Introduction

Distal radius fractures (DRF) are the most common type of fracture of the upper extremities and incidence is expected to rise due to a growing elderly population [1]. Especially white women over 60 years of age have a 15% higher lifetime risk of DRF, than men of similar age. In addition, DRF in the elderly are often associated with poor bone quality and osteoporosis [2,3,4].

Historically, DRF were conservatively treated by closed reduction and immobilization or K-wires. Following the introduction of palmar angular stable locking plates in the 2000s and the excellent results using internal fixation, a treatment shift occurred away from K-wires or external fixator to palmar plate fixation. Thus, dorsally displaced DRF can be stabilized from palmar, without the increased risk of irritation to the extensor tendons [5,6,7,8,9,10,11,12]. In addition, stabilization by palmar locking plates provides enough stability to enable early active wrist rehabilitation without immobilization. Multiple studies showed a significantly improved functional outcome compared to immobilization and an early mobilization post-surgery has no increased risk of secondary loss of reduction and complications [13,14,15].

As incidence of DRF rises and the number of patients treated by palmar locking plate increases, literature remains interested in the optimal treatment method, clinical outcomes and complication rates [13, 16]. Complication rates after palmarly stabilized DRF are reported up to 39%, whereas other studies documenting outcome after DRF showed good functional and radiological results [5, 8, 17,18,19,20,21,22].

Main aim of this study was to investigate correlation between radiological and clinical outcome with a minimum follow-up of one year.

Materials and methods

Institutional review board approval was obtained for this retrospective follow-up study. All patients treated with a palmar angular stable locking plate from 01. January 2015 to 31. December 2016, that met inclusion/excision criteria, were included in this study and invited in writing and by telephone to attend the follow-up investigation. Three invitation letters were sent to each patient. Failure to reply after the third invitations was classified as a non-responder.

Written informed consent was obtained from all participating patients. They were treated exclusively at our hospital, which is certified as a European Hand Trauma Center by the Hand Trauma Committee of the Federation of European Societies for Surgery of the Hand (FESSH).

Indications for surgery included a displaced DRF with a dorsal tilt of more than 15 degrees, an intra-articular step of more than 1 mm, a radial shortening of more than 2 mm or an incongruency in the distal radioulnar joint in the standard radiographs.

Inclusion criteria stipulated: (1) follow-up interval of at least one year (2) age 18 or older, (3) isolated acute displaced DRF, (4) surgical procedure by open reduction and stabilization using only palmar locking plates. Exclusion criteria included: (1) below 18 years of age, (2) open fractures, (3) patients who received the palmar plate fixation in another hospital, (4) polytrauma, (5) additional injuries to the involved hand, (6) corrective osteotomies, (7) bilateral DRF, (8) DRF stabilized with additional k-wires, screws or external fixator, and (9) past injuries or diseases which could influence the outcome.

From 2015 to 2016, a total of 524 patients were stabilized by palmar locking plate. Of these, 117 patients had to be excluded (not eligible), 11 lived in another country and 25 refused to attend the follow-up investigation. Another 141 could not be contacted or did not respond after three written invitations. Therefore, the final analysis totalled 230 patients.

Surgical procedure

All procedures were performed using either general or regional anesthesia in a supine position, with fluoroscopic assistance and a pneumatic arm tourniquet of 250 mmHg.

A standard palmar-radial approach over the flexor carpi radialis tendon was chosen. The flexor carpi radialis tendon was retracted ulnarly and the forearm fascia was dissected. The pronator quadratus was incised radially and elevated off the radius.

The fracture was reduced under image intensification and, when necessary, temporarily fixed with K-wires. The angular stable plate (Medartis® Aptus® 2.5 trilock distal radius locking plate) was placed on the palmar aspect of the distal radius and initially fixed with a bicortical screw through the gliding hole. After ensuring exact positioning of the plate under image intensifier, the remaining plate holes were filled with angular stable screws. Care was taken that the screws at the articular surface were placed subchondrally to prevent dorsal protrusion. Screw length was taken 2 mm shorter to prevent a dorsal protruding. The pronator quadratus was sutured if possible. Previous studies already showed no influence of a pronator quadratus refixation on the outcome [23, 24]. Thus, a refixation was not analyzed.

In 209 patients the Medartis® Aptus® (Medartis®, Basel, Switzerland) 2.5 correction plate, in 14 the flexor pollicis longus (FPL) plate, in 5 the fracture plate and in 2 the adaptive plate was used.

Carpal tunnel release was only performed if the patients showed preoperative symptoms of median nerve compression.

On the first postoperative day, the initial dressing was exchanged for either a thermoplastic splint or a nonremovable plaster cast for 5 weeks. Type of fixation was determined by the operating surgeon.

All patients started hand therapy of the free joints (shoulder, elbow, fingers) for both upper extremities on the first postoperative day. After cast or splint removal the wrist was then included in the specialized hand therapy programme.

Outcome evaluation

Each of the patients, who returned for the follow-up investigation, underwent a standard X-ray of the wrist in two planes (anteroposterior and lateral view). Range of motion (ROM) was measured in extension, flexion, supination, pronation, radial- and ulnar deviation at the follow-up investigation. Demographic data included age, gender, injured hand, interval between surgery and follow-up. In addition, pain according to the visual analog scale [VAS, ranging from 0 (no pain) to 10 (worst possible pain)], grip strength in kilograms [Jamar; Sammons Preston Rolyan, Mississauga, Ontario, Canada; minimal clinical important difference (MCID) 6.5 kg or 19.5% decrease [25]] were analyzed.

To analyze the incidence of complications and functional outcome depending on age, the patients were divided into younger than 65 (< 65 years) or equal or older than 65 years (≥ 65 years) according to the definition of the World Health Organization (WHO) for geriatric designation [26].

Self-assessment by patients was registered on the Shortened Disabilities of the Arm, Shoulder and Hand (QuickDASH) questionnaire (0–100 points, MCID 10 points [27]) [28], the Patient-Rated Wrist Evaluation (PRWE) Score (0–100 points; MCID 11.5 points [29]) [30], and modified Green O’ Brien (Mayo) Wrist Score (0–100 points) [31].

All the intraoperative and postoperative complications that were documented in the surgical write ups were recorded and each return evaluation was analyzed for complication. All types of revision surgeries were documented. Complex regional pain syndrome (CRPS) was diagnosed clinically on the basis of the Veldman’s criteria [32, 33].

Frequency and reason for hardware removal was gleaned from the medical records, although hardware removal was not regarded as a complication in itself. Complications associated with hardware removal were analyzed separately.

Radiological reviews

The primary (pre-reduction), immediate postoperative as well as final radiographs were checked for alignment and intra-articular step-off. The fractures were classified according to the Arbeitsgemeinschaft für Osteosynthesefragen (AO) classification by Müller et al. [34]. In addition, fractures of the ulnar styloid process were documented and classified into fractures of the tip and base.

An acceptable reduction was defined as 10 degrees of dorsal tilt, 15 degrees in radial inclination, 2 mm ulnar variance and 2 mm of articular incongruity [17, 35].

In the anteroposterior radiographs, radial inclination and ulnar variance according to Gelberman and in the lateral radiographs, the palmar tilt was measured [36]. Fracture healing was defined as bony bridging of the radial, ulnar, and dorsal cortical aspects of the distal part of the radius [17].

The lateral X-ray verified the plate position and was subsequently classified according to Soong et al. in Grade 0, I and II [37]. The Medartis® Aptus® FPL plate was excluded from the Soong classification. The specific design of the FPL plate (gap in the plate for the FPL tendon), allows for positioning distal to the Watershed line [38].

Statistical analyses

Shapiro-test was used to test the assumption of normal distribution of scaled parameters. Since this assumption could not be maintained, the outcome was analyzed non-parametrically using Mann–Whitney U test. Data is presented as mean ± standard deviation, significant results in addition with mean difference (MD) and 95% or 99% Confident Interval (CI).

To compare scaled parameters between AO fracture types and Soong grades (grade 0, I and II) the Kruskal–Wallis-test was used. If Kruskal–Wallis test was significant, post-hoc analysis was performed using the Mann–Whitney U test. Chi-square was used for testing categorical data.

Pearson’s bivariant correlation was run for analyzing relations between radiological parameter, ROM, grip strength and patient-reported outcome measures (PROMs) at the final check-up. Strength of the Pearson correlation coefficient (r) was classified according to Cohen [39] in small (0.1 < |r| < 0.3), medium (0.3 < |r| < 0.5) and strong (|r| > 0.5).

Due to multiple testing, p was corrected according to Bonferoni: threshold for statistical significance for clinical results was a p < 0.004 and radiological results a p < 0.006. For all other tests threshold for statistical significance was p < 0.05.

Results

A total of 230 patients (154 female, 76 male) returned for the follow-up investigation with a mean age of 59 years (range 18–83 years) and follow-up interval of 20 months (range 12–50 months).

Detailed demographic data is presented in Table 1

Table 1 Patients demographic and fracture classified according to the AO classification
Full size table

Radiological results

At the final check, all DRF and 105/123 (85%) of the ulnar styloid showed bony healing. Radial inclination was a mean of 24.8° ± 4.2°, palmar tilt 7.6° ± 7.3° and ulnar variance 1.8 ± 1.4 mm. From post-operative measures to the last follow-up there was a mean loss of reduction in radial inclination of 2.6° ± 3.4°, in palmar tilt of 1.9° ± 4.6° and ulnar variance of 0.64 ± 3.0 mm.

Bivariant correlation analysis showed a significant small positive correlation between radial inclination/flexion (r = 0.18, p = 0.01) and ulnar variance/QuickDASH (r = 0.18, p = 0.01). A significant small negative correlation was additionally registered between ulnar variance/grip strength (r = − 0.18, p = 0.04) and ulnar variance/Mayo Score (r = − 0.23, p = 0.001). No significant differences could be found between an acceptable (< 2 mm) and unacceptable (> 2 mm) ulnar variance in VAS (p = 0.74), grip strength (p = 0.23), QuickDASH (p = 0.04), PRWE (p = 0.05) and Mayo Score (p = 0.04).

Detailed functional outcome is summarized in Table 2.

Table 2 Range of motion [(mean ± SD) degree], pain, grip strength and patient-reported outcome in patients with distal radius fractures stabilized by palmar angular stable plate
Full size table

Subgroup analysis for acceptable/unacceptable ulnar variance in patient under and over 65 years of age was performed. No significant differences could be found in respect of VAS (p = 0.82; 0.65), grip strength (p = 0.67; 0.35), QuickDASH (p = 0.35; 0.03), PRWE (p = 0.33; 0.02) and Mayo Score (p = 0.10; 0.17).

Kruskal–Wallis Test showed significant differences in loss of reduction in ulnar variance for fracture type (p = 0.03). Post-Hoc analysis showed a significant higher loss of reduction for ulnar variance comparing type B3/A2 (p = 0.01), C3/A2 (p = 0.04), B3/A3 (p = 0.04), B3/C2 (p = 0.02) or C3/C2 (p = 0.03) fractures.

However, no significant difference in facture type in general [AO Type A vs. B vs. C] and loss of reduction in palmar tilt (p = 0.47), radial inclination (p = 0.66) or ulnar variance (p = 0.13) could be found.

Loss of reduction depending on AO type fracture is summarized in Table 3.

Table 3 Loss of reduction (mean ± SD) at final follow-up according to AO fracture type
Full size table

No significant differences in radial inclination, palmar tilt, ulnar variance or loss of reduction could be found between men/women, under/over 65 years of age and type of immobilization.

Detailed radiological results are summarized in Table 4.

Table 4 Radiological outcomes [Mean ± SD] in patients with distal radius fractures stabilized by palmar angular stable plate
Full size table

Clinical results

At the final follow-up mean pain according to the VAS was 0.41 ± 1.16, ROM in extension/flexion 150.0° ± 19.1°, in pro-/supination 179.9° ± 7.4° and radial-/ulnar deviation 52.2° ± 6.2°. Grip strength was a mean of 26.0 ± 11.6 kg or 91% of the uninjured hand. The mean QuickDASH was 11.2 ± 14.3 points, PRWE 9.5 ± 14.2 points and Mayo Score 85.8 ± 13.4 points.

Patients under 65 years of age showed a significantly better grip strength than patients over 65 (MD 8.0 kg, 99% CI 4.5–11.4, p < 0.001), but there was no significant difference compared to the uninjured hand (p = 0.53). For ROM, pain and PROMs no significant differences could be found between the groups.

Between women/men, with/without an additional fracture of the ulnar styloid or patients with/without complications no significant differences with regard to ROM, pain, grip strength and PROMs could be found. In addition, no differences between a tip or base of the ulnar styloid fractures were observed.

Detailed clinical results for each group are presented in Table 2.

Complications

A total of 30 complications (13.0%) occurred in 29 patients. Most common complications included CTS in 10 patients (4.3%), intra-articular screws in 7 patients (3.0%) and CRPS in 4 patients (1.7%). Complications are summarized in Table 5.

Table 5 Complications after distal radius fractures stabilized by palmar angular stable plate
Full size table

AO type C showed the highest complication rate of 20/30 (67%), whereas 14/20 (70%) were found in AO type C3 fractures. However, only 9/30 (30%) occurred in AO type B and 1/30 (3%) in AO type A. No significant differences could be found in complication rate between AO type A, B or C (p = 0.10) fractures.

Eight of the 10 patients with a CTS were treated by endoscopic carpal tunnel release and hardware removal, the remaining two were treated conservatively (splitting and physiotherapy) and showed a complete regression. All patients with CRPS were treated conservatively with splinting, hand therapy, dimethylsulphoxide (DMSO) ointment, non-steroidal anti-inflammatory drugs and vitamin E.

Intra-articular screws were seen in seven patients. In four cases screws protruded into the distal radioulnar joint and were subsequently removed. The other three presented with a loss of reduction, which necessitated plate removal.

Tendon irritation was seen in five patients. In three cases there was irritation of the extensor tendons (caused by protruding screws) the other two cases irritation of the flexor digitorum profundus tendon of the index finger. All tendon irritations were treated by plate removal.

Only one rupture of the extensor pollicis longus (EPL) tendon occurred and was treated by an extensor indicis transfer and plate removal. In the performed CT scans no protruding screws as a cause for the tendon rupture were found. No flexor tendon ruptures were observed.

A superficial infection was observed in one hand, which was treated conservatively (antibiotics and splinting).

No significant differences could be found in incidence of complications and plate type (p = 0.22), age [(over/under 65 years of age); p = 0.47], gender (p = 0.50), or post-operative immobilization [cast/thermoplastic splint (p = 0.31)].

Plate removal was performed in 52 of the 230 (23%) patients. 20/93 (22%) of the plates with Soong grade 0, 22/96 (23%) of the Soong grade I and 5/27 (19%) with Soong grade II had been removed. No significant differences in frequency of plate removal were found between the Soong groups (p = 0.84).

Discussion

Distal radius fractures are one of the most common fractures in the upper extremities and the incidence is expected to continue rising due to the growing elderly population [40,41,42,43,44,45]. Since the introduction of the palmar plates in the early 2000s plus the initial reports of low complication rates and good functional outcomes, palmar plate fixation has gained popularity in treating DRF [5, 46, 47]. Palmar locking plate fixations enables a stabilization of dorsally displaced fractures without the increased risk of tendon irritation compared to dorsal stabilization [5, 6, 48]. Thus palmar fixation of DRF provides enough stability to allow an early rehabilitation with active wrist mobilization. Thereby, better functional outcomes can be achieved in the early rehabilitation phase without the increased risk of a loss of reduction or further complications [13, 49]. Therefore, functional outcome and reported complications after operatively treated DRF remain current in the literature including a comparison of the various available treatment options.[8, 50]. Today the optimal treatment options for DRF are under debate, but a recent Network Meta-analysis concluded, that plate fixation offers the best results in terms of early functional outcome and reduction of fracture healing complications [51].

Patients in this study, with a mean age of 59 years and a minimum follow-up of one year (mean 20 months) showed a good functional and in mean an “acceptable” radiological outcome. The QuickDASH averaged 11 points, PRWE 10 points and Mayo Score 86 points.

Several studies compared relations between radiological parameters and functional outcome and suggested that an unacceptable reduction would lead to poorer results in PROMs [52, 53]. Mulders et al. [53] demonstrated in a systematic review and meta-analysis a significant correlation between an unacceptable reduction and PROMs, but all differences were minor, under the MCID, and therefore unlikely to be clinically relevant.

A positive ulnar variance after DRF is directly related to ulnar sided wrist pain, decreased grip strength and poorer PROMs, especially, in patients under 65 years of age with higher functional demands [35, 54]. Grewal and Mac Dermid’s study [55] reported significant and clinically important differences of 13 points for ulnar variance in both the DASH and PRWE. Our study also included a significant correlation between ulnar variance and grip strength, QuickDASH and Mayo Score in terms that a higher ulnar variance results in a lower grip strength and worse QuickDASH and Mayo Score. Although, an unacceptable ulnar variance showed no significant greater pain scores, lower grip strength or worse PROMs compared to the acceptable ulnar variance and all results were under the MCID. Even subgroup analyses in ulnar variance in patients under/over 65 years of age showed no significant or clinical important differences.

Complications after palmar stabilized DRF are well reported in the literature. First reports from Orbay et al. suggested a complication rate of 3% [56], respectively 4% [57], but later studies reported complications up to 60% [5, 8, 17, 19,20,21, 38, 50, 58,59,60,61]. In a recent systemic review Alter et al. [50] analyzed complications after palmarly stabilized DRF. They reported a complication rate of 15% in 3.911 operatively treated DRF with palmar locking plate. Complication rate of 13% in this study is comparable to previously published studies and the low complication rate reflects the familiarity with the implant and large numbers of treated DRF by palmar locking plate (average 262 DRF per year). Only one implant system was used, which is known to lower the risk for complications [62]. The most common complications in this study included CTS (4.3%), intra-articular screws (3.0%) and CRPS (1.7%). No significant impact on the complication rate could be found for age (over/under 65 years), gender or type of post-operative immobilization. Occurrence of a complication showed no significant or clinical important influence on the final functional outcome.

CTS is a very widespread condition, but is also known to occur in patients with DRF. Incidence after surgically treated DRF ranges from 7 to 15% [63], regardless of the treatment method. CTS after DRF is not attributed to the hardware itself, but more to the trauma to the median nerve caused by the fracture and/or fracture healing with callus bone formation or residual malunion [50]. On the other hand, because CTS is so common, it is quite possible that many patients with DRF have a pre-existing undiagnosed/asymptomatic CTS that only becomes symptomatic or to their attention after the injury [64]. We therefore, doubt that CTS is a direct complication after surgically treated DRF and it is questionable that the palmar plate can actually cause CTS.

Intra-articular screws are also frequently reported in the literature between 0.5 and 1.3% [5, 18, 62] and not only caused by malpositioning, but also due to loss of reduction or secondary fracture dislocation. Even the use of angular stable screws does not preclude secondary displacement [5]. Intra-articular screw penetration can result in a destruction of the radiocarpal joint, causing malunion, osteoarthritis and clinical failure. In these cases when conservative treatment fails (besides screw removal), salvage procedures are often necessary to relieve symptoms and improve ROM [58, 65]. Our study showed a slightly higher rate than in a previously published study with 7/230 (3%). Four out of seven were due to penetration into the distal radioulnar joint and 3 occurred due to a loss of reduction resulting in a dorsal shift of the screws into the radiocarpal joint.

CRPS is closely associated to fractures of the distal radius with an incidence between 1 and 6% [5], but is also commonly seen in injuries to the upper extremities in general [5, 8, 50, 66, 67]. However, it remains a clinical diagnosis and the pathomechanism is still not fully researched. This may, however, be related to an over excretion of cytokinins, mitochondrial dysfunction in the affected upper extremity, as well a genetic predisposition does exist [68,69,70]. We agree with Esenwein et al. that CRPS is a complication, that cannot be influenced by the surgeon [5].

Some of the limitations should be addressed before interpreting this study. Firstly, it’s based on the retrospective type of the study. The most serious drawback in retrospective studies at big trauma centers is the loss of follow-up. The study included a total of 524 operatively treated DRF in the study period. Of these 117 had to be excluded, resulting in 407 potential candidates. Of these 177 (43%) couldn’t be followed-up for several reasons. Thus, clinical results and complication rate could be biased. On the one hand, one might assume that patients who do not return have no complications and are asymptomatic, indicating that the complication rate is overestimated and the clinical results are better than reported. Alternatively, patients with complications or problems could simply have been transferred to another hospital.

In addition, there is no unique definition in the literature for an “acceptable” postoperative radiological result and a wide range for cut-off values does exist. It would be desirable, that further studies focus on specific cut-off values to determine which radiological parameters would affect PROMs.

At the follow-up radiographs were only taken of the injured wrist and not from the contralateral wrist. Therefore, no comparison with the uninjured wrist was possible. This might explain the discrepancy in the results in ulnar variance, showing a significant correlation to grip strength and PROM, but no differences between an unacceptable and acceptable ulnar variance. Further studies should consider this issue and investigate the impact of radiological differences between the injured and healthy wrist to PROMs.

Another limitation is that the study was not focused on one particular outcome parameter (e.g., QuickDASH Score at the last follow-up), resulting in multiple testing. P-value had to be corrected according to Bonferroni, and therefore, the study might be underpowered in some subgroup analyses. At the final follow-up examination, the X-rays of only 7/230, respectively, 3/230 patients showed an unacceptable palmar tilt or radial inclination. Due to the small sample size a comparison between an acceptable and unacceptable radiological result depending on functional outcome was not possible.

The minimum follow-up interval of this study was one year, with a mean of 20 months. Therefore, not all complications that typically occur later, for example, tendon rupture, are covered in this study.

Conclusion

The majority of surgically treated distal radius fractures using palmar locking plates, result in a good clinical and radiological outcome without complications. A positive ulnar variance showed a significant correlation to a decreased grip strength and worse QuickDASH or Mayo score, but an unacceptable ulnar variance (> 2 mm) is not associated with a significant or clinical important worse functional outcome. Age (< 65/> 65 years), gender, type of immobilization and fracture of the ulnar styloid showed no significant correlation to a higher risk for complications, loss of reduction or impaired functional outcome.

A total of 30 complications occurred in 230 patients (13%), most common being CTS, intra-articular screws and CRPS. It is debatable whether CTS and CRPS are direct complications due to palmar plating, thus 16/230 (7%) were procedure specific and only 14/230 (6%) were plate related.