- Email: [email protected]
The unstable colles' fracture
THE
UNSTABLE
COLLES’
FRACTURE
N. H. JENKINS From the Department of Orthopaedic and Traumatic Surgery, Universityof Wales College of Medicine, Cardifs One hundred and twenty-one
displaced Colles’ fractures were assessed radiographically until union. An acute loss of position occurred during the first week of splintage, necessitating re-manipulation in six fractures. In the remaining 115 fractures whose position had been retained after one week, chronic instability led to a mean increase of 8.22” dorsal angulation, a mean loss of 5.61” radial angle, and 3.26 mm radial shortening. The extent to which the chronic collapse of radial angle and length occurred was determined solely by the initial deformity and was not related to either intraarticular involvement or the presence of radiographically visible comminution. However, the absence of radiographically visible comminution of the dorsal radius did confer stability against mal-union in dorsal angulation. Journal of Hand Surgery (British Volume, 1989) 14B: 149-154
Despite the almost universal practice of early radiographical assessment of the manipulated Colles’ fracture followed, when necessary, by re-manipulation, mal-union remains a common complication (Stewart, Innes and Burke, 1985). In an attempt to elucidate the process of mal-union, 121 displaced Colles’ fractures were prospectively followed and particular attention paid to those radiographical features that might predispose to malunion. The study’s findings lead to a consideration of the previously implied, though not discussed, concept of the fracture’s chronic instability. Method 121 patients with unilateral displaced Colles’ fractures were studied. The age range was 17-88 years (mean 54.2
years). Following reduction, the fractures were splinted by dorsal plaster slabs and, one week later, the fracture position checked radiographically, after which the plasters were completed. Six fractures had slipped to an unacceptable position at this visit and were re-manipulated but excluded from further analysis, thus leaving 115 fractures which had retained their manipulated position after one week. A decision regarding fracture union was ma.de four weeks following injury. This was done on purely clinical grounds i.e. no fracture which was tender was considered to be united. Four fractures were considered not to have united at this visit and were splinted for a further week. At fracture union, the wrist was again X-rayed and the other wrist was also X-rayed for comparison. Radiographic measurements
The dorsal angle of the radius, the radial shortening, and the flattening of the radial angle were assessed as has been previously described (Jenkins et al., 1987). These measurements were made from the radiographs obtained at presentation, following reduction, and again at fracture union. The measurement of dorsal angulation was an absolute measurement which, unlike the measurement of radial VOL. 14-B No. 2 MAY 1989
angle and length, is without reference to the normal value. This has been done to be in concordance with the standard published works on the fracture. Therefore, in Table 7 (for example) the mean dorsal angle following reduction is 0.98” in excess of the neutral position. Fracture classification
In an attempt to study the influence of displacement and intra-articular involvement on the future development of mal-union, each of the 115 fractures was classified according to the systems described by Gartland and Werley (1951), Frykman (1967), Lidstriim (1959), and Older et al. (1965). These are summarised in Tables 1 to 4 and illustrated in Figure 1. In addition, the effect of comminution was studied by classifying the fractures by a system based upon the presence and the distribution of comminution (Table 5 and Fig. 2). Table l-Classification of Gartland and Werley (1951): based upon the presence, but not the extent, of displacement and radio-carpal involvement Group 1 Group 2 Group 3 Group 4
Extra-articular, In&a-articular, Intra-articular, Extra-articular, completeness
displaced. undisplaced. displaced. undisplaced (added for by Solgaard, 1985).
Table 2-Classification of Lidstriim (1959): based upon the presence, but not the extent, of displacement (dorsal angle and dorsal displacement), articular involvement and comminution Group 1 Group 2a Group 2b Group 2c Group 2d
Group 2e
Undisplaced. Dorsal angulation, extra-articular. Dorsal angulation, intra-articular but without gross separation of fragments. Dorsal angulation plus dorsal displacement, extraarticular. Dorsal angulation plus dorsal displacement, intraarticular but without gross separation of fragments. Dorsal angulation plus dorsal displacement, intraarticular with separation of fragments.
149
N. H. JENKINS
Table 3-Classification of Older et al. (1965): based upon the extent of displacement (dorsal angle and radial shortening) and the presence of comminution Group
“Non-displaced”-up articular surface
1
GARTLAND WERLEY (1951) 1
to 5” dorsal angulation, radial at least 2 mm above ulnar head.
Group 2
“Displaced with minimal comminution”dorsa1 angulation or displacement, radial articular surface no lower than 3 mm below ulnar head, minimal comminution of dorsal radius.
Group 3
“Displaced with comminution of dorsal radius”comminution of dorsal radius : Radial articular surface below ulnar head; Minimal comminution of distal fragment.
Group 4
“Displaced with severe comminution of radial head”-marked’comminution of dorsal and distal radius; Radial articular surface 2-8 mm below ulnar head.
2-AI
Table 4-Classification of Frykman (1967): based upon the pattern of intra-articular involvement Groups Groups Groups Groups
1 and 3 and 5 and 7 and
2 4 6 8
Groups Groups
1, 3,5 2,4,6 and 8
Extra-articular. Involve radio-carpal joint. Involve distal radio-ulnar joint. Involve both radio-carpal and distal radioulnar joints. Have an intact ulnar styloid. Have a fractured ulnar styloid.
OLDER (1965)
No radiographically visible comminution. Comminution of the dorsal radial cortex without comminution of the fracture fragment. Comminution of the fracture fragment without significant involvement of the dorsal cortex. Comminution of both the distal fragment and the dorsal cortex. As the fracture line involves the distal fracture fragment in Groups 3 and 4, intraarticular involvement is very common within these groups. Such involvement is not, however, inevitable and nor does it affect the fracture’s placement within the classification.
Group 1 Group 2 Group 3 Group 4
V?YKMAN (1967)
na
Fig.
Analyses of means, standard deviations, and regression were performed using an OXSTAT* programme on an Amstrad PCW 8256@ computer.
Results The distribution of the 115 fractures within the five classifications is shown in Table 6. The mean displacements of the three radiological parameters at injury, foiiowing reduction, and at fracture union are presented in Table 7 which shows a substantial
150
90
1982, 1983 and
1985 by Holman,
Jones,
Walter
l/:
7/a
Statistical analysis
*OXSTAT Wiggins.
1
&
Some classifications 1 to4.
F9
of Colles’ fractures:
for details,
see Tables
re-displacement of each parameter between reduction and union; the dorsal angle increasing by a mean of 8.22” to unite at 9.03”; the radial angle lost 5.61” to a deficiency of 7.75” at union; whilst 3.26 mm loss of radial length occurred from 0.92 mm radial shortening at reduction to 4.02 mm at union.
The relationship of mal-union to the initial displacement The relationship of the fracture’s initial displacement to its radiographical position at union was investigated THE
JOURNAL
OF HAND
SURGERY
THE UNSTABLE Table 6-The
COLLES’ FRACTURE
distribution of fractures within the five classification systems Comminution Type
No.
Type
1 2 3 4
17 48 8 42
1 3
Total=115
Frykman
Lidstrijm
Gartland and Werley No. 68 47
Older
Type
No.
Type
No.
2a 2b 2c 2d 2e
41 28 26 8 12 115
l/2 314 516 718
33 15 28 39
115
Type
No. 43 56 16
2 3 4
115
115
-
using a regressional analysis, the results of which are shown in Table 8. For each parameter, the analysis indicates a highly significant correlation between the fracture’s position at union and the extent of its initial displacement. The relationship of mal-union tofracture classiJication
Table 9 presents the mean displacement of each radiographical parameter at injury and again at fracture union for each group within the five fracture classifications. Within each classification system, the initial displacement necessarily varies between the groups of each system; indeed, in Older’s system the difference forms the basis of the classification.
Table 7-Mean displacements of the radiographic parameters at injury, following reduction and at fracture union (standard deviation in brackets) Injury Dorsal angle [degrees] Radial shortening
The dorsal angle at union varies according to the severity of injury in all classifications with the exception of that based upon comminution. Mal-union in dorsal angulation occurs only when the dorsal cortex is comminuted [Groups 2 and 41.
Union
25.69 6.28
(12.24) (5.37)
0.98 0.92
(10.08) (3.08)
9.03 4.02
(12.05) (3.97)
11.27
(8.94)
2.36
(5.04)
7.75
(7.23)
[mm1 Flattening of radial angle [degrees]
Table &Regressional analysis correlating the radiographic fracture union with the displacement at injury. Parameter
Dorsal angle Radial length Radial angle
Dorsal angle
Reduction
Correlation co-efficient
Probability
0.37 0.48 0.46
<0.0001 10.0001 < 0.0001
result at
according to their respective displacements at injury which are the only predictors of mal-union in either parameter. None of the classifications is capable of identifying those fractures at risk of mal-union in either radial shortening or by flattening of the radial angle.
Radial length and radial angle
The deficiencies in radial length and angle at union vary
Fig. 2
A classification of Colles’ fractures based entirely nution: for detailed explanation, see Table 5.
VOL. 14-B No. 2 MAY 1989
on commi-
Discussion
Clinical studies have consistently identified mal-union as a common complication of the displaced Colles’ fracture (Gartland and Werley, 1951; Stewart et al., 1985) although its frequency is not generally appreciated, perhaps because it is usual to make decisions regarding union upon clinical rather than radiographical grounds. It is in an attempt to prevent mal-union that manipulated fractures are routinely X-rayed at some interval around the first week after injury and, if re-displacement occurs, may be re-manipulated (Collert and Isacson, 1978; McQueen et al., 1987). This early re-displacement is relatively uncommon (accounting for 5% of the cases of the present study) and follows a macroscopic movement at the fracture site i.e., it is a consequence of acute instability. Despite these attempts to identify and prevent acute 151
N. H. JENKINS Table %-Comparison of the ability of the fracture classifications to predict mal-tion. Method of class$cation
Dorsal angle (degrees)
Mean values of the radiographic parameters
Radial shortening (mm)
Injury
Union
Injury
Union
18.06 24.00 18.00 30.48
2.31 11.46 0.50 14.10
5.25 4.82 6.88 7.44
23.26 21.89
9.31 11.36
19.54 21.00 29.81 34.13 36.17
Flattening of radial angle (degrees) Injury
Union
2.44 4.31 5.00 4.65
7.38 9.02 9.63 14.57
4.94 8.44 8.88 8.52
4.99 7.41
3.87 4.76
8.98 13.51
1.55 8.83
10.55 9.37 7.43 8.50 16.17
4.75 5.91 5.65 7.63 10.08
4.88 4.73 2.24 4.63 4.25
8.20 9.93 10.88 14.13 20.17
9.23 8.46 4.87 8.75 8.75
23.52 21.79 24.68 26.33
7.64 10.42 9.19 10.92
4.30 5.47 5.48 8.07
3.50 3.11 3.91 5.39
8.73 10.42 9.92 13.67
7.76 4.84 7.82 9.87
19.15 27.11 35.39
5.16 12.55 14.71
4.00 6.44 9.78
3.37 4.68 5.11
6.55 11.55 19.78
6.39 8.98 9.67
Comminution 1 2 3 4 Gartland & Werley 1 3 Lidstriim 2a 2b 2c 2d 2e Frykman 112 314 516 718 Older 2 3 4
Fig. 3
152
Case I A 24-year-old man sustained a badly-displaced and heavily-comminuted fracture which was treated by external fixation. The fixator frame was removed after three weeks because of persistent infection. At this time the fracture was firm and unmoveable but mildly tender, indicating that union was not complete. The fracture was splinted in a forearm cast and these X-rays show the position of the fracture in the cast.
Fig. 4
Case I A week later, whilst within the cast, the fracture had collapsed to the position shown. As the fracture site had been unmoveable a week earlier, the collapse must be a different process to the acute loss of position which occurs early in treatment and which is amenable to re-manipulation. This is clearly a collapse of uniting bone.
THE JOURNAL
OF HAND SURGERY
THE UNSTABLE
instability, mal-union is common and, in the present study, the 95% of fractures which had retained their manipulated position after one week of splintage subsequently lost some of this, with an increase of 8.22” in dorsal angle and loss of 5.61” radial angle and 3.26 mm radial length. This late deterioration in the fracture’s position occurs slowly and progressively, and is a manifestation of its chronic instability. Although chronic instability has been previously recognised (Anderson and O’Neil, 1944; Charnley, 1961; Spira and Weigl, 1969) it has never been specifically discussed and , as a result, the two types of instability are still discussed as one, implying a common cause (Solgaard, 1986). It is common experience that attempted remanipulations two weeks or more after the injury rarely improve the fracture’s position, despite continued radiographic deterioration (Figs. 3 and 4) and this is clearly a consequence of a gradual collapse of uniting cancellous bone. Charnley (1961) recognised this chronic collapse and claimed it to be an essential pre-requisite for strong union in cancellous bone, requiring control rather than prevention. Hitherto intra-articular involvement and radiographical comminution have been assumed to be the most accurate indicators of the fracture’s stability and have been widely cited as indications for operative management (Cooney et al., 1979; Grana and Kopta, 1979; Wagner and Jakob, 1985). The only previously published study aimed at identifying those fractures destined to m&union is that of Solgaard (1985) who concluded that the classification described by Older et al. (1965) was the most accurate predictor, although neither the pattern nor the extent of mal-union were studied. The classification described by Older is based upon both comminution and displacement (principally radial length, but also including dorsal angle) and thus does not allow for an assessment of the relative contribution of each factor to the
COLLES’ FRACTURE
Fig. 6
subsequent mal-union. In the present study, the effects of initial displacement, comminution and intra-articular involvement have each been studied in isolation. The results show that whilst those relatively uncommon (22%) fractures with no comminution of the dorsal cortex are capable of resisting mal-union in dorsal angulation, this pattern of mal-union regularly complicates the remaining 78% of fractures in which dorsal comminution has occurred (Figs. 5 to 7), the extent of the mal-union depending upon the severity of the initial displacement (Table 8). Mal-union in radial shortening or by flattening of the radial angle occurs independently of radiographitally-visible comminution and may be predicted only by
Fig. 7 Fig. 5
Case 2 A Colles’ fracture
VOL. 14-B No. 2 MAY 1989
sustained
by a 19-year-old
secretary.
Case 2 Anatomical reduction was attained by manipulation. This post-reduction X-ray shows the comminution of the dorsal cortex. A check X-ray a week later showed the reduction to be maintained and the plaster was completed.
Case 2 Radiographic examination four weeks after the injury shows the fracture to have mal-united in dorsal angulation but with minimal radial shortening.
153
N. H. JENKINS
the initial respective displacements (Table S), presumably reflecting comminution at trabecular level. In no instance is the development of mal-union dependent upon articular involvement. Conclusions
A minority (5%) of manipulated Colles’ fractures sustain an early loss of position which may be identified by early radiography and corrected by re-manipulation. In most cases, however, chronic instability produces a progressive collapse of the uniting cancellous bone. Those fractures which will mal-unite with radial shortening or by flattening of the radial angle are those with significant displacement of these two parameters at presentation, regardless of the radiographical pattern of comminution or the presence of articular involvement. A minority of fractures are resistant to mal-union in dorsal angulation and may be identified by a radiographically intact dorsal radial cortex. Acknowledgments I wish to thank Mrs Janice
Sharp (Department of Medical Illustration, University Hospital of Wales) and Mr Keith Bellamy (Medical Photographer, Cardiff Royal Infirmary) for their help in illustrating this paper. In addition, I extend my grateful thanks to Mrs Marjorie James for typing the manuscript.
COONEY, W. P., LINSCHEID, R. L. and DOBYNS, .I. H. (1979). External Pin Fixation for Unstable Colles’ Fractures. Journal of Bone and Joint Surgery, 61A:6: 840-845. FRYKMAN, G. (1967). Fracture of the distal radius including sequelaeshoulder-hand-finger syndrome, disturbance in the distal radio-ulnar joint and impairment of nerve function. Acta Orthopaedica Scandinavica, Supplement 108. GARTLAND, J. .I. and WERLEY, C. W. (1951). Evaluation of healed Colles’ fractures. Journal of Bone and Joint Surgery, 33A: 4: 895-907. GRANA, W. A. and KOPTA, J. A. (1979). The Roger Anderson Device in the Treatment of Fractures of the Distal End of the Radius. Journal of Bone and JointSurgery, 61A: 8: 1234-1238. JENKINS, N. H., JONES, D. G. JOHNSON, S. R. and MINTOWT-CZYZ, W. J. (1987). External fixation of Colles’ fractures. An anatomical study. Journal of Bone and Joint Surgery, 69B : 2 : 207-211. LIDSTROM, A. (1959). Fractures of the distal end of the radius. A Clinical and Statistical Study of End Results. Acta Orthopaedica Scandinavica, Supplement41. McQUEEN, M., MacLAREN, A. and CHALMERS, J. (1986). The value of remanipulating Colles’ fractures. Journal of Bone and Joint Surgery, 68B: 2: 232-233. OLDER, T. M., STABLER, E. V. and CASSEBAUM, W. H. (1965). Colles fracture: evaluation and selection of therapy. Journal of Trauma, 5: 4: 469476. SOLGAARD, S. (1985). Classification of distal radius fractures. Acta Orthopaedica Scandinavica, 56: 3: 249-252. SOLGAARD, S. (1986). Early displacement of distal radius fracture. Acta Orthopaedica Scandinavica, 57 : 3 : 229-23 1, SPIRA, E. and WEIGL, K. (1969). The Cornminuted Fracture of the Distal End of the Radius. Reconstruction Surgery and Traumatology, 11: 128-138. STEWART, H.D., INNES, A. R. and BURKE, F. D. (1985). Factors affecting the outcome of Colles’ fracture: an anatomical and functional study. Injury, 16: 5: 289-295. WAGNER, H. E. and JAKOB, R. P. (1985). Operative Treatment of Fractures of the Distal Radius Using External Fixation. Unfallchirurg, 88: 11: 473480.
References ANDERSON, R. and O’NEIL, G. (1944). Cornminuted fractures of the distal end of the radius. Surgery, Gynecology and Obstetrics, 78: 4: 434-440. CHARNLEY, J. The Closed Treatment of Common Fractures, 3rd ed. Edinburgh, ChurchillLivingstone, 1961: 128-142. COLLERT, S. and ISACSON, J. (1978). Management of Redislocated Colles’ Fractures. Clinical Orthopaedics and Related Research, 135: 183-186.
154
Accepted: 2 December 1988 892 Newport Road, Rumney, 0
1989 The British
Society
Cardiff,
for Surgery
South Clamorgan
CF3 8L.l
of the Hand
02667681/89/0014-0149/%10.00
THE
JOURNAL
OF HAND
SURGERY
UNSTABLE
COLLES’
FRACTURE
N. H. JENKINS From the Department of Orthopaedic and Traumatic Surgery, Universityof Wales College of Medicine, Cardifs One hundred and twenty-one
displaced Colles’ fractures were assessed radiographically until union. An acute loss of position occurred during the first week of splintage, necessitating re-manipulation in six fractures. In the remaining 115 fractures whose position had been retained after one week, chronic instability led to a mean increase of 8.22” dorsal angulation, a mean loss of 5.61” radial angle, and 3.26 mm radial shortening. The extent to which the chronic collapse of radial angle and length occurred was determined solely by the initial deformity and was not related to either intraarticular involvement or the presence of radiographically visible comminution. However, the absence of radiographically visible comminution of the dorsal radius did confer stability against mal-union in dorsal angulation. Journal of Hand Surgery (British Volume, 1989) 14B: 149-154
Despite the almost universal practice of early radiographical assessment of the manipulated Colles’ fracture followed, when necessary, by re-manipulation, mal-union remains a common complication (Stewart, Innes and Burke, 1985). In an attempt to elucidate the process of mal-union, 121 displaced Colles’ fractures were prospectively followed and particular attention paid to those radiographical features that might predispose to malunion. The study’s findings lead to a consideration of the previously implied, though not discussed, concept of the fracture’s chronic instability. Method 121 patients with unilateral displaced Colles’ fractures were studied. The age range was 17-88 years (mean 54.2
years). Following reduction, the fractures were splinted by dorsal plaster slabs and, one week later, the fracture position checked radiographically, after which the plasters were completed. Six fractures had slipped to an unacceptable position at this visit and were re-manipulated but excluded from further analysis, thus leaving 115 fractures which had retained their manipulated position after one week. A decision regarding fracture union was ma.de four weeks following injury. This was done on purely clinical grounds i.e. no fracture which was tender was considered to be united. Four fractures were considered not to have united at this visit and were splinted for a further week. At fracture union, the wrist was again X-rayed and the other wrist was also X-rayed for comparison. Radiographic measurements
The dorsal angle of the radius, the radial shortening, and the flattening of the radial angle were assessed as has been previously described (Jenkins et al., 1987). These measurements were made from the radiographs obtained at presentation, following reduction, and again at fracture union. The measurement of dorsal angulation was an absolute measurement which, unlike the measurement of radial VOL. 14-B No. 2 MAY 1989
angle and length, is without reference to the normal value. This has been done to be in concordance with the standard published works on the fracture. Therefore, in Table 7 (for example) the mean dorsal angle following reduction is 0.98” in excess of the neutral position. Fracture classification
In an attempt to study the influence of displacement and intra-articular involvement on the future development of mal-union, each of the 115 fractures was classified according to the systems described by Gartland and Werley (1951), Frykman (1967), Lidstriim (1959), and Older et al. (1965). These are summarised in Tables 1 to 4 and illustrated in Figure 1. In addition, the effect of comminution was studied by classifying the fractures by a system based upon the presence and the distribution of comminution (Table 5 and Fig. 2). Table l-Classification of Gartland and Werley (1951): based upon the presence, but not the extent, of displacement and radio-carpal involvement Group 1 Group 2 Group 3 Group 4
Extra-articular, In&a-articular, Intra-articular, Extra-articular, completeness
displaced. undisplaced. displaced. undisplaced (added for by Solgaard, 1985).
Table 2-Classification of Lidstriim (1959): based upon the presence, but not the extent, of displacement (dorsal angle and dorsal displacement), articular involvement and comminution Group 1 Group 2a Group 2b Group 2c Group 2d
Group 2e
Undisplaced. Dorsal angulation, extra-articular. Dorsal angulation, intra-articular but without gross separation of fragments. Dorsal angulation plus dorsal displacement, extraarticular. Dorsal angulation plus dorsal displacement, intraarticular but without gross separation of fragments. Dorsal angulation plus dorsal displacement, intraarticular with separation of fragments.
149
N. H. JENKINS
Table 3-Classification of Older et al. (1965): based upon the extent of displacement (dorsal angle and radial shortening) and the presence of comminution Group
“Non-displaced”-up articular surface
1
GARTLAND WERLEY (1951) 1
to 5” dorsal angulation, radial at least 2 mm above ulnar head.
Group 2
“Displaced with minimal comminution”dorsa1 angulation or displacement, radial articular surface no lower than 3 mm below ulnar head, minimal comminution of dorsal radius.
Group 3
“Displaced with comminution of dorsal radius”comminution of dorsal radius : Radial articular surface below ulnar head; Minimal comminution of distal fragment.
Group 4
“Displaced with severe comminution of radial head”-marked’comminution of dorsal and distal radius; Radial articular surface 2-8 mm below ulnar head.
2-AI
Table 4-Classification of Frykman (1967): based upon the pattern of intra-articular involvement Groups Groups Groups Groups
1 and 3 and 5 and 7 and
2 4 6 8
Groups Groups
1, 3,5 2,4,6 and 8
Extra-articular. Involve radio-carpal joint. Involve distal radio-ulnar joint. Involve both radio-carpal and distal radioulnar joints. Have an intact ulnar styloid. Have a fractured ulnar styloid.
OLDER (1965)
No radiographically visible comminution. Comminution of the dorsal radial cortex without comminution of the fracture fragment. Comminution of the fracture fragment without significant involvement of the dorsal cortex. Comminution of both the distal fragment and the dorsal cortex. As the fracture line involves the distal fracture fragment in Groups 3 and 4, intraarticular involvement is very common within these groups. Such involvement is not, however, inevitable and nor does it affect the fracture’s placement within the classification.
Group 1 Group 2 Group 3 Group 4
V?YKMAN (1967)
na
Fig.
Analyses of means, standard deviations, and regression were performed using an OXSTAT* programme on an Amstrad PCW 8256@ computer.
Results The distribution of the 115 fractures within the five classifications is shown in Table 6. The mean displacements of the three radiological parameters at injury, foiiowing reduction, and at fracture union are presented in Table 7 which shows a substantial
150
90
1982, 1983 and
1985 by Holman,
Jones,
Walter
l/:
7/a
Statistical analysis
*OXSTAT Wiggins.
1
&
Some classifications 1 to4.
F9
of Colles’ fractures:
for details,
see Tables
re-displacement of each parameter between reduction and union; the dorsal angle increasing by a mean of 8.22” to unite at 9.03”; the radial angle lost 5.61” to a deficiency of 7.75” at union; whilst 3.26 mm loss of radial length occurred from 0.92 mm radial shortening at reduction to 4.02 mm at union.
The relationship of mal-union to the initial displacement The relationship of the fracture’s initial displacement to its radiographical position at union was investigated THE
JOURNAL
OF HAND
SURGERY
THE UNSTABLE Table 6-The
COLLES’ FRACTURE
distribution of fractures within the five classification systems Comminution Type
No.
Type
1 2 3 4
17 48 8 42
1 3
Total=115
Frykman
Lidstrijm
Gartland and Werley No. 68 47
Older
Type
No.
Type
No.
2a 2b 2c 2d 2e
41 28 26 8 12 115
l/2 314 516 718
33 15 28 39
115
Type
No. 43 56 16
2 3 4
115
115
-
using a regressional analysis, the results of which are shown in Table 8. For each parameter, the analysis indicates a highly significant correlation between the fracture’s position at union and the extent of its initial displacement. The relationship of mal-union tofracture classiJication
Table 9 presents the mean displacement of each radiographical parameter at injury and again at fracture union for each group within the five fracture classifications. Within each classification system, the initial displacement necessarily varies between the groups of each system; indeed, in Older’s system the difference forms the basis of the classification.
Table 7-Mean displacements of the radiographic parameters at injury, following reduction and at fracture union (standard deviation in brackets) Injury Dorsal angle [degrees] Radial shortening
The dorsal angle at union varies according to the severity of injury in all classifications with the exception of that based upon comminution. Mal-union in dorsal angulation occurs only when the dorsal cortex is comminuted [Groups 2 and 41.
Union
25.69 6.28
(12.24) (5.37)
0.98 0.92
(10.08) (3.08)
9.03 4.02
(12.05) (3.97)
11.27
(8.94)
2.36
(5.04)
7.75
(7.23)
[mm1 Flattening of radial angle [degrees]
Table &Regressional analysis correlating the radiographic fracture union with the displacement at injury. Parameter
Dorsal angle Radial length Radial angle
Dorsal angle
Reduction
Correlation co-efficient
Probability
0.37 0.48 0.46
<0.0001 10.0001 < 0.0001
result at
according to their respective displacements at injury which are the only predictors of mal-union in either parameter. None of the classifications is capable of identifying those fractures at risk of mal-union in either radial shortening or by flattening of the radial angle.
Radial length and radial angle
The deficiencies in radial length and angle at union vary
Fig. 2
A classification of Colles’ fractures based entirely nution: for detailed explanation, see Table 5.
VOL. 14-B No. 2 MAY 1989
on commi-
Discussion
Clinical studies have consistently identified mal-union as a common complication of the displaced Colles’ fracture (Gartland and Werley, 1951; Stewart et al., 1985) although its frequency is not generally appreciated, perhaps because it is usual to make decisions regarding union upon clinical rather than radiographical grounds. It is in an attempt to prevent mal-union that manipulated fractures are routinely X-rayed at some interval around the first week after injury and, if re-displacement occurs, may be re-manipulated (Collert and Isacson, 1978; McQueen et al., 1987). This early re-displacement is relatively uncommon (accounting for 5% of the cases of the present study) and follows a macroscopic movement at the fracture site i.e., it is a consequence of acute instability. Despite these attempts to identify and prevent acute 151
N. H. JENKINS Table %-Comparison of the ability of the fracture classifications to predict mal-tion. Method of class$cation
Dorsal angle (degrees)
Mean values of the radiographic parameters
Radial shortening (mm)
Injury
Union
Injury
Union
18.06 24.00 18.00 30.48
2.31 11.46 0.50 14.10
5.25 4.82 6.88 7.44
23.26 21.89
9.31 11.36
19.54 21.00 29.81 34.13 36.17
Flattening of radial angle (degrees) Injury
Union
2.44 4.31 5.00 4.65
7.38 9.02 9.63 14.57
4.94 8.44 8.88 8.52
4.99 7.41
3.87 4.76
8.98 13.51
1.55 8.83
10.55 9.37 7.43 8.50 16.17
4.75 5.91 5.65 7.63 10.08
4.88 4.73 2.24 4.63 4.25
8.20 9.93 10.88 14.13 20.17
9.23 8.46 4.87 8.75 8.75
23.52 21.79 24.68 26.33
7.64 10.42 9.19 10.92
4.30 5.47 5.48 8.07
3.50 3.11 3.91 5.39
8.73 10.42 9.92 13.67
7.76 4.84 7.82 9.87
19.15 27.11 35.39
5.16 12.55 14.71
4.00 6.44 9.78
3.37 4.68 5.11
6.55 11.55 19.78
6.39 8.98 9.67
Comminution 1 2 3 4 Gartland & Werley 1 3 Lidstriim 2a 2b 2c 2d 2e Frykman 112 314 516 718 Older 2 3 4
Fig. 3
152
Case I A 24-year-old man sustained a badly-displaced and heavily-comminuted fracture which was treated by external fixation. The fixator frame was removed after three weeks because of persistent infection. At this time the fracture was firm and unmoveable but mildly tender, indicating that union was not complete. The fracture was splinted in a forearm cast and these X-rays show the position of the fracture in the cast.
Fig. 4
Case I A week later, whilst within the cast, the fracture had collapsed to the position shown. As the fracture site had been unmoveable a week earlier, the collapse must be a different process to the acute loss of position which occurs early in treatment and which is amenable to re-manipulation. This is clearly a collapse of uniting bone.
THE JOURNAL
OF HAND SURGERY
THE UNSTABLE
instability, mal-union is common and, in the present study, the 95% of fractures which had retained their manipulated position after one week of splintage subsequently lost some of this, with an increase of 8.22” in dorsal angle and loss of 5.61” radial angle and 3.26 mm radial length. This late deterioration in the fracture’s position occurs slowly and progressively, and is a manifestation of its chronic instability. Although chronic instability has been previously recognised (Anderson and O’Neil, 1944; Charnley, 1961; Spira and Weigl, 1969) it has never been specifically discussed and , as a result, the two types of instability are still discussed as one, implying a common cause (Solgaard, 1986). It is common experience that attempted remanipulations two weeks or more after the injury rarely improve the fracture’s position, despite continued radiographic deterioration (Figs. 3 and 4) and this is clearly a consequence of a gradual collapse of uniting cancellous bone. Charnley (1961) recognised this chronic collapse and claimed it to be an essential pre-requisite for strong union in cancellous bone, requiring control rather than prevention. Hitherto intra-articular involvement and radiographical comminution have been assumed to be the most accurate indicators of the fracture’s stability and have been widely cited as indications for operative management (Cooney et al., 1979; Grana and Kopta, 1979; Wagner and Jakob, 1985). The only previously published study aimed at identifying those fractures destined to m&union is that of Solgaard (1985) who concluded that the classification described by Older et al. (1965) was the most accurate predictor, although neither the pattern nor the extent of mal-union were studied. The classification described by Older is based upon both comminution and displacement (principally radial length, but also including dorsal angle) and thus does not allow for an assessment of the relative contribution of each factor to the
COLLES’ FRACTURE
Fig. 6
subsequent mal-union. In the present study, the effects of initial displacement, comminution and intra-articular involvement have each been studied in isolation. The results show that whilst those relatively uncommon (22%) fractures with no comminution of the dorsal cortex are capable of resisting mal-union in dorsal angulation, this pattern of mal-union regularly complicates the remaining 78% of fractures in which dorsal comminution has occurred (Figs. 5 to 7), the extent of the mal-union depending upon the severity of the initial displacement (Table 8). Mal-union in radial shortening or by flattening of the radial angle occurs independently of radiographitally-visible comminution and may be predicted only by
Fig. 7 Fig. 5
Case 2 A Colles’ fracture
VOL. 14-B No. 2 MAY 1989
sustained
by a 19-year-old
secretary.
Case 2 Anatomical reduction was attained by manipulation. This post-reduction X-ray shows the comminution of the dorsal cortex. A check X-ray a week later showed the reduction to be maintained and the plaster was completed.
Case 2 Radiographic examination four weeks after the injury shows the fracture to have mal-united in dorsal angulation but with minimal radial shortening.
153
N. H. JENKINS
the initial respective displacements (Table S), presumably reflecting comminution at trabecular level. In no instance is the development of mal-union dependent upon articular involvement. Conclusions
A minority (5%) of manipulated Colles’ fractures sustain an early loss of position which may be identified by early radiography and corrected by re-manipulation. In most cases, however, chronic instability produces a progressive collapse of the uniting cancellous bone. Those fractures which will mal-unite with radial shortening or by flattening of the radial angle are those with significant displacement of these two parameters at presentation, regardless of the radiographical pattern of comminution or the presence of articular involvement. A minority of fractures are resistant to mal-union in dorsal angulation and may be identified by a radiographically intact dorsal radial cortex. Acknowledgments I wish to thank Mrs Janice
Sharp (Department of Medical Illustration, University Hospital of Wales) and Mr Keith Bellamy (Medical Photographer, Cardiff Royal Infirmary) for their help in illustrating this paper. In addition, I extend my grateful thanks to Mrs Marjorie James for typing the manuscript.
COONEY, W. P., LINSCHEID, R. L. and DOBYNS, .I. H. (1979). External Pin Fixation for Unstable Colles’ Fractures. Journal of Bone and Joint Surgery, 61A:6: 840-845. FRYKMAN, G. (1967). Fracture of the distal radius including sequelaeshoulder-hand-finger syndrome, disturbance in the distal radio-ulnar joint and impairment of nerve function. Acta Orthopaedica Scandinavica, Supplement 108. GARTLAND, J. .I. and WERLEY, C. W. (1951). Evaluation of healed Colles’ fractures. Journal of Bone and Joint Surgery, 33A: 4: 895-907. GRANA, W. A. and KOPTA, J. A. (1979). The Roger Anderson Device in the Treatment of Fractures of the Distal End of the Radius. Journal of Bone and JointSurgery, 61A: 8: 1234-1238. JENKINS, N. H., JONES, D. G. JOHNSON, S. R. and MINTOWT-CZYZ, W. J. (1987). External fixation of Colles’ fractures. An anatomical study. Journal of Bone and Joint Surgery, 69B : 2 : 207-211. LIDSTROM, A. (1959). Fractures of the distal end of the radius. A Clinical and Statistical Study of End Results. Acta Orthopaedica Scandinavica, Supplement41. McQUEEN, M., MacLAREN, A. and CHALMERS, J. (1986). The value of remanipulating Colles’ fractures. Journal of Bone and Joint Surgery, 68B: 2: 232-233. OLDER, T. M., STABLER, E. V. and CASSEBAUM, W. H. (1965). Colles fracture: evaluation and selection of therapy. Journal of Trauma, 5: 4: 469476. SOLGAARD, S. (1985). Classification of distal radius fractures. Acta Orthopaedica Scandinavica, 56: 3: 249-252. SOLGAARD, S. (1986). Early displacement of distal radius fracture. Acta Orthopaedica Scandinavica, 57 : 3 : 229-23 1, SPIRA, E. and WEIGL, K. (1969). The Cornminuted Fracture of the Distal End of the Radius. Reconstruction Surgery and Traumatology, 11: 128-138. STEWART, H.D., INNES, A. R. and BURKE, F. D. (1985). Factors affecting the outcome of Colles’ fracture: an anatomical and functional study. Injury, 16: 5: 289-295. WAGNER, H. E. and JAKOB, R. P. (1985). Operative Treatment of Fractures of the Distal Radius Using External Fixation. Unfallchirurg, 88: 11: 473480.
References ANDERSON, R. and O’NEIL, G. (1944). Cornminuted fractures of the distal end of the radius. Surgery, Gynecology and Obstetrics, 78: 4: 434-440. CHARNLEY, J. The Closed Treatment of Common Fractures, 3rd ed. Edinburgh, ChurchillLivingstone, 1961: 128-142. COLLERT, S. and ISACSON, J. (1978). Management of Redislocated Colles’ Fractures. Clinical Orthopaedics and Related Research, 135: 183-186.
154
Accepted: 2 December 1988 892 Newport Road, Rumney, 0
1989 The British
Society
Cardiff,
for Surgery
South Clamorgan
CF3 8L.l
of the Hand
02667681/89/0014-0149/%10.00
THE
JOURNAL
OF HAND
SURGERY