P A T T E R N OF S C A P H O I D F R A C T U R E U N I O N D E T E C T E D BY

MACRORADIOGRAPHY
J. E. NICHOLL, J. D. SPENCER and J. C. BUCKLAND-WRIGHT
From the Department of Orthopaedics and Trauma and the Division of Anatomy and Cell Biology, United Medical and Dental
Schools of Guy's and St Thomas's Hospitals, London, UK

Stereoscopic macroradiography was used to study the pattern of union of scaphoid fractures. Of
21 patients who had partial union of a scaphoid fracture, 13 u~aited on the ulnar side and five on
the radial side only. In three it was on both sides but delayed in the centre. Where there was
initial union on the ulnar side, all progressed to complete union, including two patients with
displaced fractures. Four out of five patients with initial union on the radial side had displaced
fractures, and three of these went on to non-union. We conclude that in those patients with
partial union of a scaphoid fracture, if union is seen on the ulnar side, the fracture is likely to
unite completely, whereas if there is partial union on the radial side, there is a high risk of
non-union.
Journal of Hand Surgery (British and European Volume, 1995) 20B." 2:189-193

Most patients with a fractured scaphoid are young and
male, and want to be in plaster for the minimum length
of time. Deciding when it is safe to leave a scaphoid
fracture out of plaster remains a difficult problem
(Barton, 1992). Clinical examination and standard radi-
ography can be misleading (Dias et al, 1988) since
tenderness and/or a fracture line, visible on plain radio-
graphs, may be present despite sound union. There
appears to be very little information on the pattern of
union across the scaphoid in relation to clinical outcome,
so we undertook to examine the relationship between
the site of initial healing of scaphoid fractures detected
by radiography and their subsequent outcome in terms
of union and non-union. Standard radiography clearly
shows cortical defects and their healing, but the healing
of scaphoid fractures is endosteal (Saffar, 1990), so we
used macroradiography, which is useful for demonstrat-
ing changes in cancellous bone, to study these fractures
(Buckland-Wright and Bradshaw, 1989b).
A microfocal radiographic unit was used, with an
estimated source size of 15 txm, which allows radiographs
to be prepared with a higher magnification and spatial
resolution than that achieved with 10 x optical magnifi-
cation of standard radiographs taken on fine grain film
(Buckland-Wright, 1989a). It provides detailed images
of the structural organisation of bone comparable to
those identified in histological studies (Buckland-Wright
and Bradshaw, 1989b).
In this study 21 patients had macroradiographs of the
scaphoid fracture to determine the site of repair and its
relationship to clinical outcome.
PATIENTS A N D M E T H O D S
Patients
21 patients with fractures of the scaphoid (nine left and
12 right) were recruited into this study. 18 were male
and three female, and the mean (SD) age at the time of
fracture was 31 _+8.4 years (range 17-43.7).
The selection criteria were the presence of a scaphoid
fracture diagnosed on clinical examination and
standard radiography. The exclusion criteria were
fractures that were united, were established non-unions,
had been operated upon, or were ligamentous avulsion
fractures.
The patients had macroradiography performed on
average 14 ± 18.7 weeks, (range 10 days-87 weeks) after
they had sustained their fractures. The patient who had
his macroradiography 87 weeks after his injury had not
sought medical advice at the time of injury and presented
over a year later with a painful wrist. Patients were
subsequently kept under clinical review, with their treat-
ment being determined by the clinical findings and the
appearance of the standard and macroradiographs. Nine
patients had repeated macroradiography, five twice, two
three times, one four times and one on five occasions.
Any patient who developed an established non-union
was treated by internal fixation and bone graft.
Preparation of macroradiographs
Real-time fluoroscopic examination was undertaken of
the scaphoid to determine the site and extent of union
prior to taking the radiograph. Anteroposterior stereop-
air macroradiographs of the wrist in ulnar deviation
were prepared at a magnification of between x 5 and
x 9. The hand was placed on a platform positioned
close to the source (25 cm), and was displaced by 6 mm
between successive exposures (Buckland-Wright and
Bradshaw, 1989b). To assist in the interpretation of
partial union, additional macroradiographic views were
obtained with the wrist rotated so as to align the plane
of the fracture with the X-ray beam. The stereopair
macroradiographs obtained were examined by one
observer under a Large Format Stereoscope (Ross
Instruments, Salisbury, UK) which permitted a three-
dimensional evaluation of the fracture in the scaphoid

189
190 T H E J O U R N A L OF H A N D SURGERY VOL. 20B No. 2 APRIL 1995

and to distinguish between overlapping surfaces of the

bone and union.

Assessment

To determine the proportion of the different types of

scaphoid fracture in our sample each fracture was
classified using the Schernberg classification (Schernberg
et al, 1984), in which fractures are classified as being
through the proximal pole (type 1), proximal part of
the body (type 2), distal part of the body (type 3),
tuberosity (type 4) or distal segment (type 5), or as
partial fractures of the tuberosity (type 6); (Saffar,
1990).
The extent o f union was assessed. We defined union
to have occurred where we could see continuity in the
trabecular organization extending across and beyond
the site of pre-existing fracture in both fragments
(Fig 1). In some cases it was possible to see microcallus
formation at sites where trabeculae were united (Fig 1).
Particular care was taken in examining regions where
the plane of the fracture was not perpendicular to the
X-ray beam.
The fractures were divided into three groups: those
with union on the ulnar or radial sides only; and those
in which there was union on the ulnar and radial sides
of the scaphoid, but not in the centre.

RESULTS
The number of patients in each Schernberg group and
the relationship with the different sites of union is shown
in Table 1. Examination o f the macroradiographs under
the stereoscope revealed that in the majority of cases
(86%) union did not occur along the entire length but
occurred at either the ulnar or radial sides of the
scaphoid. At the time of the initial macroradiograph,
13 of the 21 patients (62%) had union only on the ulnar
side of the fracture (Fig 2a). In 11 of them there was
only minimal displacement of the fracture, and they all Fig 1 Macroradiograph of a uniting scaphoid fracture 2 weeks after
progressed to union with conservative treatment injury. Large arrows indicate the fracture line. Small arrows
indicate microcallus formation. (Original magnificationx 9,
(Fig 2b). The remaining two initially had a large gap at reproduced at x 5.9.)
the fracture site, but they also ultimately progressed to
full union with continued cast immobilization.
In five cases (24%) union was on the radial side only
(Fig 3), with a large gap on the ulnar side in four of
the patients. By the end of the study three of them had Table 1--Relationship between Schernberg group and initial site of
been treated operatively. The fracture united with non- union, and expected number of patients from percentages in Schernberg
et al, 1984
operative treatment in the other two.
In three cases (14%) union was seen at both the ulnar
and radial sides of the fracture line (Fig 4), but not in Site of initial union Schernberg Group
the central portion. In all three cases the fracture had 2 3 4 5
united by the end o f the study with non-operative
treatment. Ulnar 4 2 4 3
Radial 0 2 3 0
Radial and ulnar 0 0 3 0
DISCUSSION Total number 4 4 10 3
Expected number from
The enhanced magnification and spatial resolution of Schenberg's pecentages 4.2 6.5 6.5 1.7
stereoscopic micro focal radiography demonstrates
MACRORADIOGRAPHY OF SCAPHO1D
Fig 2 (a) Macroradiograph of a scaphoid showing partial union on the ulnar side of the fracture line, 3 months after injury. (Original
magnification x 7.2, reproduced at x 6.) (b) Macroradiograph of the same patient after 6 months in a cast, showing almost complete
union, (Original magnification x 7.6, reproduced at x 6.)
where union of scaphoid fractures has occurred. Before
full union has occurred across the entire length of the
fracture line, continuity in the trabecular organization
can be seen across only a portion of the fracture, and
the site of this partial union seems to influence whether
the fracture will unite or not.
Given that this is a small sample of patients, the
number in each group as described in the Schernberg
classification does not differ markedly from those that
would be expected from the percentages reported by
Schernberg et al(1984). This can therefore be considered
a fairly representative sample of scaphoid fractures.
Although union of a scaphoid fracture is influenced
by the site of the fracture, its stability, and any
displacement during treatment (Leslie and Dickson,
1981), bone healing is stimulated by axial compression.
In the scaphoid the major longitudinal compressive
force passes along the ulnar side of the bone, which is
reflected by the greatest density of longitudinal trabecu-
lae on that side (Fig 2). In an undisplaced fracture
191
initial healing is therefore likely to occur on the
ulnar side.
In our study, 13 of the 14 patients who presented
with undisplaced fractures had union on the ulnar side
on their initial macroradiographs. Three of these had
union on the radial side as well, because they were
further advanced towards complete union. Two of the
six patients with displaced fractures also showed partial
union on the ulnar side, and they both united with
conservative treatment, suggesting that once a fracture
has partial union on the ulnar side this gives it sufficient
stability to prevent secondary displacement and sub-
sequent non-union. In displaced fractures some healing
may occur on the radial side where factors such as the
direction of the displacement and the forces exerted by
the interosseous ligaments prevent union on the ulnar
side. This radial side healing often provides insufficient
stability to prevent secondary displacement. There is a
high risk of non-union following secondary displacement
(Saffar, 1990) and these patients are therefore likely to
192
Fig 3 Macroradiograph of a scaphoid with partial union on the
radial side o f the fracture line, 6.5 months after injury.
(Original magnification x 5, reproduced at x 4.2.)
develop non-union. In our study of four patients with
displaced fractures and partial radial union, only one
united with conservative treatment. Although initial
displacement of the fracture is important in predicting
ultimate union (Herbert, 1988; Taleisnik, 1985) our
results show that partial healing has a significant influ-
ence on the final outcome.
From these findings we suggest that where there is
partial union of a scaphoid fracture on the ulnar side,
it is likely to progress to complete union with continued
cast immobilization. Where a fracture has non-union on
the ulnar side after a period of conservative treatment,
with partial union on the radial side, the fracture is
likely to progress to non-union. The patient can be
advised that operative intervention is likely to be neces-
sary to achieve union, and this can be done earlier,
reducing the total time spent in plaster. Furthermore,
since macroradiography is not widely available, our
observations indicate the importance of examining the
ulnar and radial borders of the scaphoid in standard
radiographs for sites of partial union to help predict
outcome.
THE JOURNAL OF HAND SURGERY VOL. 20B No. 2 APRIL 1995
Fig 4 Macroradiograph of a scaphoid, 3.5 months after injury
showing partial union on radial and ulnar sides, but not in
the centre of the fracture line. (Original magnification x 7.2,
reproduced at x 6.)
Acknowledgment
We wish to express our gratitude for financial support received from the Bounty
Orchard Lisle Fund provided by the Special Trustees of Guy's Hospital, to Mrs
Judy Vlahovic for technical assistance and to Mr Kevin Fitzpatrick and Miss
Sarah Smith for preparation of the photographs.
References
BARTON, N. J. (1992). Twenty questions about scaphoid fractures. Journal of
Hand Surgery, 17B: 3: 289-310.
BUCKLAND-WRIGHT, J. C. (1989a). A new high-definition microfocal X-ray
unit. British Journal of Radiology, 62: 201-208.
BUCKLAND-WRIGHT, J. C. and BRADSHAW, C. R. (1989b). Clinical appli-
cations of high-definition microfocal radiography. British Journal of
Radiology, 62:209 217.
DIAS, J. J., TAYLOR, M., THOMPSON, J., BRENKEL, I. J. and GREGG,
P. J. (1988). Radiographic signs of union of scaphoid fractures: An analysis
of inter-observer agreement and reproducibility. Journal of Bone and Joint
Surgery, 70B: 2: 299-301.
MACRORADIOGRAPHY OF SCAPHOID 193

HERBERT, T. J. Scaphoid Fractures: Operative Treatment. In: Barton N. J. TALEISNIK, J. The Writ. New York, Churchill Livingstone, 1985:
(Ed.) Fractures of the Hand and Wrist. Edinburgh, Churchill Livingstone, 105 148.
1988: 220-235.
LESLIE, I. J. and DICKSON, R. A. (1981). The fractured carpal scaphoid:
Natural history and factors influencing outcome. Journal of Bone and Joint
Accepted: 25 October 1994
Surgery, 63B: 2: 225-230. Dr J. C. Bucldand-Wright,Division of Anatomy and Cell Biology,UMDS, Guy's Hospital,
SAFFAR, P. Carpal Injuries, Paris, Spring,er-Verlag, 1990:91-112. London Bridge, London SE1 9RT, UK.
SCHERNBERG, F., ELZEIN, F. and GERARD, Y. (1984). I~tude anatomo-
© 1995 The British Society for Surgery of the Hand
radiologique des fractures du scaphoide carpien. Probl6mes des cals vicieux.
Revue de Chirurgie Orthop~dique, Suppl. 2, 70: 55-63.