Clinical Outcome and Histopathologic Evaluations of Porous Tantalum Implant for the Treatment of Early-Stage Osteonecrosis

Background To evaluate the survival rate of porous tantalum rod implantation in the treatment of avascular necrosis of the femoral head, to evaluate its clinical effect and imaging results, and to analyze the reasons for its failure to return to total hip arthroplasty (THA). Methods From January 2008 to December 2013, tantalum rod implantation for avascular necrosis of the femoral head was performed in two institutions. Statistical analysis of operation data, including operation time, blood loss and blood transfusion. Harris hip score and imaging results were evaluated. Kaplan Meier survival analysis was performed with THA as the end point. Histopathological observations were performed on femoral head specimens with the tantalum rods that failed THA after tantalum rod implantation. Results 42 patients (52 hips) were followed up completely, the average follow-up time was 74.3 months (60-120 months). 24 hips turned to THA at the end of follow-up (46.2%), the average time was (44 ± 32) months, and the average Harris score before THA was (57 ± 10). The follow-up time of 28 patients without THA was 78.6 ± 13.8 months, and the average Harris score was 80.6 ± 2.8. Histopathological examination revealed the implant surface is in contact with sparse Island bone. There is limited bone tissue extending inward from the implant. In the living bone area, there was “on-growth” of new bone but no “in-growth”. In the necrotic area, there was no obvious new bone regeneration. The mid-and-long term clinical effect of tantalum rod implantation in the treatment of avascular necrosis of the femoral head is not good, and the osteogenic activity of tantalum rod in the femoral head is limited. ARCO stage, age and bone marrow edema were risk factors for the failure of tantalum rod implantation to THA.


3
Osteonecrosis of the femoral head (ONFH) is a kind of pathological state with multiple causes, which leads to the decrease of blood supply of the subchondral bone of the femoral head, the death of bone cells and the collapse of the articular surface 1 There are many methods for the early necrosis of the femoral head, including hyperbaric oxygen, shock wave and other conservative treatment [5][6][7] . Surgical operations include osteotomy, vascularized bone flap implantation, and core decompression 7-10 . At present, it is generally recognized that core decompression can reduce the pressure in the bone, remove the necrotic bone tissue, increase the growth of blood vessels in the femoral head, and promote the crawling replacement of new tissue, which can be rapidly pain relieved, but the clinical satisfaction rate of using core decompression alone is not excited. The main reason is that although simple decompression can relieve the pressure in the femoral head, it can reduce the internal supporting force of the femoral head. Therefore, bone grafting or artificial materials are needed to provide enough biological support for the joint surface of the femoral head. According to Hungerford et al. 11 , core 4 decompression combined bone grafting are suitable for the treatment of early avascular necrosis of the femoral head, which can effectively prevent the collapse of the femoral head and delay the time of hip replacement in the later stage. However, core decompression combined bone grafting often fail to provide effective bone support for the femoral head of patients. If the necrosis area is too large, the autogenous cancellous bone particles implanted after curettage of necrotic bone cannot provide effective subchondral support for the patients with core decompression combined bone grafting, and the autogenous iliac bone absorption implanted after operation is also the main reason for the collapse and deformation of the femoral head 12 .
Tantalum is a blunt metal with good biocompatibility. Bone and vascular tissue can be seen to grow rapidly in the hip joint with tantalum coating 13,14 . Core decompression combined with tantalum rod placement can release the internal pressure of the femoral head through core decompression, so as to alleviate the pain. Tantalum metal has the same elastic modulus as bone, and has the structure of bone trabecula, with high porosity.
The tantalum rod designed by Zimmer company in the United States has a cylindrical structure with an aperture of 430 μ m and a diameter of 10 mm, with a porosity of 75% -80%, and a length of 70-130 mm (5 mm increase). The elastic modulus is equivalent to that of human fibula (3GPa), which has a good support for the femoral head 15 . Tantalum rod can bear physiological load of human body, has good biocompatibility and good friction stability. The placement of tantalum rod can play a role in filling the bone at the core and supporting the femoral head, reducing the stress distribution of the surrounding bone tissue, and effectively preventing the surface of the femoral head from collapsing. At the same time, the porous structure of tantalum rod can induce osteoblasts to grow in, accelerate the regeneration of blood vessels and promote the process of vascularization, 5 which is conducive to the regeneration of femoral head and the repair of femoral head.
Many studies have reported that tantalum rod implantation has achieved good early clinical effect in the treatment of avascular necrosis of the femoral head 4,16,17 . However, the long-term clinical effect, weight-bearing time and effect of porous tantalum implant are still controversial [18][19][20] . In the past few years, the failure rate of tantalum rods in ONFH has been reported to vary from 2% to 56% 16,21 . Once the subchondral bone collapses, the progression of the disease is difficult to reverse. The collapse and deformation of the femoral head, the narrowing of the joint space, and the deterioration of the joint function occur in turn. Hip replacement has become the only treatment option for these patients.
Moreover, at present, only animal studies have confirmed the bone ingrowth of porous tantalum rods 13,22 , proving that porous tantalum rods implantation is related to rapid and extensive bone growth, but the bone ingrowth in human body is still unclear.
In this study, the survival data of patients with avascular necrosis of the femoral head treated by tantalum rod implantation were reviewed and analyzed, and the clinical and imaging results were evaluated. The effect of tantalum rod implantation on avascular necrosis of the femoral head and the related factors leading to its conversion to THA were analyzed. In addition, pathological sections of the femoral head with tantalum rod were evaluated to confirm the degree of bone growth of tantalum rod in human body, and to evaluate the mid-and-long term efficacy of porous tantalum rod in the treatment of early avascular necrosis of the femoral head.

Patient Selection
From January 2008 to December 2013, patients with avascular necrosis of the femoral head underwent tantalum rod implantation in two hospitals were analyzed retrospectively.  Table 1.

Surgical methods
After anesthesia, the patient was in supine position, the hip was sterilized and covered with towel, and the skin protection film was applied. A 4cm long incision was made 2cm below the trochanter to separate the fascial muscle layer by layer from the proximal and lateral cortex of the femur. The intersections of the horizontal line slightly above the trochanter and the lateral femoral cortex were used as the insertion points, and the center of the osteonecrosis area was used as the orientation of the insertion points. In order to eliminate the influence of the anteversion angle of the femoral neck on the position, the hip should be rotated 10 ° ~ 15 °, and the guide pin should be inserted in the central coronal plane of the femoral neck; the 3.2mm guide pin should be inserted 5 mm below the articular cartilage of the femoral head. Ream the hole along the direction of the guide pin, and use the hollow drill to gradually ream the marrow from 8mm diameter to 9mm; in the process of reaming, clean the hollow drill to prevent bone fragments from gathering in it. When the hollow drill enters the necrotic bone and does not penetrate the necrotic area, the necrotic tissue is removed through the bone tunnel with a multi-directional curette. Reaming with 10 mm hollow drill. Insert a depth gauge to determine the length of the required tantalum rod, and the measured length shall avoid the tail of the tantalum rod finally exposed outside the lateral cortex, so as to avoid the pain caused by its abrasion of surrounding soft tissue. Install the corresponding length of the tapping extension at the front end of the tapping, and rotate clockwise to make all the threads enter the femur. Screw in the tantalum rod (product of Zimmer company of the United States) so that its tail is slightly niche into the bone cortex; if there is resistance when transplanting, determine whether the tapping is in place or whether there is broken bone in the medullary canal. Fluoroscopy showed that the tantalum bar was in good position, and the incision was sutured layer by layer. The second generation cephalosporin antibiotics were given to prevent incision infection 30 minutes before operation and 24 hours after operation. In 12 weeks after operation, the patients' hip joint was fully limited in weight-bearing, and functional exercise was performed without weight-bearing. After 12 weeks, they walked with full weight-bearing.
The indications of THA were: continuous hip pain interfered with daily activities and deterioration of hip score, or radiation collapse of femoral head and intraarticular penetration of tantalum rod. The technology of tantalum rod transfer to THA after failure includes femoral neck osteotomy and implant cutting, both of which use a power saw to remove the tantalum rod from the rotor with a special ring drill. This procedure is performed as a routine hip replacement.

Clinical Assessment
The follow-up examination was arranged at 1 month, 3 months, 6 months and 12 months, 8 and then every 6 months. The evaluation parameters included Harris hip score and imaging and MR examination of the diseased hip. The X-ray films of hip joint were used to evaluate the size of lesions, the consistency of femoral head, whether there was crescent sign and the degeneration of hip joint. MR images were used to assess changes in bone marrow edema and lesion size. According to the ARCO 23 grading system, the initial stage and the degree of involvement of the femoral head were evaluated by radiology. More than 30% of necrosis is defined as large osteonecrosis. Myeloedema was defined as a low signal area on T1 weighted images, The high signal region is defined on T2-weighted image or inversion recovery image in the femoral head, neck and intertrochanteric region 24,25 . The clinical evaluation was conducted by a non surgeon observer throughout the study. Two radiologists who were not involved in the operation and did not know all the clinical information were measured radiographically. Each parameter is measured twice at appropriate intervals to prevent bias from affecting the results.

Analysis of bone implant specimens
The pathological section of the femoral head implanted with tantalum rod was completed in Shandong Key Laboratory of oral tissue regeneration by using the EXAKT hard tissue section system. The bone graft into the femoral head was fixed in 10% buffered formalin and prepared for histological analysis of thin sections without decalcification 26 . The section was mounted on a glass slide and gradually thinned by petrological grinding technology, stained by HE and Masson respectively, and examined by transmission optical microscope. This allows qualitative assessment of tissue response to tantalum implants, including calcification and fibrous tissue recognition, new bone formation, and vascular distribution and cellular structure of the femoral head.

9
The data and charts were analyzed and processed by IBMS PSS Statistical 19.0 statistical software. Continuous variables were analyzed using independent sample T test.
Categorical variables were analyzed using the Pearson chi-square or Fisher exact tests.
Kaplan-Meier survivorship analyses were used with the endpoint defined as reoperation with THA. Test level was set at both sides α =0. 05, P < 0.05 was considered statistically significant.

Clinical Results
The average follow-up time was 74.3 (60-120) months, and the preoperative score of 52 cases of hip joint was (74 ± 13) 6 months after treatment, Harris score was (80 ± 13) on average, 5.89 points higher than that before treatment, the difference was significant (P < 0.05). 12 months after treatment, the average Harris score was (84 ± 11), which was significantly different from that before treatment. 36 months after treatment, the average Harris score was (82 ± 11), and the average score was (69 ± 15) at the end of follow-up or before THA. (Figure.1) According to the stage of ARCO before operation, the average score at the end of follow-up of patients with ARCO I was higher than that of patients with ARCO II (P < 0.05); the average score at the end of follow-up of patients with age ≤ 40 years was significantly higher than that of patients with age > 40 years (P > 0.05); there was no significant difference in Harris score at the end of follow-up of different causes (P > 0.05).

Radiographic Results
At the last follow-up, 28 of 52 hip joints showed imaging progress (53.8%). Among them, 5 of 22 hips in ARCO I stage progressed to stage III and 3 to stage II; 12 of 30 hips in ARCO I stage progressed to stage III and 8 to stage IV. The imaging progress rate of hip joint in ARCO I stage (36.4%; 8/22) was significantly lower than that in ARCO Ⅱ stage (66.7%; 20/30) (x 2 = 4.690, P < 0.05). The progress rate of patients with age > 40 years was significantly higher than that of patients with age ≤ 40 years (x 2 = 4.748, P < 0.05).
Among the 28 hip joints with imaging progress, 24 (89.1%) needed THA finally, while none of the 24 hip joints without imaging progress needed THA.  There was a significant difference between the two groups (x 2 =8.421, P<0.05). The proportion of patients over 40 years old who needed THA was higher than those under 40 years old (x 2 =4.748, P<0.05). The proportion of hips with bone marrow edema to THA was higher than that without bone marrow edema (x 2 =8.421, P<0.05). 44.2% (13/30) of the male patients needed THA, and 55.6% (7/12) of the female patients needed THA. There was no significant difference between the sexes (x 2 = 0.773, P>0.05). There was no significant difference in the conversion to THA due to different etiology (hormone, alcohol, idiopathic) (x 2 =1.397, P>0.05), unilateral or bilateral (x 2 =0.120, P>0.05), preoperative Harris score ≥ 80 (x 2 =0.285, P>0.05).

Postoperative complications
One patient developed deep infection after tantalum rod implantation at 2 months postoperatively. The result of bacterial culture was Staphylococcus epidermidis infection.
The patients were treated with one-stage tantalum rod removal and antibiotic cement chain removal, and sensitive antibiotics were used after the operation, followed by twostage THA. No postoperative complications such as femoral neck fracture and intertrochanteric fracture were found in other patients.

Histopathological evaluation
When observing the gross specimen of the femoral head cut from the hip replacement, it was found that: cartilage and subchondral bone were separated, subchondral fracture and articular surface collapse; there was no repair phenomenon in the necrotic area, there was no new bone formation and blood vessel growth in the necrotic area, the front end of the tantalum rod in the necrotic area was separated from the dead bone, the articular surface collapsed, cartilage peeled off, and the front end of the tantalum rod was exposed (Figure.  The implant surface is in contact with sparse Island bone. There is limited bone tissue extending inward from the implant. Normal bone cell morphology can be observed in the rod hole of tantalum (Fig.5 f ), which is the same as that of the peripheral bone cell morphology of the porous tantalum rod (Fig.5 a, b).The bone trabecula of the regenerated bone was identified in the support area, contacting the peripheral structure of the porous tantalum rod (Fig.5 c, e). In the living bone area, there was "on-growth" of new bone but no "in-growth". In the necrotic area, there was no obvious new bone regeneration (Fig.5   d), and bone ingrowth only occurred in the implant part of femoral neck.

Discussions
Tantalum rod implantation, as a kind of hip joint preserving operation, has good biocompatibility, elastic modulus similar to fibula, subchondral bone support and core decompression effect, and has achieved good short-term effect in clinic 27 . However, the long-term effect of tantalum rod implantation is uncertain. Tanzer et al. 26 conducted histopathological analysis on 15 cases of tantalum rod removed due to clinical failure, and found that there was no new bone formation and blood vessel growth in the femoral head with large area of necrosis; the diameter of tantalum rod was only 10 mm, and its mechanical support range was limited, which could not provide enough subchondral bone support for the femoral head with large area of necrosis; in the absence of In the area supported by tantalum rods, the collapse of the femoral head will continue to progress.
After that, it has been reported that TA rod implantation failed to transfer to THA in the treatment of femoral head necrosis 28,29 .
The prognosis of osteonecrosis of the femoral head is related to age, gender, etiology and preoperative stage 30 . The results of this study show that the prognosis of tantalum rod implantation is related to the age of patients, bone marrow edema and ARCO stage before operation. The operative effect of hip joint preservation in osteonecrosis of femoral head is related to clinical stage. When the femoral head collapses before operation, the prognosis is poor 31 . Veillette et al. 16 followed up 58 patients with tantalum rod 13 implantation and found that in 49 Steinberg stage II, 6 (12%) needed THA, and in 8 Steinberg stage III, 3 (38%) needed THA. Patients in stage III were more likely to fail than those in stage II. Zhao et al. 32 also pointed out that the survival rate of patients with ARCO IV stage (63.6%) was significantly lower than that of patients with ARCO II stage (95%) and ARCO III stage (92%) (P < 0.05). The results of this study showed that the survival rate of patients with ARCO stage II was lower than that of patients with stage I, and the failure rate was higher, and the ARCO stage was a risk factor for conversion to THA. Florkemeier et al. 33 followed up 19 patients (23 hips) who underwent tantalum rod implantation combined with core decompression for an average of 1.45 years. Although the patients included in this study were patients with early necrosis of the femoral head in ARCO I stage and II stage, the overall survival rate was only 44%. It was pointed out that tantalum rod combined with core decompression did not show obvious advantages compared with core decompression alone. The advantages of early weight-bearing in this operation may be of clinical significance. As a result, they did not recommend tantalum rod implantation for osteonecrosis of the femoral head. A meta-analysis also concluded that the clinical effect of hip joint preservation surgery on patients with collapsed hip joint was poor 34 .
Osteonecrosis of the femoral head is a progressive disease that easily affects young patients with age about 35 years 35 .The young patients have a good prognosis after hip conserving surgery. The follow-up results of Nadeau et al. 28 showed that the average age of patients who failed to switch to THA for tantalum rod implantation was (50 ± 12) years old, and the average age of patients who did not fail was (37 ± 12) years old, with significant difference (P < 0.05). Age is one of the prognostic factors of tantalum rod implantation. Tsao et al. 15 also pointed out that the older patients were prone to failure after tantalum rod implantation. Liu et al. 29 followed up 44 patients (57 hips) with modified tantalum rod implantation for an average of 44.8 months, and failed 11 hips (19.3%). The preoperative age did not affect the survival rate of the patients, and had nothing to do with the prognosis of the operation. The follow-up results of this study showed that patients with age > 40 years old had higher failure rate and lower survival rate, and age was a risk factor for THA, suggesting that patients' age was one of the factors affecting the effect of tantalum rod implantation.
Previous studies have shown that bone marrow edema is a poor prognostic signal because it occurs after the onset or deterioration of hip pain and is associated with subsequent collapse of the femoral head, which may indicate progression to ONFH 24,25,36 . Iida et al. 37 reported that bone marrow edema did not exist in the initial MR imaging of osteonecrosis, and concluded that bone marrow edema should be considered as a sign of possible progression to advanced osteonecrosis. According to Ito et al. 25 , the final imaging stage of 28 hip patients with bone marrow edema was significantly more advanced than that without bone marrow edema. Bone marrow edema is closely related to necrosis volume and is the most important risk factor for the aggravation of hip joint pain.
Consistent with these studies, this study determined that bone marrow edema is the most important independent prognostic factor associated with THA demand. Although the number of bone marrow edema in the hip is limited in the present study, our results show that the survival rate of the hip with bone marrow edema is significantly lower than that without bone marrow edema. The clinical effect of tantalum rod implantation in the treatment of avascular necrosis of the femoral head is not good, the growth of tantalum rod in the femoral head is not good, and the osteogenic activity is limited. Written informed consent was obtained from all participants.

Consent for publication
Not applicable.

Availability of data and materials
We do not wish to share our data, because some of patient's data regarding individual privacy, and according to the policy of our hospital, the data could not be shared to others without permission. management, and analysis of the study. Ye Zhang, Jian-Ning Sun, Zi-Jian Hua, Shuo Feng, Xiang-Yang Chen were involved in the study design, and data analysis. All authors read and approved the final manuscript.