Sclerotic lines around the radial tuberosity of the proximal radius in patients with atypical femoral fractures associated with long-term bisphosphonate use: a preliminary report
Article information
Abstract
Purpose
Sclerotic lines have been detected at specific locations on the forearm and metaphysis of the lower extremities (femur and tibia) in some patients with atypical femoral fractures (AFFs) associated with long-term bisphosphonate (BP) use. Herein, we present our preliminary data relating to the factors associated with the occurrence of sclerotic lines in patients diagnosed with AFFs.
Methods
We inspected the clinical charts of patients who were prescribed BPs at our institution between 2010 and 2020. Fifty-six patients were finally investigated, including patients with sclerotic line lesions at the radial tuberosity (21 patients), and those without any pathognomonic lesions on either radius on a simple radiograph (35 patients).
Results
No significant between-group differences in basic demographic characteristics, bone mineral density, or the total period of BP administration were observed (p>0.05). Multivariate analyses of the variables that exhibited significant between-group differences in the univariate analyses demonstrated that the time since the last BP administration at the time of AFF diagnosis (odds ratio [OR], 0.441) and the concurrent presence of similar lines on the femur (OR, 36.00) were significantly associated with the presence of sclerotic lines on the proximal radius. The cutoff time after the last BP administration at the time of AFF diagnosis was 16.5 months, which may serve as a predictor for the development of a sclerotic line around the radial tuberosity.
Conclusion
Sclerotic lines on the proximal radius were detected more frequently in patients who had been recently treated with a BP or had similar lines on the lower extremities.
Introduction
Bisphosphonates (BPs) have become the principal treatment for managing osteoporosis during the last two decades. BPs increase bone mineral density (BMD) and prevent fragility fractures [1-3]. However, atypical fractures, particularly of the femur, have emerged as an important concern, suggesting that alternative agents are needed to manage osteoporosis. Long-term use of BPs carries a potential risk for severe suppression of bone turnover, which paradoxically weakens the ability of bone to undergo remodeling. This eventually leads to an accumulation of microdamage, reduced bone toughness, and increased brittleness [4]. Lesions associated with atypical femoral fractures (AFFs) are usually in the subtrochanteric area, and most fractures occur after various prodromal symptoms, such as pain or discomfort around the hip.
Very few BP-associated ulnar fractures were reported in the literature until 2010 [5-13]; however, studies about the treatment and risk factors have been published recently [14-17]. Due to knowledge about this specific fracture, neglect or misdiagnosis of atypical forearm fractures associated with long-term usage of BPs is expected to decrease gradually. Screening inspections of the entire skeleton of patients under long-term BP use include scintigraphy or computed tomography/magnetic resonance imaging, and simple radiography, and these have revealed new information about how BPs affect the entire body. Since 2010, very early lesions in patients under long-term BP use have been detected on simple radiographs by radiologists, and these lesions are now regarded as prodromal lesions that occur before incomplete AFF [18,19].
Interestingly, other specific sclerotic lines (zebra line or pamidronate line) are well known in the pediatric population treated with regular BPs administered for osteogenesis imperfecta, fibrous dysplasia of bone, steroid-induced osteoporosis, or Gaucher disease [20-23]. These lines have been regarded as non-pathological and are a physiological phenomenon associated with a functioning physeal plate.
We also observed the presence of sclerotic lines in the metaphysis of the forearm and lower extremities in patients with AFF following prolonged BP use. In particular, we hypothesized that the sclerotic line identified in the radial tuberosity of AFF patients could be associated with long-term BP therapy. So, we investigate a retrospective case-control study to identify relevant factors.
Methods
Ethics statement: We conducted this study in compliance with the principles of the Declaration of Helsinki. The protocol of this study was reviewed and approved by the Institutional Review Board of Chungnam National University Hospital (No. 2019-08-023). Written informed consent was obtained from all participants.
1. Patient selection
We retrospectively inspected the clinical charts of patients who were prescribed a BP at Chungnam National University Hospital between March 2010 and February 2020. Among the 706 primarily screened patients, 56 were finally investigated. The patients were selected based on the following criteria: (1) BP treatment history or ongoing administration for at least 1 year, (2) availability of simple radiographs of bilateral femurs and forearms, (3) at least one AFF on either femur, regardless of whether there was an incomplete or complete fracture, and (4) availability of complete medical records and radiological data.
The following exclusion criteria were applied: (1) no evidence of AFF, (2) variability in the agents administered, such as selective estrogen receptor modulators, receptor activator of nuclear factor kappa-B ligand antibodies, or anabolic parathyroid hormonal peptides, (3) evidence of metabolic disease in the skeleton (abnormality in the thyroid, parathyroid, or kidneys), (4) definitive traumatic forearm fracture, (5) typical osteoporotic fracture of the forearm with a low BMD as estimated by dual-energy X-ray absorptiometry (DEXA), (6) <75% medication possession ratio for adherence to oral BPs [24], (7) management of primary or metastatic malignancy using BPs as chemotherapy, such as breast cancer or multiple myeloma, and (8) pathological fracture or impending fracture due to any metastasis of a malignant tumor.
2. Demographic and clinical characteristics
We evaluated basic demographic factors, including age, sex, history of smoking, BMD, and comorbidity of diabetes mellitus or any connective tissue disease requiring steroid treatment at the time when the patient was diagnosed with AFF. Clinical variables included the total time under this treatment, and whether the BP was prescribed by a single or several physicians throughout the treatment period. BMD was measured at the time of initial administration of the BP and the first AFF event, using DEXA and Lunar Prodigy enCORE software (ver. 8.8; GE Medical Systems, Milwaukee, WI, USA). The lowest T-score of the proximal femur and lumbar spine, except that of the Ward area of the proximal femur, was recorded. We used the mean scores from the hip and spine.
3. Atypical fractures of the ulna or radius
We thoroughly checked the simple radiographs of the forearm for all patients at our institution who were already diagnosed as having an AFF associated with BP use. Concurrently, we investigated all previously treated forearm fractures that had been diagnosed as ordinary traumatic fractures. Among them, neglected fractures not known to be related to BPs at the time were selected retrospectively. We applied the same AFF characteristics to the forearm [18,19].
We followed a previous study to define the definitive atypical forearm fractures as those with a short transverse fracture pattern, a spike, a fracture that was non-comminuted or minimally comminuted, and endosteal or periosteal thickening (beaking or flaring) of one side for the cortex on a simple radiograph of the forearm, resembling an incomplete AFF, accompanied by a history of BP use [16].
4. Radiological investigation of the sclerotic margin and group designations
We checked the status of the proximal radius around the radial tuberosity in the patients with a history of AFF, regardless of whether they had suffered a displaced or complete fracture. Then, we divided the final 56 patients who met the inclusion criteria into two groups: those with any sclerotic line lesion at the radial tuberosity (Fig. 1), regardless of fracture displacement on at least one side of the femur on simple radiographs (group 1, 21 patients), and those without any pathognomonic lesions on either radius on simple radiographs (group 2, 35 patients).
(A) Multiple sclerotic lines were detected around the radial tuberosity and proximal ulnas (arrows) in a 62-year-old woman. An incomplete atypical fracture (endosteal and periosteal thickening) was found at the ulnar diaphysis (circle). The bisphosphonate (BP) had been stopped 8 months previously, with a history of atypical femoral fractures (AFFs) in both femurs. The total time of BP administration was 85 months. (B) A 72-year-old woman had sclerotic lines around the radial tuberosity and proximal ulnas (arrows). An incomplete atypical fracture (endosteal and periosteal thickening) was detected at the radial diaphysis (circle). The BP had been stopped 14 months ago, with a history of an AFF on one femur. The total time of BP administration was 73 months. Lesions of an incomplete radial fracture were detected 8 months after the AFF diagnosis.
5. Duration of bisphosphonate use
We determined the time that all oral or intravenous BP agents were administered in months. The periods between injections, regardless of the duration of the intended BP drug holiday, were not considered in the total time. Additionally, the time gap was evaluated between the final BP administration and the time that the radial tuberosity lines were detected.
6. Statistical analysis
The sample size was calculated based on the ratio of the presence of the specific line in group 1 to that in group 2 (0.6:1), assuming an effective odds ratio (OR) ≥1.2 between the two groups. A minimum of 50 patients was required for a proper statistical comparison between the two groups (logistic regression analysis with a power of 80%) [16]. The Student t-test was used to compare continuous variables, and the chi-square test or Fisher exact test was used to compare categorical variables. We used forward selection logistic regression for the multivariate and univariate analysis between the final administration of the BP after the diagnosis or AFF and similar sclerotic lines in the femur respect to the presence of a sclerotic line in the proximal radius; the likelihood ratio was calculated for the 95% confidence interval (CI). Receiver operating characteristic curves were used to determine the predictive cutoff time after the last BP administration at the time of an AFF diagnosis. The sample size was calculated using G*Power software (ver. 3.1.9.2). Data were analyzed using SPSS software for Windows (ver. 22.0; IBM Corp., Armonk, NY, USA). A p-value <0.05 was considered significant.
Results
BPs were stopped in all patients after the AFF diagnosis, even in those with incomplete fractures. Of the 21 patients in group 1, five had atypical forearm fractures, including incomplete lesions, with no significant difference from the proportion in group 2 (seven out of 35, p=0.748). All forearm fractures, including incomplete fractures, occurred later than the AFFs in both groups. The basic demographic characteristics, such as age, sex, smoking, comorbidities of diabetes mellitus or any connective tissue disease requiring steroid treatment, BMD, and the total period of BP administration were not significantly different between the groups (p>0.05) (Table 1).
Comparison of demographic, clinical, and radiographic data between groups in univariate analysis
Multivariate analyses of the variables that were significantly different between the groups were conducted: the time from final administration of BP after the diagnosis of AFF (OR, 47.749; 95% CI, 2.729–835.474; p=0.008) and the concurrent presence of similar lines in the femur (OR, 0.400, 95% CI; 0.196–0.814; p=0.012). Univariate analyses demonstrated that the time from final administration of the BP after the diagnosis of AFF (OR, 0.441; 95% CI, 0.277–0.701; p<0.001) and the concurrent presence of similar lines in the femur were significantly associated with the presence of a sclerotic line in the proximal radius (OR, 36.00; 95% CI, 7.671–168.943; p<0.001). The receiver operating characteristic curve revealed that a cutoff time from final BP administration after AFF diagnosis of 16.5 months predicted a lesion around the radial tuberosity (area under the curve, 0.954 ((95% CI, 0.907–1.000)), p<0.001) (Fig. 2).
A receiver operating characteristic curve revealing the optimal cutoff value predicting the maximal time from final bisphosphonate treatment.
Discussion
The action of BPs involves inhibition of the mevalonate enzyme pathway in osteoclasts and binding to hydroxyapatite crystals in bone. BPs exert direct effects by decreasing the recruitment and suppressing the function of osteoclasts and an indirect effect by provoking osteoblasts to produce an inhibitor of osteoclast formation [21]. Thus, BPs suppress osteoclast-mediated bone resorption, increase BMD, and prevent osteoporotic fragility fractures in the elderly with osteoporosis, but they are relatively rare in pediatric patients with specific diseases.
Sclerotic lines associated with BP are known as “pamidronate lines” or “zebra lines.” These lines are unique to the pediatric population, reflect a temporary slowdown or cessation of rapid longitudinal bone formation in the primary spongiosa, and are parallel to the contours of the contiguous physis [25]. It is now known that these sclerotic lines do not represent “frozen” bars of the growth plate cartilage but rather horizontal trabeculae undergoing turnover [26]. They represent increased bone mineralization because of an indirect response to inhibited osteoclastic activity, rather than a growth arrest phenomenon at the time of drug treatment. Several studies have revealed that growth is not affected by BP usage by showing the radiological features in the metaphysis of the lower extremities of children [27-29]. One interesting point is that the lines gradually disappear 4 years after terminating the BP treatment. Silverman et al. [30] reported that growth plate calcification (zebra line) can be detected on radiographs 3 months after BP treatment. The lesion began to resolve 5 months after discontinuing treatment.
The effect of regular BP use on bony metabolism at the cellular level has been confirmed in pediatric patients such as those with osteogenesis imperfecta. The physiology of the skeleton around the physis, where all cells composing the skeleton play their roles, is extremely active; thus, the influence of an extrinsic agent is fully manifested and well reflected on a simple radiograph. We hypothesized that sclerotic lines similar to those observed around the physis in pediatric patients might also be identifiable in the bones of elderly patients undergoing long-term BP therapy for osteoporosis, particularly in areas where bony metabolism is more active due to external forces or other factors. In fact, we identified similar lines at the metaphysis of the distal femur or proximal tibia and around radial tuberosity in some cases (Figs. 1, 3). However, these old-age lines were not parallel, which was likely a result of variations in administration of the agent, such as dosage and the related cycle in patients with osteoporosis.
(A) A typical “zebra line” in a pediatric patient. (B) Typical sclerotic lines associated with bisphosphonate (BP) use (arrows). These lines were not parallel. Additionally, unlike zebra lines in the pediatric population, the lengths of the lines were not proportional to the diameter of the metaphysis where the lines were located. Arthroplasty was performed 6 years after BP treatment.
Although sclerotic lines around the radial tuberosity could be regarded as less clinically relevant, there are three reasons why they are related to BP use. First, sclerotic lines around the radial tuberosity were found in group 1 patients, who had previous AFFs or AFFs with atypical forearm fractures, and who were undergoing long-term BP therapy. Second, in one of group 1, both an atypical ulnar fracture and a proximal radius fracture were observed (Fig. 4). Given that the fracture site was not typical by trauma and the fracture pattern was transverse with minimal comminution, it is also considered an atypical fracture. Finally, the typical location of atypical fractures associated with BP therapy is an area under tensile force, rather than the skeleton under compressive force. Although the proximal radius is an extremely rare site considering previous studies, it is believed that the supinator muscle, which attaches to the radial tuberosity, may have subjected it under tensile stress. If there is no suspicion of an atypical fracture, many orthopedic or traumatic surgeons have attempted open reduction and internal fixation with the understanding that it was an ordinary fracture originating from a Monteggia injury-like mechanism.
(A) A 70-year-old woman presented with a complete radioulnar fracture after a fall. A rare lesion at the distal ulna (circle) prompted a further evaluation of any association with long-term bisphosphonate (BP) use. The BP had been stopped 12 months ago, and an incomplete atypical femoral fracture (AFF) was detected 74 months after BP treatment. (B) At the time of the forearm fracture, an incomplete AFF was cautiously diagnosed (circle), but sclerotic lines were detected on the distal femur (arrows). The arthroplasties were performed 2 years after BP administration.
Given that atypical fractures can occur at the radial tuberosity, the observed sclerotic line at this site may be considered a consequence of both the external forces exerted on the bone and the use of BPs. Additionally, since these sclerotic lines were observed at the time of diagnosis of AFF or AFF with an atypical ulnar fracture, it suggests that the sclerotic lines were present before the occurrence of atypical fractures due to BP use. This implies a significant alteration in bone physiology caused by BP.
Based on the results, considering the timing of the AFF diagnosis and the duration of BP use, the presence of a sclerotic line at the radial tuberosity suggests that BP treatment was likely ongoing up until relatively recently. The receiver operating characteristic analysis further suggests that the sclerotic line is most likely to be observed approximately 16.5 months after the last BP use. Based on the observation that a majority of patients (17 out of 21) with a sclerotic line at the radial tuberosity also presented with sclerotic lines in the lower extremities, it is important to consider that a proximal radius fracture with an atypical pattern, particularly if accompanied by sclerotic lines in the lower extremities, may not be due to typical trauma but rather an atypical fracture induced by BP therapy.
Some limitations of the current study should be discussed. First, The subjects of this study were patients diagnosed with AFF who underwent forearm X-rays. This selection criterion may introduce bias in the patient cohort. Since 2017, our institution has taken an interest in atypical ulnar fractures resembling AFF, leading to forearm X-rays in the majority of patients diagnosed with AFF. However, In a very small number of patients, the X-rays were taken incidentally during the diagnosis of other diseases such as epicondylitis, traumatic arthritis, and De Quervain disease between 2010 and 2017. Also, if an X-ray was taken in cases of symptoms like forearm pain or even incomplete fractures, it could introduce bias into the statistics. However, since most patients in the study were those who had X-rays taken for screening purposes after 2017, the sample remains meaningful.
Second, we were unable to determine whether the lines themselves were associated with the fracture line of an atypical fracture. Most AFFs are distinct from sclerotic lines, and many AFFs do not reveal these lines concurrently. In a few patients, the lines abruptly disappeared during reaming for intramedullary fixation using interlocking nails (Fig. 5). Also, the endosteal reaction was reported as a very early finding of BPs by Mohan et al. [18]. Thus, we presumed that the lines would be located in the medulla or endosteum of the metaphysis, rather than the cortex. Third, because this study was not a prospective comparative trial, precise information about the lines associated with several kinds of agents, the time of administration, and the drug holiday was not fully understood. A pathological study based on molecular analyses should be performed to obtain more information. A micro-computed tomography study to examine the micro-architecture of the bony trabeculae or cortical geometry influenced by BPs would be helpful.
An 82-year-old woman presented with a left-side complete atypical femoral fracture (AFF) after a fall. Definitive lines were found on the distal femur (yellow circle). During reaming for interlocking nails, the lines disappeared (red circle). The right femur was fixed 7 months previously due to an AFF.
Conclusion
Conclusively, some AFF patients had sclerotic lines on the proximal radius. These lines were particularly more frequent in patients who had been treated with a BP until recently, and almost identical lines were observed in the lower extremities.
For patients with a history of long-term BP use, it is recommended to perform regular hip X-rays as part of AFF screening, as well as forearm and knee X-rays. If sclerotic lines are detected, it is important to inform the patient about the potential risk of atypical fractures and consider adjusting osteoporosis medication. However, more studies are needed to determine whether these lines are a precursor to an atypical fracture.
Notes
Conflicts of interest
The authors have nothing to disclose.
Funding
This work was supported by research funding from the 2019 Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Science, ICT & Future Planning (NRF-2019R1I1A3A01059198).