Does Bisphosphonate Delay Bone Healing After Proximal Femoral Nail Anti-Rotation

Bisphosphonates are widely used for osteoporotic patients to decrease the rate of osteoporotic fractures and they have been shown to reduce the mortality rate in clinical trials. A yearly intravenous zoledronic acid in a clinical study (The HORIZON-RFT) signicantly reduced any new clinical fracture and also secondary prevention of hip fracture while bisphosphonates are known to delay remodeling of bone raising the risk that they may delay fracture healing. However, current studies lack data demonstrating whether bisphosphonate (BPs) delays bone healing after hip fracture treatment in clinical practice. Purpose of study was to determine whether treating elderly patients with Bisphosphonate (BP’s) after proximal femoral nail xation (PFNA) for intertrochanteric fractures delays fracture healing compared to similar patients not treated with BP’s. A secondary goal was to compare the functional outcome, complications and mortality between the two treatment groups.

Osteoporosis represents a major problem in public health because it is associated with fractures from low-energy mechanisms. Hip fracture has been recognized as the most serious complication of osteoporosis because of its consequences including disability, poor quality of life, increased risk of mortality, and health care costs [7,9,15,17]. The incidence of hip fracture has been increasing around the world as a result of the aging population from 1.66 million in 1990 to 6.26 million by 2050 [8].
Bisphosphonates are widely used for osteoporotic patients to decrease the rate of osteoporotic fractures and they have been shown to reduce the mortality rate in clinical trials [4,11]. Animal studies demonstrated that a delayed single dose of zoledronic acid (1 or 2 weeks after fracture) produced signi cantly increasing bone strength and fracture repair [1] while a yearly intravenous zoledronic acid in a clinical study (The HORIZON-RFT) signi cantly reduced any new clinical fracture and also secondary prevention of hip fracture [6]. BP's are known to delay remodeling of bone raising the risk that they may delay fracture healing. However, current studies lack data demonstrating whether bisphosphonate (BPs) delays bone healing after hip fracture treatment in clinical practice. supplemental vitamin D and calcium. The primary outcome was measured prospectively by time to clinical union and radiographic union (weeks). The secondary outcome was the functional outcome (Harris Hip Score) and complications including mechanical failure and mortality rate. All outcomes were compared between the two groups.

Data collection and outcome
A total of 196 elderly patients with low energy intertrochanteric fracture underwent PFNA xation and were included in the study. Patient demographic information, comorbidities, pre-operative and postoperative status, history of prior fracture and treatment, cause of injury, medication at hospital admission and discharge, and radiological reports were also obtained from the medical record review. Patient's comorbidities were reviewed from the medical record: (1) [11]. METs were strati ed as follows: (1) 1-4 (low intensity), and (2) >4 (moderate to high intensity).
All patients received standard medications of calcium and vitamin D supplements. In this study, patients who received any bisphosphonate was classi ed as the "treated group" and those who did not receive anti-osteoporosis drugs were classi ed as the "untreated group". The enrollment of subjects and their allocation of treatment including the outcome was shown in CONSORT diagram.
All patients were prospectively identi ed for clinical union and radiological union at 2, 4, and 6 weeks, 3, Operative procedure Fractures were all xed with a titanium PFNA TM nail (Synthes). All patients were operated on the fracture table in supine position. Closed reduction was done under uoroscopy. After anatomical reduction, a guide wire was inserted into the tip of greater trochanter, proximal reaming was done, diameter of nail was measured under uoroscopy, and a standard-proximal femoral nail with 200 millimeters length was placed into the medullary canal and the guide wire was removed. Before the application of the helical blade into the femoral head, a guide wire was inserted into the femoral head and the exact position and length of helical blade in AP and lateral views was measured. The helical blade was inserted into the femoral head and it was tightened in the nal step.

Postoperative management
Appropriate pain control was provided for all patients, they were allowed weight bearing as tolerated, and deep vein thrombosis prophylaxis was applied during hospital admission.

Outcome measurement
All patients were followed up in clinic at 2 weeks, 4 weeks, 6 weeks, 3 months, 6 months, 9 months, and 1 year. Radiographic measurements were done by two orthopedic training surgeons who did not participate in the operative procedures. Mean of these measurements were calculated. Anteroposterior (AP) and lateral radiographs were assessed by PACS software and were used for assessment of quality of

Statistical analysis
All patient's information was compared between the treated and untreated groups to identify any differences. Data were summarized using descriptive statistics (mean ± SD and number of patients).
Comorbidities were compared using Fisher's exact test. Age of patient, Distance between proximal and distal fragment (gap and step in AP and lateral views) (mm), Tip and Apex Distance (TAD) (AP and lateral views, mm), and Neck Shaft Angle (degree) between groups were compared by independent T-test. ASA class, METs, and Fracture pattern (AO/OTA classi cation 31 A1, 2, and 3) were compared by Chi-Square test while the others were compared by Mann-Whitney U test. Pvalue less than 0.05 were considered a signi cant difference.

Results
Demographic data and comorbidities including Charlson Comorbidity Index (CCI) are reported in Table 1.
There was no difference in the average age between groups, but female gender in the treated group was signi cantly higher than in the untreated group (85.3% and 66.9%, p=0.004, respectively). For ASA classi cation, METs, any underlying diseases, and CCI there were no signi cant difference between groups.
Fracture pattern assessed by modi ed AO/OTA 2017 are demonstrated in Table 2. There were no signi cant differences in fracture pattern (Stable and Unstable type) between groups. There was comparable fracture displacement (gap in lateral view and step in both AP and lateral views) between both groups although the gap in AP view in treated group was almost signi cantly higher than untreated group (1.42±2.17 and 0.89±1.85, p=0.079). There was no signi cant difference in the neck shaft angle (NSA), NSA ratio, and Tip Apex Distance (TAD) following PFNA xation between groups as shown in Table3. The average TAD was 20.9 mm in the treated group and 21.0 mm in untreated group which fall within the recommended range (20-30 mm). There was no signi cant correlation between TAD and mechanical failure.
The average time to clinical union and radiographic union (weeks) were similar in both groups. Even though pre-fracture HHS and HHS at one year follow up were not signi cantly different between groups, in most patients in both groups the HHS was lower when compared with pre-fracture HHS (67.7% in treated group and 74.0% in untreated group). Mechanical failure after PFNA xation was not signi cantly different between groups including mortality rate in both groups (4.0% vs 6.6%, p=0.537) ( Table 4).

Discussion
Bisphosphonates (BP's) are widely used for osteoporotic patients and they have been shown to decrease the fracture rate and reduce mortality in clinical trials [4,11]. However, there is concern that BP's delay bone healing and this may be relevant in clinical practice after hip fracture repair a scenario in which BP's are frequently utilized. The purpose of this study was to assess whether elderly patients treated with BPs after proximal femoral nails (PFNA) for intertrochanteric fractures have delayed fracture repair compared to similar patients not treated with BP's and to compare the functional outcomes between the two groups.
The characteristics of the fracture and the treatment were similar between the two groups of patients.
The fracture pattern assessed by modi ed AO/OTA 2017 was not signi cantly different (Stable and Unstable type) between groups. The fracture displacement (gap in lateral view and step in both AP and lateral views) between both groups was comparable even though the gap in the AP view in the treated group was almost signi cantly higher than in the untreated group (1.42±2.17 and 0.89±1.85, p=0.079).
There was no signi cant difference in the neck shaft angle (NSA), NSA ratio, and Tip Apex Distance (TAD) following PFNA xation in both groups as shown in Table 3. The average TAD in our study was 20.9 mm in the treated group and 21.0 mm in the untreated group which was well within an acceptable range (20-30 mm) as de ned in previous studies [2,3,14,16]. had lower functional outcome than pre-injury and walking ability was comparable with previous studies demonstrating a high rate of dependence (80% were using a walking aids at 1 year follow up, 16% were institutionalized, and 11% were bedridden postoperatively) [5,16,18]. Mechanical failure after PFNA xation was not signi cantly different between treated and untreated groups further supporting the contention that BP's do not inhibit bone healing after hip fracture xation to an extent that is detectable in actual clinical practice. The mortality rate in this study was not signi cantly different (4.0% vs 6.6%, p=0.537) even though several studies have been demonstrated reduction of mortality rate after BP's use [4,11,13]. Even though we prospectively measured the time to clinical and radiographic union because these are the primary outcome of this study, some parameters were measured retrospectively, for instance, the fracture pattern assessed by modi ed AO/OTA 2017 classi cation that established later after this study had been started. Another limitation was time to bone union evaluated by clinical and plain radiography. These are semiqualitative assessment for complete bone healing, the better tool should be a Computerized Tomography. In addition, the rate of mechanical failure after PFNA xation was so small. Therefore, there was no actual power to detect a signi cant difference of mechanical failure between groups.

Conclusion
The data in this study suggests that BP's can be used after surgical treatment of an osteoporotic hip fracture without concern that fracture healing will be delayed.