A radiological study of bone remodeling with two different types of porous β-tricalcium phosphate in humans

In this study we compared the bone remodeling of unidirectional (UDPTCP) and spherical porous β-tricalcium phosphate (SPTCP) radiologically in humans. We performed a retrospective analysis of the data of 14 patients (sex, nine men and five women; age, 37–70 years) who underwent medial opening-wedge high tibial osteotomy (MOWHTO) and were followed up for 12 months after surgery. Two wedge-shaped β-TCPs (one UDPTCP and one SPTCP) were cut and placed parallel to each other in the gap. In Group A (eight knees), UDPTCP was implanted anteriorly and SPTCP posteriorly, while in Group B (six knees), SPTCP was implanted anteriorly and UDPTCP posteriorly. Computed tomography (CT) was performed at 1 week, 6 months, and 12 months after surgery, with the CT attenuation values calculated for UDPTCP and SPTCP. In Groups A and B, the CT attenuation values for UDPTCP were significantly lower at 6 and 12 months after surgery compared to those at 1 week (P < 0.05); nevertheless, no statistical difference in the comparison with SPTCP was observed. After a short-term follow-up of 12 months following MOWHTO, UDPTCP provided earlier bone remodeling than SPTCP. This outcome was achieved regardless of the position, anterior or posterior, in the MOWHTO gap.


Results
The mean ages of Groups A and B were 57.8 (range, 50-70) and 57.0 (range, 37-64) years, respectively. Relevant demographics for the two groups [sex, body mass index, pre-and post-operative femorotibial angle (FTA), and the mean opening-gap after osteotomy] are reported in Table 1, with no between-group differences noted. The time-dependent change in computed tomography (CT) attenuation values in Groups A and B at 1 week, 6 months, and 12 months after surgery were as follows: Group A [SPTCP, 1303 ± 101, 1317 ± 145, and 1279 ± 136 Hounsfield units (HU), respectively; UDPTCP, 1575 ± 137, 1136 ± 205, and 955 ± 170 HU, respectively]; Group B (SPTCP, 1300 ± 89, 1379 ± 55, and 1329 ± 42 HU, respectively; UDPTCP, 1520 ± 68, 1009 ± 149, and 930 ± 196 HU, respectively). For UDPTCP, the CT attenuation values were significantly lower at 6 and 12 months than at 1-week post-surgery (P < 0.05); however, there was no statistical difference in the time-dependent CT attenuation values for SPTCP (Figs. 2, 3). The comparison of the CT attenuation values between the two artificial bones at each time-point were as follows. In Group A, the CT attenuation value was significantly higher in UDPTCP compared to that in SPTCP (P < 0.05) at 1-week post-surgery, with no difference at 6 months. Moreover, the CT attenuation value was significantly lower in UDPTCP compared to that in SPTCP at 12 months (P < 0.05) was observed. In Group B, the CT attenuation value was significantly higher in UDPTCP compared to that in SPTCP (P < 0.05) at 1-week post-surgery. Moreover, the CT attenuation value was significantly lower in UDPTCP compared to that in SPTCP at 6 and 12 months (P < 0.05).  www.nature.com/scientificreports/

Discussion
This study showed that CT attenuation values decreased earlier for UDPTCP than for SPTCP when these two materials were used as a bone substitute for MOWHTO in the same individual. In previous reports, the CT attenuation value of the proximal tibia after OWHTO was 100-200 HU 12,13 , and the fact that the UDPTCP reached that value earlier compared to SPTCP indicated that UDPTCP promoted faster bone remodeling. The strength of our study was that the two types of artificial bone were compared in the same individual and examined at different locations, namely an anterior or posterior position in the MOWHTO gap. Therefore, we assessed the bone remodeling potential of UDPTCP and compared it with that of SPTCP in the same environment. Considering that the sagittal mechanical axis of the lower extremity is not located at the center of the proximal tibia 14 , the loading conditions on the two artificial bone spacers could differ depending on whether they are placed anteriorly or posteriorly in the MOWHTO gap. In our study, we completed an in vivo assessment of the two artificial bone types for the anterior and posterior positions in the MOWHTO gap. Regardless of the location and, thus, loading, the CT attenuation values decreased faster for UDPTCP than for SPTCP. UDPTCP and SPTCP have a different pore structure but have almost the same porosity (57% and 60%, respectively). UDPTCP is characterized by an interconnected network of pores that are aligned in a single direction. Makihara et al. 6 demonstrated that this characteristic of UDPTCP allowed the rapid penetration of the tissue into the material when implanted in the tibia of rabbits. Similarly, using a rat model, Murayama et al. 15 reported a faster infiltration of the bone into UDPTCP than into other types of β-TCP, resulting in a more rapid formation of capillaries throughout the UDPTCP. To the best of our knowledge, this was the first study that evaluated β-TCP remodeling in different structures and compared the CT attenuation values in humans. Interestingly, our findings were consistent with those of animal studies. However, we should consider that a simple comparison could not be made, as a previous report has suggested that bone remodeling in humans and animals was different 16 .
Although no clinical trial has directly compared SPTCP and UDPTCP, the benefits of UDPTCP in practice have been reported in a few studies. For calcaneal fractures, UDPTCP was replaced with autogenous bone at 6 months post-surgery, with favorable clinical outcomes obtained 17 . UDPTCP also provided good clinical potential as a bone substitute to fill gaps after the treatment of benign bone tumors of the hand and vertebral fractures 18 . Additionally, in an image-based assessment of lateral interbody fusion in the lumbar spine, UDPTCP was not inferior to the autologous bone graft 19 . In this study, we focused only on bone remodeling, but clinical outcomes should also be evaluated in the future.
However, the limitations of our study should be acknowledged. First, the sample size of the study was small. Second, our last assessment was performed at 12 months post-surgery; thus, longer-term differences between SPTCP and UDPTCP on bone remodeling were not considered. Third, we measured the CT attenuation values at only one point (i.e., at the center of the osteotomy). A previous work indicated that bone remodeling began from the TCP and the bone contact area in OWHTO 20 ; therefore, the CT attenuation values may have been measured at the last area to be remodeled, and decrease in the CT attenuation values at this area implied that other areas had also been remodeled. Fourth, in this study, SPTCP and UDPTCP were implanted in the same gap, which makes it impossible to avoid their mutual influence. Finally, we did not perform any histological assessment of the bone union site. Therefore, further research is needed to clarify the remodeling process in detail.
In conclusion, over a short-term follow-up of 12 months after MOWHTO, UDPTCP provided earlier bone remodeling than SPTCP, as quantified by the CT attenuation findings. This outcome was achieved regardless of the position, anterior or posterior, in the MOWHTO gap.

Methods
Study group. This was a retrospective comparative study of 14 patients who underwent MOWHTO, using a locking compression plate. Of these, eight patients were enrolled between January and December 2017, in whom UDPTCP (Affinos) and SPTCP (Osferion 60, Olympus Terumo Biomaterials, Tokyo, Japan) were implanted in the anterior and the posterior spacer (Group A), respectively. The other six patients were enrolled between January 2018 and March 2019; in these patients, SPTCP and UDPTCP were implanted anteriorly and posteriorly, respectively (Group B).
Description of the surgery. The surgery was performed by two surgeons using the standard MOWHTO protocol, after accurate pre-operative planning of the osteotomy on standing full-limb anterior-posterior (AP) radiographs 21 . A biplanar osteotomy was performed with a horizontal cut of the posterior two-third of the proximal tibia and a 100° angulated cut ascending anteriorly. The osteotomy site was opened and the gap created was filled with bone substitutes. From a rectangular block of SPTCP or UDPTCP, we created two wedgeshaped spacers, implanted parallel into the anterior and posterior parts of the opening gap (Fig. 4). In Group A, UDPTCP and SPTCP were used in the anterior and the posterior spacer, respectively. In contrast, in Group B, SPTCP and UDPTCP were used in the anterior and the posterior spacer, respectively.
The biomechanical characteristics of the SPTCP and UDPTCP are shown in Table 2. UDPTCP was implanted with the pores within the material oriented parallel to the tibial axis. A locking compression plate (TriS Medial HTO Plate System, Olympus Terumo Biomaterials, Tokyo, Japan) was used to fix the MOWHTO site. The patients were non-weight-bearing for 1 week after surgery, followed by partial weight-bearing, with full-body weight-bearing permitted at 4 weeks after surgery.   CT computed tomography, HU Hounsfield units, SPTCP spherical porous β-tricalcium phosphate, UDPTCP unidirectional porous β-tricalcium phosphate.