Cortical Bone Thickness of the Distal Part of the Tibia Predicts Bone Mineral Density

SLR - October 2016 - Samuel Hall

Reference: Patterson J, Rungprai C, Den Hartog T, Gao Y, Amendola A, Phisitkul P, Femino J. Cortical Bone Thickness of the Distal Part of the Tibia Predicts Bone Mineral Density. J Bone Joint Surg Am. 2016 May 4;98(9):751–60.

Scientific Literature Review

Reviewed By: Samuel Hall, DPM
Residency Program: St. Vincent Hospital/WMC

Podiatric Relevance: The age as well as the Medicare beneficiaries across the United States is increasing. Projections are as high as 70 percent from 2010 to 2030, and a large portion of these patients are affected by osteoporosis. Just considering the lower extremity, bone quality can affect fracture management as well as implant failure risks. The gold standard today for assessing bone mineral density is the DEXA/DXA (dual x-ray absorptiometry) scan of the proximal femur, hip and lumbar spine. CBT (cortical bone thickness) can be used as an alternative when DXA is unavailable. Studies have shown positive correlation between CBT of the humerus compared with bone mineral density of the hip and femur via DXA scans. Similar results have been shown at the hip, knee and proximal tibia. This is the first study to look at CBT of the distal tibia compared with DXA scans. Since we take ankle radiographs of nearly every new patient in clinic, this information could be useful in determining bone quality for preoperative planning purposes. The purpose of this study was to investigate the correlation between CBT of the distal tibia using three standard ankle films and bone mineral density assessed with DXA scans.

Methods: A diagnostic level II study of all consecutive adult patients seen between 2006 and 2014 who had standard ankle radiographs as well as a DXA scan performed within six months of each other. Exclusions included any trauma or deformity, dysplasia or implants that would interfere with cortical measurements. All CBT measurements were performed by one orthopaedic foot and ankle surgeon. Measurements were obtained at two levels on the AP, 30mm and 50mm proximal to the distal tibial joint line perpendicular to a line from the midpoint of the ankle joint and the midpoint of the tibial shaft. The difference between the thickness of the entire bone and the thickness of the medullary canal was obtained with a digital caliper. The CBT avg was obtained by averaging the two measurements. The CBT avg was divided by the total bone diameter to obtain a CBTg (the gauge method). On the lateral, using center arc angles to find the center of the ankle joint CBT measurements were taken at 25mm and 45mm above the joint line. The hindfoot alignment view provides more cortical bone visualization, so 50mm and 70mm were used similarly to the AP. Interrater reliability was tested with 25 measurements by four separate blinded surgeons.

Results: One hundred sixty-seven patients were included in the cohort (145 female and 22 male). Mean CBT average values on the lateral, AP and hindfoot alignment views were 2.77, 3.04 and 5.15mm respectively. The lateral and AP CBT average correlated strongly with aBMD of the hip and proximal femur and lumbar spine. The hindfoot alignment view CBT correlated moderately. The CBT avg correlated more strongly than compared to the CBTg on all views. Both had excellent interrater and intrarater reliability. Twenty-eight patients were in the osteoporotic subgroup and had a mean T score of -2.70 in proximal femur, -2.28 in the hip, and -2.33 in the lumbar spine. The nonosteoporotic group (139 patients) were -1.00, -0.77 and -0.65, respectively. Mean CBTavg in the nonosteoporotic group on the AP was 3.18mm, 2.88mm on the lateral and 5.37mm on the hindfoot alignment view. For the osteoporotic group, the mean CBTavg was 2.18, 2.25 and 3.98mm, respectively. Prediction of osteoporosis calculations were carried out, and on the AP view, a threshold of 3.5mm resulted in sensitivity of 100 percent, specificity of 25 percent, accuracy of 33 percent, positive predictive value of 19 percent and negative predictive value of 100 percent. Lateral view calculations were sensitivity of 96 percent, specificity of 39 percent and negative predictive value of 98 percent. Hindfoot alignment view calculations with 5mm threshold resulted in sensitivity of 86 percent, specificity of 60 percent and negative predictive value of 95 percent.

Conclusions: CBT of the distal tibia correlated strongly with aBMD measured on DXA scans. This is consistent with the findings in other anatomical sites even though there is no aBMD of the distal tibia using DXA scans. The study indicates that a threshold of 3.5mm on the AP ankle view could be useful in screening for osteoporosis. Using four simple measurements on a standard set of ankle films, the treating physician can be provided information regarding the patient aBMD. Correlation of the CBT with aBMD of the distal extremity would confirm this. This information would be useful to assess fracture risk, implant failure and postoperative rehabilitation protocols.