Titanium wear from magnetically controlled growing rods (MCGRs) for the treatment of spinal deformities in children

Abstract

Magnetically controlled growing rods (MCGRs) are an effective treatment method for early-onset scoliosis (EOS). In recent years, increasing titanium wear was observed in tissue adjacent to implants and in blood samples of these patients. This study aims to investigate the potential correlation between amount of metal loss and titanium levels in blood during MCGR treatment as well as influencing factors for metal wear. In total, 44 MCGRs (n = 23 patients) were retrieved after an average of 2.6 years of implantation and analyzed using a tactile measurement instrument and subsequent metal loss calculation. Titanium plasma levels (n = 23) were obtained using inductively coupled plasma-mass spectrometry (ICP-MS). The correlation of both parameters as well as influencing factors were analyzed. Titanium abrasion on MCGRs was observed in the majority of implants. There was no correlation of metal implant wear or titanium plasma values to the duration of MCGR implantation time, number of external lengthening procedures, patient's ambulatory status, gender, weight or height. Material loss on the MCGRs showed a positive correlation to titanium blood plasma values. The present study is one of the first studies to analyze retrieved MCGRs using high-precision metrological techniques and compare these results with ICP-MS analyses determining blood titanium values.

Figure 1

Bilateral MCGR system for spinal deformity correction in children. Anterior–posterior (a,c) and lateral radiographs (b,d) of a 6-year old boy with neuromuscular scoliosis. The main scoliotic curve was corrected from 71° (a) to 32° (c) with additional improvement of the sagittal profile (b,d). Intraoperatively, discoloration of the adjacent soft tissue was found (e).

Figure 2

Tactile measurement of MCGRs. Selection of MCGRs with completely expanded inner segments, displaying one rod without detectable abrasion (left rod) and five rods with metal abrasion (a). Display of an individual MCGR fixed within the MarSurf LD260 machine (Mahr GmbH, Germany) with one segment facing upwards at a time (b). The diamond-tip of the probe arm was set to record longitudinal tactile traces along this segment of the MCGR (c). Screenshot of the software MarWin (Mahr GmbH, Germany), which was used to measure the depth, width and area of individual notches, representing metal abrasion along the tactile traces (d).

Figure 3

Volume of abrased material calculated from tactile traces. Schematic representation of the inner segment of a MCGR with five representative notches on one angular segment; with b being the width of the segment (a). Nine longitudinal traces (A–I) with a distance of 0.5 mm from each other were recorded longitudinally along the segment, so that the entire width b (approximately 5 mm) of the segment was covered by regular traces. In (b) two representative notches (1) and (2) are shown. Schematic representation of a notch as displayed in the software MarWin. Using the software, the 2D-area of each individual notch (here F_C2) on each individual trace was determined (c). The volume was calculated from the areas by numerical integration. Therefore, the area of each notch on a trace (e.g. F_A1, F_B1 etc.) was multiplied by 0.5 mm, which is the distance between the traces (d).

Figure 4

Titanium abrasion in mm³ for all measured rods of the 23 patients. 21 patients had two rods implanted (concave, convex), whereas two patients (patient number 13 and 18) only had one implant on the concave side. The convex rods of patient 6 and 12 displayed no and very little abrasion (0.00016 mm³) respectively, while patient 10 showed maximum values (a). No significant difference was detected between abrasion of MCGRs implanted on the concave or convex side of the scoliotic spine (paired t test; p = 0.9913) (b).

Figure 5

Titanium plasma analysis. Significant difference in titanium wear in MCGR-patients (n = 23) in comparison to age-matched controls (n = 9; p = 0.020). Unpaired t test with Welch’s correction (different SD assumed).

Figure 6

Correlation of abrasion and titanium analysis in plasma. For patients, whose implants at the time of blood sampling were the initial ones (n = 12), linear regression analysis between abrasion on implants in mm³ (x-axis) and titanium in plasma samples in ng/mL (y-axis) display a positive correlation (R² = 0.3743; deviation of slope from zero significant p = 0.0345) (a). When extrapolating the measured titanium concentrations to the estimated entire plasma volume of each patient, the positive relationship remained (R² = 0.2041; deviation of slope from zero not significant p = 0.1404) (b).