A Transfusion Epidemic in Dialysis Patients? Time to Reassess
The DOPPS study has been monitoring key benchmarks of care for patients with ESRD since the change in the bundled payment system in January 2011, and noted that mean hemoglobin levels declined by 0.10 g/dL between August 2010 and July 2011. However, between July and October 2011, mean levels declined by an additional 0.29 g/dL. The timing and magnitude of this change points toward the label change as a greater influence on this decline than the change in the bundled payment system.
Transfusions have long raised concern in the nephrology community, owing to their potential to increase a patient's alloantigen sensitization, which can prolong the time to receipt of a transplant. Given that the product label for ESAs now instructs prescribers to use the smallest dose sufficient to avoid transfusion, the existing research needs to be rigorously evaluated and thoroughly vetted in order to identify the hemoglobin threshold above which transfusion would not be needed.
The few studies that are available in the literature provide interesting conclusions. Seliger and colleagues examined transfusion rates and outcomes in a large population of persons with CKD and anemia receiving healthcare within the Veterans' Administration. Each of the 1837 patients in the dataset was identified as having had their ESA therapy initiated early (at a hemoglobin level between 10 and 11 g/dL) or initiated late (at a hemoglobin level between 8.0 and 9.9 g/dL). Those who initiated ESA therapy early had a lower risk for receiving a subsequent red blood cell (RBC) transfusion (hazard ratio [HR], 0.71; 95% confidence interval [CI], 0.59-0.97; P = .001).
Foley and colleagues examined the dataset of a completed trial in which 596 patients with ESRD had been randomized to 1 of 2 hemoglobin targets (9.5-11.5 g/dL or 13.5-14.5 g/dL). The transfusion rates were 0.66 per patient per year in the lower-target arm and 0.26 per patient per year in the higher-target arm (P < .001). In the multivariate model, an epoetin dosage of greater than 6000 U/wk and a transferrin saturation value of 32% or less were both associated with a greater risk for transfusion. After these factors were controlled for, the lower-target arm continued to be significantly associated with transfusion (HR, 0.42; 95% CI, 0.26-0.67; P < .001 favoring the higher-target arm).
Toward understanding the clinical scenario that prompted the transfusion, all adverse events in the 4 weeks before the transfusions were examined. One can debate the extent to which specific adverse events seem to have a greater or lesser likelihood of an association with the need for transfusion. For example, hemorrhage, infection, vascular access problems, and surgery could be considered as likely to be related; cardiovascular and respiratory events as potentially related; and musculoskeletal, skin, and neurologic events as unlikely to be related. The only significant difference in the proportions of participants who experienced adverse events was seen for events described only as gastrointestinal (11.2% vs 1.3% in the lower- and higher-target groups, respectively). Although the limited information precludes any strong conclusions to truly define the event and its relationship to the transfusion, one could hypothesize that if these were gastrointestinal bleeding events, then the lower target does not provide a significant buffer against transfusion.
In light of these studies, 2 important questions emerge after considering Dr. Allan Collins' research, discussed in the New York Times.First, at what hemoglobin value and in what clinical circumstances is it appropriate to consider a transfusion in a person with chronic anemia, such as anemia associated with CKD? And second, what is the impact of transfusions on sensitization before transplant using the more modern methods of white cell filters?
Some guidelines provide suggestions regarding the use RBC transfusions. The American Red Cross' compendium of clinical practice guidelines states that patients with asymptomatic chronic anemia should be treated with pharmacologic agents, such as ESAs, on the basis of the specific diagnosis and transfused to minimize symptoms and risks associated with the anemia. The compendium further states that transfusion is usually required when the hemoglobin level is less than 6 g/dL.
Similar to this and specific to anemia due to kidney disease, the 2008 guidelines from the Japanese Society for Dialysis Therapy state that clinical symptoms due to anemia do not appear until hemoglobin levels are less than 7 g/dL. They further provide specific guidance on the clinical scenarios in patients with CKD that might prompt transfusion (Table).
Table. Representative Patients Who Require Red Blood Cell Transfusion
ESA = erythropoiesis-stimulating agent
The downsides of transfusion include the concern that patients will become sensitized to antigens, thereby increasing their panel-reactive antibody levels and negatively affecting the possibility of kidney transplantation. However, some investigators have suggested that the use of modern white cell filters may minimize sensitization and obviate this concern.
Perkins and colleagues examined this possibility in a contemporary retrospective dataset analyzing all adult patients listed between 2004 and 2008 for first deceased donor kidney transplantation at 2 transplant centers in central Pennsylvania. Of the 407 patients listed for transplant and included in the analysis, 84 received a deceased donor kidney during the period of study. After periods during which the patients were inactive on the wait list were controlled for, transfusion independently predicted longer transplant wait time (HR 0.27; 95% CI, 0.11-0.69; P =.01 favoring those who had not been transfused). ESA therapy, however, was not associated with a longer wait time. This supports that even in a small population, the influence of RBC transfusion on transplant wait time remains an important consideration for clinicians and policymakers.
Transfusions and the Evidence
The DOPPS study has been monitoring key benchmarks of care for patients with ESRD since the change in the bundled payment system in January 2011, and noted that mean hemoglobin levels declined by 0.10 g/dL between August 2010 and July 2011. However, between July and October 2011, mean levels declined by an additional 0.29 g/dL. The timing and magnitude of this change points toward the label change as a greater influence on this decline than the change in the bundled payment system.
Transfusions have long raised concern in the nephrology community, owing to their potential to increase a patient's alloantigen sensitization, which can prolong the time to receipt of a transplant. Given that the product label for ESAs now instructs prescribers to use the smallest dose sufficient to avoid transfusion, the existing research needs to be rigorously evaluated and thoroughly vetted in order to identify the hemoglobin threshold above which transfusion would not be needed.
The few studies that are available in the literature provide interesting conclusions. Seliger and colleagues examined transfusion rates and outcomes in a large population of persons with CKD and anemia receiving healthcare within the Veterans' Administration. Each of the 1837 patients in the dataset was identified as having had their ESA therapy initiated early (at a hemoglobin level between 10 and 11 g/dL) or initiated late (at a hemoglobin level between 8.0 and 9.9 g/dL). Those who initiated ESA therapy early had a lower risk for receiving a subsequent red blood cell (RBC) transfusion (hazard ratio [HR], 0.71; 95% confidence interval [CI], 0.59-0.97; P = .001).
Foley and colleagues examined the dataset of a completed trial in which 596 patients with ESRD had been randomized to 1 of 2 hemoglobin targets (9.5-11.5 g/dL or 13.5-14.5 g/dL). The transfusion rates were 0.66 per patient per year in the lower-target arm and 0.26 per patient per year in the higher-target arm (P < .001). In the multivariate model, an epoetin dosage of greater than 6000 U/wk and a transferrin saturation value of 32% or less were both associated with a greater risk for transfusion. After these factors were controlled for, the lower-target arm continued to be significantly associated with transfusion (HR, 0.42; 95% CI, 0.26-0.67; P < .001 favoring the higher-target arm).
Toward understanding the clinical scenario that prompted the transfusion, all adverse events in the 4 weeks before the transfusions were examined. One can debate the extent to which specific adverse events seem to have a greater or lesser likelihood of an association with the need for transfusion. For example, hemorrhage, infection, vascular access problems, and surgery could be considered as likely to be related; cardiovascular and respiratory events as potentially related; and musculoskeletal, skin, and neurologic events as unlikely to be related. The only significant difference in the proportions of participants who experienced adverse events was seen for events described only as gastrointestinal (11.2% vs 1.3% in the lower- and higher-target groups, respectively). Although the limited information precludes any strong conclusions to truly define the event and its relationship to the transfusion, one could hypothesize that if these were gastrointestinal bleeding events, then the lower target does not provide a significant buffer against transfusion.
In light of these studies, 2 important questions emerge after considering Dr. Allan Collins' research, discussed in the New York Times.First, at what hemoglobin value and in what clinical circumstances is it appropriate to consider a transfusion in a person with chronic anemia, such as anemia associated with CKD? And second, what is the impact of transfusions on sensitization before transplant using the more modern methods of white cell filters?
Some guidelines provide suggestions regarding the use RBC transfusions. The American Red Cross' compendium of clinical practice guidelines states that patients with asymptomatic chronic anemia should be treated with pharmacologic agents, such as ESAs, on the basis of the specific diagnosis and transfused to minimize symptoms and risks associated with the anemia. The compendium further states that transfusion is usually required when the hemoglobin level is less than 6 g/dL.
Similar to this and specific to anemia due to kidney disease, the 2008 guidelines from the Japanese Society for Dialysis Therapy state that clinical symptoms due to anemia do not appear until hemoglobin levels are less than 7 g/dL. They further provide specific guidance on the clinical scenarios in patients with CKD that might prompt transfusion (Table).
Table. Representative Patients Who Require Red Blood Cell Transfusion
| • Patients with severe anemia and signs/symptoms specific to anemia • Patients with acute blood loss associated with unstable hemodynamics • Patients with severe angina pectoris • Intraoperative patients with a large volume of blood loss • Patients with a very low response to ESA |
ESA = erythropoiesis-stimulating agent
The downsides of transfusion include the concern that patients will become sensitized to antigens, thereby increasing their panel-reactive antibody levels and negatively affecting the possibility of kidney transplantation. However, some investigators have suggested that the use of modern white cell filters may minimize sensitization and obviate this concern.
Perkins and colleagues examined this possibility in a contemporary retrospective dataset analyzing all adult patients listed between 2004 and 2008 for first deceased donor kidney transplantation at 2 transplant centers in central Pennsylvania. Of the 407 patients listed for transplant and included in the analysis, 84 received a deceased donor kidney during the period of study. After periods during which the patients were inactive on the wait list were controlled for, transfusion independently predicted longer transplant wait time (HR 0.27; 95% CI, 0.11-0.69; P =.01 favoring those who had not been transfused). ESA therapy, however, was not associated with a longer wait time. This supports that even in a small population, the influence of RBC transfusion on transplant wait time remains an important consideration for clinicians and policymakers.
SHARE
