Identifying Complete Atypical Femur Fractures in Adults with Bisphosphonate Exposure

COMMENTARY

Background and Context

Atypical femur fractures (AFF) are uncommon adverse events that affect patients receiving potent antiresorptive osteoporosis therapy, including bisphosphonate drugs and denosumab.^1,2 Among bisphosphonate-treated women, the risk of AFF increases with bisphosphonate duration^3–5 and varies by race and ethnicity, with higher risk among Asian compared to White women.^3,6 The earliest bisphosphonate-associated cases reported in 2005 included three women with femoral shaft fractures, where histomorphometry showed severely suppressed bone turnover attributed to long-term alendronate therapy.⁷ A 2007 Singapore report described nine alendronate-treated women with subtrochanteric fracture (eight transverse or short oblique), including some with lateral cortex hypertrophy, prodromal pain, and atypical contralateral femur findings.⁸ In 2008, Lenart et al.⁹ described “a unique radiographic pattern” of transverse or oblique femoral shaft fractures with cortical beak and cortical thickening in women with long-term bisphosphonate use. This “beak” pattern at fracture origin due to focal cortical hypertrophy is an important finding^4,9,10 that may be subtle and not seen in all image views. Recognition of AFF has implications for osteoporosis treatment and is essential for ensuring appropriate care in ethnically diverse populations.

Historically, AFF identification has been challenged by evolving consensus definitions informed by clinical experience. A 2010 American Society for Bone and Mineral Research (ASBMR) Task Force defined the major features of AFF as a noncomminuted transverse or short oblique diaphyseal femur fracture occurring with minimal to no trauma.¹¹ Fractures were classified as complete when involving both cortices and incomplete when involving only the lateral cortex.¹¹ However, localized hypertrophy of the lateral cortex, resulting in the beak pattern of AFF and depicted in the ASBMR report, was only a minor feature.¹¹ An updated 2013 ASBMR Task Force report listed five major features of AFF, notably (1) a substantially transverse fracture originating in the lateral cortex, (2) localized periosteal or endosteal hypertrophy of the lateral cortex, (3) incomplete fractures could be AFF if they originated in the lateral cortex (complete fractures traverse both cortices), (4) minor comminution was permitted, and (5) fracture occurred with minimal or no trauma (a non-radiographic feature).¹ The 2013 ASBMR definition of AFF required only “four of five major features”,¹ resulting in some case variability depending on which four criteria were used.¹² LeBlanc and colleagues¹² found that discrepancy in AFF adjudication was primarily due to whether periosteal cortical findings (a key feature according to expert groups)^4,5 was included. Those using “any four of five” major features to classify AFF might not require focal periosteal (or endosteal) hypertrophy of the lateral cortex.

Currently, AFF recognition in clinical practice is variable, a concern for optimizing skeletal health. Diagnoses of subtrochanter (ST) and femoral shaft (FS) fractures identify potential AFF, but review of radiographs is necessary to confirm diaphyseal location^13,14 and to differentiate “typical” versus “atypical” fractures. Clinical factors, such as prodromal pain and lack of trauma help discern potential AFF, but diagnostic codes for AFF (ICD-10 M84.75x since 2016) remain underutilized. To support greater recognition of AFF in hospital and primary care settings, this report from a team with 15 years’ experience^5,6,13 aims to guide AFF identification using the major features endorsed by the 2013 ASBMR Task Force.¹ Drawing from epidemiologic research and clinical observations, this commentary addresses: (1) identification of diaphyseal femur fractures, (2) classification of AFF among diaphyseal femur fractures, and (3) implications for clinical practice and surveillance.

Identifying Diaphyseal Femur Fractures

Radiograph review of all hospitalized femur fractures (>90% proximal femur/hip and <10% diaphyseal femur among older women, excluding distal femur fractures) was an early labor-intensive approach to AFF ascertainment⁴ that established orthopedic expertise and recognition of defining AFF features. However, since most diagnoses of proximal femur (femoral neck or pertrochanter) fractures correctly localize to the proximal femur,¹³ efforts now center on review of diaphyseal femur fractures (Figure 1), which occur in the ST (within 5 cm below the lower border of the lesser trochanter) and FS (below the ST and up to but not including the distal metaphyseal flare).^6,13 In studies of adults with bisphosphonate exposure, ST- and FS-diagnosed fractures are reviewed to confirm diaphyseal femur location (non-diaphyseal fractures are excluded),^13,14 followed by AFF adjudication.^3,5 Included in this AFF ascertainment framework^3,5,14 are diagnoses of “unspecified” femur fracture (a subset are diaphyseal),¹⁴ pathologic fracture, and femur stress fracture, the latter used by some to identify incomplete AFF prior to availability of specific codes for AFF (Figure 1). Whether diaphyseal fractures might be mis-coded as distal femur fractures is uncertain, but AFFs do not generally occur below the distal metaphyseal flare. Once diaphyseal femur fracture is confirmed in a patient with potent antiresorptive drug exposure, the fracture pattern should be examined for atypical features.

Figure 1.

A Clinical Approach to Diaphyseal Femur Fracture and Atypical Femur Fracture Ascertainment

A Working Case Definition for (Complete) AFF

Because most experts agree that a primarily transverse fracture with localized periosteal (or endosteal) hypertrophy of the lateral cortex at fracture origin comprise key features of AFF,^4,5,10,12 we recommend that all four major radiographic features of AFF in the 2013 ASBMR report be used to identify complete AFF (Figure 1) when there is complete access to fracture images. At times, the “beak” may be subtle or poorly visualized due to positioning/views or when cortical hypertrophy is endosteal. Hence, in addition to AFF versus non-AFF classification, a third category of “equivocal AFF” should be used for fractures that meet most but not all AFF radiographic criteria. As more knowledge is gained regarding the spectrum of AFF, it is likely that a nontraumatic, noncomminuted, transverse fracture in a patient on long-term bisphosphonate therapy, especially with prodromal pain, would be considered an AFF (and managed as such). Simple transverse diaphyseal femur fractures (AO Foundation/Orthopedic Trauma Association fracture classification 32A3; www.aofoundation.org)¹⁵ outside of potent antiresorptive drug-associated AFFs uncommonly present during orthopedic trauma and more likely result from stress fracture. Other “equivocal” cases include pathologic and periprosthetic fractures with AFF features. Once AFF/equivocal AFF is identified radiographically, records should be examined to determine whether the fracture occurred spontaneously (e.g., minimal lateral impact or twisting) or with minor trauma (e.g., fall from standing height). While the ASBMR Task Force requires absence of major trauma, AFFs typically do not result from generalized bone fragility and low bone density does not appear to be a risk factor for AFF.³

Fracture radiographs in emergency settings are usually sufficient for case adjudication when multiple views are provided. However, emergent positioning and oblique or lateral views may limit visualization of cortical beaking. In such cases, fluoroscopic images during operative repair can show periosteal beaking (or endosteal hypertrophy) more clearly. Cortical findings can also be seen when the femur is re-aligned after repair, but only immediate post-operative images since callous formation occurs with fracture healing. Some patients have femur radiographs prior to complete AFF (for evaluation of prodromal pain or other indications); these earlier images can document focal lateral cortex findings at the site of future AFF. Indeed, the sequence of focal cortical hypertrophy or femur stress fracture (lateral cortical reaction, often with focal scintigraphy uptake) followed by development of a horizontal translucent line in the lateral cortex (the “dreaded black line” of incomplete AFF)¹⁶ may be evident.^1,2 A spectrum of such findings can present in the contralateral femur (up to one-third of cases before/after the index AFF),⁵ which should be imaged to assess contralateral AFF risk.¹

Implications for Clinical Care

Diaphyseal femur fractures are uncommon. Hence, when patients receiving potent antiresorptive therapy experience diaphyseal fracture, fracture morphology should be examined to exclude AFF during fracture evaluation or follow-up, especially when osteoporosis treatment is being considered. Identified AFF cases should be coded (ICD-10 M84.75x). It is crucial that fracture liaison services, endocrinologists, and physicians involved in osteoporosis care understand AFF identification (and the need for bilateral femur imaging in patients with AFF), so that these patients are not inadvertently restarted on potent antiresorptive therapy for secondary fracture prevention. Although beyond the scope of this report, clinicians should also be aware of AFF risk factors.^1,2 US Asian women have half the risk of hip fracture compared to White women, but their AFF risk with bisphosphonate therapy, albeit lower than hip fracture, is five- to six-fold higher than White women.^3,6 Femur fracture in an Asian woman with bisphosphonate exposure should prompt image review to exclude AFF.

Unlike typical diaphyseal femur fractures, post-surgical management of AFFs can present challenges, due to greater risk of delayed healing and fracture non-union. Teriparatide has been used to expedite AFF healing, but large clinical trials are lacking. Hence there is an important role for AFF prevention, including bisphosphonate treatment holidays to limit the years of uninterrupted therapy, particularly since treatment benefit beyond the first 3-5 years is less clear. Physicians prescribing bisphosphonates or denosumab for osteoporosis should survey patients about thigh/hip or groin pain during long-term therapy, which if present should prompt imaging to identify early signs of (incomplete) AFF^1,2 so that measures to prevent devastating complete fracture can be implemented. In the meantime, femur surveillance using “single-energy” x-ray absorptiometry during follow-up bone density testing (dual energy x-ray absorptiometry) has been proposed¹⁷ and is an option for monitoring higher risk patients.

In summary, AFF recognition is essential among bisphosphonate- (and denosumab)-treated populations with diaphyseal femur fracture. Although these fractures are uncommon, timely recognition can improve outcomes in affected patients, for both femurs.