ARTICLE IN PRESS

The Journal of Foot & Ankle Surgery 000 (2018) 1−5

Contents lists available at ScienceDirect

The Journal of Foot & Ankle Surgery

journal homepage: www.jfas.org

Case Reports and Series

Osteomyelitis of the Calcaneus With Pathologic Fracture

D1X XMaryellen P. Brucato, D2X XDPM, AACFAS1, D3X XMatthew F. Wachtler, D4X XDPM, FACFAS1,
D5X XEllianne M. Nasser, D6X XDPM, FACFAS2
1
Attending Physician, Atlantic Health System, Morristown, NJ
2
Attending Physician, Geisinger Health System, Danville, PA

A R T I C L E I N F O A B S T R A C T

Level of Clinical Evidence: 4 Pathologic fractures of the calcaneus secondary to osteomyelitis (OM) have rarely been reported in the literature.
Keywords: This case series describes 5 patients who were treated in our institution for chronic OM of the calcaneus and sub-
amputation sequently suffered a fracture of the involved calcaneus in the absence of trauma. All 5 patients had a history of
calcaneus insulin-dependent diabetes mellitus and were treated with a range of surgical treatments including open reduc-
death tion and internal fixation, external fixation, and excision of the fracture fragment. Three (60%) of the patients
diabetes mellitus required a below-the-knee amputation of the ipsilateral limb, 1 (20%) expired, and 1 (20%) experienced healing of
pathologic fracture the fracture and the associated heel wound. Pathologic fracture of the calcaneus secondary to OM is a recognized
probe to bone entity, although case descriptions have rarely been reported for this challenging condition.
ulcer
© 2018 by the American College of Foot and Ankle Surgeons. All rights reserved.

The calcaneus plays a crucial role in normal ambulation and lower compression fracture; and type III, a wedge-type fracture. The study
extremity function, making 2 types of possible ailments particularly dev- identified 22 total patients, 4 of whom had a history of a heel ulcer (2).
astating: fracture and osteomyelitis (OM). Calcaneal fractures are consid- In this report, we describe 5 diabetic patients who were treated at
ered a severe foot and ankle injury, often requiring extensive surgical our institution for chronic OM of the calcaneus and who subsequently
intervention and postoperative care with an elevated risk of complica- suffered a pathologic fracture of the involved calcaneus in the absence
tions. OM of the calcaneus is another devastating pedal disorder that is of an acute traumatic event. The purpose of this report is to highlight
often associated with multiple complications. Independently, OM of the this condition in terms of its morbidity and mortality, with the hopes of
calcaneus can result in the need for long-term intravenous antibiotic instigating future observational studies and further discussion of its
therapy, surgical debridement and/or reconstruction, difficulty with evaluation and management.
ambulation, loss of function, amputation, and even death (1). When a
fracture of the calcaneus occurs in the setting of OM, the foot and ankle Case Series
surgeon is presented with a condition that is exceptionally difficult to
treat successfully (salvage of a weightbearing foot), and major amputa- Identification of Patients
tion is often the resultant outcome. Importantly, diabetes mellitus is a
common comorbidity of this patient population, and it is associated with A chart review was conducted by 2 of the authors (M.P.B. and
various lower extremity risk factors such as neuropathy, peripheral arte- M.F.W.) on all patients who presented to the podiatric surgery service
rial disease (PAD), a suppressed immune response, and neuroarthrop- between January 2010 and December 2012. We inspected each record
athy, all of which predispose the patient to infection and decrease the on a consecutive master registry to look for patients who fit the inclu-
likelihood of salvaging a functionally weightbearing foot. sion criteria. Inclusion criteria were defined as patients presenting with
The Hedlund classification pertains to diabetic patients with calca- a calcaneal fracture found on plain radiographs with a prior history of
neal fractures (2). The report, published in 1998, described 3 basic frac- OM in the same calcaneus. The presence of OM needed to be docu-
ture patterns: type I, an avulsion fracture; type II, a midcalcaneal mented preceding the fracture via positive wound culture combined
with either histological evidence in the bone biopsy or magnetic reso-
nance imaging (MRI) findings suggestive of infection, namely increased
Financial disclosure: None reported. signal intensity on T2 images showing bone marrow edema within the
Conflict of Interest: None reported. calcaneus (3,4). Additionally, to be included, the patient had to have a
Address correspondence to: Maryellen P. Brucato, DPM, AACFAS, Morristown Memo-
rial Hospital, Atlantic Health Podiatric Surgery, 103 Colfax Avenue, First Floor, Pompton
current ulceration of the ipsilateral heel at the time of the calcaneal
Lakes, NJ 07442. fracture, with a positive probe-to-bone sign, suggestive of OM (5). The
E-mail address: marebeary1@gmail.com (M.P. Brucato). presence of diabetes mellitus was not a distinct inclusion criterion, and

1067-2516/$ - see front matter © 2018 by the American College of Foot and Ankle Surgeons. All rights reserved.
https://doi.org/10.1053/j.jfas.2018.09.016
 ARTICLE IN PRESS
2 M.P. Brucato et al. / The Journal of Foot & Ankle Surgery 00 (2018) 1−5

any patient treated for presumed OM without documentation (culture with external fixation, which failed 2 days postoperatively, and subse-
plus biopsy or MRI findings) was excluded. Based on these criteria, a quently underwent a resection of the fracture fragment (Fig. 7). All
total of 5 diabetic patients who were treated at the primary (M.P.B.) and patients continued to receive local wound care after the procedures,
secondary (M.F.W.) authors’ institution, namely the Atlantic Health Sys- which included negative-pressure wound therapy and biosynthetic
tem, in Morristown, New Jersey, and who were initially diagnosed with grafting. Bone cultures that were recorded revealed 13 different types
calcaneal OM and subsequently sustained a fracture of the involved cal- of organisms, with a mean of 3.2 organisms per patient. The most prev-
caneus were identified and included in this review. All 5 of these alent organisms were b-hemolytic group B Streptococcus, Diphtheroid
patients presented with no obvious history of acute trauma localized to bacillus, and Enterococcus faecalis.
the involved heel. The first treatment on a patient started in November In total, 3 (60%) of the cases resulted in BKA of the affected limb sec-
2010 and the last treatment date was September 2012, so the duration ondary to recalcitrant OM, nonhealing ulceration, and inability to salvage
of the observation period for inclusion in this review was »23 months. the limb. One (20%) case that was initially treated with the excision of
Although not a requirement for inclusion in our review, all 5 of the the fracture fragment remained unhealed, with an open ulceration of
patients were treated with 6 to 8 weeks of intravenous antibiotics and the ipsilateral heel until the patient expired secondary to cardiac arrest.
partial calcanectomy at the time that the OM was diagnosed. At our Finally, the patient who underwent fixation with a percutaneous screw
hospital, this investigation (N < 25) qualified for expedited review, and experienced healing of the ulceration and the fracture within 52 weeks.
our sole requirement was that protected health information had to Additional treatment for the wound included negative-pressure wound
remain confidential. therapy and biosynthetic grafting. Also of note, the fracture pattern was
Demographic exposures that we recorded included patient age, sex, type III. Follow-up with the patient continued for 11 months after the
body mass index, side involved, duration of OM at the time of fracture ulcer healed without incidence of reulceration or recalcitrant OM. The
(days), fracture pattern type based on the diabetic calcaneal fracture patient was partially ambulatory owing to ulceration on her other foot.
classification (2), presence of insulin-dependent diabetes, presence of After 11 months, the patient was lost to follow-up.
anemia (hemoglobin < 10 g/dL), smoking history, history of end-stage
renal disease, and history of PAD. We also recorded wound culture, Discussion
bone biopsy, and MRI results. Wound cultures from the heel ulcers
were recorded from the time that the patients were diagnosed with, Pathologic fracture of the calcaneus secondary to OM is a rare, limb-
and subsequently treated for, OM. The results of the probe-to-bone test threatening condition. These 2 distinct ailments alone typically present
were also logged for each patient (5). The end point (outcome) for each a challenge to the foot and ankle surgeon; when combined, they create
patient was determined to be healing of the fracture and ulcer, amputa- a unique condition that is exceptionally difficult to treat.
tion, or death. Follow-up examined in this study stopped when any of Literature regarding diabetic calcaneal fractures is insufficient. The
the 3 end points was reached, which ranged from 10 months to 2 years. majority of what is written is extrapolated from the diabetic ankle frac-
ture literature and focuses on hemoglobin A1C as well as prolonged
Results immobilization.
In 1998, Hedlund et al (2) identified 22 diabetic patients who pre-
A statistical description of the 5 patients is shown in the Table. The sented with calcaneal fractures in a 25-year period without any signifi-
mean patient age was 55 § 6.9 years, and the mean body mass index cant contributing trauma. The study described 3 different basic fracture
was 29.39 § 6.16 kg/m2. Only 1 (20%) of the patients was male, and the patterns: types I, II, and III. Type I is an avulsion fracture pattern, type II
right side was involved in 2 (40%) of the cases. All of the patients had was described as a midcalcaneal compression fracture, and a wedge-
insulin-dependent diabetes and a history of cigarette smoking, and in type fracture was branded as a type III fracture pattern. Only 1 patient
each case the probe-to-bone clinical test suggested the presence of OM in their review had a diagnosis of OM prior to fracture. Ten patients in
of the calcaneus (5). In this series, moreover, 4 (80%) were anemic, 1 total (45.5%) had a history of preceding chronic heel ulcer on the ipsilat-
(20%) was on dialysis for end-stage renal disease, and 2 (40%) had PAD. eral foot. The study cited multiple contributing factors to the etiology of
Four (80%) of the fractures were type III as seen in Fig. 1, and 1 (20%) the calcaneal fractures, including decreased bone mineralization (sec-
was type I as seen in Fig. 2. The mean duration of OM prior to fracture ondary to diabetes, long-term steroid use, neuropathic hyperemia,
was 100 § 51.1 days (range 30 to 165), which is summarized in Fig. 3. decreased weightbearing) and alteration of gait (secondary to neuropa-
In addition to clinical history and appearance, including the presence of thy, retinopathy, injury, amputation). Other factors that have been pre-
the probe-to-bone sign (5), the diagnosis of OM was based on MRI find- viously identified as risk factors for diabetic pathologic calcaneal
ings in 3 (60%) and biopsy in 2 (40%) of the cases. Bone cultures revealed fractures include tight Achilles tendon and renal disease (2). The
13 different types of organisms including Enterococcus faecalis (40%), patients in our study suffered from a variety of these different stressors,
b-hemolytic group B Streptococcus (40%), and Diphtheroid bacillus (40%), with the addition of OM of the calcaneus, a previously unidentified risk
with the following organisms cultured only once: Escherichia coli, Enter- factor.
obacter cloacae, Enterococcus faecium, Candida parapsilosis, Clostridium The type I fracture pattern was initially described in 1991 by Kathol
cadaveris, Klebsiella pneumoniae, methicillin-resistant Staphylococcus et al (6) as a calcaneal insufficiency avulsion fracture. The authors noted
aureus, methicillin-sensitive Staphylococcus aureus, Proteus mirabilis, that these fractures tend to be extra-articular and limited to the poste-
and Pseudomonas aeruginosa. During the follow-up period, 3 (60%) rior one third of the calcaneus. They hypothesized that these types of
patients went on to a below-the-knee amputation (BKA), 1 (20%) fractures occurred owing to the Achilles tendon overpowering a weaker
healed, and 1 (20%) died (the cause of death was determined to be car- area of the calcaneus in the absence of trauma. In the 1991 article, 14
diac arrest). diabetic patients who presented with calcaneal insufficiency avulsion
Initial fracture management varied on a case-by-case basis. Three fractures were identified. Of these patients, all were insulin dependent,
patients were initially treated with excision of the fracture fragment as all had neuropathy, and 4 were noted to have associated heel ulceration
seen in Fig. 4. In 1 (20%) patient, in addition to excision of the fragment, at the ipsilateral foot (6). Kathol et al (6) described a sequence of events
the Achilles tendon was able to be salvaged and was transferred to the through which the calcaneal insufficiency fracture occurs. The process
remaining posterior calcaneus with anchor fixation (Fig. 5). One (20%) begins with microfracture to the calcaneus from daily activity. This pro-
patient was treated with a percutaneous 8-mm cannulated screw and gresses to a nondisplaced insufficiency fracture, which may show radio-
Achilles tendon lengthening (Fig. 6). Last, 1 (20%) patient was treated graphic changes. Fragmentation occurs to the insufficiency fracture,
 ARTICLE IN PRESS
M.P. Brucato et al. / The Journal of Foot & Ankle Surgery 00 (2018) 1−5 3

Table
A statistical description of the case series

Case 1 2 3 4 5 Summary Statistics, Mean § SD

(median [range]) or count (%)

Age, y 49 61 48 63 58 55.8 § 6.9 (58 [48, 63])

Sex Female Female Male Female Female 4 (80%) female; 1 (20%) male
Body mass index, kg/m2 33.1 24.77 25.06 38.5 25.5 29.39 § 6.16 (25.5 [24.77, 38.5])
Side involved Left Left Left Right Right 3 (60%) left; 2 (40%) right
Duration of OM, d 30 165 95 130 80 100 § 51.1 (95 [30, 165])
Fracture category* III III III III I 4 (80%) III (wedge); 1 (20%) I (avulsion)
Diabetes mellitus Yes Yes Yes Yes Yes 5 (100%) yes
Anemiay Yes Yes Yes No Yes 1 (20%) no; 4 (80%) yes
Smoking history Yes Yes Yes Yes Yes 5 (100%) yes
End-stage renal disease No No No No Yes 4 (80%) no; 1 (20%) yes
Peripheral arterial disease No No No Yes Yes 3 (60%) no; 2 (40%) yes
Probe to bone Yes Yes Yes Yes Yes 5 (100%) yes
Calcaneal culture Yes Yes Yes Yes Yes 5 (100%) yes
Biopsy indicates OM No No No Yes Yes 3 (60%) no; 2 (40%) yes
MRI indicates OM Yes Yes Yes No No 2 (40%) no; 3 (60%) yes
Treatment of fracture Excision of Excision of External Percutaneous Excision of 3 (60%) excision fragment; 1 (20%) external fixation;
fragment fragment fixation screw + TAL fragment 1 (20%) percutaneous screw + TAL
Outcome BKA BKA BKA Healed Died 3 (60%) BKA; 1 (20%) healed; 1 (20%) died
Abbreviations: BKA, below-the-knee amputation; MRI, magnetic resonance imaging; OM, osteomyelitis; SD, standard deviation; TAL, tendo-Achilles lengthening.
* Hedlund et al (2).
y
Anemia defined as hemoglobin < 10 g/dL.

based on anatomic location. Type I fractures correspond to a Brodsky

type IIIb (7). Brodsky described the type IIIb Charcot deformity as a
pathologic fracture of the posterior tubercle of the calcaneus leading to
deformity and collapse of distal joints (7).
There have been minimal reports of pathologic calcaneal fractures in
patients with a history of chronic ulcerations of the heel, and none of
the reports showed documented OM before the time of fracture. In
2004, Roldan (8) reported a single case study involving a pathologic
fracture of the calcaneus from presumed OM following a chronic heel
ulceration of 2 months. The OM was not documented or diagnosed
before the fracture. Athans and Stephens (9) reported a case series of 3
diabetic patients who sustained open calcaneal fractures referred to as
pathologic fractures secondary to diabetes, not OM. All 3 had preceding
heel ulcerations, and 1 had confirmed OM.
Fig. 1. Radiograph of case no. 1, presentation of type III wedge fracture. In 1979, Coventry and Rothacker (10) published a case report that
documented an insulin-dependent diabetic female with bilateral stress
fractures of the calcaneus. The patient did not have significant trauma
before the fractures; however, there were altered biomechanical
stresses combined with a history of polyneuropathy (10). Platts-Mills et
al (11) later reported a diabetic patient who presented with a calcaneal
avulsion fracture in the complete absence of trauma, which was then
treated with a single 6.5-cm screw.
Although all 5 patients in the current study had been treated previ-
ously with a partial calcanectomy, there are no prior reports that show
this would weaken the calcaneus to the point of fracture with regular
load bearing. In 2002, Bollinger and Thordarson (12) reported that the
partial calcanectomy is a viable alternative option to BKA for treatment
of OM.
Any type of foot and ankle fracture in the setting of OM presents a
distinct challenge regarding treatment. This is compounded by the fact
that 25% of all calcaneal fractures have wound complications (13).
When these 2 types of injuries occur simultaneously in the calcaneus, it
Fig. 2. Radiograph of case no. 5, presentation of type I avulsion fracture. creates a perfect storm of an extremely complicated surgical challenge.
Traditional open reduction and internal fixation is not a completely
viable option because bone quality is most likely compromised. Screw
holding power is directly proportional to bone density. Traditionally,
typically resulting in a secondary fracture line. Finally, the fracture frag- the surgeon wants to avoid placing hardware inside of infected bone for
ments displace (6). fear of seeding the hardware with bacteria. The single patient in our
The type I fracture pattern has a corresponding collocation in the series who was ultimately treated with open reduction and internal fix-
Brodsky classification, a classification system of Charcot deformities ation of the calcaneus was first treated with long-term intravenous
 ARTICLE IN PRESS
4 M.P. Brucato et al. / The Journal of Foot & Ankle Surgery 00 (2018) 1−5

Fig. 3. Graph presentation of duration (days) of osteomyelitis at the time of the calcaneal fracture.

including the infected fracture fragment. Partial calcanectomy or exci-

sion of the fracture fragment is a choice that is complicated by the inser-
tion of the Achilles tendon. If the Achilles tendon is not addressed, the
patient may become at risk for calcaneal gait. Greenhagen et al (15)
reported a case study in which a diabetic calcaneal insufficiency fracture
was treated with excision of the fracture fragment with a gastrocne-
mius recession and reattachment of the Achilles tendon to the remain-
ing calcaneus. It is important to note that OM was not present.
Last, BKA may present the most functional option for these
patients. Although the goal is for limb salvage, the decision may be
made at the time of the infected calcaneal fracture to proceed with a
BKA and plan for prosthetic fitting and extensive rehabilitation once
Fig. 4. Postoperative radiograph of case no. 2, treated with excision of fracture fragment. the incision is healed. It is important to note that some studies have
found decreased functional ability following BKA. In 1976, Waters et
al (16) found that patients had a gait velocity significantly decreased
from normal ambulation. If the decision is made to go forward with
BKA, the surgeon and patient should be in agreement that this would
indeed provide the best result based on the patient’s needs and func-
tional ability (1,17).
The outcomes for the 5 patients included in this case series were sig-
nificantly worse compared with previous reports that did not involve
OM. Hedlund et al (2) reported that all 22 diabetic calcaneal fractures
healed; however, 50% healed with significant calcaneal deformity. Fur-
thermore, the patients in that study did not have documented OM.
When a diabetic patient presents with a nonhealing heel ulcer in the
setting of OM, the patient should be properly offloaded to avoid a path-
ologic fracture. In this case series, all of the patients underwent previous
partial calcanectomy, which could have further weakened the infected
Fig. 5. Postoperative radiograph of case no. 5, treated with excision of fracture fragment bone. A recent study by Oliver et al (18) evaluated functional results in
and reattachment of Achilles tendon with anchor fixation. patients after undergoing partial calcanectomy to treat OM. They
hypothesized that with a plantar calcanectomy leaving much of the
antibiotics for a course of 6 weeks, and the treating surgeon believed Achilles tendon intact, the patient is at more of a risk to fracture the cal-
that this was sufficient. In 2017, a study by Yildirum et al (14) demon- caneus biomechanically with the pull of the Achilles tendon overpower-
strated successful results in pediatric fractures secondary to OM treated ing the dorsal cortex. They looked at 2 cohorts: patients who had < 50%
with wide debridement, irrigation, bone grafting, and anatomic locking of the calcaneus resected versus those who had > 50% of the calcaneus
plate and screws to reduce the fracture. Their conclusion was that a resected. However, their results showed a BKA rate of 29% in all patients
pathologic fracture must be stabilized in addition to conventional treat- who underwent partial calcanectomy regardless of the surgical tech-
ment for OM. nique (18).
External fixation may be warranted to reduce and stabilize the frac- Limitations of the current study include a small study population
ture; however, that does require structural integrity of the bone, and the retrospective nature of our case series. Additionally, this is
 ARTICLE IN PRESS
M.P. Brucato et al. / The Journal of Foot & Ankle Surgery 00 (2018) 1−5 5

In conclusion, looking to the future, there is no definitive treatment

algorithm for calcaneal fractures in the setting of OM. The outcomes in
this study highlight the limb-threatening nature of this combined con-
dition. A prospective study would be warranted to develop recommen-
dations and guidelines for best treatment practices.

Acknowledgment

The primary author thanks all of the librarians at the Morristown

Medical Center library for their unwavering and diligent support.

References

Fig. 6. Postoperative radiographs of case no. 4, treated with open reduction and internal 1. Fukuda T, Reddy V, Ptaszek AJ. The infected calcaneus. Foot Ankle Clin N Am
2010;15:477–486.
fixation.
2. Hedlund LJ, Maki DD, Griffiths HJ. Calcaneal fractures in diabetic patients. J Diabetes
Complications 1998;12:81–87.
3. Butalia S, Palda VA, Sargeant RJ, Detsky AS, Mourad O. Does this patient with diabetes
have osteomyelitis of the lower extremity? JAMA 2008;299:806−813.
4. Dinh MT, Abad CL, Safdar N. Diagnostic accuracy of the physical examination and
imaging tests for osteomyelitis underlying diabetic foot ulcers: meta-analysis. Clin
Infect Dis 2008;47:519–527.
5. Grayson ML, Gibbons GW, Balogh K, Levin E, Karchmer AW. Probing to bone in
infected pedal ulcers: a clinical sign of underlying osteomyelitis in diabetic patients.
JAMA 1995;273:721–723.
6. Kathol MH, El-Khoury GY, Moore TE, Marsh JL. Calcaneal insufficiency avulsion frac-
tures in patients with diabetes mellitus. Musculoskelet Radiol 1991;180:725–729.
7. Coughlin MJ, Mann RA, Saltzman CL. Surgery of the Foot and Ankle. Mosby, Maryland
Heights, MO; 2007.
8. Roldan CJ. A pathologic fracture: underestimated mechanism in a patient with risk
factors. J Emerg Med 2004;26:207–208.
9. Athans W, Stephens H. Calcaneal fractures in diabetic patients with neuropathy: a
report of three cases and literature review. Foot Ankle Int 2008;29:1049–1053.
10. Coventry MB, Rothacker GW. Bilateral calcaneal fracture in a diabetic patient: a case
report. J Bone Joint Surg 1979;61:462–464.
11. Platts-Mills TF, Burg MD, Pollack ZT. Calcaneal avulsion fracture. Emerg Med J 2006;24:231.
12. Bollinger M, Thordarson DB. Partial calcanectomy: an alternative to below knee
Fig. 7. Postoperative radiographs of case no. 4, treated with open reduction and internal amputation. Foot Ankle Int 2002;23:927–932.
fixation with screw, after screw removal and healed fracture. 13. Folk AW, Starr AJ, Early JS. Early wound complications of operative treatment of cal-
caneus fracture: analysis of 190 fractures. J Orthop Trauma 1999;13:369–372.
14. Yildirum A, Kapukaya A, Ramazan A, Mertsoy Y, Yigit S, Cacan M. The use of an “inter-
a purely observational study. Each patient was treated by a different nal fixator technique” to stabilize pathologic fractures developing secondary to osteo-
myelitis. J Pediatr Orthop 2017;37:222–226.
primary surgeon at our institution, so it was impossible to deter-
15. Greenhagen RM, Highlander PD, Burns PR. Double row anchor fixation: a novel tech-
mine the exact surgical technique of the initial partial calcanectomy nique for a diabetic calcaneal insufficiency avulsion fracture. J Foot Ankle Surg
that could have contributed to the fracture of the calcaneus. Ideally, 2012;51:123–127.
16. Waters RL, Perry J, Antonelli D, Hislop H. Energy cost of walking amputees: the influ-
a bone biopsy at the time of fracture would have been performed
ence of level of amputation. J Bone Joint Surg 1976;58:42–46.
as well to confirm the diagnosis of OM for a second time. Moreover, 17. Evran M, Sert M, Tetiker T, Akkus G, Bicer OS. Spontaneous calcaneal fracture in
we recognize that identifying patients by the diagnosis of calcaneal patients with diabetic foot ulcer: four cases report and review of literature. World J
fracture on a master patient list can be troublesome, because it is Clin Cases 2016;4:181–186.
18. Oliver NG, Steinberg JS, Powers K, Evans KK, Kim PJ, Attinger CE. Lower extremity
possible that the record keeping was not done in a rigorous or uni- function following partial calcanectomy in high-risk limb salvage patients. J Diabetes
form fashion. Res 2015;2015:432164.