The head of the femur protrudes into the pelvis beyond the iliac seating line and causes pain or dysfunction is known as acetabular entropion. Primary acetabular invagination is rare, and most of the clinical cases are secondary invaginations caused by other diseases, such as rheumatoid arthritis, ankylosing spondylitis, infection, trauma and metabolic genetic diseases. The management of acetabular invagination has always been a challenge for joint surgeons due to the presence of acetabular wall defects, low bone strength and internal displacement of the hip joint center. In recent years, total hip arthroplasty has gradually become the main treatment method for adult acetabular impingement, but there are still controversies regarding the efficacy and treatment methods. In recent years, total hip arthroplasty has been used in our department to treat 31 cases and 35 hips with acetabular impingement, and this paper discusses its surgical operation method and efficacy. I. General data From June 2003 to May 2008, total hip arthroplasty was used in our department to treat 31 cases of 35 hips with acetabular impingement, 16 cases of 18 hips in men and 15 cases of 17 hips in women; age ranged from 36 to 71 years old, with an average of 52.2 years old. [m1] The primary diseases included rheumatoid arthritis in 18 cases and 20 hips, ankylosing spondylitis in 6 cases and 8 hips, abrasive invagination after artificial femoral head replacement in 3 cases and 3 hips [m2] (Stryker prosthesis in the United States in 2 cases and the specific model in 1 case in China is unknown), infection in 2 cases and 2 hips (tuberculosis and septic arthritis in 1 case and 1 hip each), trauma and tumor in 1 case and 1 hip each. [m3] The relative position of the acetabular floor and Kohler’s line (iliac seating line) on the pelvic orthopantomograph was used to determine whether there was acetabular invagination, specifically using the Sotello-Garza and Charnley grading method. If the acetabular floor is located 1~5 mm medial to Kohler’s line, it is considered as mild acetabular impingement, 6~15 mm as moderate acetabular impingement, and >15 mm as severe acetabular impingement. In this group, there were 12 h of mild acetabular impaction, 8 h of moderate acetabular impaction, and 15 h of severe acetabular impaction. The center of rotation of the affected side was determined symmetrically by measuring the center of the hip joint on the healthy side on the orthopantomogram of the pelvis. In the case of bilateral lesions, the midpoint of the line connecting the outer upper and inner lower edges of the acetabular ring on the pelvic orthopantomograph is used as the center of hip rotation. The same surgeon should be used for all surgeries. A posterior-lateral approach to the hip joint was used, and the external rotator muscle group and joint capsule were incised to expose the hip joint. [m4] The femoral head was gently dislodged, and if dislodgement was difficult, the femoral head was removed after osteotomy 0.5-1 cm above the lesser trochanter. The acetabulum was carefully filed and the defect was filled with autologous cancellous bone particles made from the amputated femoral head, while allogeneic lyophilized cancellous bone particles (Shanghai Anjiu Biotechnology Co., Ltd.) were used for infected, tumor and revision patients. The acetabular floor was filled and compacted by back filing, and a non-cemented prosthesis of appropriate size was implanted by press-fitting [m5] (11 all-ceramic and 12 plain prostheses from Stryker, USA, 7 plain and 2 revision prostheses from Johnson & Johnson Depuy, USA, and 2 all-ceramic and 1 revision prosthesis from Wright, USA). For bilateral patients, the decision of simultaneous replacement was made according to the systemic condition. Postoperative antibiotics were routinely applied for 3 to 5 days to prevent infection [m6] (10 to 14 days for infected patients), and partial weight-bearing walking was possible on postoperative day 5 to 7, and full weight-bearing walking was possible after 8 to 12 weeks. Outpatient follow-up was performed at 1, 3, 6 months and 1 year postoperatively, and annually thereafter. The Harris hip score [m7] was used to evaluate hip function. The change in the vertical distance from the center of the acetabular cup to the line of the tear drop on the orthopantomogram of the pelvis of more than 2 mm was identified as loosening of the acetabular prosthesis[m8] ; a shortening of the horizontal distance from the center of the acetabular cup to the tear drop was identified as re-integration of the acetabulum[m9] . The healing of the graft was judged according to the presence or absence of continuous bone trabeculae through the interface of the graft and the host bone. All cases were followed up[m12] for 19 to 152 months, with a mean of 46.5 months. The average operative time was 82 min, the average bleeding volume was 265 ml, and there was no medically induced vascular nerve injury or fracture. 1 patient developed deep vein thrombosis in the right lower extremity on the second postoperative day, which healed after muscle contraction exercise and anticoagulation with low-molecular heparin sodium [m13]. 1 case developed [m14] mild thigh pain 2 months after surgery, which did not affect function without special treatment, and the thigh pain disappeared six months after surgery. No complications such as infection, dislocation, prosthetic stem fracture, osteolysis and wear occurred in the whole group of patients. The mean preoperative Harris hip score was 48.9±6.5, which increased to 91.2±5.7 at the last follow-up [m15], with a mean improvement of 42.3 points compared with the preoperative score. There were 27 excellent hips and 8 good hips, with an excellent rate of 100%. Total hip mobility increased by 122±6.2°; including 54.3±4.6° in flexion, 22.7±3.8° in abduction, 19.6±2.8° in internal rotation, and 19.1±2.7° in external rotation. Patient subjective satisfaction was good[m16] . Radiographs at the final follow-up showed that bony stability was obtained in all prostheses, and discontinuous periprosthetic transillumination lines were seen in acetabular regions I and II at 29 and 34 months postoperatively in 2 hips, respectively, but no acetabular loosening or re-integration occurred by the end of the follow-up period. The average postoperative healing signs were seen at 6 months after surgery, with continuous bone trabeculae passing through the junction of the host bone-graft bone on X-ray. I. Reconstruction methods of the invaginated acetabulum Compared with the ordinary hip joint, the invaginated acetabulum mainly has the following characteristics: (1) the femoral head is embedded in the acetabulum with a small mouth and a large base, which causes the center of rotation of the hip joint to shift inward and the range of motion of the joint to be significantly limited, resulting in the femoral head to be prolapsed and the anatomical center of the hip joint to be difficult to locate; (2) the invaginated acetabulum generally has poor bone quality and low strength, and is prone to fracture during surgery; (3) the inner wall of the acetabulum is defective and the (3) the inner wall of the acetabulum is defective and the socket ring is weak, which is not strong enough to support the prosthesis and is prone to loosening and re-integration after surgery. The above characteristics of the invaginated acetabulum determine the complexity of its management. A study by Bayley et al. showed that the loosening of the acetabular prosthesis was significantly reduced when the center of the head was closer to the anatomic center of the hip joint than 10 mm after total hip arthroplasty. The loosening rate of the acetabular prosthesis will be as high as 50% if the center of rotation is restored to the anatomic position, while the loosening rate will be only 8% if the center of rotation is restored to the anatomic position. Therefore, in the case of an invaginated acetabulum, it should be moved out to the anatomic center of the hip joint as much as possible to reduce the occurrence of loosening. Outward migration of the prosthesis to the anatomic center of the hip joint will inevitably result in a defect in the medial wall of the acetabulum, which must be repaired to provide adequate medial support for the prosthesis in order to maintain stability. In the early stages, the bone defect was usually filled with bone cement and a cemented acetabular prosthesis was directly installed, which had a high incidence of loosening and re-integration after surgery. Because it is difficult to provide long-term stable support for the prosthesis with large pieces of bone cement, the polymerization heat generated during curing leads to bone necrosis of the thin medial wall, resulting in further bone defects; moreover, cemented prostheses have difficulty in accurately restoring the anatomic center of the hip joint and have poor long-term results. Therefore, for medial wall defects, bone graft repair should be used to provide long-term stability to the prosthesis through bony healing of the grafted bone to the host, thereby reducing the incidence of loosening and re-integration. The use of cemented or uncemented acetabular prostheses after repair of medial acetabular wall defects with bone grafting is controversial. In the early stage, cemented prostheses are preferred, which have good early and mid-term outcomes, but unsatisfactory long-term outcomes, such as a 22% failure rate in Schulte’s study with a 20-year follow-up. The failure rate was 44% at 10-14 years. Thus, the use of cemented prostheses in younger patients under 50 years of age has resulted in higher rates of loosening. In recent years, there has been an increasing number of studies using uncemented prostheses, and their long-term results are better than those of cemented prostheses, especially in younger patients. In 79 patients younger than 50 years of age with acetabular reconstruction, no loosening occurred 7 to 11 years after surgery. Although most scholars choose the uncemented fixation method, some still insist on cemented fixation and achieve better results. We believe that patients with invaginated acetabulum are generally young and active, such as the average age of 52.2 years in this group, and should choose non-cemented prosthesis to reconstruct the acetabulum as much as possible in order to obtain better biological fixation and long-term stability. In conclusion, the reconstruction strategy for the invaginated acetabulum is: outward displacement of the acetabulum to the anatomical center, bone grafting to repair the medial defect, and reconstruction of the acetabulum with a non-cemented prosthesis. Second, the surgical technique The characteristics of the invaginated acetabulum determine that its treatment and surgical techniques are different from those of ordinary total hip replacement, and we appreciate that the following aspects should be noted during the surgical operation. (a) Removal of the femoral head: the acetabulum of acetabular invagination is shaped like a small mouth and large bottom “beaker”, the femoral head is embedded in it and difficult to get out, coupled with the thin and loose acetabular bone, it is easy to cause fracture. Therefore, when dislodging the head of the femur, you should try to dislocate it with gentle force, do not use rough force, and try to avoid forcible dislodgement with metal aids to avoid fractures of the acetabulum or femur side. If dislocation is not possible, the femoral neck should be decisively truncated at 0.5~1cm above the lesser trochanter, and the femoral head should be cut into 3~4 pieces with a chainsaw or bone knife and removed in stages, i.e., by the retrograde removal method. The removed femoral head is placed properly for bone grafting. The retrograde extraction method is easy and fast, which can effectively avoid damaging the acetabulum, protect the integrity of the acetabular ring, and create favorable conditions for subsequent reconstruction. (ii) Acetabular preparation: The bone strength of the invaginated acetabulum is low, and the acetabular ring and floor are weak and easily damaged. Unlike ordinary total hip replacement, we prepare the socket in two steps: acetabular ring preparation and acetabular floor preparation. Note that the acetabular ring is treated first and then the acetabular floor. Acetabular ring preparation: In order for the acetabular ring to provide stable support for the prosthesis, it is important to maintain the maximum amount of bone and the integrity of the ring. According to the preoperative X-ray measurements and the actual intraoperative situation, an acetabular file [m23], which is one to two sizes smaller than the true acetabular ring, is selected for grinding and filing the ring, with the base of the file [m24] not in contact with the acetabular floor and only the cartilage of the ring removed. Gradually increase the diameter of the acetabular file until the subchondral bone is exposed. Care should be taken to be gentle during filing, not to damage the acetabular ring, and to ensure that at least 60% to 70% of the prosthesis is covered to provide initial mechanical stability. Preparation of the acetabular floor: The acetabular floor is often surrounded by a membrane that needs to be scraped away. If there is a cartilaginous surface, a file smaller in diameter than the acetabular ring can be used to lightly file it or just scrape it away with a spoon, using a kerfing needle to drill until fresh blood oozes out if necessary. Be careful not to file the acetabular floor excessively, as this may cause further bone loss and bone defects. (iii) Bone grafting technique The socket ring of the invaginated acetabulum is weak, and the mechanical support provided by the prosthesis is weaker than that of a normal total hip replacement; after the acetabular prosthesis is placed externally, the internal stress it bears increases, and it is easy to loosen and invaginate again. In this case, the bone grafting technique should be very important. Bone grafting not only fills the bone defect, but also assists the socket ring to provide mechanical support for the prosthesis and provides bony stability in the medium to long term. Proper bone grafting techniques can improve the strength of the implant, promote early healing, and prevent re-integration. In our group, we mainly made cancellous granular bone from the intercepted autologous femoral head as bone graft material, and the size of the granular bone should be as consistent as possible, with the optimal diameter range being 0.5~1 cm. If the diameter of the granular bone is less than 0.5 cm, the ability to prevent re-integration of the acetabulum decreases; if it is larger than 1 cm, the granular bone is not easily compacted and is prone to early loosening. We used a homemade granular bone production device to produce a more uniform granular bone with a diameter of 0.7~1 cm. In order to reduce the loss of osteogenic factors, the granulated bone was not washed with saline and not overly rinsed in the acetabular floor bed, but simply wiped with a gauze ball. After the above treatment, the granular bone was implanted in the acetabular floor and compacted by back filing and, if necessary, by using a compressor to achieve a certain strength of the implant and to provide solid medial support for the acetabular prosthesis. After the bone grafting was completed, an acetabular prosthesis with a diameter of 2 mm was used according to the angle of the expanded and filed acetabulum, which was installed by the press-fit technique. In this group of cases, all implants healed after surgery without loosening or re-integration of the acetabulum.