Hidden Blood Loss of Total Knee FVIII Prophylaxis in Hemophilia Arthritis : An Analysis of Inuencing Factors


 BackgroundTotal knee arthroplasty is the leading way to treat hemophilia arthritis. At present, there is a lack of research on the influencing factors of blood loss in total knee arthroplasty for hemophilia arthritis. This study comprehensively explores the definite factors affecting the hidden blood loss in total knee arthroplasty for hemophilia patients.Materials and methodsNinety-two hemophilia A patients who underwent total knee arthroplasty in our center were included. Demographic characteristics, laboratory data, surgical data, and complications were collected. The Gross equation and Sehat equation were used to calculate the estimated value of hidden blood loss. Multivariate stepwise linear regression analysis was used to determine the influencing factors of hidden blood loss.ResultThe hidden blood loss of hemophilia patients undergoing total knee arthroplasty was 1069.51±341.99mL, and the age was positively correlated with the hidden blood loss (P<0.001), while tranexamic acid, FVIII prophylaxis, and incremental invivo recovery were negatively correlated with the hidden blood loss (P<0.001, P=0.008, P=0.017).ConclusionElderly patients have a greater risk of blood loss, and additional preventive measures can be appropriately added. Intraoperative intra-articular injection of tranexamic acid is recommended to reduce hidden blood loss, FVIII prophylaxis is recommended for every patient. We recommend that all HA patients measure the incremental invivo recovery and develop a personalized coagulation factor infusion regimen.

However, there is still a lack of research on the in uencing factors of TKA hidden blood loss in HA patients. In this study, 92 patients with HA who underwent unilateral TKA in our center from January 2015 to October 2020 were included. Multivariate linear regression was used to explore the in uencing factors of postoperative hidden blood loss, including age, Tranexamic acid(TXA), FVIII prophylaxis, incremental invivo recovery(IVR), etc. The purpose of this study is to comprehensively explore the exact factors in uencing the hidden blood loss of HA patients undergoing total knee arthroplasty.

Ethical review
The clinical data of HA patients who underwent TKA in the Orthopedic Center of the First A liated Hospital of Zhejiang Chinese Medicine University from January 2015 to October 2020 were retrospectively collected and analyzed. This study was approved by the Ethics Review Committee of the First A liated Hospital of Zhejiang Chinese Medicine University and obtained the unique identi cation number of research registration (research registration number is 2020047). Each patient signed a written informed consent form.

Inclusion and exclusion criteria
The inclusion criteria include: A clear history of hemophilia A and the knee joint lesions meet the "Improved Arnold-Hilgartner Classi cation of Joint Lesions" Class IV diagnostic criteria [7] . One-stage unilateral total knee arthroplasty was performed in the Orthopedic Center of the First A liated Hospital of Zhejiang Chinese Medicine University.
Exclusion criteria include Lack of information such as height, weight, etc. With Other diseases such as AIDS, tumor, etc. Postoperative complications such as infection and nerve and blood vessel injury occurred. The positive inhibitors and/or Classical IVR of coagulation factorVIII is less than 66%.

Operation method
All the operations were performed by the same team in our center, and all patients were given general anesthesia and a tourniquet for the whole process. We adopted standardized methods: the median knee approach, with the incision length ranging from 10cm to 15cm, opened the patella, separated the soft tissue layer by layer, opened the joint cavity, exposed the femoral end and tibial plateau, and deployed the prosthesis sample to determine the suitable prosthesis model. All the prostheses used were surface prostheses provided by Zimmer Biomet ® and then xed with bone cement. It can be divided into two groups according to whether or not TXA was injected into the joint cavity during the operation. Before the tourniquet was released, TXA solution consisting of 1gTXA and 50ml normal saline was injected into the joint cavity in the user group, and then the joint cavity was closed. Without TXA, the joint cavity was closed directly.No drainage tube was placed in all patients during the operation.

Perioperative management
After admission, the level of FVIII was measured for the patients, then 2500 international units (IU) of recombinant human coagulation factorVIII was injected into the patients experimentally, and 2500IU was divided by the bodyweight to obtain the infusion dose per kilogram of body weight (iu·kg -1 ). Re-measure the level of coagulation factors in patients at 1 hour, 2 hours, and 8 hours after injection to obtain the peak level and valley level of coagulation factors, according to the calculation formula: Classical IVR [8] =[ the increase of actual coagulation factor (%or iu·dl -1 /the increase of expected coagulation factor (%or iu·dl -1 ] × 100%, and calculate the Classical IVR. The expected increase of coagulation factor (%or iu·dl -1 )= infusion dose per kilogram of body weight (iu·kg -1 )× 2 (iu·dl -1 /iu·kg -1 ). The Classical IVR was used as the basis of perioperative infusion dose of coagulation factors. According to the Expert Consensus on Perioperative Management of Hemophilia Orthopaedic Surgery in China [9] and Management Guide for Hemophilia [4] , the FVIII level should be kept at 80%-100% on the operation day, 60%-80% on 1-3 days after the operation, 40%-60% on 4-6 days after the operation and 30%-50% on 7-14 days after the operation.
Bayer Health Care LLC ® provided recombinant human coagulation factorVIII for all patients during hospitalization. All patients had no autologous blood transfusion. If the patient has HB<70g/L or HCT<25%, it is considered as an indication that allogeneic blood transfusion is needed [10] .
Patients had routine preoperative X-ray lms of the anteroposterior position of the knee joint and fulllength standing position of lower limbs, and Arnold-Higartner arthrosis classi cation was performed by the same professional radiologist [11] . Assess the patients according to the standard to determine the indications of total knee arthroplasty. Routine preparation before the operation also includes CT and MRI examination of the knee joint, routine ECG examination, B-ultrasound examination, lung function examination, liver and kidney function examination.
Cephalosporins antibiotics were routinely used to prevent infection 1 day before operation and 3 days after operation. On the rst day after the operation, ice packs were applied around the knee joint for 24 hours to reduce postoperative bleeding. Because of the particularity of hemophilia patients, we use intermittent lower limb pneumatic pumps to prevent Venous thromboembolism(VTE) of lower limbs instead of anticoagulants. From the second day after the operation, the same rehabilitation team in the center will perform postoperative rehabilitation training for patients, including passive and active knee joint movements, plantar exion and extension movements, getting out of bed, and walking with walking auxiliary equipment.

Quantitative and variable calculation methods
Gender, height, age, weight, HCT, HB, operation time,and FVIII peak level on the operation day were obtained directly from the medical records, in which the data of HCT and HB before operation were taken within 2 days before the operation, and those after the operation were taken at around 7 o'clock in the morning.
The level of FVIII valley is the level of FVIII measured after admission, and the peak level of FVIII is the level measured 1 hour after experimental infusion of FVIII. The Classical IVR is calculated according to the above formula, and the incremental IVR (IU·DL -1 /IU·kg -1 )= [FVIII peak level (%or IU·DL -1 -baseline FVIII There are multiple time measurement points for postoperative HCT. Compare the HCT values of multiple time measurement points, take the lowest value as postoperative HCT, and use the Gross equation formula perfected by Gross.J B to calculate [12] . Gross equation speci c calculation method: total blood loss = preoperative blood volume (PBV)× (preoperative HCT-postoperative HCT)/ (preoperative HCT+ postoperative HCT)×2. PBV=kl×height(m)^3+k2×weight(kg) +k3.(male kl=0.3669, k2=0.03219, k3=0.6041; Female kl=0.3561, k2=0.03308, k3=0.1833). According to Sehat equation [13] , Subtract the intraoperative blood loss from the total blood loss to get the hidden blood loss.
According to the operation records, the usage of TXA was obtained, including intra-articular infusion of 1gTXA+50ml normal saline; TXA is not used.
According to medical records or telephone follow-up, the FVIII prophylaxis of FVIII was obtained, including irregular FVIII prophylaxis of FVIII in the past year; In recent one year, there was a low dose regimen for prevention and treatment: 10IU/kg per dose, which was given to hemophilia A patients twice times a week; In recent one year, there was a middle dose regimen for prevention and treatment: 15~30 IU/kg per dose, which was given to hemophilia A patients three times a week; In recent one year, there were high-dose FVIII prophylaxis s: 25~40 IU/kg per dose, which was given to hemophilia A patients three times a week.
According to the medical records, Infection with hepatitis B virus(HBV) was obtained, including No Infection with HBV ; Infection with HBV.

Multivariate linear regression analysis
Describe the statistical results, in which the continuous variables are shown as means±standard deviation; Classi cation is shown as a percentage. Continuity variables can be directly used as dependent variables of multi-factor linear regression analysis, and classi ed variables are assigned by numerical values: see Table 1 for speci c treatment of variables. There are 19 factors as independent variables, Multi-factor linear regression analysis was carried out with the hidden blood loss after the operation as the dependent variable. 3 Results

General information
Summarize all data, and distinguish them according to quantitative or qualitative data, in which the quantitative data are shown as means ± standard deviation; See Table 2 for details. Qualitative information is shown in percentage; See Table 3 for details.  According to the professional knowledge, we analyzed the independence of each variable and preliminarily screened out eight variables: BMI, age, incremental IVR, operation duration, TXA, HBV, FVIII prophylaxis, and FVIII peak level on the operation day. Single-factor linear regression was conducted for the above variables, and the results are shown in Table 4.
The variables with P values less than 0.05 were included in the multivariate linear regression analysis. There were four variables: age, incremental IVR, TXA, and FVIII prophylaxis. b. Without FVIII prophylaxis as control.
c. Without HBV as control.

Results of multi-factor linear regression analysis
In this study, multivariate linear regression was used to predict hidden blood loss according to TXA, age, incremental IVR, and FVIII prophylaxis. It is judged that there is a linear relationship between independent variables and dependent variables by drawing partial regression scatter plots and scatter plots of studentbased residual and predicted values. It has been veri ed that the research observations are independent of each other (Durbin-Watson test value is 1.665); By drawing the scatter diagram between the student residual and the non-standardized predicted value, it is proved that the variance of the residual is equal.
The regression tolerance is greater than 0.1, and there is no multicollinearity. In the abnormal value test, there is no observed value with student deletion residual greater than three times the standard deviation, the data leverage value is less than 0.2, and there is no value with Cook distance greater than 1. P-P diagram indicates that the residual is an approximately normal distribution.  b. without FVIII prophylaxis as control.

Discussion
In our study, the total blood loss of patients was 1221.79±349.89mL, which was similar to the results of Huang [14] . But unlike expected, there is no signi cant difference in blood loss between HA patients and patients without coagulation disorder [11], [15] . Therefore, we believe that hemophilia patients do not have an extremely high risk of blood loss as expected by strictly implementing the perioperative blood management plan [4] for HA patients to ensure the required level of coagulation factors. Due to the application of tourniquet in the whole process, the intraoperative blood loss is rare. In our study, intraoperative blood loss of patients is 152.33±67.65mL, and we didn't analyze intraoperative blood loss because, in our operation, intraoperative blood loss is closely related to the pressure of tourniquet, but the effects of other factors are di cult to analyze, and the proportion of intraoperative blood loss to the total blood loss is low, and more blood loss comes from hidden blood loss. The study of Hu et al. [16] showed that the hidden blood loss of TKA in osteoarthritis(OA) patients was 465mL, and in our study, the hidden blood loss of patients was 1069.51±341.99mL. Zhai [17] and other studies also show that there is a signi cant difference between HA and OA in TKA (1326mL vs.854mL), so we think that hidden blood loss is the focus of HA blood management. Multivariate stepwise linear regression showed that TXA, FVIII prophylaxis, age and incremental IVR were related to hidden blood loss. However, the length of operation, the peak level of FVIII during operation, HBV, and BMI did not show any correlation.
Many research experiences [18], [19], [20] show that TXA in TKA can effectively reduce blood loss, and the research of Huang [14] shows that TXA in TKA can also achieve a good hemostatic effect for HA patients. In our study, patients injected with 1gTXA before the closure of the joint cavity will reduce the hidden blood loss (b=-395.806). TXA is a synthetic derivative of amino acid lysine. It inhibits brinolysis by reversibly blocking lysine binding sites on plasminogen. It also inhibits the activation of plasminogen by plasminogen activator. As the level of plasmin is lowered, the brinolytic activity is weakened, and brin is not decomposed, thus reducing bleeding [21] . TXA has no effect on other coagulation parameters such as activated partial thromboplastin time, prothrombin time, and platelet count due to different ways of hemostasis. Using TXA in TKA to reduce blood loss has been a common choice of clinicians, but there is no uni ed view on the administration mode and dosage of TXA. Sarzaeem et al. [22] compared three ways of TXA administration: intravenous injection, articular cavity injection, and drainage tube injection. The results showed that all three ways could reduce blood loss, while the intravenous injection was better than the other two ways. A meta-analysis of Alshryda et al. [23] showed that the hemostatic effect of local injection was better than that of the intravenous route. Keyhani et al. [24] shows that local injection and intravenous injection have similar effects. In addition, although VTE is rare in HA patients, we are concerned about the increased risk of VTE in HA patients who have high postoperative levels of FVIII while using only physical prophylaxis, if TXA is administered intravenously. Therefore, we carefully choose joint cavity local injection as the TXA administration method.
FVIII prophylaxis is considered as the main treatment of HA [25] , played an important role in preventing bleeding and arthrosis. However, we found that the hidden blood loss of patients treated with FVIII prophylaxis was lower (b=-130.529). We suspect that FVIII prophylaxis can effectively reduce the frequency and severity of joint bleeding, thus reducing the damage degree of muscles, bones, and blood vessels, so the bleeding in these tissues will decrease after TKA. FVIII prophylaxis was advocated, and it is recommended to adopt the medium-dose regimen or the high-dose regimen [26] , but it also means a huge economic burden. In the United States, the annual medical expenditure of each HA patient is 250,000 US dollars [27] , in developing countries, this is unrealistic. Among our patients, only 35 patients were treated with the low-dose regimen, and no patients were treated with medium and high-dose regimen. Therefore, we failed to explore the in uence of different dosage schemes on TKA hidden blood loss. We suggest that every patient should make an individualized FVIII prophylaxis plan, combining economic factors, On the premise of ensuring the quality of treatment, the minimum dose of FVIII prophylaxis should be maintained.
Multivariate linear regression showed that age was a positive in uencing factor of hidden blood loss (b=8.884), Dong [28] and others found that the activity of chaperone-mediated autophagy will decrease with the increase of age, thus leading to the decline of the function of hematopoietic stem cells. It is more di cult for patients to self-correct anemia. However, age has little in uence on hidden blood loss, and the age of patients with HA is generally young (the average age of patients in our study is 38.72±8.74 years old). But for older patients, We need to be alert to the greater risk of bleeding and take appropriate additional preventive measures.
Multivariate linear regression showed that the incremental IVR was the negative in uencing factor of hidden blood loss (b=-164.983). Theoretically, the dosage and frequency of FVIII infusions for each patient should be individualized according to the incremental IVR. Actually, according to the instructions of FVIII, the value of incremental IVR is usually regarded as 2 IU dL -1 /IU kg -1 . Our study also used this method to calculate the dose of FVIII. In the post-analysis, the incremental IVR of our patients was 2.75±0.32 IU dL -1 /IU kg -1 . This indicates that the actual FVIII level will be higher than expected, and fortunately, it will reduce the hidden blood loss. However, the higher FVIII does not mean a better result. Extremes of hemostasis are thrombosis, if the FVIII level is higher than the ideal level for a long time, VTE may occur [29] .In our study, VTE was not observed, so it was impossible to evaluate the in uence of high FVIII levels on VTE. All in all, the individualized difference of incremental IVR will lead to the deviation of perioperative FVIII level control. Only by calculating the infusion dose of FVIII based on the individual incremental IVR will we not go to the extreme of Massive blood loss or thrombosis.
We also considered the in uence of HBV on hidden blood loss. Before the emergence of FVIII products, HA patients were usually treated with plasma when bleeding, and transfusion-related hepatitis became a vital complication of hemophilia [30] . Among our patients, 49 have HBV, the proportion is as high as 53.3%, which may be close to the actual situation of the whole, and the damage of liver function may affect the coagulation function, so HBV must be taken into account as a factor of hidden blood loss. All of our patients had been given the formal liver protect and antiviral therapy before the operation. Therefore, their liver function is within the normal range; otherwise, the operation will not be allowed. Under this premise, the in uence of HBV on hidden blood loss was not statistically signi cant (P=0.793). Three of our patients were infected with HCV, but they were excluded because they were simultaneously infected with HIV, and no patients were infected with HDV. Therefore, we only discussed the effect of HBV on hidden blood loss. AIDS is also a severe complication of hemophilia [31] ; however, in China, the patients with HIV are not as common as those with hepatitis. In our center, there are three HA patients with HIV who have undergone TKA, but their blood loss is much higher than that of ordinary patients, and the perioperative management of HA patients with HIV, including the surgical plan, is more complicated [32] , it is necessary to formulate a more detailed treatment plan for HA patients with HIV. Therefore, we exclude HA patients with HIV from the study.

Conclusion
The blood loss of TKA mainly comes from hidden blood loss, which is as high as 1069.51±341.99mL.
The main in uencing factors are age, TXA, FVIII prophylaxis, and incremental IVR. Elderly patients have a higher risk of blood loss, so perioperative blood management should be strengthened. TXA can effectively reduce occult blood loss, and local injection into the articular cavity is recommended. FVIII prophylaxis can not only improve the joint condition but also reduce blood loss. It is recommended that every patient take FVIII prophylaxis. The incremental IVR affects the accuracy of perioperative FVIII level, which easily leads to serious consequences of blood loss or thrombosis. It is recommended that each patient measure the individual incremental IVR and formulate the FVIII infusion scheme, which can not only ensure the treatment quality but also greatly reduce the economic burden.

Availability of data and materials
Data cannot be provided due to identifying information of participants but are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests. Funding