Discussion
RhEPO is widely recognized to be effective in perioperative autologous blood donation during TKA (6-11). Despite the common application of rhEPO in total joint arthroplasty during the last decade, with benefits including attenuated Hb drop, decreased blood loss, and reduced need for transfusion, no final consensus has been reached with respect to the optimal dosing schedule, which has yet to be investigated.
Previous studies have focused on the standard high-dose protocol (300-600 IU/kg) of rhEPO starting 2-4 weeks before surgery in TJA patients (10, 17). However, the long-term preoperative dosing schedule required patients to return to the hospital every week for injections, which was not only inconvenient to the patients but also increased the preoperative waiting time and medical costs. In contrast, some investigators reported that more frequent perioperative application of a small dose of rhEPO might be more effective. It was reported that high weekly doses of rhEPO might be considerable wasteful due to limited EPO receptors on progenitor cells in the bone marrow, which are easily saturated. When the receptors are saturated, no amount of rhEPO works until the receptors are free to bind again, but by then, the serum level of rhEPO has dropped. Therefore, frequent application of a small dose of rhEPO could maintain a more constant low but more effective serum rhEPO level (11). Similarly, another study found that repeated multiple doses of rhEPO were more effective than a single dose in stimulating the reticulocyte response, even when the total amount of rhEPO was the same (12). However, even if applied in repeated small doses, there is still no consensus on many dosing schedules.
In this study, we summarized and compared the main three types of perioperative rhEPO dosing schedules and found no significant difference in Hb levels among the three groups until the day after TKA, when the Hb level in group A was markedly higher than that in groups B and C (the Hb level in group B was also significantly higher than that in group C). There was no need for transfusion in any group. Although the three groups had comparable intraoperative blood loss, the total blood loss in group A was still the lowest among the groups (P<0.05), while the total blood loss in group C was the highest (P<0.05). No treatment-related adverse events, such as DVT, PE or anaphylactic reaction, occurred throughout the trial, suggesting that the three dosing schedules of rhEPO were well tolerated, and no significant differences in other complication rates were observed among the three groups. Therefore, recommend initiating the dosing schedule of rhEPO starting from preoperative day 5.
rhEPO reduces the rate of Hb decline shortly after application in the postoperative period (9), with Hb reaching a recovery peak approximately 8-10 days later (18). We found that patients in group A had markedly higher Hb levels than those in groups B and C on the day after surgery, with higher Hb levels in group B than in group C, indicating that the dosing schedule of rhEPO from preoperative day 5 was better than that from day 3 and the day of surgery in reducing the rate of Hb decline in the early postoperative period. When the Hb level reached the lowest concentration on postoperative day 3, the day of which transfusions mostly occurred (19), patients in group A also had significantly higher Hb levels than those in groups B and C, which suggested that the drop in Hb level in group A was the mildest. The application of rhEPO in group A started 5 days before surgery, which meant that it was 8 days until postoperative day 3 when the erythropoiesis estimated by rhEPO reached the peak, which was comparable with prior studies (7). However, it was reported that patients who suffered a large blood loss in a short time would have a strong endogenous erythropoietin feedback (20). Another study also investigated that a linear–logarithmic relationship was found between the change in Hb level and endogenous erythropoietin feedback, which meant that the more the Hb dropped, the stronger the endogenous erythropoietin feedback, even if the effect is small when compared with the extraerythropoietin (21). Similarly, in this study, on postoperative day 3, the quick decline in Hb level in group C might be accompanied by a sharp increase in endogenous erythropoietin feedback, which could therefore stimulate erythropoiesis quickly, while the decline in Hb level in group B was not so fast that the exogenous rhEPO in group B still needed more time to stimulate erythropoiesis and promote the Hb level after surgery. This may be why the Hb levels between groups B and C became comparable on postoperative day 3.
Many studies have demonstrated that regardless of which kind of rhEPO dosing schedule is used, intraoperative blood loss is not influenced (7, 8), which is consistent with our study. In terms of total blood loss, we calculated the blood loss with the preoperative haematocrit (HCT) and the HCT tested on postoperative day 3 (19). The trend of HCT was consistent with the Hb level, so it was not surprising to find that total blood loss was significantly less in group A than in groups B and C, which suggested that the application of rhEPO 5 days before surgery could achieve the least total blood loss. However, the blood loss was net instead of actual because when losing blood, the body was still producing blood under the effect of haemopoiesis. Therefore, blood loss was calculated to compare the blood saving effect of different dosing schedules of rhEPO. With respect to transfusion requirements, prior studies all reported the application of rhEPO. In our study, no patients needed transfusion, which might be mostly due to the effective blood management in our centre, including the perioperative use of tranexamic acid (22), the intraoperative controlled hypotension and the application of rhEPO, leading to little blood loss and an extremely low transfusion rate.
Reticulocytes are generally released from the marrow 18 to 36 hours before their final maturation into erythrocytes, and they are regarded as a real-time assessment of erythropoiesis (20). It was reported that the blood reticulocyte count peaked 72 hours to day 5 after application of rhEPO (18). In this study, on preoperative day 1, the reticulocyte counts in groups A and B were markedly larger than those in group C, which was consistent with prior studies. We found that on postoperative day 1, the reticulocyte count in group C did not increase as much as those in groups A and B in group C, which may be because of the relatively late application of rhEPO from the day of surgery. However, on postoperative day 3, the reticulocyte count in group C caught up with the other two groups. After postoperative day 14, no significant difference was observed in the reticulocyte count among the three groups, which suggested that the marrow haematopoietic phase mobilized by exogenous rhEPO was over.
According to the instructions, rhEPO had the potential effect of increasing platelet count and blood viscosity, leading to hypercoagulability, thus increasing the risk of DVT and PE (23, 24). It was also reported that patients who underwent TJA were at high risk for developing DVT and PE, which relatively limited the clinical administration of rhEPO (23, 24). However, in this study, no episodes of DVT or PE occurred in any patient, which was attributed to not only the well-tolerated dosing schedules of the three groups but also the good perioperative thromboembolism prophylaxis in our centre.
However, there are also several limitations in our study. First, the sample size was calculated according to the decline in Hb level, which might not be sufficient to identify a significant difference in other indexes. In addition, we did not record the platelet count or any quantitative data related to coagulation function, such as D-dimer or fibrinogen/fibrinogen degradation products (FDPs), so it might be inaccurate to judge whether different dosing schedules had any influence on patients’ risk of DVT or PE. Finally, it was reported by prior studies that the application of rhEPO perioperatively could significantly increase medical costs (25). However, in this study, we did not collect health economics indexes.