Full Length Article

Journal of Hand Surgery

(European Volume)
Robot-assisted double screw fixation 0(0) 1–6
! The Author(s) 2020
of minimally displaced scaphoid Article reuse guidelines:
sagepub.com/journals-permissions
waist fracture nonunions or delayed DOI: 10.1177/1753193420944546
journals.sagepub.com/home/jhs

unions without bone graft

Yang Guo1, Wei Ma2, Dedi Tong1, Kun Liu1, Yaobin Yin1 and
Chen Yang1

Abstract
We retrospectively reviewed 12 minimally displaced fractures of the scaphoid waist in 12 patients who
developed delayed or nonunions with or without conservative treatment. Mean time between injury and sur-
gery was 6 months (range 3–12). The fractures were stabilized with double screws, which were percutan-
eously inserted with robot assistance, and without bone grafting. All fractures united at a mean of 8 weeks
(range 6–10) after surgery. The patients were followed-up at 6 months and 1 year. The patients recovered
good wrist function. No major postoperative complications were reported, and the patients returned to their
usual level of activity. Robot assistance gave a high degree of accuracy when placing the cannulated screws
since only two attempts were needed for correct placement of the guide wires. We explain the high union
incidence by patient selection, good stabilization and not disturbing the vascular supply.
Level of evidence: IV

Keywords
Scaphoid fractures, delayed union, nonunion, percutaneous fixation, robotic surgical procedures

Date received: 29th March 2020; revised: 9th June 2020; accepted: 4th July 2020

Introduction Tian et al., 2017; Wang et al., 2017). The use of two
Open reduction and internal fixation with bone graft- screws inserted with robotic assistance has been
ing is considered the reference treatment of scaphoid proven to be safe and accurate for acute non-
nonunions and delayed unions (Chim et al., 2011; displaced scaphoid fractures (Liu et al., 2019).
Merrell et al., 2002; Radford et al., 1990). Recently, The purpose of this study was to evaluate the out-
however, percutaneous screw fixation without bone comes of percutaneous double screw fixation of
grafting has been proposed for stable and non- delayed unions or nonunions of scaphoid waist frac-
displaced cases (Capo et al., 2012; Hegazy, 2015). tures using robot assistance and without bone
The advantages of this minimal invasive technique grafting.
include preservation of the vascularity and minimal
scar formation (Slade and Dodds, 2006). Optimal
screw placement is an essential part of the percu-
taneous fixation technique (Luchetti et al., 2018; Suh 1
Department of Hand Surgery, Beijing Jishuitan Hospital, Beijing,
and Grewal, 2018), but it is challenging even for China
2
experienced hand surgeons (Adams et al., 1988; Department of Orthopedics, Air Force Special Medical Center,
Beijing, China
Dias et al., 2005; Suh and Grewal, 2018; Walsh
et al., 2009). In recent years, robotic-assisted screw Corresponding Author:
Yang Guo, Department of Hand Surgery, Beijing Jishuitan Hospital,
insertion has been used in upper extremity and spinal The Fourth Clinical College of Peking University, 31st Xinjiekou
operations to reduce surgery time, improve accuracy East Street, Western District, Beijing 100035, China.
and reduce costs after initial setup (Le et al., 2018; Email: [email protected]
2 Journal of Hand Surgery (Eur) 0(0)

Methods The inner surface of this model matched with the

surface of the hand, wrist and forearm. Then we cre-
Patients ated a template with a 3-D printer (3D Artisan,
We retrospectively reviewed 12 consecutive patients Beijing, China) in which the hand and a positioning
with delayed unions or nonunions of the scaphoid in jig could be placed (Figure 1). The template was
12 patients treated between January 2018 and ready to use after a 30-minute hydrogen peroxide
January 2019 with double screw fixation and without sterilization process (Guo et al. 2019).
bone graft. The study was approved by the local eth-
ical committee. We defined delayed union as no union
3 months after injury and nonunion as no union 6
Operative technique
months after injury in plain X-ray films (Inaparthy The operation was performed under brachial plexus
and Nicholl, 2008; Mack et al., 1998). We included block with the patient in a supine position. The hand
only minimally displaced scaphoid waist fractures was placed in the template with the wrist maximally
without substantial cystic degeneration or sclerosis, extended. The positioning jig was fixed to a sterilized
namely Grade I, II and III according to the classifica- base with Shantz pins (Figure 2).
tion of Slade et al. (2003). These fractures were man- TiRobot, a commercially available robotic naviga-
aged initially without treatment or with conservative tion system (TINAVI Medical Technologies, Beijing,
treatment. Exclusion criteria were humpback
deformity, dorsal intercalated segmental instability
(DISI) deformity, avascular necrosis of the proximal
pole and treatment delay more than 1 year.

Radiographic examinations
Preoperatively, computed tomographic (CT) scans of
the scaphoid were obtained with the wrist maximally
extended to evaluate the fracture site and to generate
data for jig positioning. Postoperatively radiographs
were obtained at 6 weeks and then every 2 weeks
until radiographic union was noted. CT scans were
obtained to finally confirm consolidation at the frac-
ture site.

Outcome measures
Figure 1. A three-dimensional printed positioning jig
All outcome measures were obtained preoperatively derived from CT data.
and postoperatively. Wrist pain was assessed using a
visual analogue scale. Grip strength was measured
with a Jamar dynamometer (Distributor: Shanghai
RuiShi Biological Technology Co, Shanghai, China)
(Manufacturer: SI Instruments, Hilton, SA 5033,
Australia), and an independent hand therapist mea-
sured range of motion with a goniometer. The results
are presented as percentages of the contralateral
wrist. Wrist function was assessed by the modified
Mayo wrist score (Cooney et al., 1994).

Planning of the positioning jig

The Digital Imaging and Communications in Medicine
data of the CT scans were imported into a surgical
planning platform (Zhouge Vxwork, Beijing, China).
The software was adapted to the soft tissues of the Figure 2. The jig is fixed to a base by Shantz pins, and
wrist. After calculation and Boolean subtraction, we the patient marker is fixed onto the three-dimensional
obtained a three-dimensional (3-D) surface model. printed jig.
Guo et al. 3

China) was used to insert the scaphoid screws. Radiographic anteroposterior (AP), lateral and
The TiRobot consists of a surgical planning worksta- semi-supinated oblique images of the wrist were
tion, a stereotactic robotic arm and an optical track- obtained by a C-arm and transmitted to the worksta-
ing device (Figure 3). The optical tracking system has tion. The surgeon selected the entry and exit points
an infrared stereo camera and two reference mar- for the screw trajectories on the AP and lateral
kers: the patient marker and the robotic marker. The images. These trajectories were confirmed on the
patient marker was fixed onto the 3-D printed jig. The oblique view. One trajectory was planned just
robotic marker was attached to the robotic arm. medial to the central axis of the scaphoid and 2 mm
lateral to the medial cortex on the AP view. The other
trajectory was planned just lateral to the central axis.
Ideally the two trajectories were parallel in the AP
view and overlapped in the lateral view (Figure 4).
After the trajectories were accepted, the robot arm
with the sleeve attached to its distal end moved pre-
cisely to the trajectory positions under control from
the workstation. The surgeon inserted the guide wire
into the scaphoid through the sleeve attached to the
robot arm (Figure 5). The positions of the guide wires
were verified by the C-arm, and the length of the

Figure 3. The TiRobot consists of a surgical planning

workstation, a stereotactic robotic arm and an optical Figure 5. The guide wire is inserted into the scaphoid
tracking device. through the sleeve attached to the robot arm.

Figure 4. Trajectory selection steps on the workstation screen showing virtual guide wires on anteroposterior, supinated
oblique and lateral views of the wrist.
4 Journal of Hand Surgery (Eur) 0(0)

Figure 6. The length and position of the screws are confirmed in different views after we removed the positioning jig.

Table 1. Preoperative measurement and postoperative outcome measures 1 year after surgery.

Assessment Preoperative Postoperative p-value

Grip strength (% of contralateral) 63 (7) 92 (9) <0.01

Wrist extension ( ) 58 (13) 72 (4) <0.01
Wrist flexion ( ) 61 (11) 76 (3) <0.01
Modified Mayo wrist score 59 (6) 79 (4) <0.01
Pain VAS score 6 (1) 2 (1) <0.01
Data in this table are shown as mean (SD).
VAS: visual analogue scale.

screws to be inserted were determined by a program and surgery was 6 months (range 3–12). Five hands
in the robotic system. Two cannulated Mini and Micro were dominant. Two patients were classified as Slade
Acutrak screws (Acumed, Hillsboro, OR, USA) were Grade I, three as Grade II and seven as Grade III.
inserted into the scaphoid from distal to proximal. There were four delayed unions and eight nonunions.
After removal of the positioning jig, the length and The mean duration of surgery was 68 minutes (range
position of the screws were confirmed radiographic- 55–78), including a mean setup time of 15 minutes
ally (Figure 6). (range 12–18), a mean guide wire insertion time of
30 minutes (range 20–35) and a mean screw insertion
time of 20 minutes (range 15–30). Only two guide wire
Postoperative care and follow-up insertion attempts were needed in each patient. All
We did not immobilize the wrist, and the patients patients attended the follow-up, and all fractures
were immediately allowed to gently move the wrist achieved union at a mean of 8 weeks (range 6–12).
and to hold small objects, such as chopsticks and No perforation of the surface of the scaphoid at either
cups. When radiographic union was confirmed, daily the proximal pole or the tubercle was noted on the
activities and non-contact sports were resumed follow-up radiographs.
gradually. All patients were followed-up at 6 Preoperative measurement and final follow-up
months and 1 year after the surgery. outcome measures 1 year after surgery are listed
in Table 1. In general, the patients recovered good
wrist function. Grip strength improved by about
Results
30%, and wrist flexion–extension improved by about
There were 11 men and one a woman. Mean age was 20% compared with preoperative measurements.
29 years (range 16–48). Mean time between injury Pain decreased to minimal. No patient reported any
Guo et al. 5

major complications, and two patients complained of able to precisely drill the guide wire in the desired
minor discomfort at the palmar incisional scar at position. Several articles have demonstrated that the
final follow-up. All patients returned to their positioning accuracy of a guide wire is around 0.5 mm
normal activities. (Tian et al., 2017). The positioning jig prevents the
wrist from moving throughout the procedure. Our
results indicate a high degree of accuracy since
Discussion
only two attempts were needed for correct place-
Our method achieved union without using bone graft in ment of the wires. The robot-assisted method
12 minimally displaced scaphoid waist nonunions or increases the overall costs of the surgical procedure.
delayed unions that were less than 1 year old. In the but we hope that the costs will be compensated by
1980s several authors described the technique of higher accuracy and hence less expense resulting
screw or K-wire fixation of scaphoid nonunions with- from malpositioned screws.
out bone grafting (Cosio and Camp, 1986; Leyshon Patient selection is important. Although Mahmoud
et al, 1984; Parkinson et al., 1989). In the past 10 and Koptan (2011) claimed that some scaphoid non-
years, some authors have reported achieving union unions with gaps of more than 5 mm could be treated
in as much as 100% of the cases (Capo et al., 2012; with fixation without grafting, most hand surgeons
Hegazy, 2015; Mahmoud and Koptan., 2011; Saint-Cyr still believe that scaphoid nonunions with substantial
et al., 2011). They believed that well-aligned non- bone loss or sclerosis should be treated by open
unions without vascular disturbance only require reduction, bone grafting and internal fixation (Ernst
rigid fixation and that percutaneous reaming provided et al., 2018). Poor vascularity of the proximal pole
sufficiently debridement at the nonunion site. For a may be an important cause of poor success of the
scaphoid fracture, primary bone healing depends on fixation without grafting (Capo et al., 2012; Hegazy,
rigid stabilization of the fracture fragments. The treat- 2015; Somerson et al., 2015). Thus, the current,
ment rationale for fixation without grafting is that widely accepted indications for screw fixation without
instability is the main factor disturbing the healing grafting are delayed unions and nonunions without
process and that the stability provided by screws substantial bone loss, DISI, hump-back deformity
results in healing of minimally displaced fractures and necrosis of the proximal pole. We used CT
(Capo et al., 2009; Ernst et al., 2018; Slade et al., 2003). scans to guide us in selecting the patients, although
A recent cadaveric study showed significantly the value of the CT scan to identify osteonecrosis of
higher fixation stability for the double screw com- the scaphoid proximal pole is still under debate
pared with single screw fixation (Mandaleson et al., (Smith et al., 2009).
2018). This is the reason why we selected two There are some of limitations in our study. The
screws. More rigid fixation may also allow partial patient numbers were small, and there were no trad-
loading of the wrist soon after surgery. Accordingly, itional freehanded percutaneous fixations to compare
our current postoperative regimen allows the with the efficacy and accuracy of robot-assisted
patients to bear some weight, and no fixation failure placement of the screws. Nor can we compare the
occurred. Mean healing time was 8 weeks, which is use of two screws with a single screw. Double screw
shorter than in previous reports treated by percutan- placement obviously increases the costs of the sur-
eous fixation only. However, we cannot definitely con- gery and the overall operative time. Another limita-
clude that two screw fixation shortens the healing tion is that the technique was applied for a rather
process because other factors also may play a role, limited indication: minimally displaced nonunions
including treatment delay, patient age, bone quality, and delayed unions. Large scale prospective clinical
comorbidities and smoking. trials are needed to compare the healing time and
Optimal placement of one cannulated screw in the healing incidence of single screw and double screw
scaphoid is technically demanding, more so with two fixation.
screws. In recent years, navigation-assisted technol-
ogy has been developed to provide accurate calcula-
tion and guidance for volar or dorsal percutaneous Declaration of conflicting interests The authors
declare no potential conflicts of interest with respect to
scaphoid screws (Kam and Greenberg, 2014; Walsh
the research, authorship, and/or publication of this article.
et al., 2009). However, it still needs to rely on the
surgeon’s hand–eye coordination. The surgeon
must hold the drill and try to follow the trajectory Funding The authors disclosed receipt of the following
shown in the navigation system. In our robotic financial support for the research, authorship, and/or pub-
system, the robot arm with six degrees-of-freedom lication of this article: this work was supported by Beijing
firmly holds the sleeve, through which the surgeon is Talents Fund [2015-3-305].
6 Journal of Hand Surgery (Eur) 0(0)

References Mack GR, Wilckens JH, McPherson SA. Subacute scaphoid frac-
tures: a closer look at closed treatment. Am J Sports Med.
Adams BD, Blair WF, Reagan DS, Grundberg AB. Technical factors
1998, 26: 56–8.
related to Herbert screw fixation. J Hand Surg Am. 1988, 13:
Mahmoud M, Koptan W. Percutaneous screw fixation without bone
893–9.
grafting for established scaphoid nonunion with substantial
Capo JT, Orillaza NS Jr, Slade JF III. Percutaneous management of
bone loss. J Bone Joint Surg Br. 2011, 93: 932–6.
scaphoid nonunions. Tech Hand Up Extrem Surg. 2009, 13:
Mandaleson A, Tham SK, Lewis C, Ackland DC, Ek ET. Scaphoid
23–9.
fracture fixation in a nonunion model: a biomechanical
Capo JT, Shamian B, Rizzo M. Percutaneous screw fixation without
study comparing 3 types of fixation. J Hand Surg Am. 2018,
bone grafting of scaphoid non-union. Isr Med Assoc J. 2012, 14:
43: 221–8.
729–32.
Merrell GA, Wolfe SW, Slade JF III. Treatment of scaphoid non-
Chim H, Malkoc F, Tay SC, Yam A, Teoh L. Technique of olecranon
unions: quantitative meta-analysis of the literature. J Hand
bone grafting for surgical fixation of scaphoid fractures. J Hand
Surg Am. 2002, 27: 685–91.
Surg Am. 2011, 36: 1220–3.
Parkinson RW, Hodgkinson JP, Hargadon EJ. Symptomatic non-
Cooney WP, Linscheid RL, Dobyns JH. Triangular fibrocartilage
union of the carpal scaphoid: Matti-Russe bone grafting
tears. J Hand Surg Am. 1994, 19: 143–54.
versus Herbert screw fixation. Injury. 1989, 20: 164–6.
Cosio MQ, Camp RA. Percutaneous pinning of symptomatic scaph-
Radford PJ, Matthewson MH, Meggitt BF. The Herbert screw for
oid nonunions. J Hand Surg Am. 1986, 11: 350–5.
delayed and non-union of scaphoid fractures: a review of fifty
Dias JJ, Wildin CJ, Bhowal B, Thompson JR. Should acute scaphoid
cases. J Hand Surg Br. 1990, 15: 455–9.
fractures be fixed? A randomized controlled trial. J Bone Joint
Saint-Cyr M, Oni G, Wong C, Sen MK, LaJoie AS, Gupta A. Dorsal
Surg Am. 2005, 87: 2160–8.
percutaneous cannulated screw fixation for delayed union and
Ernst SMC, Green DP, Saucedo JM. Screw fixation alone for scaph-
nonunion of the scaphoid. Plast Reconstr Surg. 2011, 128:
oid fracture nonunion. J Hand Surg Am. 2018, 43: 837–43.
467–73.
Guo Y, Tian G, Zlotolow DA et al. A cadaveric study on the accuracy
Slade JF 3rd, Dodds SD. Minimally invasive management of scaph-
of an individualized guiding template to assist scaphoid fixation
oid non- unions. Clin Orthop Relat Res. 2006, 445: 108–19.
using computed tomography and 3-dimensional printing. J
Slade JF 3rd, Geissler WB, Gutow AP, Merrell GA. Percutaneous
Hand Surg Am. 2019, 44: 2511–6.
internal fixation of selected scaphoid nonunions with an
Hegazy G. Percutaneous screw fixation of scaphoid waist fracture
arthroscopically assisted dorsal approach. J Bone Joint Surg
non-union without bone grafting. J Hand Microsurg. 2015, 7:
Am. 2003, 85: 20–32.
250–5.
Smith ML, Bain GI, Chabrel N, Turner P, Carter C, Field J. Using
Inaparthy PK, Nicholl JE. Treatment of delayed/nonunion of scaph-
computed tomography to assist with diagnosis of avascular
oid waist with Synthes cannulated scaphoid screw and bone
necrosis complicating chronic scaphoid nonunion. J Hand
grsaft. Hand. 2008, 3: 292–6.
Surg Am. 2009, 34: 1037–43.
Kam CC, Greenberg JA. Computer-assisted navigation for dorsal
Somerson JS, Fletcher DJ, Srinivasan RC, Green DP. Compression
percutaneous scaphoid screw placement: a cadaveric study.
screw fixation without bone grafting for scaphoid fibrous non-
Journal of Hand Surgery. 2014, 39: 613–20.
union. Hand (NY). 2015, 10: 450–3.
Le X, Tian W, Shi Z et al. Robot-assisted versus fluoroscopy-
Suh N, Grewal R. Controversies and best practices for acute
assisted cortical bone trajectory screw instrumentation in
scaphoid fracture management. J Hand Surg Eur. 2018, 43:
lumbar spinal surgery: a matched-cohort comparison. World
4–12.
Neurosurg. 2018, 120: 745–51.
Tian W, Fan M-X, Liu Y-J. Robot-assisted percutaneous pedicle
Leyshon A, Ireland J, Trickey EL. The treatment of delayed union
screw placement using three-dimensional fluoroscopy: a pre-
and non-union of the carpal scaphoid by screw fixation. J Bone
liminary clinical study. Chin Med J. 2017, 130: 1617–8.
Joint Surg Br. 1984, 66: 124–7.
Walsh E, Crisco JJ, Wolfe SW. Computer-assisted navigation of
Liu B, Wu F, Chen S, Jiang X, Tian W. Robot-assisted percutaneous
volar percutaneous scaphoid placement. J Hand Surg Am.
scaphoid fracture fixation: a report of ten patients. J Hand Surg
2009, 34: 1722–8.
Eur. 2019, 44: 685–91.
Wang JQ, Wang Y, Feng Y et al. Percutaneous sacroiliac screw
Luchetti TJ, Hedroug Y, Fernandez JJ, Cohen MS, Wysocki RW.
placement: a prospective randomized comparison of robot-
The morphology of proximal pole scaphoid fractures: implica-
assisted navigation procedures with a conventional technique.
tions for optimal screw placement. J Hand Surg Eur. 2018, 43:
Chin Med J. 2017, 130: 2527–34.
73–9.