU), or standard care (oral iron or no iron), who had admission Hb levels between 90 and 130 g/L (n = 544).14 As depicted in Fig. 1, for Hb levels of 90 to 119 g/L (n = 239) there were no differences in ABT rates irrespective of treatment. For those presenting with Hb concentrations between 120 and 130 g/L (n = 305), who represent 20% of all HFs in this study (n = 1443), a significant reduction of ABT rate was observed among those who received IV iron sucrose with or without EPO. Very-short-term administration of IV iron and/or EPO has also been shown to reduce transfusion rates in other surgical series.14, 23-25 Transfusion rates in HF repair according to admission Hb and very-short-term treatment with intravenous iron sucrose (IVI), with or without recombinant human EPO (reanalyzed data from Muñoz et al.14). Data suggest that a Hb level of 120 g/L should be considered suboptimal in surgical settings in which significant blood loss is expected, and a higher cutoff level may be desirable (Hb > 130 g/L). In our opinion, the authors included patients who probably will not benefit from treatment, while excluding others who clearly would have avoided ABT. The "restrictive transfusion protocol" implemented for this study is misleading. "Patients with any Hb level <70 g/L received 3 units of packed red cells (PRC) independent of symptoms, and those with any Hb level ranging from 71 to 89 g/L in the presence of severe symptoms received 2 PRC."15 We do have two major concerns with this transfusion protocol. First, according to the largest randomized trial of surgical HF patients26 taking into account the high burden of comorbidity in an elderly cohort, a transfusion trigger Hb level of less than 70 g/L appeared unsafe. This transfusion trigger has been used in the ICU where patients are closely monitored27 but this situation is absolutely different than an orthopedic ward where care and monitoring may be less intense. Second, according to recommendations from most scientific societies and consensus statements, "single unit red cell transfusions should be standard for non-bleeding, hospitalized patients. Additional units should only be administered after re-assessment of related symptoms and Hb concentrations."28 Transfusion of fixed numbers of red blood cell units is inappropriate and makes observation of benefit of erythropoiesis-stimulating therapy on the transfusion index (units per transfused patient) problematic. Finally, assessment of HRQoL by the Short-Form 36 Version 2 for acute patients (SF-36v2) at enrollment and 60 days after discharge was a secondary endpoint. No significant difference between groups was found despite higher Hb concentrations and rates of anemia recovery among those receiving treatment.15 Phosphate plays an important role in patients and hypophosphatemia, especially when severe (<2 mg/dL), may be associated with various complications including fatigue, myocardial depression, proximal myopathy, or rhabdomyolysis.29 Hypophosphatemia is frequent after parenteral FCM injection, lasting for weeks or months, and may have clinical consequences, including persistent fatigue.29, 30 Therefore, FCM-induced hypophosphatemia may have affected responses to the SF36v2 questionnaire affecting the observation of lack of HRQoL improvement after IV FCM administration. Unfortunately, phosphate levels were not measured in this trial. This otherwise well-conducted, multicenter randomized trial missed an opportunity to ascertain whether very-short-term stimulation of erythropoiesis with IV iron and EPO may contribute to reduced ABT rate and improved HRQoL in HF patients presenting with low preoperative Hb levels. Further appropriately designed trials are urgently needed to assess the role of these therapies in patients with HF, particularly women. MM has received honoraria for lectures and/or consultancy from Vifor Pharma (Saint Gallen, Switzerland), Vifor Pharma España (Barcelona, Spain), Pharmacosmos A/S (Holbaek, Denmark), Zambon (Barcelona, Spain), and Sandoz (Barcelona, Spain); MA for consultancies from AMAG Pharmaceuticals (Waltham, MA), Pharmacosmos A/S (Holbaek, Denmark), and Luitpold Pharmaceuticals, Inc. (Shirley, NY). SGR has disclosed no conflicts of interest. Manuel Muñoz, MD, PhD1 e-mail: [email protected] Susana Gómez-Ramírez2 Michael Auerbach, MD3,4 1Perioperative Transfusion Medicine School of Medicine University of Málaga 2Internal Medicine University Hospital Virgen de la Victoria Málaga, Spain 3Auerbach Hematology and Oncology Baltimore, MD 4Georgetown University School of Medicine Washington, DC
Referência(s)