- Email: [email protected]
Conventional Versus Computer-Assisted Corrective Osteotomy of the Forearm: a Retrospective Analysis of 56 Consecutive Cases
SCIENTIFIC ARTICLE
Conventional Versus Computer-Assisted Corrective Osteotomy of the Forearm: a Retrospective Analysis of 56 Consecutive Cases David Ephraim Bauer, MD,* Stefan Zimmermann, MD,* Alexander Aichmair, MD,* Andreas Hingsammer, MD,* Andreas Schweizer, MD,* Ladislav Nagy, MD,* Philipp Fürnstahl, PhD†
Purpose Accuracy and feasibility of corrective osteotomies using 3-dimensional planning tools and patient-specific instrumentation has been reported by multiple authors with promising results. However, studies describing clinical outcomes following these procedures are rare. Therefore, the purpose of this study was to compare the results of computer-assisted corrective osteotomies of the diaphyseal and distal radius with a conventional nonecomputer-assisted technique regarding duration of surgery, consolidation of the osteotomy, and complications. Also, subjective and objective clinical outcome parameters were assessed. Methods We retrospectively compared the results of 31 patients who underwent a corrective osteotomy performed conventionally with 25 patients treated with a computer-assisted method (CA) using patient-specific instrumentation. Baseline data were similar among both groups. The duration of surgery, bony consolidation, complications, gain in range of motion, and subjective outcome were recorded. Results The mean operating time was significantly shorter in the CA group compared with the conventional group. After 12 weeks, significantly more osteotomies were considered healed in the CA group compared with the conventional group. Two patients in the CA group required revision surgery to treat nonunion of the osteotomy. Otherwise clinical results were similar among both groups. Conclusions The results demonstrate that the computer-assisted method facilitates shorter operation times while providing similar clinical results. (J Hand Surg Am. 2017;-(-):-e-. Copyright Ó 2017 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Therapeutic IV. Key words Malunion, distal radius, patient specific instrumentation, rapid prototyping.
M
forearm fractures may lead to decreased range of motion, reduced grip strength, or painful instability of the distal radioulnar joint (DRUJ).1,2 ALUNION OF DISTAL RADIUS OR
Several authors have reported that adequate realignment and accurate restoration of the normal anatomy is associated with an overall positive clinical outcome.3e5 However, quantification of a complex
From the *Orthopedic Department; and the †Computer Assisted Research and Development Team, Balgrist University Hospital, University of Zurich, Zurich, Switzerland.
Corresponding author: David Ephraim Bauer, MD, Orthopedic Department, University Zurich Balgrist, Forchstrasse 340, Zurich 8008, Switzerland; e-mail: [email protected].
Received for publication May 18, 2016; accepted in revised form March 20, 2017.
0363-5023/17/---0001$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2017.03.024
A.S., L.N., and P.F. are shareholders of a company offering services for computer-assisted preoperative planning (Balgrist CARD AG, Zurich, Switzerland).
Ó 2017 ASSH
r
Published by Elsevier, Inc. All rights reserved.
r
1
2
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
DRUJ (n ¼ 27, 15 patients in the conventional group, 12 in the CA group), reduced range of motion (ROM) (n ¼ 27, 15 patients in the conventional group vs 12 in the CA group), ulnocarpal impaction syndrome (n ¼ 11, 4 in patients in the conventional group vs 7 in the CA group) after nonoperatively or operatively treated extra-articular fracture of the distal radius (n ¼ 40), diaphyseal radius fractures (n ¼ 4), or both bone forearm fractures (n ¼ 12). Five fractures in the CA group and 4 fractures in the conventional group had previously been treated with open reduction and internal fixation. Regarding the preoperative ROM, there were 8 patients with decreased pro-/supination and 2 patients with decreased flexion/extension alone in the conventional group and 7 patients with decreased pro-/ supination and 3 patients with decreased flexion/ extension alone in the CA group. A combination of decreased flexion/extension and pro-/supination was recorded in 6 patients in the conventional group and 3 patients in the CA group (Table 1). The analysis of clinical outcome parameters was performed after stratifying patients into the above-mentioned 3 groups of preoperative complaints. Clinical assessment was performed by the surgeon and was therefore not blinded. Range of motion was assessed with a standard goniometer. We calculated the increase in ROM after surgery by subtracting the postoperative from the preoperative arc of ROM. A distinct cutoff value for reduced ROM, as an indication for surgery, was not set. The indication for operation was based on the patient’s complaint. Instability was diagnosed by a clinical examination of increased translation of the distal ulna relative to the distal radius compared with the uninjured side. The indication for surgical intervention was pain on loading the unstable joint, provoking the instability pattern. Ulnar-sided wrist pain, exacerbated by ulnocarpal loading in combination with positive ulnar variance, was considered diagnostic for ulnocarpal impaction syndrome. The subjective outcomes of patients operated because of painful instability of the DRUJ were classified as (1) persisting pain, (2) improved but remaining symptoms, and (3) complete resolution, respectively. In addition to assessing self-reported parameters, a retrospective chart review was performed to determine hospital stay, duration of surgery, and bony consolidation. Regarding the duration of surgery, the tourniquet time was considered as an adequate surrogate.
multiplanar malunion and subsequent determination of the required correction by means of conventional radiographs or computed tomography (CT) images remain challenging.6e8 Recent advancements in computer-assisted 3-dimensional (3D) planning and 3D printing technology can support the surgeon in enabling a more accurate preoperative simulation combined with easy-to-use navigation aids so that the surgery can be performed as preoperatively planned. In the context of a malunion of the radius, computer-assisted corrective osteotomy has been described as a promising technique for accurate reconstruction if patientspecific guides are used.9,10 Although the accuracy of the reduction has been demonstrated, studies investigating clinical outcome parameters following these procedures are rare.11 MATERIALS AND METHODS The aim of this study was to compare the results of computer-assisted corrective osteotomies of the diaphyseal and distal radius using 3D planning and patient-specific guides with a conventional, nonecomputer-assisted technique. We evaluated the duration of surgery, healing of the osteotomy, and complications. In addition, we assessed subjective and objective clinical outcomes. Patient sample After institutional review board approval was obtained, a convenience sample of 56 patients with symptomatic, posttraumatic deformities of the forearm treated with either a nonecomputer-assisted method (conventional group) or a computer-assisted corrective osteotomy9 (CA group) was available for final analysis. The conventional group was treated between January 2003 and October 2008 and the CA group was treated between December 2009 and April 2015. Patients with congenital deformities of the forearm and patients who underwent concomitant interventions during the index procedure (scapholunate ligament reconstruction or repair of the triangular fibrocartilage complex12) were excluded from final analysis. Thirty-one patients in the conventional group (15 males and 16 females; mean age, 31 y; range, 10e66 y) and 25 patients in the CA group (17 males and 8 females; mean age, 28 y; range, 11e71 y) were finally included. Preoperative clinical and radiological assessment Indications for surgical intervention were one or more of the following: painful instability of the J Hand Surg Am.
r
Vol. -, - 2017
3
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
TABLE 1.
Baseline Data Conventional (n ¼ 31) Variable
N
Mean or %
CA (n ¼ 25) SD
n
Mean or %
36
25
133
35
24
140
12.6
18
26.4
SD
Preoperative clinical function Pro-/supination ( )
31
128
Flexion/extension ( )
30
124
Grip strength (kg)
20
26.1
21
67.7%
18
72.0%
Both bone forearm fracture
8
25.8%
5
20.0%
Radius shaft fracture
2
6.5%
2
8.0%
38 40 12.6
Preoperative radiological findings Distal radius fracture
kg, kilogram; SD, standard deviation.
calculating the optimal fit between the fragments proximal to the fracture line using the iterative closest point surface registration functionality14 of the CASPA software. The osteotomy was then simulated on the pathological bone model, resulting in a proximal and distal fragment. The reduction was achieved by aligning the distal fragment to the reconstruction template using the iterative closest point method. The relative transformation of the distal fragment from initial to reduced position quantified the amount of required reduction in all 6 degrees of freedom (ie, 3 values each for rotation and translation, respectively). Based on the calculated reduction, the patient-specific guides were designed with the CASPA software. Detailed descriptions of the applied planning and guide design techniques can be found in previously published studies.15 The patient-specific guides were manufactured based on the preoperative simulation by selective laser-sintering using biocompatible polyamide PA 2200 (Medacta SA, Castel San Pietro, Switzerland). Preoperatively, the guides were verified using a true-size replica of the injured and reduced bone and sterilized with conventional steam sterilization. A volar approach to the radius was performed in all cases. Closing or opening wedge osteotomies were performed if an angular deformity was predominant and single-cut osteotomies were performed if a rotational deformity was predominant. All the osteotomies performed in the CA group were planned in a way such that angular and rotational deformities are simultaneously corrected. An opening wedge osteotomy was performed in 13 cases, a closing wedge osteotomy in 2 cases, and a 3D single-cut osteotomy in 10 cases.
Radiographic analysis of consolidation of the osteotomy was performed by a board-certified radiologist using CT scans or plain radiographs (28 plain radiographs vs 3 CT scans in the conventional group compared with 13 plain radiographs vs 12 CT scans in the CA group). Patients were routinely scheduled for clinical and radiological examinations every 6 weeks after the index procedure until the osteotomy was considered healed. Consolidation was assumed when the radiolucent osteotomy line was no longer visible on CT scans or plain x-rays. Computer-assisted osteotomy Both the injured forearm and the contralateral uninjured limb of all 25 patients who underwent the computer-assisted method were evaluated preoperatively based on bilateral CT scans (slice thickness, 1 mm; 120 kV; CT device: Philips Brilliance 40; Philips Healthcare, Best, the Netherlands). The CT data were segmented using a commercially available segmentation software package (Materialise, Loewen, Belgium) in a semiautomatic fashion by applying thresholding and region-growing algorithms to obtain 3D surface models of both the injured and the contralateral bone. The 3D models were imported into the planning software CASPA (Balgrist CARD AG, Zurich, Switzerland) to perform the 3D preoperative planning according to a technique described by Murase et al13 and further developed by Schweizer et al8 (Fig. 1). In brief, the method can be summarized as follows: the 3D model of the malunited bone was divided into a region proximal and distal to the previous fracture zone. The 3D bone model of the injured side was then superimposed with the mirror model of the contralateral uninjured side by J Hand Surg Am.
r
Vol. -, - 2017
4
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
FIGURE 1: Types of 3D planned osteotomies. The rotation axis (r,C) and the osteotomy plane(s) are shown in red and gray, respectively. The pathological (orange) and contralateral bone (green) are given before (first column) and after (second column) simulated reduction. The reduced bone fragment is depicted in violet. A Distal radius opening wedge osteotomy. B Closing wedge osteotomy of the radius. C Single cut osteotomy of the tibia. (Reprinted with permission from Fürnstahl P, Schweizer A, Graf M, et al. Surgical treatment of long-bone deformities: 3D preoperative planning and patient-specific instrumentation. In: Zheng G, Li S, eds. Computational Radiology for Orthopaedic Interventions. Cham, Switzerland: Springer International Publishing; 2016:123e149.15)
If an opening wedge osteotomy was performed, cancellous bone autograft harvested from the iliac crest was interposed between the osteotomy fragments.
to the true angle of deformity and the corresponding orientation in space. A trigonometric calculation of the deformity allowed for prebending and fixation of the plate to either the proximal or distal fragment before performing the osteotomy to control correct reduction. An opening wedge osteotomy was performed in 23 cases, a closing wedge osteotomy in 7 cases, and a transverse osteotomy, for simultaneous correction of combined angular and rotational deformity, in 1 case. Cancellous bone autograft harvested from the iliac crest was interposed between the osteotomy fragments if an opening wedge osteotomy was performed. The correction was maintained with a conventional 3.5-mm plate (Synthes, Westchester, PA) for shaft osteotomies or a 2.4/ 2.7 mm distal Radius Locking compression plate (Synthes). Intraoperative radiographs were obtained to verify proper spatial reconstruction and implant placement.18
Conventional osteotomy Preoperative planning was performed using anteroposterior and lateral plain radiographs of both forearms. The contours of both the uninjured and the deformed bone were copied on separate sheets of tracing paper. The location and extent of the maximal deformity were evaluated by superposition of the tracing paper sheets.16 The true 3D angle of deformity and its corresponding orientation in space were determined as described by Nagy et al.16 Crosssectional CT scans were obtained only if a rotational deformity was suspected. To assess rotational deformity using CT slices, the anatomic relationship of the radial styloid process and the bicipital tuberosity was evaluated according to the method described by Bindra et al.17 A standard volar approach to the radius was performed in all cases. Intraoperatively, Kirschner wires were placed under fluoroscopic guidance in different planes to mark the orientation and extent of malposition of the proximal and distal fragment. After performing the osteotomy, proper anatomic reduction was achieved by aligning the previously placed Kirschner wires in a parallel fashion. The osteotomy was performed with respect J Hand Surg Am.
Statistical analysis Data are presented as mean and standard deviation for continuous variables and as proportion (%) for categorical variables, if not stated otherwise. A 2tailed Kolmogorov-Smirnov test was used for testing a normal distribution; if P .05, the data were considered as normally distributed. The chisquared test was applied for testing differences of distribution of categorical variables between groups r
Vol. -, - 2017
5
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
TABLE 2.
Surgical Details Conventional (n ¼ 31)
Variable
CA (n ¼ 25)
n
%
n
%
P Value
Distal radius
21
67.7%
18
72.0%
n.a.
Midshaft
10
30.3%
7
28.0%
Opening wedge
23
74.2%
13
52.0%
Closing wedge
7
22.6%
2
8.0%
Single cut
1
3.2%
10
40.0%
Localization of osteotomy
Type of osteotomy <.05
n.a., not applicable.
as appropriate. For testing differences between means, the unpaired t test was applied for normally distributed data and the Mann-Whitney U test for nonnormally distributed data. A P value of < .05 was considered statistically significant.
the CA group, 2 patients underwent revision surgery because of nonunion of the osteotomy, 1 single cut and 1 open wedge osteotomy, respectively. For the initial open wedge osteotomy, nanocrystalline hydroxyapatite embedded in a silica gel matrix (NanonBone, ARTOSS, Rostock, Germany) was used as a bone substitution graft. Iliac crest autograft was used for revision surgery in both cases. Considering cases not undergoing revision surgery, all osteotomies except for one in the conventional group were consolidated within 36 weeks after the index procedure (Fig. 2). The analysis of clinical outcome parameters was performed after stratifying patients into different groups based on the preoperative complaint. The average postoperative gain of ROM in patients in whom the arc of pro-/supination was the indication for surgical intervention was 41 39 in the conventional group compared with 44 27 in the CA group. The average postoperative gain of ROM in patients in whom amplitude of flexion/extension was the indication for operative intervention was 28 42 in the conventional group compared with 25 36 in the CA group. The results of subjective outcome parameters in patients in whom painful instability of the DRUJ was the primary complaint are illustrated in Table 3. Sixteen patients in the conventional group (51.6%) and 12 patients in the CA group (48.0%) underwent implant removal. The indications for implant removal were either prophylactic to prevent a chronic contact between the plate and the flexor tendons, discomfort caused by the implant, or reduced ROM attributed to the implant. The average gain of pro-/ supination after implant removal was 1 34 in the conventional group compared with 10 15 in the CA group.
RESULTS The median time from the initial trauma until corrective osteotomy was 18.0 months (interquartile range, 100.5 mo) in the conventional group compared with 28.0 months (interquartile range, 33.0 mo) in the CA group. The average follow-up was significantly longer in the conventional group compared with the CA group (25.4 20.1 mo vs 13.6 7.4 mo, P < .05). Preoperative clinical assessment Baseline data were similar amongst both groups as illustrated in Table 2. The average arc of flexion/ extension in the group of patients in whom reduced flexion/extension was the indication for surgery was 96 29 in the conventional group compared with 83 35 in the CA group. The average arc of pro-/supination in the group of patients in whom reduced pro-/supination was the indication for surgery was 105 33 in the conventional group compared with 96.0 32 in the CA group. Surgical intervention and outcome The mean operating time was significantly shorter in the CA group compared with the conventional group (140 37 vs 108 26 min, P < .05). Twelve weeks after the index procedure significantly more cases in the CA group were considered healed compared with the conventional group (14 cases [60.9%] vs 10 cases [32.3%], P < .05). However, in J Hand Surg Am.
r
Vol. -, - 2017
6
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
This study has several limitations. Because of the retrospective design, we were not able to consistently record potentially meaningful parameters such as preoperative ROM of the contralateral uninjured side, pain measured by the visual analog scale, or smoking status if the required documentation was missing. As of 2009, we changed the surgical technique in our institution from the conventional to the computerassisted method. This may be one factor contributing to the shorter follow-up time in the CA group compared with the historic conventional group. As a result of the shorter follow-up time, this study may have failed to demonstrate adverse long-term outcomes in the CA group. Further, the size of the study sample of each subgroup is small, thereby reducing the power. Therefore, statistical comparisons were not possible for many of the parameters of interest. In addition, clinical evaluations of DRUJ instability and ulnar abutment were performed by the surgeon in a nonblinded fashion potentially introducing the risk of a biased outcome. As opposed to the CA group, preoperative planning in the conventional group was performed using plain radiographs of both forearms. Hence, correct execution of the preoperative plan was also verified using plain radiographs, and patients in the conventional group therefore inconsistently received postoperative CT scans. The duration of surgery among cases operated using the computer-assisted method was significantly shorter compared with the conventional method. This may be attributable to the preoperative simulation and the advantage of having the patientspecific instrument as a reliable reference. Therefore, the previously described technique of placing Kirschner wires under fluoroscopic guidance to mark the orientation and extent of malposition of the proximal and distal fragments in conventional osteotomies is not required thereby saving operation and fluoroscopy time. The authors’ experience is also that the guides enable the surgeon to perform the surgery in a more standardized fashion compared with the nonecomputer-assisted method. From an economic standpoint, manufacturing of the patient-specific instruments entails additional time and costs, and requires a CT scan, including the contralateral side, for planning of the correction. In our institution, the CT scan required for planning of the osteotomy includes both forearms and therefore does not incur additional costs for scanning the contralateral limb. Depending on the complexity of the osteotomy, planning of the correction requires 2 to 4 hours. In this case series, production of the laser sintered
FIGURE 2: Time to consolidation of the osteotomy. Significant difference in consolidation of the osteotomy site after 12 weeks. *Indicates statistical significance (P < .05).
Complications One patient in the conventional group experienced rupture of the extensor pollicis longus tendon 8 weeks after the index procedure, 1 patient required an additional corrective osteotomy because of persisting ulnar impaction, 1 patient a repair of the triangular fibrocartilage complex because of persisting instability of the DRUJ, and 2 patients underwent a Sauvé-Kapandji salvage procedure19 because of persisting ulnar-sided wrist pain. In both cases, a DRUJ-preserving procedure was attempted even though signs of DRUJ arthritis were apparent. One patient experienced a transient palsy of the ulnar nerve with spontaneous, complete resolution within 1 year after surgery. In the CA group, 2 patients required an additional ulna shortening osteotomy because of persisting ulnocarpal impaction syndrome. In 1 case, the initial correction was lost after a prolonged time to bony consolidation, and in the second case, correction could not be executed as preoperatively planned. DISCUSSION Three-dimensional preoperative planning and the use of patient-specific instrumentation have been described as a promising technique for various indications in hand surgery. Prior studies demonstrated that the use of custom-designed osteotomy guides enables a more accurate reduction compared with free hand techniques.9,13,20,21 In this study, we described our experience with 56 patients who underwent a corrective osteotomy, in 25 of whom, it was performed via a computer-assisted approach. J Hand Surg Am.
r
Vol. -, - 2017
7
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
TABLE 3.
Postoperative Outcome Data Conventional (n ¼ 31) Variable
CA (n ¼ 25)
n
Mean
SD
n
Mean
SD
14
41
39
10
43
27
8
28
41
6
25
36
Clinical outcome Gain pro-/supination ( )* †
Gain flexion/extension ( ) Gain grip strength (kg) Stability: asymptomatic Stability: improved‡ ‡
Stability: persisting
‡
18
3.1
9 4 2
9.8
17
4.1
60.0%
7
58.3%
26.7%
4
33.3%
13.3%
1
8.3%
7.7
Radiological outcome Consolidation (wk)
31
23.9
13.3
25
20.9
16.6
kg, kilogram; SD, standard deviation. *In the reduced pro-/supination group only. †In the reduced flexion/extension group only. ‡In the instability group only.
intraoperative guides was performed by a third-party company and required approximately 4 weeks. For the cases presented in this study, the additional costs including CT scan, and time required for planning and production of the guide, amount to $2,415 USD. Two patients in the CA group required revision surgery because of nonunion of the osteotomy (1 diaphyseal, 1 metaphyseal). Given the fact that achieving perfect fit of the patient-specific cutting guides is paramount for the correct and precise execution of the preoperative plan, extensive stripping of the periosteum is often necessary. However, stripping of the periosteum, a primary source of osteogenic stem cells giving rise to callus formation and promoting the initial phase of bone healing, as well as decreased blood supply to the bone, may therefore be considered as a contributing factor to these complications.22 However, significantly more osteotomies in the CA group were considered healed after 12 weeks compared with the conventional group. This might be accounted to the lower number of patients requiring open wedge osteotomies or to the fact that osteotomy union was evaluated with a CT scan more often in the CA group. Although not specifically analyzed, more rapid union might facilitate quicker rehabilitation and return to work. Two patients in the conventional group underwent palliative Sauvé-Kapandji procedures because of secondary osteoarthritis of the DRUJ. In both cases, the indication for corrective osteotomy was painful subluxation of the DRUJ because of posttraumatic deformity after a distal radius fracture. Both patients were young manual workers, and an attempt to J Hand Surg Am.
perform a joint-preserving procedure was undertaken although signs of osteoarthritis of the DRUJ were already apparent. Similarly, 2 patients in the CA group required additional ulnar shortening. In both cases, the indication for the corrective osteotomy was reduced ROM. One patient presented with exacerbated ulnar abutment after the index procedure, and the other patient experienced loss of the planned reduction with a prolonged time to consolidation potentially affected by a history of heavy smoking. In our study, the mean arc of pro-/supination improved from 105 to 147 in the conventional group and from 95 to 139 in the CA group. Previously published studies on conventional and computer-assisted correction of malunion of the radius report slightly worse preoperative forearm motion compared with our study cohort. Also, in our study, the gain of pro-/supination in both intervention groups was less than reported in the literature for conventional osteotomies.13,23e25 Similar results have been observed in one of our previously published cohorts describing the conventional technique.16 The smaller reported gain in ROM may be explained by the higher patient age and the longer delay until the corrective osteotomy was performed in our cohort.23,25 However, as a result of the small number in these subgroups, we were not able to detect a significant difference considering these parameters. In summary, computer-assisted corrective osteotomy of the radius enables shorter operation times by providing a reliable intraoperative reference for the planned correction thereby replacing previously r
Vol. -, - 2017
8
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
FIGURE 3: Single-cut multiplanar correction. A Injured (orange) compared with uninjured contralateral side (green). B Planning of the correction plane and rotation axis. C Correction plane with cutting guide. D After reduction with reduction guide. E Cutting guide in situ. F Reduction with guide in situ.
required fluoroscopic referencing steps. Considering clinical outcome parameters, further studies with larger cohorts are warranted, to evaluate whether a more accurate reduction yields a better clinical outcome. However, the computer-assisted method may be particularly helpful for complex, rotational deformities, which are difficult to plan and implement using the conventional technique. In particular, a multiplanar correction by rotating around a single axis can only be achieved by 3D analysis and preoperative simulation of the correction (Fig. 3). From an economic standpoint, the total direct costs for segmentation, 3D analysis, and planning and production of the osteotomy guides per case amount to $2,415 USD. These additional expenses should be compared against potential financial benefits including reduced operation time, quicker rehabilitation, and return to work. However, large-scale costeffectiveness analyses of the CA method require prospective documentation of outcome measures, such as health-related quality of life and qualityadjusted life years, in combination with the assessment of both direct and indirect costs. A long-term study reporting these parameters is currently being planned.
2. Schemitsch EH, Richards RR. The effect of malunion on functional outcome after plate fixation of fractures of both bones of the forearm in adults. J Bone Joint Surg Am. 1992;74(7):1068e1078. 3. McQueen M, Caspers J. Colles fracture: does the anatomical result affect the final function? J Bone Joint Surg Br. 1988;70(4):649e651. 4. Fernandez DL. Malunion of the distal radius: current approach to management. Instr Course Lect. 1993;42:99e113. 5. Matthews LS, Kaufer H, Garver DF, Sonstegard DA. The effect on supination-pronation of angular malalignment of fractures of both bones of the forearm. J Bone Joint Surg Am. 1982;64(1):14e17. 6. Dumont CE, Nagy L, Ziegler D, Pfirrmann CW. Fluoroscopic and magnetic resonance cross-sectional imaging assessments of radial and ulnar torsion profiles in volunteers. J Hand Surg Am. 2007;32(4): 501e509. 7. Dumont CE, Pfirrmann CW, Ziegler D, Nagy L. Assessment of radial and ulnar torsion profiles with cross-sectional magnetic resonance imaging. A study of volunteers. J Bone Joint Surg Am. 2006;88(7): 1582e1588. 8. Schweizer A, Furnstahl P, Harders M, Szekely G, Nagy L. Complex radius shaft malunion: osteotomy with computer-assisted planning. Hand (NY). 2010;5(2):171e178. 9. Vlachopoulos L, Schweizer A, Graf M, Nagy L, Furnstahl P. Threedimensional postoperative accuracy of extra-articular forearm osteotomies using CT-scan based patient-specific surgical guides. BMC Musculoskelet Disord. 2015;16(1):336. 10. Schweizer A, Furnstahl P, Nagy L. Three-dimensional correction of distal radius intra-articular malunions using patient-specific drill guides. J Hand Surg Am. 2013;38(12):2339e2347. 11. Farshad M, Hess F, Nagy L, Schweizer A. Corrective osteotomy of distal radial deformities: a new method of guided locking fixed screw positioning. J Hand Surg Eur Vol. 2013;38(1):29e34. 12. Palmer AK, Werner FW. The triangular fibrocartilage complex of the wrist—anatomy and function. J Hand Surg Am. 1981;6(2):153e162. 13. Murase T, Oka K, Moritomo H, Goto A, Yoshikawa H, Sugamoto K. Three-dimensional corrective osteotomy of malunited fractures of the upper extremity with use of a computer simulation system. J Bone Joint Surg Am. 2008;90(11):2375e2389.
REFERENCES 1. Jupiter JB, Fernandez DL, Levin LS, Wysocki RW. Reconstruction of posttraumatic disorders of the forearm. Instr Course Lect. 2010;59: 283e293.
J Hand Surg Am.
r
Vol. -, - 2017
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
20. Schweizer A, Furnstahl P, Nagy L. [Three-dimensional planing and correction of osteotomies in the forearm and the hand]. Ther Umsch. 2014;71(7):391e396 [in German]. 21. Schweizer A, Mauler F, Vlachopoulos L, Nagy L, Furnstahl P. Computer-assisted 3-dimensional reconstructions of scaphoid fractures and nonunions with and without the use of patient-specific guides: early clinical outcomes and postoperative assessments of reconstruction accuracy. J Hand Surg Am. 2016;41(1):59e69. 22. Nakahara H, Bruder SP, Haynesworth SE, et al. Bone and cartilage formation in diffusion chambers by subcultured cells derived from the periosteum. Bone. 1990;11(3):181e188. 23. Trousdale RT, Linscheid RL. Operative treatment of malunited fractures of the forearm. J Bone Joint Surg Am. 1995;77(6):894e902. 24. Price CT, Knapp DR. Osteotomy for malunited forearm shaft fractures in children. J Pediatr Orthop. 2006;26(2):193e196. 25. van Geenen RC, Besselaar PP. Outcome after corrective osteotomy for malunited fractures of the forearm sustained in childhood. J Bone Joint Surg Br. 2007;89(2):236e239.
14. Chen Y, Medioni G. Object modeling by registration of multiple range images. Image Vision Comput. 1992;10(3):145e155. 15. Fürnstahl P, Schweizer A, Graf M, et al. Surgical treatment of longbone deformities: 3D preoperative planning and patient-specific instrumentation. In: Zheng G, Li S, eds. Computational Radiology for Orthopaedic Interventions. Cham, Switzerland: Springer International Publishing; 2016:123e149. 16. Nagy L, Jankauskas L, Dumont CE. Correction of forearm malunion guided by the preoperative complaint. Clin Orthop Relat Res. 2008;466(6):1419e1428. 17. Bindra RR, Cole RJ, Yamaguchi K, et al. Quantification of the radial torsion angle with computerized tomography in cadaver specimens. J Bone Joint Surg Am. 1997;79(6):833e837. 18. Meyer DC, Siebenrock KA, Schiele B, Gerber C. A new methodology for the planning of single-cut corrective osteotomies of mal-aligned long bones. Clin Biomech (Bristol, Avon). 2005;20(2): 223e227. 19. Lluch A. The Sauve-Kapandji procedure. J Wrist Surg. 2013;2(1):33e40.
J Hand Surg Am.
9
r
Vol. -, - 2017
Conventional Versus Computer-Assisted Corrective Osteotomy of the Forearm: a Retrospective Analysis of 56 Consecutive Cases David Ephraim Bauer, MD,* Stefan Zimmermann, MD,* Alexander Aichmair, MD,* Andreas Hingsammer, MD,* Andreas Schweizer, MD,* Ladislav Nagy, MD,* Philipp Fürnstahl, PhD†
Purpose Accuracy and feasibility of corrective osteotomies using 3-dimensional planning tools and patient-specific instrumentation has been reported by multiple authors with promising results. However, studies describing clinical outcomes following these procedures are rare. Therefore, the purpose of this study was to compare the results of computer-assisted corrective osteotomies of the diaphyseal and distal radius with a conventional nonecomputer-assisted technique regarding duration of surgery, consolidation of the osteotomy, and complications. Also, subjective and objective clinical outcome parameters were assessed. Methods We retrospectively compared the results of 31 patients who underwent a corrective osteotomy performed conventionally with 25 patients treated with a computer-assisted method (CA) using patient-specific instrumentation. Baseline data were similar among both groups. The duration of surgery, bony consolidation, complications, gain in range of motion, and subjective outcome were recorded. Results The mean operating time was significantly shorter in the CA group compared with the conventional group. After 12 weeks, significantly more osteotomies were considered healed in the CA group compared with the conventional group. Two patients in the CA group required revision surgery to treat nonunion of the osteotomy. Otherwise clinical results were similar among both groups. Conclusions The results demonstrate that the computer-assisted method facilitates shorter operation times while providing similar clinical results. (J Hand Surg Am. 2017;-(-):-e-. Copyright Ó 2017 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Therapeutic IV. Key words Malunion, distal radius, patient specific instrumentation, rapid prototyping.
M
forearm fractures may lead to decreased range of motion, reduced grip strength, or painful instability of the distal radioulnar joint (DRUJ).1,2 ALUNION OF DISTAL RADIUS OR
Several authors have reported that adequate realignment and accurate restoration of the normal anatomy is associated with an overall positive clinical outcome.3e5 However, quantification of a complex
From the *Orthopedic Department; and the †Computer Assisted Research and Development Team, Balgrist University Hospital, University of Zurich, Zurich, Switzerland.
Corresponding author: David Ephraim Bauer, MD, Orthopedic Department, University Zurich Balgrist, Forchstrasse 340, Zurich 8008, Switzerland; e-mail: [email protected].
Received for publication May 18, 2016; accepted in revised form March 20, 2017.
0363-5023/17/---0001$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2017.03.024
A.S., L.N., and P.F. are shareholders of a company offering services for computer-assisted preoperative planning (Balgrist CARD AG, Zurich, Switzerland).
Ó 2017 ASSH
r
Published by Elsevier, Inc. All rights reserved.
r
1
2
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
DRUJ (n ¼ 27, 15 patients in the conventional group, 12 in the CA group), reduced range of motion (ROM) (n ¼ 27, 15 patients in the conventional group vs 12 in the CA group), ulnocarpal impaction syndrome (n ¼ 11, 4 in patients in the conventional group vs 7 in the CA group) after nonoperatively or operatively treated extra-articular fracture of the distal radius (n ¼ 40), diaphyseal radius fractures (n ¼ 4), or both bone forearm fractures (n ¼ 12). Five fractures in the CA group and 4 fractures in the conventional group had previously been treated with open reduction and internal fixation. Regarding the preoperative ROM, there were 8 patients with decreased pro-/supination and 2 patients with decreased flexion/extension alone in the conventional group and 7 patients with decreased pro-/ supination and 3 patients with decreased flexion/ extension alone in the CA group. A combination of decreased flexion/extension and pro-/supination was recorded in 6 patients in the conventional group and 3 patients in the CA group (Table 1). The analysis of clinical outcome parameters was performed after stratifying patients into the above-mentioned 3 groups of preoperative complaints. Clinical assessment was performed by the surgeon and was therefore not blinded. Range of motion was assessed with a standard goniometer. We calculated the increase in ROM after surgery by subtracting the postoperative from the preoperative arc of ROM. A distinct cutoff value for reduced ROM, as an indication for surgery, was not set. The indication for operation was based on the patient’s complaint. Instability was diagnosed by a clinical examination of increased translation of the distal ulna relative to the distal radius compared with the uninjured side. The indication for surgical intervention was pain on loading the unstable joint, provoking the instability pattern. Ulnar-sided wrist pain, exacerbated by ulnocarpal loading in combination with positive ulnar variance, was considered diagnostic for ulnocarpal impaction syndrome. The subjective outcomes of patients operated because of painful instability of the DRUJ were classified as (1) persisting pain, (2) improved but remaining symptoms, and (3) complete resolution, respectively. In addition to assessing self-reported parameters, a retrospective chart review was performed to determine hospital stay, duration of surgery, and bony consolidation. Regarding the duration of surgery, the tourniquet time was considered as an adequate surrogate.
multiplanar malunion and subsequent determination of the required correction by means of conventional radiographs or computed tomography (CT) images remain challenging.6e8 Recent advancements in computer-assisted 3-dimensional (3D) planning and 3D printing technology can support the surgeon in enabling a more accurate preoperative simulation combined with easy-to-use navigation aids so that the surgery can be performed as preoperatively planned. In the context of a malunion of the radius, computer-assisted corrective osteotomy has been described as a promising technique for accurate reconstruction if patientspecific guides are used.9,10 Although the accuracy of the reduction has been demonstrated, studies investigating clinical outcome parameters following these procedures are rare.11 MATERIALS AND METHODS The aim of this study was to compare the results of computer-assisted corrective osteotomies of the diaphyseal and distal radius using 3D planning and patient-specific guides with a conventional, nonecomputer-assisted technique. We evaluated the duration of surgery, healing of the osteotomy, and complications. In addition, we assessed subjective and objective clinical outcomes. Patient sample After institutional review board approval was obtained, a convenience sample of 56 patients with symptomatic, posttraumatic deformities of the forearm treated with either a nonecomputer-assisted method (conventional group) or a computer-assisted corrective osteotomy9 (CA group) was available for final analysis. The conventional group was treated between January 2003 and October 2008 and the CA group was treated between December 2009 and April 2015. Patients with congenital deformities of the forearm and patients who underwent concomitant interventions during the index procedure (scapholunate ligament reconstruction or repair of the triangular fibrocartilage complex12) were excluded from final analysis. Thirty-one patients in the conventional group (15 males and 16 females; mean age, 31 y; range, 10e66 y) and 25 patients in the CA group (17 males and 8 females; mean age, 28 y; range, 11e71 y) were finally included. Preoperative clinical and radiological assessment Indications for surgical intervention were one or more of the following: painful instability of the J Hand Surg Am.
r
Vol. -, - 2017
3
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
TABLE 1.
Baseline Data Conventional (n ¼ 31) Variable
N
Mean or %
CA (n ¼ 25) SD
n
Mean or %
36
25
133
35
24
140
12.6
18
26.4
SD
Preoperative clinical function Pro-/supination ( )
31
128
Flexion/extension ( )
30
124
Grip strength (kg)
20
26.1
21
67.7%
18
72.0%
Both bone forearm fracture
8
25.8%
5
20.0%
Radius shaft fracture
2
6.5%
2
8.0%
38 40 12.6
Preoperative radiological findings Distal radius fracture
kg, kilogram; SD, standard deviation.
calculating the optimal fit between the fragments proximal to the fracture line using the iterative closest point surface registration functionality14 of the CASPA software. The osteotomy was then simulated on the pathological bone model, resulting in a proximal and distal fragment. The reduction was achieved by aligning the distal fragment to the reconstruction template using the iterative closest point method. The relative transformation of the distal fragment from initial to reduced position quantified the amount of required reduction in all 6 degrees of freedom (ie, 3 values each for rotation and translation, respectively). Based on the calculated reduction, the patient-specific guides were designed with the CASPA software. Detailed descriptions of the applied planning and guide design techniques can be found in previously published studies.15 The patient-specific guides were manufactured based on the preoperative simulation by selective laser-sintering using biocompatible polyamide PA 2200 (Medacta SA, Castel San Pietro, Switzerland). Preoperatively, the guides were verified using a true-size replica of the injured and reduced bone and sterilized with conventional steam sterilization. A volar approach to the radius was performed in all cases. Closing or opening wedge osteotomies were performed if an angular deformity was predominant and single-cut osteotomies were performed if a rotational deformity was predominant. All the osteotomies performed in the CA group were planned in a way such that angular and rotational deformities are simultaneously corrected. An opening wedge osteotomy was performed in 13 cases, a closing wedge osteotomy in 2 cases, and a 3D single-cut osteotomy in 10 cases.
Radiographic analysis of consolidation of the osteotomy was performed by a board-certified radiologist using CT scans or plain radiographs (28 plain radiographs vs 3 CT scans in the conventional group compared with 13 plain radiographs vs 12 CT scans in the CA group). Patients were routinely scheduled for clinical and radiological examinations every 6 weeks after the index procedure until the osteotomy was considered healed. Consolidation was assumed when the radiolucent osteotomy line was no longer visible on CT scans or plain x-rays. Computer-assisted osteotomy Both the injured forearm and the contralateral uninjured limb of all 25 patients who underwent the computer-assisted method were evaluated preoperatively based on bilateral CT scans (slice thickness, 1 mm; 120 kV; CT device: Philips Brilliance 40; Philips Healthcare, Best, the Netherlands). The CT data were segmented using a commercially available segmentation software package (Materialise, Loewen, Belgium) in a semiautomatic fashion by applying thresholding and region-growing algorithms to obtain 3D surface models of both the injured and the contralateral bone. The 3D models were imported into the planning software CASPA (Balgrist CARD AG, Zurich, Switzerland) to perform the 3D preoperative planning according to a technique described by Murase et al13 and further developed by Schweizer et al8 (Fig. 1). In brief, the method can be summarized as follows: the 3D model of the malunited bone was divided into a region proximal and distal to the previous fracture zone. The 3D bone model of the injured side was then superimposed with the mirror model of the contralateral uninjured side by J Hand Surg Am.
r
Vol. -, - 2017
4
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
FIGURE 1: Types of 3D planned osteotomies. The rotation axis (r,C) and the osteotomy plane(s) are shown in red and gray, respectively. The pathological (orange) and contralateral bone (green) are given before (first column) and after (second column) simulated reduction. The reduced bone fragment is depicted in violet. A Distal radius opening wedge osteotomy. B Closing wedge osteotomy of the radius. C Single cut osteotomy of the tibia. (Reprinted with permission from Fürnstahl P, Schweizer A, Graf M, et al. Surgical treatment of long-bone deformities: 3D preoperative planning and patient-specific instrumentation. In: Zheng G, Li S, eds. Computational Radiology for Orthopaedic Interventions. Cham, Switzerland: Springer International Publishing; 2016:123e149.15)
If an opening wedge osteotomy was performed, cancellous bone autograft harvested from the iliac crest was interposed between the osteotomy fragments.
to the true angle of deformity and the corresponding orientation in space. A trigonometric calculation of the deformity allowed for prebending and fixation of the plate to either the proximal or distal fragment before performing the osteotomy to control correct reduction. An opening wedge osteotomy was performed in 23 cases, a closing wedge osteotomy in 7 cases, and a transverse osteotomy, for simultaneous correction of combined angular and rotational deformity, in 1 case. Cancellous bone autograft harvested from the iliac crest was interposed between the osteotomy fragments if an opening wedge osteotomy was performed. The correction was maintained with a conventional 3.5-mm plate (Synthes, Westchester, PA) for shaft osteotomies or a 2.4/ 2.7 mm distal Radius Locking compression plate (Synthes). Intraoperative radiographs were obtained to verify proper spatial reconstruction and implant placement.18
Conventional osteotomy Preoperative planning was performed using anteroposterior and lateral plain radiographs of both forearms. The contours of both the uninjured and the deformed bone were copied on separate sheets of tracing paper. The location and extent of the maximal deformity were evaluated by superposition of the tracing paper sheets.16 The true 3D angle of deformity and its corresponding orientation in space were determined as described by Nagy et al.16 Crosssectional CT scans were obtained only if a rotational deformity was suspected. To assess rotational deformity using CT slices, the anatomic relationship of the radial styloid process and the bicipital tuberosity was evaluated according to the method described by Bindra et al.17 A standard volar approach to the radius was performed in all cases. Intraoperatively, Kirschner wires were placed under fluoroscopic guidance in different planes to mark the orientation and extent of malposition of the proximal and distal fragment. After performing the osteotomy, proper anatomic reduction was achieved by aligning the previously placed Kirschner wires in a parallel fashion. The osteotomy was performed with respect J Hand Surg Am.
Statistical analysis Data are presented as mean and standard deviation for continuous variables and as proportion (%) for categorical variables, if not stated otherwise. A 2tailed Kolmogorov-Smirnov test was used for testing a normal distribution; if P .05, the data were considered as normally distributed. The chisquared test was applied for testing differences of distribution of categorical variables between groups r
Vol. -, - 2017
5
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
TABLE 2.
Surgical Details Conventional (n ¼ 31)
Variable
CA (n ¼ 25)
n
%
n
%
P Value
Distal radius
21
67.7%
18
72.0%
n.a.
Midshaft
10
30.3%
7
28.0%
Opening wedge
23
74.2%
13
52.0%
Closing wedge
7
22.6%
2
8.0%
Single cut
1
3.2%
10
40.0%
Localization of osteotomy
Type of osteotomy <.05
n.a., not applicable.
as appropriate. For testing differences between means, the unpaired t test was applied for normally distributed data and the Mann-Whitney U test for nonnormally distributed data. A P value of < .05 was considered statistically significant.
the CA group, 2 patients underwent revision surgery because of nonunion of the osteotomy, 1 single cut and 1 open wedge osteotomy, respectively. For the initial open wedge osteotomy, nanocrystalline hydroxyapatite embedded in a silica gel matrix (NanonBone, ARTOSS, Rostock, Germany) was used as a bone substitution graft. Iliac crest autograft was used for revision surgery in both cases. Considering cases not undergoing revision surgery, all osteotomies except for one in the conventional group were consolidated within 36 weeks after the index procedure (Fig. 2). The analysis of clinical outcome parameters was performed after stratifying patients into different groups based on the preoperative complaint. The average postoperative gain of ROM in patients in whom the arc of pro-/supination was the indication for surgical intervention was 41 39 in the conventional group compared with 44 27 in the CA group. The average postoperative gain of ROM in patients in whom amplitude of flexion/extension was the indication for operative intervention was 28 42 in the conventional group compared with 25 36 in the CA group. The results of subjective outcome parameters in patients in whom painful instability of the DRUJ was the primary complaint are illustrated in Table 3. Sixteen patients in the conventional group (51.6%) and 12 patients in the CA group (48.0%) underwent implant removal. The indications for implant removal were either prophylactic to prevent a chronic contact between the plate and the flexor tendons, discomfort caused by the implant, or reduced ROM attributed to the implant. The average gain of pro-/ supination after implant removal was 1 34 in the conventional group compared with 10 15 in the CA group.
RESULTS The median time from the initial trauma until corrective osteotomy was 18.0 months (interquartile range, 100.5 mo) in the conventional group compared with 28.0 months (interquartile range, 33.0 mo) in the CA group. The average follow-up was significantly longer in the conventional group compared with the CA group (25.4 20.1 mo vs 13.6 7.4 mo, P < .05). Preoperative clinical assessment Baseline data were similar amongst both groups as illustrated in Table 2. The average arc of flexion/ extension in the group of patients in whom reduced flexion/extension was the indication for surgery was 96 29 in the conventional group compared with 83 35 in the CA group. The average arc of pro-/supination in the group of patients in whom reduced pro-/supination was the indication for surgery was 105 33 in the conventional group compared with 96.0 32 in the CA group. Surgical intervention and outcome The mean operating time was significantly shorter in the CA group compared with the conventional group (140 37 vs 108 26 min, P < .05). Twelve weeks after the index procedure significantly more cases in the CA group were considered healed compared with the conventional group (14 cases [60.9%] vs 10 cases [32.3%], P < .05). However, in J Hand Surg Am.
r
Vol. -, - 2017
6
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
This study has several limitations. Because of the retrospective design, we were not able to consistently record potentially meaningful parameters such as preoperative ROM of the contralateral uninjured side, pain measured by the visual analog scale, or smoking status if the required documentation was missing. As of 2009, we changed the surgical technique in our institution from the conventional to the computerassisted method. This may be one factor contributing to the shorter follow-up time in the CA group compared with the historic conventional group. As a result of the shorter follow-up time, this study may have failed to demonstrate adverse long-term outcomes in the CA group. Further, the size of the study sample of each subgroup is small, thereby reducing the power. Therefore, statistical comparisons were not possible for many of the parameters of interest. In addition, clinical evaluations of DRUJ instability and ulnar abutment were performed by the surgeon in a nonblinded fashion potentially introducing the risk of a biased outcome. As opposed to the CA group, preoperative planning in the conventional group was performed using plain radiographs of both forearms. Hence, correct execution of the preoperative plan was also verified using plain radiographs, and patients in the conventional group therefore inconsistently received postoperative CT scans. The duration of surgery among cases operated using the computer-assisted method was significantly shorter compared with the conventional method. This may be attributable to the preoperative simulation and the advantage of having the patientspecific instrument as a reliable reference. Therefore, the previously described technique of placing Kirschner wires under fluoroscopic guidance to mark the orientation and extent of malposition of the proximal and distal fragments in conventional osteotomies is not required thereby saving operation and fluoroscopy time. The authors’ experience is also that the guides enable the surgeon to perform the surgery in a more standardized fashion compared with the nonecomputer-assisted method. From an economic standpoint, manufacturing of the patient-specific instruments entails additional time and costs, and requires a CT scan, including the contralateral side, for planning of the correction. In our institution, the CT scan required for planning of the osteotomy includes both forearms and therefore does not incur additional costs for scanning the contralateral limb. Depending on the complexity of the osteotomy, planning of the correction requires 2 to 4 hours. In this case series, production of the laser sintered
FIGURE 2: Time to consolidation of the osteotomy. Significant difference in consolidation of the osteotomy site after 12 weeks. *Indicates statistical significance (P < .05).
Complications One patient in the conventional group experienced rupture of the extensor pollicis longus tendon 8 weeks after the index procedure, 1 patient required an additional corrective osteotomy because of persisting ulnar impaction, 1 patient a repair of the triangular fibrocartilage complex because of persisting instability of the DRUJ, and 2 patients underwent a Sauvé-Kapandji salvage procedure19 because of persisting ulnar-sided wrist pain. In both cases, a DRUJ-preserving procedure was attempted even though signs of DRUJ arthritis were apparent. One patient experienced a transient palsy of the ulnar nerve with spontaneous, complete resolution within 1 year after surgery. In the CA group, 2 patients required an additional ulna shortening osteotomy because of persisting ulnocarpal impaction syndrome. In 1 case, the initial correction was lost after a prolonged time to bony consolidation, and in the second case, correction could not be executed as preoperatively planned. DISCUSSION Three-dimensional preoperative planning and the use of patient-specific instrumentation have been described as a promising technique for various indications in hand surgery. Prior studies demonstrated that the use of custom-designed osteotomy guides enables a more accurate reduction compared with free hand techniques.9,13,20,21 In this study, we described our experience with 56 patients who underwent a corrective osteotomy, in 25 of whom, it was performed via a computer-assisted approach. J Hand Surg Am.
r
Vol. -, - 2017
7
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
TABLE 3.
Postoperative Outcome Data Conventional (n ¼ 31) Variable
CA (n ¼ 25)
n
Mean
SD
n
Mean
SD
14
41
39
10
43
27
8
28
41
6
25
36
Clinical outcome Gain pro-/supination ( )* †
Gain flexion/extension ( ) Gain grip strength (kg) Stability: asymptomatic Stability: improved‡ ‡
Stability: persisting
‡
18
3.1
9 4 2
9.8
17
4.1
60.0%
7
58.3%
26.7%
4
33.3%
13.3%
1
8.3%
7.7
Radiological outcome Consolidation (wk)
31
23.9
13.3
25
20.9
16.6
kg, kilogram; SD, standard deviation. *In the reduced pro-/supination group only. †In the reduced flexion/extension group only. ‡In the instability group only.
intraoperative guides was performed by a third-party company and required approximately 4 weeks. For the cases presented in this study, the additional costs including CT scan, and time required for planning and production of the guide, amount to $2,415 USD. Two patients in the CA group required revision surgery because of nonunion of the osteotomy (1 diaphyseal, 1 metaphyseal). Given the fact that achieving perfect fit of the patient-specific cutting guides is paramount for the correct and precise execution of the preoperative plan, extensive stripping of the periosteum is often necessary. However, stripping of the periosteum, a primary source of osteogenic stem cells giving rise to callus formation and promoting the initial phase of bone healing, as well as decreased blood supply to the bone, may therefore be considered as a contributing factor to these complications.22 However, significantly more osteotomies in the CA group were considered healed after 12 weeks compared with the conventional group. This might be accounted to the lower number of patients requiring open wedge osteotomies or to the fact that osteotomy union was evaluated with a CT scan more often in the CA group. Although not specifically analyzed, more rapid union might facilitate quicker rehabilitation and return to work. Two patients in the conventional group underwent palliative Sauvé-Kapandji procedures because of secondary osteoarthritis of the DRUJ. In both cases, the indication for corrective osteotomy was painful subluxation of the DRUJ because of posttraumatic deformity after a distal radius fracture. Both patients were young manual workers, and an attempt to J Hand Surg Am.
perform a joint-preserving procedure was undertaken although signs of osteoarthritis of the DRUJ were already apparent. Similarly, 2 patients in the CA group required additional ulnar shortening. In both cases, the indication for the corrective osteotomy was reduced ROM. One patient presented with exacerbated ulnar abutment after the index procedure, and the other patient experienced loss of the planned reduction with a prolonged time to consolidation potentially affected by a history of heavy smoking. In our study, the mean arc of pro-/supination improved from 105 to 147 in the conventional group and from 95 to 139 in the CA group. Previously published studies on conventional and computer-assisted correction of malunion of the radius report slightly worse preoperative forearm motion compared with our study cohort. Also, in our study, the gain of pro-/supination in both intervention groups was less than reported in the literature for conventional osteotomies.13,23e25 Similar results have been observed in one of our previously published cohorts describing the conventional technique.16 The smaller reported gain in ROM may be explained by the higher patient age and the longer delay until the corrective osteotomy was performed in our cohort.23,25 However, as a result of the small number in these subgroups, we were not able to detect a significant difference considering these parameters. In summary, computer-assisted corrective osteotomy of the radius enables shorter operation times by providing a reliable intraoperative reference for the planned correction thereby replacing previously r
Vol. -, - 2017
8
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
FIGURE 3: Single-cut multiplanar correction. A Injured (orange) compared with uninjured contralateral side (green). B Planning of the correction plane and rotation axis. C Correction plane with cutting guide. D After reduction with reduction guide. E Cutting guide in situ. F Reduction with guide in situ.
required fluoroscopic referencing steps. Considering clinical outcome parameters, further studies with larger cohorts are warranted, to evaluate whether a more accurate reduction yields a better clinical outcome. However, the computer-assisted method may be particularly helpful for complex, rotational deformities, which are difficult to plan and implement using the conventional technique. In particular, a multiplanar correction by rotating around a single axis can only be achieved by 3D analysis and preoperative simulation of the correction (Fig. 3). From an economic standpoint, the total direct costs for segmentation, 3D analysis, and planning and production of the osteotomy guides per case amount to $2,415 USD. These additional expenses should be compared against potential financial benefits including reduced operation time, quicker rehabilitation, and return to work. However, large-scale costeffectiveness analyses of the CA method require prospective documentation of outcome measures, such as health-related quality of life and qualityadjusted life years, in combination with the assessment of both direct and indirect costs. A long-term study reporting these parameters is currently being planned.
2. Schemitsch EH, Richards RR. The effect of malunion on functional outcome after plate fixation of fractures of both bones of the forearm in adults. J Bone Joint Surg Am. 1992;74(7):1068e1078. 3. McQueen M, Caspers J. Colles fracture: does the anatomical result affect the final function? J Bone Joint Surg Br. 1988;70(4):649e651. 4. Fernandez DL. Malunion of the distal radius: current approach to management. Instr Course Lect. 1993;42:99e113. 5. Matthews LS, Kaufer H, Garver DF, Sonstegard DA. The effect on supination-pronation of angular malalignment of fractures of both bones of the forearm. J Bone Joint Surg Am. 1982;64(1):14e17. 6. Dumont CE, Nagy L, Ziegler D, Pfirrmann CW. Fluoroscopic and magnetic resonance cross-sectional imaging assessments of radial and ulnar torsion profiles in volunteers. J Hand Surg Am. 2007;32(4): 501e509. 7. Dumont CE, Pfirrmann CW, Ziegler D, Nagy L. Assessment of radial and ulnar torsion profiles with cross-sectional magnetic resonance imaging. A study of volunteers. J Bone Joint Surg Am. 2006;88(7): 1582e1588. 8. Schweizer A, Furnstahl P, Harders M, Szekely G, Nagy L. Complex radius shaft malunion: osteotomy with computer-assisted planning. Hand (NY). 2010;5(2):171e178. 9. Vlachopoulos L, Schweizer A, Graf M, Nagy L, Furnstahl P. Threedimensional postoperative accuracy of extra-articular forearm osteotomies using CT-scan based patient-specific surgical guides. BMC Musculoskelet Disord. 2015;16(1):336. 10. Schweizer A, Furnstahl P, Nagy L. Three-dimensional correction of distal radius intra-articular malunions using patient-specific drill guides. J Hand Surg Am. 2013;38(12):2339e2347. 11. Farshad M, Hess F, Nagy L, Schweizer A. Corrective osteotomy of distal radial deformities: a new method of guided locking fixed screw positioning. J Hand Surg Eur Vol. 2013;38(1):29e34. 12. Palmer AK, Werner FW. The triangular fibrocartilage complex of the wrist—anatomy and function. J Hand Surg Am. 1981;6(2):153e162. 13. Murase T, Oka K, Moritomo H, Goto A, Yoshikawa H, Sugamoto K. Three-dimensional corrective osteotomy of malunited fractures of the upper extremity with use of a computer simulation system. J Bone Joint Surg Am. 2008;90(11):2375e2389.
REFERENCES 1. Jupiter JB, Fernandez DL, Levin LS, Wysocki RW. Reconstruction of posttraumatic disorders of the forearm. Instr Course Lect. 2010;59: 283e293.
J Hand Surg Am.
r
Vol. -, - 2017
THREE-DIMENSIONAL GUIDED RADIUS RECONSTRUCTION
20. Schweizer A, Furnstahl P, Nagy L. [Three-dimensional planing and correction of osteotomies in the forearm and the hand]. Ther Umsch. 2014;71(7):391e396 [in German]. 21. Schweizer A, Mauler F, Vlachopoulos L, Nagy L, Furnstahl P. Computer-assisted 3-dimensional reconstructions of scaphoid fractures and nonunions with and without the use of patient-specific guides: early clinical outcomes and postoperative assessments of reconstruction accuracy. J Hand Surg Am. 2016;41(1):59e69. 22. Nakahara H, Bruder SP, Haynesworth SE, et al. Bone and cartilage formation in diffusion chambers by subcultured cells derived from the periosteum. Bone. 1990;11(3):181e188. 23. Trousdale RT, Linscheid RL. Operative treatment of malunited fractures of the forearm. J Bone Joint Surg Am. 1995;77(6):894e902. 24. Price CT, Knapp DR. Osteotomy for malunited forearm shaft fractures in children. J Pediatr Orthop. 2006;26(2):193e196. 25. van Geenen RC, Besselaar PP. Outcome after corrective osteotomy for malunited fractures of the forearm sustained in childhood. J Bone Joint Surg Br. 2007;89(2):236e239.
14. Chen Y, Medioni G. Object modeling by registration of multiple range images. Image Vision Comput. 1992;10(3):145e155. 15. Fürnstahl P, Schweizer A, Graf M, et al. Surgical treatment of longbone deformities: 3D preoperative planning and patient-specific instrumentation. In: Zheng G, Li S, eds. Computational Radiology for Orthopaedic Interventions. Cham, Switzerland: Springer International Publishing; 2016:123e149. 16. Nagy L, Jankauskas L, Dumont CE. Correction of forearm malunion guided by the preoperative complaint. Clin Orthop Relat Res. 2008;466(6):1419e1428. 17. Bindra RR, Cole RJ, Yamaguchi K, et al. Quantification of the radial torsion angle with computerized tomography in cadaver specimens. J Bone Joint Surg Am. 1997;79(6):833e837. 18. Meyer DC, Siebenrock KA, Schiele B, Gerber C. A new methodology for the planning of single-cut corrective osteotomies of mal-aligned long bones. Clin Biomech (Bristol, Avon). 2005;20(2): 223e227. 19. Lluch A. The Sauve-Kapandji procedure. J Wrist Surg. 2013;2(1):33e40.
J Hand Surg Am.
9
r
Vol. -, - 2017