Pediatric forearm fractures with in situ intramedullary implants

Fractures of the forearm are one of the most common injuries seen in childhood [1, 2]. Most of these fractures can be treated by closed means with reduction and cast immobilization; however, unstable or open injuries often require surgical treatment to maintain adequate alignment [3, 4]. Intramedullary (IM) fixation with titanium elastic nails (TENS) or Kirchner wires (K-wires) has emerged as the most common method for fixation of forearm fractures in skeletally immature patients [5, 6]. While practices regarding removal of these implants vary considerably, implant removal is typically performed 6 months to a year after the index procedure [7‐9]. Refractures occur in 4–8 % of patients treated non-operatively, which has historically dictated the timing of removal of these implants [10, 11].

There are several case reports in the literature of fractures that occur after radiographic healing but prior to removal of the IM implants as well as description of this complication in larger series; however, no conclusions can be drawn about the incidence of this complication, optimal treatment after refracture with in situ implants, or possible risk factors leading to refracture [12‐21]. The purpose of this investigation therefore was to review our institutional experience with forearm fractures with in situ IM fixation and present the characteristics, treatment, and outcomes of these complications, as well as to estimate the frequency of this complication. We hypothesize that these patients will require a second operative procedure but will go on to uncomplicated union.

Descriptive statistics, including mean and proportion, were used to present the outcomes of our review.

Of the 485 patients treated with intramedullary implants for forearm fractures, six patients (1.2 %) with fractures with in situ IM implants were eligible for inclusion in this review (Table 1). There were 4 males and 2 females included, with a mean age of 10.6 years (Range 4–16 years) at the time of initial injury. All six patients had an injury to both forearm bones in the middle 1/3rd of the diaphysis. For patients 2, 3, and 6, surgery was performed within 48 h of injury. Patient 1 was taken to the OR 2 weeks after the injury, patient 5 1 month after injury, and patient 4, 9 days after injury, all after a loss of initial closed reduction was noted in clinic. Three patients (Patients 2, 3, and 6) had grade 1 open fractures, with patient 2 sustaining a segmental fracture of the radius and patients 2 and 6 developing ulnar nerve palsies at the time of injury that resolved over the course of follow-up. An open reduction was performed through the open wound in these three patients. Patient 4 required an open reduction at the initial surgery to successfully pass the IM implants. Another patient (Patient 5) required removal of his ulnar rod with retention of his radial implant 3 months after the initial surgery due to implant prominence and discomfort.

All six patients suffered a fracture with in situ implants after radiographic confirmation of healing but before implant removal at a mean time of 13.0 months from index procedure (Range 3–45 months) and 11.1 months from healing (Range 0–43 months). Each sustained a second traumatic event that was similar in energy to the original mechanism leading to refracture. The mean angulation of the refracture was 28.4° (Range 4–51°). One patient (Patient 5) had a fracture distal to the radial implant that was adequately aligned. He was treated with closed casting without manipulation in the clinic and did not require a trip to the operating room for treatment.

The remaining five patients (83 %) had unacceptable alignment of their fractures and returned to the operating room for a second surgical procedure. One patient (Patient 6) had an attempt at closed reduction under conscious sedation in the emergency department due to concern for skin tenting, but had continued unacceptable alignment of the fracture. The remaining four patients were taken to the OR without an attempt at closed reduction in the emergency room due to surgeon preference. One patient (Patient 1, Fig. 1) had successful closed bending of the radial implant under general anesthesia in the OR with residual deformity of the ulnar implant and subsequent replacement of the ulnar implant. Two patients (Patients 3, 4) had removal of their intramedullary implants with replacement of new nails (Patient 3, Fig. 2). Patient 3 had an attempt at closed reduction in the OR prior to implant exchange with continued unacceptable alignment and underwent single bone fixation of the ulna with acceptance of the residual radial deformity because of the patient’s age. Patient 4 underwent replacement without an attempt at closed reduction. One patient (Patient 2, Fig. 3) had removal of nails with plate osteosynthesis because of his age and skeletal maturity utilizing stacked 1/3rd tubular plates for both the radius and ulna and one patient (Patient 6) with stacked 1/3rd tubular plates for the ulna and 3.5 mm LC-DCP plate for the radius (Synthes, West Chester, PA ). Patients 2 and 6 did not undergo an attempt at closed reduction prior to plating.

Patient 4 was diagnosed with type III osteogenesis imperfecta (OI), and suffered three forearm refractures with in situ implants. Two of these refractures were minimally displaced and were treated with immobilization without reduction. The third fracture was treated with removal and replacement of nails in the OR because of unacceptable alignment without an attempt at reduction. This patient presented with penetration of the ulnar implant approximately 1.5 months after revision surgery, requiring a return to the OR for removal with retention of the radial implant.

All patients went on to uneventful healing of their injuries at a mean of 3.6 months from revision surgery. There were no other complications during the follow up period for any of the patients after their refracture or revision surgery.

In our series of patients with forearm fractures treated with IM implants, we identified six patients who suffered a fracture while implants were still in place. This represented approximately 1.2 % of the patients treated with IM implants during the study period. As we hypothesized, all patients went on to successful healing of their fractures, but only five of six patients required a second trip to the operating room for revision surgery.

The timing for removal of IM implants for pediatric forearm fractures has historically been based on the rate of refracture, which can be as high as 4–8 % in patients treated non-operatively [10, 11]. The initial investigations into the use of titanium elastic nails for pediatric forearm fractures highlighted the risk of refracture after implant removal [10]. In the series from Nancy, France, implants were buried and removed after an average of 4.25 months for their first 50 patients. After observing 3 refractures in this group, they altered their practice and began removing implants at 10 months to 1 year following the index procedure, after which no further refractures were seen. Due to this experience, intramedullary implants are typically left in place for at least 6–12 months prior to removal to provide mechanical protection against refracture [23]. However, refracture prior to explant of these IM devices is a rare complication that has been recognized and described.

There are case reports in the literature of refractures occurring before removal of the implants and their treatment. Mittal et al. reported on one case in a 14 year old boy with 2.0 mm titanium nails in place [12]. The refracture occurred after radiographic confirmation of healing and at 5 months after the index surgery. An attempt at closed reduction was made, but this resulted in breakage of the ulnar nail and ultimately the nails required replacement in the operating room for adequate realignment. Shahid et al. described another successful technique in a 10 year old girl who refractured 3 months after fixation [13]. They returned to the operating room and withdrew the nails a short distance without removing the nail completely so a straight portion of the nail crossed the fracture site, resulting in improved alignment. She went on to uncomplicated healing.

Muensterer and Regauer describe a 13 year old boy with 2.5 mm titanium nails who suffered a refracture with 21° of angulation 1 month after initial fixation [14]. This patient underwent successful closed reduction with rebending of the implants, and the nails were removed 5 months later after healing. They went on to test the mechanical properties of both titanium elastic nails and stainless steel nails bent to 21° in vitro. They determined that the force required for permanent deformation of previously bent nails decreased 37 %, and there was no evidence of metal fracture or fatigue after one cycle of reversed bending. None of the nails tested fractured after five cycles of bending and reversed bending. On the basis of their experience and results, they suggest that closed reduction and re-bending of in situ intramedullary implants is a mechanically viable option for forearm fractures with in situ implants.

The patients in our series were treated by several different methods, which include closed casting, removal and replacement of nails, in situ bending of nails, and removal of nails with plate osteosynthesis. There is currently no consensus as to the best method of treatment of forearm fractures with in situ IM implants. On the basis of our series, these differing methods are all viable options for treatment, and the surgeon should be able to expect that uncomplicated healing will occur with whatever method is chosen. This is in agreement with the previously published case reports. Based on this study and the current literature, it is our preference to treat these injuries with removal of bent implants and revision fixation in the operating room. Closed reduction and rebending of the initial implants maybe attempted initially as a temporizing measure in setting of skin tenting and soft tissue compromise; however residual deformity of retained implants may lead to suboptimal bony alignment. When performing revision fixation, a low threshold for open reduction is needed, as often percutaneous passage of new IM implants is difficult. In patients at or nearing skeletal maturity, plate fixation should be considered.

This study has several limitations. First, the retrospective design is subject to selection and treatment biases. Second, due to the very rare nature of this complication, a small number of patients were available for this review. As such, it is not possible to recommend a specific technique for the treatment of or to draw any conclusions about predictors of refractures with in situ implants. Generalizability of the results may be limited, as all patients were from a single, high volume pediatric tertiary care center. Finally, one patient in the series was diagnosed with OI, which could be considered separately. However, given the rare nature of fractures with in situ implants, this patient was included to provide further information about the complication. As most implants will be left in place long term for patients with OI, it is important to keep in mind that normal growth of the arm can make removal of implants difficult or impossible, and treatment of fractures with in situ implants may need to be adjusted accordingly.

In conclusion, refracture of the forearm in pediatric patients with IM implants in situ is a rare but recognized complication, occurring in approximately 1.2 % of patients treated with IM implants. Despite the numerous options for treatment of these injuries, all refractures in our series went on to uncomplicated healing following appropriate bony realignment and stabilization.