• <- Back to List

Foot & Ankle Trauma: Calcaneus Fractures

 

Calcaneus fractures have been a particular interest of mine. Historically the pendulum has swung between non-operative and operative treatment. Despite the advances made over the past 40 years making operative treatment safer and more effective, Henderson’s famous quote, “The man who breaks his heel bone is done” still looms large in my head. This axiom unfortunately applies in many situations where increased fracture severity portends a poor outcome. While the pendulum may have started to swing back toward non-operative treatment when Buckley’s randomized prospective study was published in 2002, I think the pendulum is swinging in a new direction that benefits patients sustaining calcaneus fractures: minimally invasive percutaneous approaches which help mitigate potential complications from operative management while allowing restoration of calcaneal architecture and better outcomes (Figures 1 & 2).

     

As my experience grows with the treatment of these difficult problems, questions have arisen as to how to approach these fractures differently. This has sparked several research studies. One such idea was sparked by patient A.M.

 

A.M. is a 26 year old male who sustained a closed calcaneus fracture after a fall from height. He was a smoker and sustained a Sanders 2A calcaneus fracture amenable to percutanous fixation (Figure 3). He underwent uneventful percutanous fixation using a minimally invasive sinus tarsi approach (Figure 4).

   

After obtaining a Harris heel view at his first postoperative appointment the sustentaculum screw was noted to be misplaced inferiorly irritating the flexor hallucis longus tendon (and irritating me…) (Figure 5) How did I make this mistake? While it is recognized as a difficult screw to place (as it is done from lateral to medial without direct visualization and the sustentaculum is a relatively small bony structure), I was left pondering this for months. Intraoperative Harris heel views I obtained during the procedure appeared to show that I had placed the screw correctly but clearly I had not. Still it was unsettling for me to chalk this up to “technical error” when I wasn’t sure why I made such an error.

 

To try to answer this question Ida Leah Gitajn, MD, Rull James Toussaint, MD and myself performed a research study to determine if the Harris heel view could accurately determine placement of a sustentaculum screw. On a cold winter Saturday we performed an anatomic study using a cadaver foot. Lateral and medial dissection was performed to remove the skin and subcutaneous tissues. A 4.0 cancellous screw was placed from lateral to medial in 5 different configurations using the same starting point: a screw placed (1) anatomically within the sustentaculum; (2) misdirected inferior to the sustentaculum; (Figure 8), (3) misdirected superior to the sustentaculum; (4) misdirected anterior to the sustentaculum; and (5) misdirected posterior to the sustentaculum. A large C-arm was utilized to obtain Harris heel views at five different angulations (10 – 50 degrees) for each screw configuration. We then analyzed the C-arm images to determine at which angulation correct screw placement could be confirmed for each cadaver. As there are important anatomic structures surrounding the sustentaculum our ability to determine accurate screw placement was important.

   

We found that a screw placed anatomically was noted to be radiographically within the sustentaculum in all five views. An inferiorly misdirected screw, which can endanger the flexor hallucis longus or the neurovascular bundle (the error I made in patient A.M.), appeared to be radiographically within the sustentaculum using the standard Harris heel view at 30, 40 & 50 degrees (Figure 6) but was confirmed to be misplaced inferiorly on modified 10 & 20-degree views (Figure 7). A superiorly misdirected screw, which can penetrate the subtalar joint or flexor digitorum longus, was confirmed to be misplaced superiorly on all five views. An anteriorly misdirected screw appeared to be radiographically within the sustentaculum on the 10-degree view but was confirmed to be misplaced on all other views. A posteriorly misdirected screw, which can penetrate the tarsal canal, was confirmed to be misplaced on all five views. Our inter-observer agreement was 100%.

   

We concluded that certain angulations when obtaining the Harris heel view are important to obtain to ensure accurate screw placement. The standard Harris heel view has been previously described as an angulation of 45 degrees toward the midline of the heel. Inferiorly misdirected sustentaculum screw placement can’t be verified on the standard Harris heel view. However, 10 & 20-degree modified heel views should be obtained to verify screw positioning. The answer was found! The screw was removed several months postoperatively and patient AM is doing well.

 

Patient J.H. is a 61 year old male who also sustained a closed calcaneus fracture after falling from a hot air balloon. He sustained a displaced intra-articular calcaneus fracture which first underwent percutanous reduction (his posterior skin was acutely threatened) followed by staged open reduction internal fixation. I noticed during postoperative visits that the surgical restoration of Bohler’s angle looked poor on certain x-rays (Figure 9) and better on others (Figure 10). Certainly his fixation was rigid and he was compliant with non-weightbearing. Had I done a poor job or was there something else going on? I hypothesized that oblique lateral x-rays would change the appearance and measurement of Bohler’s angle. Since this angle is used to determine the amount of posterior facet displacement and guide the need for additional imaging studies such as CT but more importantly our surgical decision making, I appeared to have stumbled upon an important question to answer. As lateral x-rays in the trauma setting are often oblique due to difficulties of obtaining perfect lateral radiographs, I pondered whether we were over or underestimating Bohler’s angle and making incorrect management decisions? Back to the cadaver lab with James and Leah…

   

A cadaver specimen was imaged using a large C-arm to obtain multiple fluoroscopic images. First, a perfect lateral x-ray was obtained of the ankle. Next, a series of oblique lateral x-rays were taken with the beam directed anteriorly, posteriorly, cephalad and caudad. The images were taken in 5-degree increments off the perfect lateral x-ray, from 0-25 degrees in each direction. 41 orthopaedic staff and residents volunteered to measure Bohler’s angles using the oblique lateral x-rays we obtained (Figure 11). Next, to define the true Bohler’s angle, metallic markers were placed on the anatomic bony landmarks that define Bohler’s angle: the anterior calcaneal process, the superior most portion of the posterior facet and the superior posterior tuberosity (Figure 12). The same series of oblique images were obtained to determine the true Bohler’s angles using the metallic beads placed on the appropriate anatomic landmarks (Figure 13). We then compared the true Bohler’s angle off the oblique x-rays (measuring off the metallic beads) to the measurements obtained by our volunteers using the same oblique x-rays (without the metallic beads) to see how accurately they could measure.

     

The mean value for the observed Bohler’s angles was significantly different (p < 0.05) from the true Bohler’s angles for all series of images except a posteriorly directed x-ray beam at 20 degrees from the horizontal (p = 0.43). The mean value for observed Bohler’s angles deviated further from the true Bohler’s angles with increasing image obliquity for all series except the posteriorly directed x-ray beam. The true Bohler’s angle on a perfect lateral image was 35 degrees. The true Bohler’s angle was found to vary based on the obliquity of the fluoroscopic image.

 

The study findings revealed that the ability of orthopaedic staff and resident physicians to accurately measure Bohler’s angle significantly decreased with increasing obliquity of the lateral radiograph. The true Bohler’s angle also varied with image obliquity of lateral radiographs. We concluded that understanding of these changes with oblique lateral x-rays taken in the trauma setting should decrease reliance on Bohler’s angle to determine management and need for additional imaging.

 

Patient J.H. taught me more than one lesson. On preoperative CT scan I noticed that his peroneal tendons had dislocated (Figure 14). At the time of his calcaneal open reduction internal fixation, performed through a modified sinus tarsi approach, the peroneal tendons were relocated and the superior peroneal retinaculum was repaired (Figure 15).

   

While this is a known association with intra-articular calcaneus fractures, I wanted to explore this further. Upon performing a literature search I realized that the association was estimated to be as high as 25%. While peroneal tendon dislocation (PTD) appeared to have a relatively high association with intra-articular calcaneus fractures, the two previous studies had been limited by a relatively small number of cases making the determination of true incidence unreliable. In addition, these previous studies failed to examine which types of fractures may have a higher association with PTD which could aid diagnosis for treating clinicians. Were we missing the boat on these? The clinical significance of missed PTD in the setting of intra-articular calcaneus fractures is not known although one could extrapolate from outcomes studies performed on PTD (not associated with calcaneus fractures) that additional morbidity would result. While PTD could be corrected at the time of surgical fixation of operative intra-articular calcaneus fractures, we hypothesized that these injuries are often missed radiographically and untreated. The goals of our study were: 1. determine the incidence of PTD as demonstrated on CT and plain radiographs 2. determine if certain fracture types have a higher association with PTD 3. determine if increased heel width after intra-articular calcaneus fractures influences the rate of PTD and finally 4. determine the rate of missed radiographic diagnosis (and subsequent lack of treatment) of PTD at two level I trauma centers.

 

With the help of James Toussaint, MD, Darius Lin, MD and Lauren Erhlichman, MD, an IRB-approved retrospective review was undertaken from 6/30/06 to 6/30/11 of all calcaneus fractures presenting at our institutions. Radiographic imaging was reviewed and those demonstrating intra-articular calcaneus fractures with CT imaging were included in the present study. Fractures were classified by the Essex-Lopresti classification based on plain radiographs and by the Sanders classification based on coronal CT images. Maximal heel width was measured on the axial CT scan. CT scans were examined for PTD using the technique described by Ho et al. and others. Plain radiographs were examined for radiographic suspicion of PTD (ie. “fleck” sign, distal fibular avulsion fracture). Radiology reports were reviewed for cases demonstrating PTD to see if this finding was captured by the reviewing radiologist. The incidence of PTD was noted and correlated with fracture classification and heel width. Of the cases that demonstrated PTD, medical records were examined to determine which calcaneus fractures were treated operatively and if PTD was specifically addressed at the time of surgery or at a later setting.

 

354 calcaneus fractures, of which 269 (76%) were intra-articular, presented at our institutions during the study period. 63.2% were classified as joint depression with the rest being tongue type per Essex-Lopresti classification. There was 13.8% Sander I type, 37.1 % Sanders II type, 27.5% Sanders III type and 17.1 % Sanders IV type calcaneus fractures. 31.7% of intra-articular calcaneus fractures were found to have peroneal tendon dislocations. There was a statistically significant correlation between heel width (p: < 0.001), increasing fracture severity as classified per Sanders classification and joint depression fractures as classified per Essex Lopresti classification with PTD. Only 9.2 % of PTD was correctly identified by the radiologist reviewing the CT scans. 0% of the intra-articular calcaneus fractures with associated PTD that were taken for surgical fixation had the peroneal tendons addressed at the time of surgery or at a later date. The “fleck” sign was only seen in 5.6% of patients with PTD but had a specificity of 100%.

 

We concluded that peroneal tendon dislocation is an under-recognized and under-treated complication of intra-articular calcaneus fractures (Figure 16). The results of our study demonstrated a high incidence as associated with intra-articular calcaneus fractures and statistically significant association with fracture severity. Increased awareness of PTD with intra-articular calcaneus fractures is required by radiologists and treating orthopaedic surgeons. Further outcome studies are required to determine if increased morbidity is associated with untreated peroneal tendon dislocation in the setting of intra-articular calcaneus fractures.

   

Our patients are our best teachers and I’ve found that my experiences in taking care of these patients have driven my research interests. As my experience grows with taking care of patients with calcaneus fractures I continue to be fascinated, yet humbled, by this very difficult problem. With the help from several very dedicated residents, I hope to continue to find answers on how we can treat this difficult problem better so as to improve outcomes for our patients.

 

John Y. Kwon, M.D.