The Use of 3-Dimensional Modeling and Printing in Corrective Osteotomies of the Malunited Pediatric Forearm: A Systematic Review and Meta-Analysis

Abstract

Introduction:

Forearm fractures contribute up to 40% of all pediatric fractures, with ≤39% of conservatively managed fractures resulting in malunion. Surgical management of malunion is challenging as precise calculation of multiplanar correction is required to obtain optimal outcomes. Advances in 3D computer modeling and printing have shown promising results in orthopaedics, reducing surgical time, blood loss, and fluoroscopy. This systematic review and meta-analysis are the first to explore the accuracy and functional outcome of 3D techniques in pediatric diaphyseal forearm malunion correction.

Methods:

A systematic review was carried out according to PRISMA guidelines.

Results:

Sixteen studies (44 patients) were included. Average 2D residual deformity was 1.84° (SD=1.68°). The average gain in range of movement (ROM) was 76.08° (SD=41.75°), with a statistically significant difference between osteotomies ≤12 months from injury and >12 months (96.36° vs. 64.91°, P = 0.027). Below a 2D residual deformity of 5.28°, no statistically significant difference on gain of ROM was found, indicating this as a nonconsequential residual deformity (P = 0.778). Multivariate regression analysis showed that 2D residual deformity and time to osteotomy only account for 6.3% gain in ROM, indicating that there are more factors to be researched.

Conclusion:

This study found superior accuracy of 3D techniques, reporting lower residual deformities than published standard osteotomy data; however, the volume of literature was limited. Larger studies are required to explore additional factors that influence accuracy and ROM, such as 3D residual deformity and the effect of particular 3D printed adjuncts. This will aid clarity in determining superiority and improve cost-effectiveness.

Diaphyseal forearm fractures are a common injury, estimated to contribute up to 40% of all pediatric fractures.¹ Definitive management consists of conservative or surgical intervention depending on the age of the patient, the extent of displacement, the proximity of the fracture to the physis, and, in turn, the remodeling potential.²

Radiographic malunion occurs in up to 39% of conservatively managed pediatric forearm fractures, with symptomatic malunion occurring in 0.5% due to the notable remodeling capacity in children.³ The latter can cause notable reduction in forearm kinematics, grip strength, and result in painful instability of adjacent joints.^4-6 Surgical management of symptomatic malunion is challenging because multiplanar correction is required to obtain optimal outcomes. Current standard methods for pediatric forearm malunion comprise corrective osteotomies with internal or external fixation.³ 2-dimensional (2D) orthogonal radiographs of the contralateral limb are used for preoperative planning, and intraoperative accuracy relies on the transferability of the plan to the operating room. Intraoperative execution of the plan necessitates obtaining fluoroscopic “maximal and no deformity” views, identifying bony landmarks and using measuring triangles with Kirchner wires.⁷ These steps are subject to inter- and intrauser variability, and a small error may result in unwanted secondary deformities.

Recent advancements in 3-dimensional (3D) computer planning software have seen the introduction of preoperative 3D modeling for surgical planning and 3D printing of patient-specific instruments, such as osteotomy guides, drill guides, and plates. These are based on CT imaging, providing detailed understanding of the multiplanar malunion for deformity analysis and planning of a theoretically precise corrective osteotomy. This concept strives to ease surgical procedures, improve accuracy, provide superior functional outcomes, improve safety, and reduce complications.^8,9 Current literature has shown its use in multiple anatomical locations for elective orthopaedic and trauma surgery, with benefits of reduced surgical time, intraoperative fluoroscopy, and blood loss.^10,11

Pediatric forearm fracture malunion correction is the most common use of 3D modeling and printing in pediatric orthopaedics.⁸ This study aims to perform a systematic review and meta-analysis evaluating the use of 3D computer modeling and 3D printing, on the accuracy and functional outcome of pediatric forearm malunion correction.

Methods

Systematic Review

A systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.¹² A comprehensive search was carried out through PubMed, Medical Literature Analysis and Retrieval System Online (MEDLINE), Excerpta Medica Database (Embase), Cumulative Index to Nursing and Allied Health Literature (CINAHL), and the Cochrane Library, from inception to January 31, 2024. The search term used was ((upper limb OR upper extremity OR forearm OR radius OR ulna) AND (deformity OR correction OR osteotomy OR maluni*) AND (3D OR 3-Dimension*) AND (Print* OR model* OR patient specific OR custom* OR tailor*)). After the search, the articles were filtered according to PRISMA guidelines by two independent reviewers.

The inclusion criteria comprised original published articles reporting the use of 3D computer modeling and printing techniques, for corrective osteotomies of pediatric malunited forearm fractures, who were ≤18 years old at the time of osteotomy. The exclusion criteria consisted of non-English language, no full-text availability, narrative reviews, animal studies, cadaveric studies, adult-only population, only grouped data, nontraumatic deformity, malunion of the distal radius, radial head/neck, olecranon, and complex injuries, such as Monteggia, Galeazzi, and Essex-Lopresti injuries.

Quality Assessment and Risk of Bias

The Joanna Briggs Institute critical appraisal checklists were used for quality assessment and to evaluate for the risk of bias of the selected studies.^13,14

Data Extraction

Individual patient data were extracted from the articles and collected onto a Microsoft Excel spreadsheet as follows:

• Demographic data: age and sex.

• Deformity and osteotomy data: location of malunion, preoperative deformity, time to osteotomy, and reason for osteotomy.

• 3D computer modeling and printing: the use of 3D computer modeling preoperatively and the use of 3D printing pre- and intraoperatively.

• Outcome measures: gain in range of movement (ROM), radiographic residual deformity after correction (2D and 3D), and grip strength.

In studies with mixed populations (adults and pediatrics) or mixed anatomical location of malunion correction, only individual patient data that fulfilled the inclusion criteria were extracted and used. If only grouped data was available within heterogenous or homogenous populations, then this was discounted. In both bone forearm fractures, if studies only provided data on the maximally deformed bone, then this was counted as one bone in the analysis and not two.

Preoperative deformity and postoperative residual deformity was reported as either 2D or 3D deformities depending on whether pre- and postoperative radiographs or CT scans were used. For studies that reported the 2D and 3D deformity angles, these were extracted directly. Other studies reported raw data, which was used to calculate 2D deformity using the following formula: $\mathrm{{\rm A}}=\left(\sqrt{C^2}+S^2\right)$, where C is the coronal angular deformity, and S is the sagittal angular deformity.^15,16 Similarly, 3D deformity was calculated from raw data, if not available directly, using the following formula¹⁶: $\beta =\text{arc}\hspace{0.17em}\cos \left[\frac{1}{2}\left(\cos \hspace{0.17em}\mathrm{{\rm T}}+\cos \hspace{0.17em}\mathrm{{\rm A}}+\cos \hspace{0.17em}\mathrm{{\rm T}}\hspace{0.17em}\cos \hspace{0.17em}\mathrm{{\rm A}}-1\right)\right].$.

In cases of both bone forearm fractures, when assessing residual deformity and how it is affected by osteotomy guide type, data from individual bones were used. However, when assessing the effect of residual deformity on ROM and grip strength, only the bone with the largest residual deformity was used, due to the forearm working as a construct to perform these actions.

Statistics

T-tests, linear regression, and multivariate linear regression statistical methods were used to evaluate the effect of certain factors on common outcome measures: residual deformity, ROM, and grip strength. A P value < 0.05 was deemed statistically significant.

Results

Systematic Review

Seven hundred and eighty articles were identified, which were subsequently filtered, resulting in 16 published articles that complied with the inclusion and exclusion criteria of this review (Figure 1).¹² Quality assessment and risk of bias performed on each study showed that three studies were of medium quality and 13 were of high quality (Appendix 1, http://links.lww.com/JG9/A366).^13,14

Figure 1.

Flowchart showing the systematic literature search according to PRISMA guidelines.¹²

Data Extraction

Forty-four patients with pediatric diaphyseal forearm malunions, secondary to fractures, were extracted (Figures 2 and 3). The mean age was 13.7 years old (SD=2.5). There were 27 males, 14 females and 3 did not specify gender. Twenty-four patients had both bone diaphyseal forearm corrective osteotomies, 9 other patients had both bone correction, but data was only provided for the bone with maximum deformity, 8 patients had radial shaft malunion correction alone and 3 patients had ulna shaft malunion correction alone, totaling 68 bones. Preoperative 2D deformity requiring corrective osteotomy was extracted from 11 studies^{17,18,19,20,21,22,23,24,25,26,27} (5 directly and 6 calculated from raw data), showing an average of 18.2° (SD=7.54°; 37 patients/56 bones). Preoperative 3D deformity was extracted from six studies^{17,18,20,22,25,26} (2 directly and 4 calculated from raw data), showing an average of 23.86° (SD=18.59°; 9 patients/13 bones). The most common reason for corrective osteotomy was reduced ROM in 34 patients (77.3%).

Figure 2.

Table showing the 16 studies and a summary of the data extracted. NS = not specified, ROM = range of movement, P = pain, DRUJ = distal radioulnar joint instability, PRUJ = proximal radioulnar joint instability, (c) = calculated from raw data in study, ✓d = grouped data or heterogenous data that was discarded.

Figure 3.

Table showing the results of data collected from the 16 studies.

All 16 studies (100%) reported the use of 3D computer modeling for osteotomy planning. Nine studies (23 patients) used 3D printing to create a model of the malunited bone, either for preoperative rehearsal or plate contouring ^27-29 or for intraoperative orientation or plating contouring.^{17,20-22,24,30} All 16 studies used 3D printed osteotomy guides, three (14 patients) using osteotomy guides alone^19,26,27 and the remaining 13 (30 patients) using osteotomy guides combined with drill guides for their intended plate.^{3,17,18,20,21,22,23,24,25,28,29,30,31} Two studies (5 patients) used patient-specific plates,^24,28 and one study (2 patients) used patient-specific hydroxyapatite wedges for opening wedge osteotomies.²³

Outcome Measures

Accuracy

The accuracy of deformity correction was determined by the residual deformity postoperatively. 2D residual deformity was extracted from nine studies (3 directly and 6 calculated from raw data) with a total of 26 patients/34 bones^{17-19,22,23,24,25,26,27} (Figure 4). The mean 2D residual deformity was 1.84° (SD=1.68°); 3D residual deformity was extracted from six studies (3 directly and 3 calculated from raw data) with a total of nine patients, 13 bones.^{17,18,22,25-27} The mean 3D residual deformity was 4.56° (SD=3.67°).

Figure 4.

Forest plot showing the average 2D residual deformity per study. |-•-| = average ± SD.

Function

The pre- and postoperative ROM was reported in 14 studies (39 patients), with a mean gain in ROM of 76.08° (SD=41.75°)^{3,17,19,20,21,22,23,24,25,27,28,29,30,31} (Figure 5).

Figure 5.

Forest plot showing the average gain in range of movement per study. |-•-| = average ± SD.

Grip Strength

Grip strength was reported in six studies; however, because of grouped or heterogenous data, only three studies were used.^19,27,31 The mean grip strength improvement was 17% (SD=16%) compared as a percentage of contralateral grip strength.

Statistics

The Effect of 2D Residual Deformity on Gain in Range of Movement

21 patients had 2D residual deformity and gain in ROM reported. The maximum residual deformity was 5.28°; 42.9% (9) of patients had a residual deformity of 0°, indicating complete resolution of alignment through corrective osteotomy. Below the maximal residual deformity of 5.28°, no statistically significant correlation was found between residual deformity and amount of ROM gained (Figure 6; R² = 0.0028, P = 0.8184).

Figure 6.

Graph showing the correlation between 2D residual deformity and gain in ROM. R² = 0.0028. 95% CI = −10.67 to 13.34, P = 0.8184.

The Effect of 2D Residual Deformity on Grip Strength

Twelve patients had residual deformity and grip strength measured. Linear regression (R² = 0.0239) showed that the more residual deformity, less of an increase in grip strength is achieved, although not statistically significant (P = 0.2446; Figure 7).

Figure 7.

Graph showing the correlation between 2D residual deformity and increase in grip strength. R² = 0.0239, CI = −8.443 to 5.376, P = 0.2446.

3D Adjuncts

No difference was observed between the type of osteotomy guide used (osteotomy guide only vs. osteotomy and drill guide) for 2D residual deformity and ROM gained (Figure 8).

Figure 8.

Table showing t-test results for the effect of variables on 2D residual deformity and gain in ROM (* = statistically significant).

The Effect of Time to Osteotomy on Gain in Range of Movement

Thirty-seven patients had time to osteotomy and gain in ROM results. A statistically significant difference in gain in ROM was found through a t-test, in patients undergoing corrective osteotomy ≤12 months post injury, compared with >12 months (96.36° vs. 64.91°, P = 0.027; Figure 8).

Multivariate Regression Analysis

Multivariate regression analysis was used to evaluate the effect of residual deformity and time to osteotomy on gain in ROM, in further detail. The R² value is 0.063, indicating that these two variables only account for 6.3% of the gain in ROM and that other variables are influential.

Discussion

Throughout this systematic review and meta-analysis, the 16 published articles were most consistent at reporting accuracy of the technique, as 2D residual deformity and functional outcomes, as the gain in ROM.

Residual Deformity

First, accuracy of the 3D modeling and printing technique has been reported within studies as residual deformity after osteotomy, calculated through radiographs (2D deformity) or CT scans (3D deformity). The mean 2D residual deformity was 1.84° (SD=1.68°), and the mean 3D residual deformity was 4.56° (SD=3.67°; Figure 4). The largest study in the meta-analysis, by Miyake et al,¹⁹ reported 2D residual deformity with 11 patients (11 bones) achieving a mean residual deformity of 0.82°, using 3D modeling and printing of osteotomy guides. Comparing our results with standard osteotomy techniques, only one study was found reporting 2D residual deformity in forearm malunion in pediatrics.³² Nagy et al³² reported on 17 patients of whom 12 (24 bones) were ≤18 years old at the time of osteotomy and comparable to our data. The standard osteotomy technique resulted in a larger 2D residual deformity of 2.5°, compared with our meta-analysis. It is well known that standard techniques have less accuracy because they are based on 2D radiographs and have a freehand nature^{7,32,33,34,35}; however, there is no sufficient literature to directly and reliably compare the results of standard and 3D modeling and printing techniques for 2D residual deformity.

3D residual deformity was reported less commonly (9 patients) in our meta-analysis, as postoperative CT scans were not regularly sought. No studies exist reporting 3D accuracy of standard osteotomy techniques on forearm malunions for comparison.²²

The Effect of Residual Deformity on Functional Outcomes

Range of Movement

No studies reviewed the effect of residual deformity on gain of ROM; however, it is known that there is a statistically significant correlation between 3D residual deformity, particularly rotation, on functional outcome.⁷ In our study, the maximal 2D residual deformity was 5.28°. Linear regression analysis showed a weak correlation between residual deformity and gain in ROM that was not found to be statistically significant (R² = 0.0028, P = 0.8184; Figure 6). This could have been due to the small sample size of 22 patients who had matched data available for analysis. It can also represent that when using 3D corrective osteotomy techniques, a residual deformity of 5.28° does not cause a statistically significant effect on ROM, implying that a residual deformity of 5.28° is an acceptable outcome.

Grip Strength

Reviewing the effect of residual deformity on grip strength, this review found a mean increase in grip strength of 17% (SD=16%), as compared as a percentage of the contralateral arm. Six studies reported grip strength; however, because of grouped data or heterogenous data mixed with adults or different anatomical locations, the data were only used from three of these studies.^19,27,31 Twelve patients had residual deformity and grip strength reported. Regression analysis showed a weak correlation between the two (R² = 0.0239, P = 0.2446), with less gain in grip strength with increasing residual deformity (Figure 7). This would be consistent with the impact of malunion on tendons and ligaments affecting lever arms, soft-tissue tension, and compliance.

Range of Movement

The most reported outcome measure in our meta-analysis was gain in ROM, after corrective osteotomy. This was reported in 14 of 16 studies (39 patients), with a mean gain of 76.08° (SD=41.75°; Figure 5). Studies investigating standard osteotomy techniques have reported variable results in gain in ROM in pediatric forearm malunion correction. Nagy et al³² showed a lower average gain in ROM of 32.92° in 12 patients (average 16 years), Price and Knapp³⁶ found a higher average gain of 102° in nine patients (average 6.99 years), and Van Geenen and Besselaar³⁷ reported an average gain in ROM of 85° in 20 patients (average 12 years).The variability in gain of ROM in standard methods could be related to the difference in age at the time of osteotomy, with younger children gaining more ROM secondary to further remodeling and growth potential after corrective osteotomy. On the other hand, it can be due to the variability in surgical execution of the proposed plan. A systematic review by Roth et al³⁸ has shown a statistically significant difference in gain of ROM through multiple regression analysis, favoring 3D osteotomy techniques compared with standard techniques in pediatric forearm malunion (84° vs. 75° respectively, P = 0.042).

Multivariate Linear Regression Analysis

Multivariate regression analysis was used to investigate the effect of 2D residual deformity and time to osteotomy on gain in ROM. The studies and data available in the pediatric population do not strongly support the anticipated functional gains of deformity correction with 3D techniques, as only 6.3% of the improvement in ROM can be explained by the two variables: time to surgery and residual deformity. This indicates that there are other variables accountable and could be due to underestimating the true extent of the residual deformity when it is assessed as 2D, rather than 3D, given the data available in the studies. Another variable may be due to gradual bone remodeling following injury and before osteotomy that transforms an acute deformity with a single center of rotation of angulation (CORA), to a smooth, long, curved deformity with multiple CORAs, hence a degree of residual deformity after a single or even 2-level osteotomy. These confounding factors might all contribute to the weak association or the presence of nonlinear correlation between residual deformity and time to surgery on gain in ROM.

3D Printed Adjuncts

The use of osteotomy guides alone vs. osteotomy with drill guides were reviewed for their effect on residual deformity or gain in ROM. We found less 2D residual deformity in osteotomy guides alone, compared with osteotomy with drill guides; however, this was not a statistically significant difference (P = 0.3; Figure 8). Interestingly, Walenkamp et al³⁵ investigated the use of osteotomy guides vs. no guides for distal radius osteotomy correction but was unable to draw conclusive benefit of using a guide.³⁵ Although the osteotomy itself may have been more accurate with a guide, their fixations were carried out freehand in both groups, which is thought to have affected postosteotomy reduction.³⁵ They suggested the additional use of 3D printed reduction guides to aid accuracy. Osteotomy guides with plate-specific drill guides could be thought of as a reduction aid because the predrilled screw holes only allow the plate to sit in one position. Furthermore, patient-specific plates could be used to aid osteotomy reduction and fixation. No statistically significant difference was found in gain in ROM, when an osteotomy guide alone or osteotomy with drill guide was used (P = 0.8573; Figure 8). Although not statistically significant, the mean gain in ROM with osteotomy guide alone was 77.92°, and combined with drill guide was 75.26°. As our residual deformity and gain in ROM was found to have better results than published standard methods, this would suggest that any additional 3D printed guide would offer a superior outcome, irrespective of it being an osteotomy guide alone or when combined with drill guides.³² A systematic review investigating standard and 3D techniques in malunion correction support this idea because they found a statistically significant difference between gain of ROM in 3D osteotomy techniques and standard methods (75° vs. 84°, P = 0.042).³⁸

No further analysis of different 3D printed adjuncts was possible as only two patients had custom-made hydroxyapatite wedges, five had patient-specific plates, and three had 3D printed bones used preoperatively. This did not provide enough data to reliably evaluate the effect of these factors on corrective osteotomy outcomes for pediatric forearm fracture malunion. Given the cost associated with 3D modeling and printing techniques, future work could focus on investigating which 3D printed adjuncts have the most notable effect on residual deformity.

Cost

Costs associated with 3D modeling and patient-specific printing is higher than the cost for standard osteotomy techniques. The studies in this systematic review reported a range of costs for 3D modeling, planning, and printing between $600 and $4300 per case, depending on the complexity.^21,22,24-26 However, using commercially available 3D planning software and medical grade resin, Miyake et al¹⁹ was able to design and print osteotomy guides for $50 per case. In time, as 3D modeling and printing techniques become more established, this could see hospitals owning their own 3D printer, as well as 3D planning software becoming more readily available, reducing overall costs.

Limitations of This Study

Despite pediatric forearm malunion being the most common use of 3D printed adjuncts in pediatrics, there is little definitive literature highlighting whether it is superior to standard osteotomies and whether the increased cost and time is beneficial. In this review, the sample size was a limitation, with 16 studies and a total of 44 patients. When evaluating certain 3D modeling and printing techniques, there was not a large enough sample size to analyze 3D printed hydroxyapatite wedges, patient-specific plates, or use of 3D printed bones preoperatively. Another limitation in sample size was that only nine patients had 3D residual deformity after osteotomy reported, whereas 26 patients had 2D residual deformity reported from radiographs. Given that malunion is often multiplanar, 2D residual deformities do not capture all aspects of deformity, such as rotation. Therefore, future research should focus on measuring residual deformity with 3D imaging to enable comparisons between standard and 3D modeling and printing techniques.

The type of studies from the systematic review were mainly level VI and VII evidence, with one study being level IV (cohort study). Higher levels of evidence are required to reliably evaluate the use of 3D modeling and printing compared with standard methods. Comparative prospective cohort studies and randomized control trials would offer more reliable information.

Conclusion

3D computer modeling and printing is an intriguing, upcoming adjunct for corrective osteotomy surgery, with many applications in malunion correction. Published literature has previously highlighted its superiority in reducing surgical times, blood loss, and intraoperative fluoroscopy throughout orthopaedics; however, this is the first systematic review and meta-analysis evaluating its use in pediatric forearm malunion, regarding accuracy of correction and functional outcome. Our study found superiority in accuracy of 3D techniques, reporting lower residual deformities than published standard osteotomy data; however, the volume of literature was limited. Regarding functional outcomes, as ROM, time to surgery had a notable benefit, and ROM was not affected by residual deformities below 5.28%, indicating that this as an acceptable outcome of deformity correction. Furthermore, multivariate regression analysis showed that only 6.3% of gain in ROM was due to 2D residual deformity and time to surgery, indicating that there are more factors to be researched. Further research is also required to determine the effect of particular components, such as the type of 3D printed guides and plates through larger comparative cohort studies or RCTs, to provide a clearer picture of the superiority of the technique and improve the cost-effectiveness.