In patients with hemophilia, poor musculature and reduced bone mineral density may predispose them to the risk of fractures[9]. HPT-related fractures are a severe complication of HPTs and may be caused by minimal trauma or have no obvious cause[13]. Jensen et al. demonstrated that bone may be affected by HPTs through pressure necrosis and he femur is the most common site of involvement[11]. Usually, HPTs progress asymptomatically until pathological bone fractures or neurovascular compression occurs[4]. Literatures recommend surgical treatment for HPTs combined with bone erosion or fracture[12, 4]. There is no consensus on the standard surgical protocol because of the complexity and variety of HPTs and fractures. Thus, surgery is very challenging for surgeons.
Seven patients received internal fixation: 1 with screws and 6 with plates. Plate breakage occurred in 1 patient because of HPT recurrence, who eventually underwent amputation. Bone union was observed in 5 patients and visible adequate callus was observed in 1 patient. Intramedullary nails are reported to be preferable fixation method for HPTs combined with fractures[14], but the locking plate was used in the most of patients in our study. Because HPT resection, fracture reduction and internal fixation could be completed with a single incision. Moreover, because pseudotumors usually erode the bone cortex, the extramedullary blood supply is affected during pseudotumor resection, and the reaming required for intramedullary fixation may affect the intramedullary blood supply[15], which would aggravate the damage to the bony blood supply. Although studies have illustrated the effect of metal internal fixation on stress shielding, which will cause peri-implant fractures and nonunion[12], in order to provide support for fractured bone to facilitate union at an appropriate position and to allow patients to perform early functional exercises, rigid internal implants are necessary. The use of long plate can reduce risk for fixation failure and spread the stress to the whole bone[16, 17].
When the pseudotumor is large and difficult to remove or the bone is massively eroded, amputation can be an effective treatment option. The advantage of amputation is that it can completely eliminate pseudotumors and reduce the cost of surgery and the risk of readmission. In fact, when severe bone deconstruction occurs, pseudotumors almost completely erode the bone, and the affected limb loses its bony support, making it very difficult to reconstruct the limb. Jacob et al. reported that reconstruction with a custom total femoral prosthesis is a valuable alternative to amputation in massive pseudotumors of the femur and soft tissues of the thigh[18]. However, the long-term outcome of custom total femoral prostheses is not clear. In our study, 3 patients with pathological fractures caused by pseudotumors ultimately chose amputation. No HPT recurrence was observed among them. Due to the presence of pseudotumors, the surrounding tissues including blood vessels and nerves are anatomically abnormal, and there is a risk of excessive bleeding during amputation. Accordingly, the operation should be performed delicately to avoid iatrogenic vascular damage and excessive bleeding, especially when tourniquets are not applied. In our study, patients who undergo amputation usually use crutches to walk because poor economic situation. Attention should be paid to the additional bleeding risk of the upper limbs caused by the long-term use of crutches.
EF procedures are typically used for temporary fracture fixation, deformity correction, limb lengthening, etc. For fracture treatment, a major benefit of external fixators is that they can stabilize a fracture without the need for open reduction or invasive surgery at the fracture site[19]. In theory, EF can stabilize the fracture in a minimally invasive way and simultaneously avoid interfering with the pseudotumor. In this study, two fracture patients underwent CREF, and both developed pin infections. Similar to conservative treatment, EF does not fundamentally solve the cause of the fracture; that is, EF does not remove the pseudotumor. Even when fractures have been stabilized, it is difficult to prevent HPT progression through conservative treatment, and uncontrolled progression will affect both the pins and bone, resulting in pin infection, bleeding and bone erosion. In addition, to avoid pseudotumors, the entry points of the pins are located far from the fracture site, which means that fixation may not be powerful enough to control fracture displacement. Therefore, we do not recommend EF for patients with HPT-related fractures. However, when a patient's overall condition is not suitable for open surgery, EF may be able to provide temporary fixation. After the situation is corrected, the external fixation should be replaced with an internal fixation.
For HPT-related fractures, the main purpose of surgery is to remove the pseudotumors while providing stable conditions for bone union. Zhai reported the use of structural bone grafts for bone defects > 5 cm caused by HPTs[12]. Similarly, we used structural internal fixation with bone graft for patients with massive bone defects. Graft incorporation was observed in all patients. When autologous grafts are not sufficient, allogeneic structural bone can also be used. Bastiaan C et al. used allogeneic strut bone grafts for the treatment of fibrous dysplasia of the proximal femur with a mean follow-up of 13 years. They argued that cortical allografts were less prone to be affected by pathological fibrous dysplasia bone and therefore less prone to resorption and failure[20].
For patients without actual fracture, we recommend that if the HPT has eroded more than one-third of the bone diameter, struct grafts are necessary for mechanical stability. However, whether prophylactic internal fixation is performed should be based on radiographic and intraoperative findings according to the surgeon’s decision. Mirels et al. analyzed 78 metastatic long bone lesions from 28 patients, and the results showed that when the size of the lesion was measured as more than two-thirds of the diameter, the rate of fracture significantly increased[21].
There are some limitations in this study that should be noted. This study had a retrospective design and involved only a small number of patients. In addition, because the patients in this study were from different cities, sometimes it took several months for them to return to the hospital for review, which made it difficult for us to accurately evaluate the time of bone union. Furthermore, our follow-up period was relatively short.