While AO type A and B fractures have been treated with palmar plates for an extended period of time, highly comminuted fractures have either been treated with a dorsal or combined approach, but not regularly through a palmar approach only. Only the invention of angle stable implants has opened the door to consider this approach in severely comminuted fractures. While palmar fixed angle implants could be the future for treatment of most Colles' fractures, the dorsal approach remains a good choice in highly comminuted fractures with a metaphyseal defect, when a bone graft is also required . A previous study from our workgroup was able to show a faster functional recovery with palmar plates, but it remained unclear if these advantages would persist over an extended period of time . Results show an advantage with the palmar plating system within the 12 month period. Patients regain most function and strength during the first 6 months, afterwards, only a slow, yet still measurable progress can be anticipated [11–13]. Function and strength were significantly improved in the group with palmar plates. This finding was present in flexion/extension, pro-/supination and radial/ulnar abduction. Although dorsal plates did show a comparable increase in range of motion, palmar plates had better function from 6 weeks on and this advantage was not lost over 12 months. Also postoperative grip strength was significantly better at 12 months after use of palmar plates, with similar results to those of other studies using palmar plates [1, 2, 6, 7, 11, 12, 14–17]. Interestingly the early postoperative functional advantage prevailed and thus can be regarded as the cornerstone to successful rehabilitation. This shows that functional recovery with palmar implants is not only faster but also more complete within 12 months. It can be hypothesized that longterm results may be similar as plate removal does improve functional outcome in patients with Pi-plates . Patients with dorsal plates benefited from hardware removal more than patients with palmar plates. They experienced a major reduction of pain levels so that no significant difference between both groups was present after 12 months. Likewise an increase in grip strength after plate removal was seen in the dorsal group. Nevertheless, this group did not completely catch up in terms of strength and function, where statistical significance prevailed after 12 months. Hardware removal cannot be expected to substantially increase ROM or decrease pain levels in patients with palmar plates. The Gartland-Werley score, which incorporates subjective data from patients among objective data, verified a faster recovery in the palmar plate group with significantly lower scores in the palmar plate group. On the other hand, the DASH score, which is subjective and based solely on patient perception, failed to show statistical significance between both groups. This means that patients with dorsal implants adapt to the situation comparable to patients with palmar plates and return to activities of daily living in the same manner.
Radiological results did not show statistically significant difference in regard to radial inclination. Palmar tilt did show a slightly increased result in the dorsal group compared to the palmar group. This proved to be statistically significant. It seems that if increased palmar tilt is desired, some degree of "overcorrection" has to be obtained. This may prove difficult since palmar plates would need additional bending. Nevertheless the palmar tilt of 9 degrees in palmar plates is sufficient in terms of function. A steeper palmar tilt in dorsal plates may be the result of the surgical technique, when the whole distal part of the fracture is reduced to the palmar side with the dorsal plate rigourosly preventing dorsal displacement. This aspect also can be noticed in the carpal sag. The carpal sag was increased in the dorsal plate group, which shows that a complete palmar shift of the fragment and the carpus occurs when the fracture is stabilized from dorsally. This may be due to a straight plate being applied to a curved dorsal aspect of the radius, especially as features like the Lister's tubercle would require a dorsal bend of the plate. A palmar overcorrection will, in some cases, lead to palmar dislocation of the fractures which has occurred in three patients. It has to be stated clearly that, in terms of anatomical restoration, a dorsal implant is by no means inferior to a palmar implant, and in certain aspects such as palmar tilt proved more effective in our group. Nevertheless, radiographs are not the only decisive aspect to judge a plating system as function is the most important aspect for the patient.
The complications encountered in the patients are similar to those described in the current literature and our earlier articles [4, 5]. For dorsal implants, complications rates as high as 60% have been described . The most troublesome complication of all, secondary dislocation resulting in a loss of reduction, only occurred with dorsal implants. Especially in AO-type C3 fractures, the lunate facet fragments are very challenging to reduce and retain in position with a dorsal plate only. This may also be due to the design of the Pi-plate, which only offers angle stability when pins are used. In a biomechanical evaluation, Martineau et al. have shown that smooth pegs provide less stability than screws in palmar plates . When dorsal plates are used, the palmar, proximal angulation of the pins and screws will further increase the possibility of dislocation, opposed to the palmar buttressing effect in palmar plates. Because pins are smooth, they do not develop comparable retaining power as screws may let fragments slide palmarly thus causing secondary displacement. Especially in multifragmentary situations it is impossible to secure every fragment with a pin and an additional screw. Palmar buttressing proves advantageous for this problem, by stabilising the fracture from the side with the thicker cortex and avoiding the dorsal comminution zone altogether. Although not present in our patients we are well aware of the fact that some palmarly treated fractures may also show dorsal dislocation and may require dorsal buttressing. With regard to extensor tendon irritations, the second most frequent complication of the dorsal plate, this concurs with a previous study of ours . Despite meticulous development of retinacular flaps to protect the transverse aspect of the plate from irritating extensor tendons, this will not be successful in all cases. Tendon problems cannot be blamed solely on the Pi-Plate, but are due to the presence of any hardware in the dorsal extensor compartments, as shown by tendon ruptures due to protruding screws in palmar plates [19, 20]. Therefore no estimation about possible lower complication rates with other dorsal implants that differ in shape and/or diameter can be given although it can be hypothesized that similar complications will be encountered regardless of the shape or material of the implant. It also has to be noticed that no complications were encountered after hardware removal.
There are limitations to this report. The cohort of patients was collected at a tertiary care center with expertise in both plating systems. It is not known whether these results can be generalised, as there is a referral-bias in our patient population. Also our inclusion criteria were deliberately strict to limit our patient population to elderly patients in which plate fixation may be problematic and produce less than ideal results. These patients require stable fixation for poor bone quality to allow quick rehabilitation . This again may increase complication rates compared to other studies when younger patients are included. Another weakness may be the follow-up period of one year. It can be righteously argued that such a period cannot be appropriately regarded as long term. Still, it has been shown by Kreder et al. that one year follow-up is sufficient . Nevertheless our follow up is insufficient to allow conclusions about "true longterm" occurrences such as posttraumatic osteoarthritis which may arise after more than a decade .
Overall we have found an advantage of palmar over dorsal plating in comminuted fractures of the distal radius, which supports our earlier findings . While it can be hypothesized that differences in function and radiological outcome will level out over the course of several years, the study has shown the key advantage of the palmar plate to be the faster recovery time. A faster recovery will not only reduce the cost after this injury, but especially in the elderly patient population may restore individual independence thus possibly preventing nursing home placement. While late complications of different plating systems may be similar, it is especially the elderly population who benefits most from faster recovery.