While preoperative hemoglobin levels are a known risk factor for peri- and postoperative transfusion needs in knee arthroplasty [15
], this is the first study to assess the effect of preoperative epoetin-α
use on the mildly anemic revision knee patient. We report that our patients did not require transfusion, had higher postoperative and discharge hemoglobin levels, and had a shorter length of stay. Epoetin-α
successfully increased the preoperative hemoglobin levels (from 11.97
g/dL to 13.93
g/dL, resp.). One patient had a deep venous thrombosis (3.6%), postoperatively.
may be costly, its use in selected patient groups may actually reduce total cost (direct and indirect) during knee arthroplasty [18
]; however, Moonen et al. [9
] reported that epoetin-α
injection, supplemented by ferrofumerate tablets, increased the direct cost per patient when compared with a retransfusion system. They did note that without the collection of indirect costs a true cost-effectiveness analysis could not be performed. Although the retrospective nature of our study prevented a worthwhile cost analysis, our experience working with epoetin-α
in the mildly anemic patient has been safe and efficacious, and we hypothesize that the elevated postoperative hemoglobin level may increase the patient's short-term outcome [20
]. A Future randomized study should include direct and indirect cost data as part of their analysis.
Cushner et al. conducted a retrospective review of 100 consecutive patients who had a revision TKA. Some of the patient's received either a closed suction or reinfusion type drain. Fifty-two percent of the patients were female, which was lower than the 75% in our cohort. Fifty-eight percent of their patients participated in the preoperative autologous donation program (at least one unit of blood), and a significant decrease in preoperative Hgb levels was found. They noted that the preoperative autologous donation resulted in an “orthopaedic-induced anemia.” They also concluded that autologous donation may not be appropriate because it predisposed patients to transfusion. All control patients included in our study had mild anemia (10–13
g/dL) prior to surgery (whether preoperative autologous donation was utilized or not). Greater than half of our control patient's were transfused (17% allogeneic) with an average of 256cc's of blood.
De Andrade et al. compared epoetin-α
to a placebo in a primary knee arthroplasty double-blind study. They found that those with a mild anemia (10–13
g/dL) who received epoetin-α
IU/kg or 100
IU/kg) had a lower allogeneic transfusion rate relative to placebo [21
]. Additionally, Stowell et al. found that epoetin-α
weekly doses of 40,000 units raised hemoglobin levels from 12.3
g/dL to 13.8
g/dL, preoperatively. Their patients maintained higher levels peri- and postoperatively compared to a cohort who had preoperative autologous donation [22
]. Our revision patients responded with an increased hemoglobin level from pre- to postop of 11.97
g/dL to 13.93
g/dL, respectively. We believe this change was clinically significant and most likely prevented peri- or postoperative transfusions, decreased length of stay, and allowed patients to more actively participate in physical therapy sooner [20
According to Sehat et al., there may be hidden blood loss into the soft tissue and joint of an arthroplasty patient [14
]. In addition, two studies using radiolabeled RBC's showed unexplained peri-operative blood loss likely into tissue compartments [23
]. Our study was not powered to detect such a difference in hidden and total blood loss between patient groups; however, our average total and hidden blood losses were elevated when compared to the primary knee patient [14
]. In addition, when we attempted to evaluate total and hidden losses, we noted a few important imperfections. For example, the estimated blood loss (EBL) is a nonstandardized subjective measure that likely varies as much inter- as intrainstitutionally. Moreover, the median value has often been used to report hidden and total blood loss findings because the range of losses had tremendous variability [13
]. However, an interesting finding derived from such calculations is that a total knee arthroplasty, on average, has a higher hidden blood loss than total hip arthroplasty [14
]. The change in hemoglobin from pre- to postop in our study cohort was 3.24
g/dL. Such a change was similar to those recorded by Sehat et al. who noted that a TKA without reinfusion had a change of 3.3
g/dL and 2.8
g/dL after re-infusion [14
Our study resulted in not only an increase in preoperative hemoglobin levels, but also a higher hemoglobin level at time of discharge. This is substantial considering the control patients discharge counts included blood transfused.
We performed all cases under tourniquet control, which is thought to cause fibrinolysis [25
]. Furthermore, postoperative fibrinolysis is also thought to occur transiently [27
]. This increased activity may elevate blood loss after revision total knee arthroplasty [28
]. Interestingly, epoetin-α
has been found to transiently increase the number of circulating platelets as well as improve their function. This may potentially decrease total blood loss for the revision knee patient [30
]. There is also an antiapoptotic activity that in preclinical and small clinical studies has been shown to protect cells from hypoxic and ischemic events [31
]. However, cancer and chronic renal failure trial patients had an increased risk of thrombotic complications and death [34
]. Our study population developed one uncomplicated deep venous thrombosis (3.6%). However, one control patient developed a pulmonary embolism (2.3%). The control patient was a 93 yo female with an ASA score of 3 and BMI of 31, while the study patient was a 63 yo male with an ASA score of 4 and BMI of 31; neither patient had a history of deep venous thrombosis or pulmonary embolus in their past. All patients received the same postoperative treatment course that consisted of antithromboprophylaxis, early ambulation, and physical therapy. Additionally, no study patient had an ischemic event.
No study is without limitation. First, we collected all data retrospectively and there are inherent limitations in such study design. Therefore, we attempted to reduce potential confounding by patient-matching based on age, gender, BMI, and American Society of Anesthesiology (ASA) scores. We also included a consecutive series of case patients. Second, spinal anesthesia is known to be associated with less perioperative blood loss in TKA when compared to general anesthesia [37
]. The majority of our patients underwent spinal anesthesia but a few patients required general anesthesia. Third, the percentage of patients who received pre-op iron, folic acid, vitamin B12, and multivitamins was elevated in the study group which may have enhanced the effects of our epoetin-α
dosing regimen. Fourth, our patient selection was limited to those without a thromboembolic history. Fifth, the triggers for transfusion are physician dependent. To better control for this, we only included cases performed by two surgeons who utilize similar transfusion protocols. Furthermore, those patients treated with epoetin-α
did not require any transfusions. Therefore it is unlikely that a true clinical difference in transfusion criteria occurred between our cohorts.
In conclusion, the present study may suggest that epoetin-α
has a role in reducing the need for blood transfusion in the mildly anemic patient who undergoes revision knee surgery. It may also decrease patient length of stay allowing for earlier readiness to resume normal activities or meet short-term milestones [20
]. Its use may also be an attractive alternative to autologous donation. A randomized study to evaluate the direct and indirect costs of such a treatment methodology in the mildly anemic revision patient may be warranted.