Atypical femoral diaphyseal and subtrochanteric fractures and their association with bisphosphonates

Abstract

Antiresorptive bisphosphonate agents are the mainstay of treatment for osteoporosis in both men and postmenopausal women. However, recent studies have raised concerns about the oversuppression of bone turnover related to the long-term use of bisphosphonates. Cases of atypical femoral diaphyseal and subtrochanteric fracture were reported recently in patients on long-term alendronate, and oversuppression of bone turnover was postulated to be the cause. We retrospectively reviewed all patients with femoral diaphyseal and subtrochanteric fracture presented between July 2003 and June 2008, and identified 10 patients who reported prior bisphosphonate use. Bone formation markers of all these patients were in the low range. Although the incidence of bisphosphonate-related atypical fracture accounts for an extremely low percentage of the total number of femoral diaphyseal and subtrochanteric fractures, we observed a steady increase from 0% in 2003 to 2004 to 25% in 2007 to 2008.

BACKGROUND

Antiresorptive bisphosphonate agents are the mainstay of treatment for osteoporosis in both men and postmenopausal women. Alendronate was the first bisphosphonate shown to be effective in reducing osteoporotic fractures in randomised placebo-controlled trials, and has been used extensively for the prevention and treatment of osteoporosis in the past decade. In the past year, a number of cases with femoral insufficiency fractures associated with the long-term use of alendronate have been reported.^1–6 Among the cases with bone markers measured, suppressed biochemical markers of bone formation and resorption were observed. Bone biopsy in some patients showed features of “adynamic bone disease”. This raised the possibility of bisphosphonate-induced oversuppression of bone turnover resulting in “frozen bone”. In view of the sketchy and incomplete information on these cases, the incidence and pathogenesis of these fractures are unclear.

CASE PRESENTATION

Femoral insufficiency fractures are unusual fractures that are associated with high-impact trauma, and their incidence is unclear. Based on the Clinical Data Analysis Reporting System (CDAS) of the Hospital Authority of the Hong Kong Government, we evaluated all cases of subtrochanteric fracture and fracture of diaphysis of the femur presented to Queen Mary Hospital between July 2003 and June 2008 and compared them to osteoporotic femoral fractures using the International Classification of Disease, Tenth Revision (ICD-10.0). Only the head count of acute admissions was included in the analysis to avoid duplication and multiple entries due to hospital transfer and repeated admissions. Only fractures sustained in low-trauma injury were included. Fractures associated with motor vehicle accident, a fall from height, malignancy, or infiltrative bone diseases were excluded.

Among the 88 cases of subtrochanteric and 66 cases of diaphyseal femoral fractures identified from 1 July 2003 to 30 June 2008, 10 patients had received prior bisphosphonate therapy. Clinical information, markers of bone turnover (formation marker bone specific alkaline phosphatase and resorption marker urinary N-telopeptide (NTx)), parathyroid hormone, and serum 25(OH) vitamin D levels of these patients were reviewed (tables 1 and 2). Approval was obtained from the hospital board for retrospective analysis of patient records.

Table 1.

Clinical information for patients with diaphyseal and subtrochanteric fracture of the femur who received bisphosphonate therapy

Case	Sex/age (years)	Indication for BP	Duration of BP	Mechanism of injury	Site of fracture	Fracture configuration	Beaking of lateral cortex at fracture site	Cortical thickening (cortical thickness/femoral diaphysis diameter), cm	Healing time (weeks)	Prodominal symptoms	Comorbidities and medications
1	F/73	Partial vertebral fracture T12, L4	Aln × 0.5 years	Slipped and fell	Right middle 1/3 of femoral diaphysis fracture	Transverse	+	+ (0.9/2.8)	16	Nil	Hypertension, hyperlipidaemia, thalassaemia
2	F/92	S/F: fracture neck of femur	Aln × 3 months	Slipped and fell	Left subtrochanteric fracture	Oblique	–	− (0.5/3.0)	18	Nil	Hypertension, dementia, COPD on prednisone
3	F/87	Vertebral fracture T12–L4, fracture of right ulna, distal radius, and left clavicle	Aln × 1.5 years	Slipped and fell	Left subtrochanteric fracture	Transverse	–	− (0.4/3.2)	14	Nil	Hypertension, dementia, neurosis
4	M/55	Prevention of steroid induced osteoporosis	Aln × 2 years	Slipped and fell	Right middle 1/3 of femoral diaphysis	Short oblique (<30°)	–	− (0.9/3.2)	20	Progressive pain for 3–4 months over right thigh	Chronic alcoholic, Graves opthalomopathy on warfarin, prednisone, ciclosporin
5	F/78	S/F: right intertrochanteric fracture	Aln × 5 years	Slipped and fell	Right proximal 1/3 of femoral diaphysis	Transverse	+	+ (0.8/2.9)	14	Nil	Hypertension, diabetes mellitus with diet control
6	F/88	S/F: fracture neck of femur	Aln × 10 years	Spontaneous fracture	Distal 1/3 of femoral diaphysis	Cortical stress fracture	Nil	+ (0.9/2.9)	6	Lower limb weakness for 6 months, pain for 3 months	Nil
7	F/79	Vertebral fracture T10–L5	Aln × 4 years	Slipped and fell	Middle femoral diaphysis	Transverse	+	+ (0.7/3.2)	9	Bilateral thigh pain for 6 months	Nil
8	F/81	Partial vertebral fracture T12	Aln × 5 years	Slipped and fell	Middle 1/3 of femoral diaphysis	Oblique	+	+ (0.9/2.8)	18	Nil	Hypertension, hyperlipidaemia
9	F/80	S/F: left middle femoral diaphysis fracture	Aln × 1 year, Ibn × 2 years	Spontaneous fracture	Middle 1/3 of femoral diaphysis	Transverse	+	+ (0.9/2.9)	14	Nil	Hypertension, hyperlipidaemia, asthma, hysterectomy
10	F/74	Vertebral fracture T10	Aln × 6 years	Slipped and fell	Proximal 1/3 of femoral diaphysis	Transverse	+	+ (0.8/2.9)	10	Nil	Nil

Aln, alendronate; BP, bisphosphonates; Ibn, ibandronate; S/F, slipped and fell.

Table 2.

Biochemical data and bone mineral density (BMD)

		Patient										Reference range
		1	2	3	4	5	6	7	8	9	10
Serum	Total calcium (mg/dl)	2.42	2.31	2.44	2.27	–	2.53	2.3	–	2.21	1.92	2.11–2.55
	Adjusted calcium (mmol/litre)	2.57	2.49	2.48	2.31	–	2.44	2.38	–	2.23	2.16	2.11–2.55
	Albumin (g/litre)	36	33	40	40	30	41	38	38	41	43	38–48
	Creatinine (mg/dl)	67	73	63	63	42	72	57	96	131	128	49–82
	Parathyroid hormone (pg/dl)	38	–	–	23	–	5.1	46	–	–	61	11–54
	Phosphate (mmol/litre)	1.11	1.47	0.95	1.09	–	0.96	1.06	–	1.16	1.26	0.88–1.45
	25-OH-vitamin D (nmol/ml)	–	–	–	69.36	–	60.3	52.38	–	42.8	46.27	8–55
	Total ALP (U/litre)	31	76	72	111	81	58	68	41	58	42	49–138
	BSAP (U/litre)	–	–	–	21.7	–	14.4	42.2	–	28.2	20.2	14.2–42.7
Urine	NTx (nmol BCE/mmol Cr)	–	–	–	24	–	* 956	87	–	61	28	14–74
BMD (pretreatment)	L1–4	−1.3	–	–	−2.2	–	−0.7	−3.5	−2.6	–	−1.2	>−2.5
	Femoral neck	−1.9	–	–	–	–	−2.5	−2.3	–	–	−1.6	>−2.5

*Serum C-telopeptide: reference range for premenopausal women <250–4500 nmol/litre.

ALP, alkaline phosphatase; BCE, bone collagen equivalent; BSAP, bone-specific alkaline phosphatase; Cr, creatinine; NTx, N-telopeptide.

OUTCOME

Subtrochanteric and femoral diaphyseal fractures contributed to 3.86% and 2.86% respectively of all osteoporotic fractures. Among the 88 and 66 cases of subtrochanteric and femoral diaphyseal fractures presented in the study period, 10 reported prior use of bisphosphonates. The incidence of subtrochanteric and femoral diaphyseal fractures associated with the use of bisphosphonates increased from 0 out of 22 total cases (0%) in 2003/4, to 2/33 (6%) in 2004/5, 3/23 (8.6%) in 2006/7, and 5/20 (25%) in 2007/8. All 10 cases were associated with the use of alendronate that ranged from 3 months to 10 years; 1 received alendronate in the first year followed by ibandronate for 2 years (table 1).

Six patients showed femoral diaphyseal cortical thickening and lateral cortex beaking at the fracture site on x ray. Only one patient agreed to a tetracycline-labelling bone biopsy, which revealed the absence of tetracycline label and markedly decreased osteoblasts and osteoclasts, suggestive of severe suppression of bone turnover. Three patients (patients 6, 8 and 9) reported prodromal symptoms of pain and weakness 3 to 6 months prior to fracture, while two patients had spontaneous fracture without trauma. Almost all had subnormal or low normal alkaline phosphatase (ALP) levels, whereas the bone resorption markers (NTx) were within the normal range. None of the patients had delayed bone healing.

DISCUSSION

This retrospective audit documented an increasing trend of prior use of bisphosphonates among patients presenting with subtrochanteric and femoral diaphyseal fractures from 0% in 2003/4 to 25% in 2007/8, despite a steady number of subtrochanteric and femoral diaphyseal fractures in this period. The percentage of subtrochanteric fractures among femoral fractures in our population is similar to the 4.3% reported in Caucasians,⁷ while no data have been reported on the incidence of femoral diaphyseal fracture. The increasing trend of incidence over the years may be associated with the increase use of bisphosphonate. Reports of femoral insufficiency fractures associated with the use of alendronate have only been recorded in the CDAS system since 2004, despite alendronate being available in Hong Kong for 12 years. The association with alendronate may simply reflect the longer availability and wider use of this drug.

Whether bisphosphonate use increases subtrochanteric or diaphyseal femoral fractures is unclear. Based on a national registry, Abrahamsen et al⁸ reported an increased risk of subtrochanteric fracture with alendronate, but this was attributed to an increased risk of hip fracture. The risk of diaphyseal fracture was not statistically significant. Long-term use of bisphosphonate carries a potential risk of oversuppression of bone turnover and may impair the biomechanical properties of bone.⁹ Bone biopsies from dogs that have received high doses of alendronate have shown severe suppression of bone turnover and the accumulation of microdamage.¹ While there is no reduction in bone strength, a twofold to sevenfold increase in the accumulation of microdamage reduces bone toughness by 20%, hence lowering the ability for the bone to sustain deformation without breaking.^10,11 Also, continuation of secondary mineralisation while bone turnover was oversuppressed by bisphosphonate may result in hypermineralised bone that is more brittle and less tough.^12,13

Apart from the absence of double-tetracycline labelling and the minimal presence or absence of osteoblasts and osteoclasts, other features of severe bone turnover suppression, including marked reduced bone volume as well as severely reduced bone formation and resorption surfaces of cancellous, endocortical and intracortical bone, were seen in human quantitative bone histomorphometric studies.¹ Biochemical bone formation and resorption markers were not as profoundly suppressed as that reflected by the histomorphometric features in bone biopsy, as the latter represent local changes at the involved skeletal site. According to Wasserman et al,¹⁴ the subtrochanteric or diaphyseal cortical bone is most susceptible to the inhibition of bone remodelling and microdamage accumulation. The microcracks tend to colocalise within highly mineralised regions of cortical bone tissue on physiological loads. This leads to stress fracture in these regions, which eventually propagate to clinical fracture with minimal trauma. Although bisphosphonates may potentially suppress bone turnover, there are other secondary causes and patient-related factors (eg, two patients were on glucocorticoid). Other causes include the use of other antiresorptive agents, hypoparathyroidism, and hypothyroidism.

There were a few limitations in this study. As this is a retrospective audit, we could only report the biochemical bone markers closest to the time of fracture in a subset of patients on bisphosphonates, and we did not have information from subjects who did not receive bisphophonates. In addition, without the total number of patients and duration of treatment, we were unable to determine the actual incidence of atypical femoral diaphyseal and subtrochanteric fractures associated with bisphosphonate agents. However, our data are unlikely to be biased, as the government’s CDAS of the Hospital Authority has been collecting clinical data on 94% of the Hong Kong population for more than a decade, and Queen Mary Hospital is one of the biggest emergency hospitals in Hong Kong, serving over 1 million residents in the Hong Kong southwest region.

LEARNING POINTS

• Atypical femoral fractures may be associated with long-term use of bisphosphonates.

• Low bone markers of bone turnover may signify the onset of oversuppression of bone formation in bisphosphonate users.