The association of intra-articular distal radial fractures with the development of osteoarthrosis1-4 and diminished wrist function5,6 in several studies suggests that even a well-reduced or stable single sagittal-plane fracture line between the scaphoid and lunate facets of the distal aspect of the radius may be associated with a poorer outcome than that associated with an extra-articular distal radial fracture. However, osteoarthrosis at the radiocarpal joint is not necessarily correlated with a diminished outcome or substantial impairment7,8, and restoration of articular congruity does not guarantee good function3.
Volar plate fixation of distal radial fractures is now commonplace9,10. A previous study of arm-specific disability after volar plate fixation indicated that articular incongruity, dorsal angulation, and lower income were the strongest predictors of disability three months after fracture, but only age and income were associated with disability one year after fracture11. Another study indicated that patients with a complex intra-articular fracture and those with an extra-articular fracture had comparable wrist motion and upper-extremity-specific health status one year after open reduction and volar plate fixation12.
The current study was designed to test the hypothesis that wrist function and health status did not differ between patients with a stable or well-reduced single sagittal intra-articular fracture and those with an extra-articular fracture of the distal aspect of the radius after open reduction and volar plate fixation.
We retrospectively analyzed a subset of the patients in a large, prospective, multicenter cohort study of open reduction and plate and screw fixation of fractures of the distal aspect of the radius that had been performed for another purpose13. That study enrolled 420 patients with a distal radial fracture between 2001 and 2004; inclusion criteria were an age of at least eighteen years, open reduction and internal fixation with plate and screws performed within ten days of injury, and an absence of prior open reduction and internal fixation treatment of the wrist. Exclusion criteria included general or local conditions that adversely affected bone physiology (such as tumor, hyperparathyroidism, and osteogenesis imperfecta), polytrauma (Injury Severity Score >1614), co-enrollment in other studies, or a history of drug or alcohol abuse.
We excluded forty-one of the 420 patients from our current study because of a lack of adequate preoperative radiographs, and three patients because they had bilateral fractures. The remaining 376 fractures were classified by two of the authors (J.S.S. and D.R.) according to the AO/OTA (Arbeitsgemeinschaft für Osteosynthesefragen/Orthopaedic Trauma Association) Comprehensive Classification of Fractures with use of standard posteroanterior and lateral radiographs15. Discrepancies in classification were discussed within a group of hand surgeons16. Computed tomography was not used to identify the articular fracture line. Fifty-seven of the fractures were Type A2, eighty-seven were Type A3, three were Type C1, and fifty-nine were Type C2.
We identified all patients who had a dorsally displaced fracture treated with a volar locking plate and who had been followed for at least one year. We then used a matched cohort design to select two groups with specific fracture patterns from among these patients. Thirty-seven patients had a simple intra-articular compression fracture (a single intra-articular fracture line in the sagittal plane only; Type C1 or C2; Fig. 1). Each of these patients was matched by age and sex with two patients with an extra-articular fracture (Type A2 or A3) (Table I).
Evaluation
Thirty-three (89%) of the patients with a single intra-articular fracture completed the two-year evaluation compared with sixty-three (85%) of the patients with an extra-articular fracture. Each patient was evaluated according to the modified system of Gartland and Werley17 at six months, one year, and two years after initial surgery. Patients also completed the Disabilities of the Arm, Shoulder and Hand (DASH)18 and the Short Form-36 (SF-36)19 questionnaire twelve and twenty-four months after surgery. Pain at rest and in motion was rated on a 10-point visual analog scale (VAS), on which 0 is no pain and 10 is the worst possible pain.
Radiographic Evaluation
The alignment of the distal aspect of the radius was measured on posteroanterior and lateral radiographs with use of commercial software (eFilm, Milwaukee, Wisconsin). Ulnar inclination, palmar tilt, ulnar variance, and articular congruity were measured with use of the standardized techniques described by Kreder et al.20. Because the films were digital and unscaled, ulnar variance and articular step-off were expressed with reference to the measured length of the capitate21 rather than in millimeters. Radiographic signs of osteoarthrosis were rated according to the system of Knirk and Jupiter1.
Statistical Analysis
Twenty-two parameters (i.e., outcome measures [Table II], radiographic measures [Table III], arthrosis, return to work, and complications) were assessed at each examination and regressed on two indicator variables, the presence of an intra-articular fracture line and the duration of follow-up, as well as on an interaction term that evaluated the collinearity of these two variables. All data for each outcome measure were pooled and analyzed together in one overall linear regression model, taking the repeated measurements of each patient into account.
The likelihood ratio test was used to study the overall effect of an intra-articular fracture line on each outcome parameter. The maximum-likelihood estimate of the full regression model (including the presence of an intra-articular fracture, the follow-up time point, and the interaction term) was compared with that of a null regression model (involving time only) in order to determine whether adding the intra-articular fracture line variable would significantly improve the model22. In a similar fashion, the full regression model was compared with a model without the time variable in order to determine the overall effect of time on outcome. Statistical significance was adjusted for multiple testing with use of a Bonferroni correction; specifically, a p value of <0.0023 (0.05 divided by twenty-two separate comparisons) was considered significant. This approach reduces the probability of identifying a spurious difference between the two groups simply because of the large number of parameters analyzed. Each outcome that exhibited a significant overall effect was analyzed further with use of the full regression model to estimate and test the difference between the two groups (effect size) at each follow-up time point22.
A post hoc power analysis showed that the initial sample size of 111 patients with a 1:2 allocation ratio was sufficient to provide 97% power to detect a minimum difference of 10 points in the DASH score, given a standard deviation of 13 points in this score. A total of eighty patients would provide 90% power to detect such a difference.
Source of Funding
Some of the authors are employees of AO Clinical Investigation and Documentation.
Patients with a Single Sagittal Intra-Articular Fracture (AO Type C1 or C2)
The thirty-seven patients with a single intra-articular fracture included thirty-one women and six men with an average age of fifty-six years (range, twenty-one to seventy-nine years) (Table I). The initial injury was the result of a fall from a standing height in eighteen patients (49%) and a higher-energy injury (e.g., fall from a height, motor-vehicle collision, or fall during sports) in nineteen patients. The left wrist was involved in sixteen patients (and was dominant in seven of these), and the right wrist was involved in twenty-one patients (and was dominant in eight). According to the Comprehensive Classification of Fractures15, one fracture was Type C1.2, one was C1.3, nineteen were C2.1, seven were C2.2, and nine were Type C2.3. Twenty-three patients (62%) had a concomitant fracture of the distal aspect of the ulna, which involved an avulsion of the tip of the ulnar styloid in two patients, a fracture of the base of the styloid in eighteen patients, and a fracture of the ulnar neck in three patients. One patient had a grade-2 open wound; all other fractures were closed. No patient had an injury of the ipsilateral upper limb. Sixteen patients were employed prior to the injury.
The mean time (and standard deviation) between the injury and the index procedure was 4.9 ± 3.4 days (range, zero to nine days). Open reduction and volar plate and screw fixation was performed with use of a 2.4-mm LCP plate (Synthes, Paoli, Pennsylvania) in twelve patients, a 3.5-mm LCP plate in twenty-one, and a DVR plate (DePuy, a Johnson & Johnson company, Warsaw, Indiana) in four patients. Prophylactic carpal tunnel release was performed in one patient. Ancillary stabilization included Kirschner wires in six patients and lag screws in three patients. A bone graft was used in three patients.
Seven patients (19%) experienced adverse events or had additional surgery, which included irritation of the volar plate requiring removal in two patients; backing out of an interfragmentary screw in one patient; irritation of the ulnar nerve requiring Kirschner wire removal in one patient; a tendon adhesion treated with tenolysis and carpal tunnel release in one patient; and disproportionate pain and disability in two patients, one of whom also had a loss of reduction of the volar lunate facet treated with a second surgical procedure that was unsuccessful in maintaining realignment of the lunate facet.
Patients with an Extra-Articular Fracture (AO Types A2 and A3)
The seventy-four matched control patients with an extra-articular fracture included sixty-two women and twelve men with an average age of fifty-six years (range, twenty-one to seventy-nine years) (Table I). The initial injury was the result of a fall from a standing height in thirty-one patients (42%) and a higher-energy injury (e.g., fall from a height, motor-vehicle collision, or fall during sports) in forty-three patients. The left wrist was involved in thirty-nine patients (and was dominant in fourteen of these), and the right wrist was involved in thirty-five patients (and was dominant in twenty-seven). Two fractures were Type A2.1, fourteen were A2.2, eleven were A2.3, forty-three were A3.2, and four were Type A3.3. Forty-nine patients (66%) had a concomitant fracture of the distal aspect of the ulna, which involved an avulsion of the tip of the ulnar styloid in six patients, a fracture of the base of the styloid in thirty-six, and a fracture of the ulnar neck in seven patients. One patient had a grade-1 open fracture; all other fractures were closed. No patient had an injury of the ipsilateral upper limb. Thirty-two patients were employed prior to the injury.
The mean time between the injury and the index procedure was 3.6 ± 3.5 days (range, zero to nine days). Open reduction and volar plate and screw fixation was performed with use of a 2.4-mm LCP plate in thirty patients, a 3.5-mm LCP plate in thirty-seven, and a DVR plate in seven patients. A prophylactic carpal tunnel release was performed in five patients. Ancillary stabilization included Kirschner wires in one patient and lag screws in three patients. A bone graft was used in four patients.
Fourteen patients (19%) experienced adverse events or had additional surgery, which included an extension deficit in one patient, rupture of the extensor pollicis longus tendon in one patient, numbness at the base of the thumb in one patient, a superficial infection treated with operative debridement in one patient, plate irritation in four patients (two of whom underwent implant removal during the study period), carpal tunnel syndrome requiring release in three patients, a regional pain syndrome in two patients, and excision of a ganglion of the extensor pollicis longus in one patient.
Effect of Articular Involvement
The patients with a single sagittal intra-articular fracture did not differ significantly from the patients with an extra-articular fracture on any of the outcome measures. However, compared with patients with an extra-articular fracture, patients with a single intra-articular compression fracture had a trend toward less pronation (95% compared with 98% of that of the contralateral arm; p = 0.02) and grip strength (76% compared with 81% of that of the contralateral arm; p = 0.03) at six months and a trend toward a smaller flexion-extension arc (118° compared with 128°; p = 0.03) at one year after surgery. These small differences did not result in a significant difference in any health status or functional outcome measure at any follow-up time point. However, although these findings did not reach significance with the number of patients available, they could be clinically relevant, indicating slightly slower recovery in the presence of an intra-articular fracture (Table II).
Improvement with Time
Patients in both groups improved significantly between the six-month and the one-year follow-up evaluations with respect to the arc of wrist flexion and extension, wrist flexion, wrist extension, the arc of forearm rotation, pronation, supination, the arc of radioulnar deviation, radial deviation, ulnar deviation, grip strength, and the DASH and Gartland and Werley scores.
Radiographic Evaluation
There were no significant differences in volar angulation, radial inclination, or ulnar variance at any follow-up time point (Table IIITable III). Articular step-off was seen in four (11%) of the thirty-seven patients with a single sagittal fracture at the time of final follow-up. Fourteen (38%) of the patients with a single sagittal intra-articular fracture had signs of radiocarpal arthrosis compared with six (8%) of the seventy-four patients with an extra-articular fracture of the distal aspect of the radius (p = 0.36, Fisher exact test).
We were unable to demonstrate a significant difference in wrist motion and function scores between patients with a stable or well-reduced single sagittal intra-articular fracture and patients with an extra-articular fracture of the distal aspect of the radius at any of the follow-up time points. Our finding of a trend toward less pronation and less grip strength at six months and toward a smaller flexion-extension arc at one year in patients with a single intra-articular fracture may suggest that these patients take slightly longer to recover than patients with an extra-articular fracture but eventually do equally well.
Articular incongruity has been identified by several authors as a predictor of radiographic evidence of osteoarthrosis1,3; however, there is limited correlation between the presence of radiographically evident degenerative changes and impairment or disability after operative treatment7,8. The prevalence of radiographic evidence of radiocarpal arthrosis was considerably higher in the patients with a well-reduced or stable intra-articular fracture (38%) than in those with an extra-articular fracture (8%). Since we did not find a difference in pain, wrist function, or health status at any follow-up time point, the radiographic diagnosis of radiocarpal arthrosis appeared to have little influence on the short-term outcome following the distal radial fracture.
Chung et al. found similar results in their study on the predictors of outcome following volar plate fixation of the distal aspect of the radius11. Articular incongruity (defined as a >1-mm articular step-off or gap), greater volar tilt, and decreasing income were the strongest predictors of outcome in sixty-six patients at three months postoperatively. However, age and income were the only predictors of arm-specific disability at one year after volar plate fixation of a fracture of the distal aspect of the radius. Catalano et al. found a strong correlation between residual articular displacement and the development of radiocarpal osteoarthrosis7. At an average of 7.1 years postoperatively, 76% of their patients had evidence of osteoarthrosis, but this finding was not associated with lower functional outcome.
The prevalence of arthrosis was higher in the patients with an intra-articular fracture than in the patients with an extra-articular fracture in our study even though most of the intra-articular fractures were well reduced, which is consistent with prior studies1-4. Since damage to the articular surface would be expected to lead to increased impairment and therefore greater disability, the absence of such findings in our study and in the studies by Chung et al. and Catalano et al. is counterintuitive. Rather than concluding that the intra-articular fracture and the arthrosis have no functional effect, one should also consider the possibility that the effect is simply much smaller than those of other factors23, and thus difficult to measure.
One of the limitations of our study is that the fracture classification and the radiographic measurements were made with use of only posteroanterior and lateral radiographs. In a separate study involving the patients from the same multicenter clinical trial, we identified variations in fracture classification between surgeons16, and such variation may have affected the results of our current study. Furthermore, since evaluation of residual step-off on standard radiographs is also subject to interobserver and intraobserver variability, the results of radiographic measurements should be interpreted with caution20. It is possible that other researchers analyzing the same data might find slightly different results because they might classify and measure the fractures differently than we did. Another shortcoming is the relatively short (two-year) duration of follow-up. It is unclear whether radiocarpal arthrosis is correlated with decreased outcome in the long term for patients with a single sagittal fracture. Another shortcoming is that the treating surgeon, who was not blinded to fracture type, made the diagnosis of arthrosis. Finally, our study was not designed and powered to test for equivalence. It was a retrospective analysis of prospectively collected data that passed a post hoc test showing that it had sufficient power to demonstrate noninferiority. Failure to reject the null hypothesis of noninferiority should not be misinterpreted as a demonstration of equivalence.
These data suggest that well-reduced or stable intra-articular distal radial fractures with a single sagittal intra-articular fracture line treated with open reduction and volar plate fixation have outcomes comparable with those of extra-articular fractures. A prospective cohort study designed to address this question specifically would provide stronger support for this hypothesis.