Surgical Infections II – A. Osteomyelitis, Acute and Chronic
This Review is part of a series discussing the relationship of Human Infection and Surgery. Future Reviews in this series will deal with Septic Arthritis, Tuberculosis of Bones and Joints, Infections of specific body cavities, the surgical manifestations of HIV infection, wound infection and others.
| Clinical cases | |
| Acute Hematogenous Osteomyelitis | |
| Chronic Osteomyelitis | |
| Post-trauma Osteomyelitis | |
| References | |
| Questions | |
Introduction
This Review must begin with a disclosure. I am not an orthopaedic surgeon and
my orthopaedic experience has been limited to my surgical experiences in Africa.
However these experiences have taught me that bone and joint infections are
common in Africa; that they present late, often inadequately treated (1);
are a major source of disability in all age groups, but primarily the young;
present a challenge for the treating surgeon with limited means of diagnosis
and with perhaps, like me, limited orthopaedic expertise. My experience is born
out by the, unfortunately sparse, literature available from Africa. Bickler’s
study (2) from the Gambia showed that, of all surgical conditions,
cases of osteomyelitis alone accounted for 15% of total inpatient days, second
only to burns. I have written this Review with that situation in mind.
I have relied heavily on the standard surgical literature on bone and joint
infections (3 - 5), which is derived mostly
from the developed world, but have focussed on the more common conditions seen
in Africa. I discuss acute hematogenous osteomyelitis (AHO) and the chronic
osteomyelitis (CO) that derives from it, as well as osteomyelitis that follows
trauma and open fractures. I have omitted any discussion of infected prostheses
and implants, although these represent a major problem in the developed world
(6), or osteomyelitis involving the diabetic foot, which has
specific characteristics. Finally I include photographs from my own experience,
which illustrate some of the variety of bone infections.
General
Principles of Bone and Joint Infections
Osteomyelitis can be classified into three
major etiologic categories: 1. hematogenous, 2.
exogenous from a contiguous infection or direct inoculum as in open fracture
and 3. related to vascular insufficiency as in CO and when
associated with diabetes. The disease is divided into acute, subacute or chronic
based on the duration of symptoms and the presence of previous episodes. Major
patient-related characteristics as in sickle cell anemia or HIV infection alter
the response and character of infection. The Cierny-Mader classification incorporates
both the anatomic extent of disease and patient characteristics and has been
particularly useful in CO. (7)
The pathophysiology of osteomyelitis develops
as a result of the particular anatomy and physiology of long bones. (8)
The medullary cavity of bones, where hematogenous infection begins, is encased
in a rigid structure, which does not allow for the expansion of the inflammatory
process. Progression of the infection restricts medullary blood supply. Passage
of pus through the cortex elevates the periosteum and the resulting sub-periosteal
abscess causes bony infarction as the cortical bone is supplied by end-arteries
from the periosteum. The necrotic bone acts as a persistent foreign body. Healing
by periosteal new bone formation forms an involucrum, which further isolates
the sequestrum from antibiotic therapy. Poor blood supply and necrotic bone
account for the need for prolonged antibiotic therapy in AHO and the resistance
of CO to therapy. Inflammation induces osteoclastic activity, which is partially
responsible for the osteoporosis seen in x-rays of infected bone. The pathophysiology
of post-trauma osteomyelitis shares some but not all of these characteristics.
(9)
The growth and development of long bones from the epiphyseal plate also determines
the characteristics of bony infection. Prior to 2 years of age, blood vessels
cross the physis allowing the initial metaphyseal infection to spread into the
epiphysis or even the joint. After 2 years, the developing plate restricts this
and the infection tends to spread into the diaphysis of long bones. A recent
study (10) found simultaneous septic joint involvement adjacent
to an Osteomyelitis in as many as 1/3 of cases and also questioned this age
distribution. Since the neck of the femur is essentially intra-articular, osteomyelitis
here is always associated with septic arthritis. Untreated septic arthritis,
involving particularly the intra-articular proximal femoral and humeral growth
plates, can cause profound growth disturbances. With plate closure after puberty,
hematogenous osteomyelitis becomes uncommon in adults, only to recur again in
the elderly. (11)
Particular biochemical features also determine the bacteriology
of osteomyelitis. Staphylococcus aureus possesses receptors, which
facilitate its adherence to and invasion of bone and joint tissues. These may
account for the fact that staph aureus constitutes the majority of endogenous
infections. Osteomyelitis from staph aureus carrying the pvi gene appears to
result in higher complications. (12) Streptococci are the
second most common organisms, while specific vaccines against haemophilus influenzae
have restricted this organism in those countries where they are used. (13)
While most hematogenous infections are single agent, post-traumatic infections
often yield multiple bacteria. Pseudomonas is a common pathogen, after puncture
wounds of the foot. Patients with sickle cell anemia have an increased incidence
of osteomyelitis caused by more frequently by salmonella agents. (14)
Differentiating osteomyelitis from bone infarction, which requires only hydration
and analgesia, is challenging. (15)
The diagnosis of bone or joint infection
may be obvious or obscure. Fever, swelling and pain are the classic clinical
triad; pain is the most persistent. Laboratory tests may or may not be positive.
These include WBC count, ESR and C-reactive protein. The latter is probably
the most sensitive, if not specific. (16) Blood culture may
be positive in over 30% of cases of AHO. Aspiration of the medullary cavity,
joint or direct bone biopsy, with culture, are the mainstays of certain diagnosis,
are crucial in identifying the infecting agent and guide anti-microbial therapy.
Gram stain may be positive in 30% of aspirates, but only culture is definitive.
If no pus is identified a deep bone biopsy is necessary. (17)
Aspiration cytology may be used to identify neoplastic lesions. (18)
In acute infection blood or tissue culture is positive in 65% of cases. (19)
Culture of sinus tracts in CO is misleading and only cultures of infected bone
and soft tissue will reveal the causative agent. The fluid in septic arthritis
generally has a cell count of more than 80,000 with more than 75% neutrophils,
distinguishing septic arthritis from other causes of joint effusion.
A wide variety of imaging modalities
have been developed for bone and joint infections. (20; 21)
The plain x-ray may show bone destruction or periosteal new bone formation but
not until 10-21 days after infection. Nuclear medicine scans are valuable for
detecting inflammation. Technetium based bone scans are most sensitive and can
detect osteomyelitis 48 hours after infection. Gallium and indium WBC scans
have also been used. Trauma, recent surgery, the presence of prostheses; all
lower the specificity of these tests. Ultrasound is particularly useful in identifying
fluid collections in joints and around bone and is more likely to be available
in African hospitals. (22) It may be used for aspirating infected
areas. (23) CT scan and MRI are particularly useful in assessing
degree of bony destruction, soft tissue extension and for specific regions.
(24)
Acute
Osteomyelitis
Acute hematogenous osteomyelitis (AHO) is a specific clinical entity affecting
primarily children. Males are affected twice as often as females. While this
disease is decreasing in frequency in the developed world (25),
it is common throughout the developing world where it causes significant morbidity.
Australian Aborigines and New Zealand Maoris experience a 7-10 fold increase
in comparison to other children in the same countries. (26)
A report from Lithuania (27) shows AHO increasing in frequency.
Poverty, malnutrition, chronic disease, perhaps anemia, trauma and immuno-compromise
are all compounding factors.
In Africa, where late presentation is the norm, the clinical picture is usually
a febrile child with a swollen painful extremity s/he is resistant to use. Bony
tenderness is characteristic. The differential diagnosis is cellulitis or pyomyositis,
septic arthritis, bony infarction in a patient with sickle cell anemia, bony
neoplasm, or other infectious diseases such as brucellosis, yaws, syphilis or
leprosy. Septic arthritis, either alone or in combination with osteomyelitis
is the most important differential diagnosis and is evaluated by aspiration
of the joint.
X-rays and laboratory tests, including blood culture, should be done recognizing
their limitations. U/S or nuclear scans may be useful if available.
The single most important investigative tool is aspiration of the involved tissue
looking for pus and sending any material for culture and histology. A subperiosteal
abscess represents a more advanced form of the disease. In its absence medullary
aspiration or bone biopsy should be carried out.
In the absence of culture and sensitivity reports, empiric antibiotic therapy
must be directed against all likely agents. Since staph aureus is the most common
pathogen, a semi-synthetic penicillin, resistant to B-lactamase, such as cloxacillin
is appropriate. (28) In situations where MRSA bacteria are
common vancomycin, clindamycin, or rifampicin and an aminoglycoside may be necessary.
Clindamycin achieves high bone concentrations. In children under 4 years, who
have not been immunized against haemophilus influenzae, ampicillin should be
added or ceftriaxone substituted.
The chief controversies in the management of AHO consist in the route and duration
of antibiotic therapy (29) and the need for mandatory drainage
of the involved bone.
The standard recommendation is for 4-6 weeks of antibiotic therapy, but this
has been significantly reduced recently to 21 days. Unfortunately a meta-analysis
showed that there are no high quality studies on which to base decisions. (30)
One small RCT from Iran showed that in patients with AHO, who were responding
to treatment and who continued on oral antibiotics for an additional 4 weeks,
there were no differences between 10 or 21 days of intravenous therapy. (31)
It should be further pointed out that all these patients had surgical drainage.
While it is well established that medullary pus and necrotic bone need to be
surgically removed, the literature in developed countries recommends antibiotics
as sole therapy, if initiated before the disease has progressed to this more
advanced stage. Whether these recommendations are suitable for Africa is debatable.
There is general agreement however that, if pus is found, the bone must be opened.
Since blood supply derives from the periosteum, dissection should be limited.
No less an authority than Primary Surgery is adamant on the need for drilling.
(32) It states: “There is little point in aspirating
it first, except sometimes to localise the site, because you will have to drill
it anyway, even if the aspiration is negative”. King’s method is
drilling as a diagnostic and therapeutic procedure and he gives a very good
description of the technique.
Drilling requires the placement of 3 to 4 holes staggered obliquely down the bone to prevent weakening.
When intra-medullary pus is found, the Western
literature recommends the creation of a cortical window with evacuation of pus
and necrotic material. The limb is then splinted in plaster and mobilized after
the wound has healed.
The point here is that pus from osteomyelitis must never go undrained and that,
where possible, drainage should occur before the development of a subperiosteal
abscess. While drilling and cortical windows have been associated with an increased
risk of fracture, this may be due to the more advanced stage of the disease
rather than the treatment itself. It is unlikely that the early cases described
in Western literature are common in Africa. Under treatment or delayed treatment
of AHO is associated with significant complications, the most significant being
chronic osteomyelitis. (33) However with early treatment only
3% of patients will have long-term sequela.
AHO in adults more commonly involves the vertebrae than long bones. (34)
Surgical treatment, which was required in over 60% of patients, requires specialized
capability.
Sub-acute Osteomyelitis
Sub-acute osteomyelitis is a discrete clinical entity seen in the developed
world and with increasing frequency. (35) It is felt to be
a result of more indolent infection and is associated with specific radiographic
abnormalities such as Brodie’s Abscess. (36) It must
be distinguished from bony neoplasms. (37) In some cases these
may indeed be indolent forms of infectious osteomyelitis (38
- 40); others may be the manifestations of chronic recurring
multifocal osteomyelitis CRMO, which is a non-infectious condition of unknown
etiology. (41) Aspiration or biopsy, culture and histology
will resolve problems of diagnosis.
Chronic Osteomyelitis
The end stage of delayed or poorly treated AHO is chronic osteomyelitis (CO).
Here, a chronic infection of the involved bone is perpetuated by the presence
of a sequestrum, a necrotic piece of bone devascularized by the pathologic process
described above. The sequestrum is surrounded by the involucrum, new bone formed
by the periosteum. Indolent, recurrent, chronic infection is associated with
sinus formation, pain and prolonged disability.
The diagnosis of CO is usually fairly obvious with a prior history of recurrent
infections, disability and deformity arising after an initial event. However,
Museru has stressed the necessity of distinguishing CO from bony neoplasms and,
in this regard, bone biopsy is again the gold standard. (42)
All the standard imaging techniques have been used to assess the extent of CO
and to plan treatment. Sinogram is a simple technique, which can give information
on the extent of disease. Bacteriologic diagnosis is important and can only
be made with culture of bone. Effluent from sinus tracks is not a reliable indicator
of the nature of the underlying infection.(43; 44)
Interestingly staph aureus remains the main pathogen. (45)
The treatment of CO is multifaceted and complex. It involves: 1. the excision
of all necrotic bone, 2. if necessary stabilization of the limb, 3.the application
of appropriate, prolonged antibiotics, and 4.the obliteration of dead space
through the closure and reconstruction of the soft tissue defect. (46)
Multiple procedures are required. This applies equally to the treatment of posttraumatic
osteomyelitis in the following section.
The patient's general condition should be improved by correction of anemia or any nutritional deficiencies. There is seldom need for urgent intervention.
Maurice King recommends delaying intervention
until an adequate involucrum has formed to stabilize the limb. The recent literature
from the developed world with its advanced orthopaedic techniques does not consider
an inadequate involucrum a contraindication to debridement. The Cierny-Mader
classification has been useful in assessing the degree of disease and the ability
of the patient to respond to the treatment.
Certainly some patients are better left without major intervention.
A team approach has been associated
with improved results. (47). The role of reconstructive efforts
to improve local blood supply is important. (48)
There is a discrepancy between recent and older literature on the value of antibiotic
therapy. Clearly, antibiotics are an adjunct to surgery and must be prolonged.
The parenteral route is favoured. (49; 50)
Various new oral agents including quinolones and linezolid are being investigated.
(51; 52) Ciprofloxacin is a fluoroquinolone
agent with excellent bone and soft tissue penetration and a broad spectrum of
activity against gram positive organisms including staphylococci and gram negatives
including pseudomonas. Extensive clinical use in pediatric populations with
cystic fibrosis has not resulted in observed clinical toxicities and in particular
the potential cartilage toxicity suggested by animal studies does not appear
to be important in children. Because of its long half life once daily ceftriaxone
has been used in an ambulatory setting. (53)
Surgical treatment begins with excision of all necrotic, infected bone. (54)
The extent of excision has been correlated with success of treatment. (55)
After adequate debridement, the medullary cavity is filled with a harvested
autogenous bone graft. This is the Papineau technique. Posterior iliac bone
grafts have lower complication rates than anterior grafts. (56)
Alternatively local antibiotics in the form of bone cement or beads have been
used. (57 - 59) Individual reports of the
use of these methods show reasonable results. (60 - 62)
No comparison can be made between techniques. When debridement results in instability
of the limb temporary stabilization is achieved through immobilization in plaster,
external fixation and even intramedullary nailing. Because the periosteum has
been damaged, bony defects may have to be repaired through bone transfer (63)
or the Ilizarov technique. (64 - 66)
Finally, soft tissue coverage of the defect is necessary to cover the bone and
bring new blood supply to the region. (67) Sherman provides
very good descriptions of some of the standard flaps, which are necessary to
achieve closure of these wounds. For smaller indolent wounds with apparently
healthy bone in their depths, I have found sugar to be a useful stimulant to
closure. Hyperbaric oxygen has been used as an adjunctive method but the information
on its efficacy is limited. (68)
The success rates for these techniques in eliminating infection and providing
a functional limb vary between 50 and 90%. Considering the extent and difficulties
of treatment, certain patients, in certain settings, are probably best left
untreated. Amputation may occasionally be appropriate. Because of these difficulties,
all attempts should be made to prevent CO by early and adequate treatment of
AHO.
Certain regions and bones merit specific consideration. King gives a very good
description of the approach to the major long bones and of the infection of
specific regions.(32) Puncture wounds of the foot commonly
give rise to osteomyelitis of tarsal and metatarsal bones, which usually require
drainage or resection. (69) The calcaneus is a cancellous
bone and so seldom forms a sequestrum or involucrum. Resection of all involved
bone can be accomplished through a posterior midline incision with very good
functional results. The fibula can be resected in its entirety if there is no
infection of the tibia. (4) Infection of the distal third of
the femur is difficult to treat and should be approached laterally with avoidance
of the knee joint and neurovascular bundle. Conservative debridement is advised
with splinting of the knee postoperatively. The ischial tuberosities are often
involved in decubitus ulcers and prevent healing of these. Excision of involved
bone and flap closure is required. When infected the entire ilium may be involved
and may have to be resected. Jaw and skull infection have specific features.
Post-trauma
Osteomyelitis
Post-trauma osteomyelitis (PTO) usually arises as a complication of open fracture.
High velocity injuries associated with multiple bony devascularized fragments
combined with major soft tissue loss are the frequent antecedents. (70;
71) Because of its poor soft tissue coverage, open fractures
of the tibia are particularly susceptible to this complication. (72;
73) Delay in initiating treatment is a potent cause of complications.
In Africa, traditional methods may delay modern treatment and are associated
with complications. (74) In certain cases of mangled limbs,
primary amputation may be appropriate.
The principles of open fracture treatment include: peri-operative short term
antibiotics, urgent debridement of devitalized bone and soft-tissue, particularly
muscle, copious irrigation of the wound, reduction and immobilization of fracture
site, primary or delayed primary closure depending on the degree of contamination
and soft tissue injury, early soft tissue closure, if necessary with flaps.
(75) Various controversies exist and are beyond the scope
of this Review. These include the degree of bony debridement, the use or avoidance
of internal fixation, timing of wound closure, etc. Much of this is dependent
on surgical expertise. A recent Cochrane Review confirmed the utility of peri-operative
antibiotics in preventing infection with open fracture, but could not draw further
conclusions. (76)
Once PTO has complicated the injury, treatment proceeds along similar lines
discussed for CO.
Because of the nature of the bony injury and its associated complications such
large bone defect, non- or mal-union of the fracture, advanced stabilization
techniques, particularly external fixation, are more often required. Similarly
advanced reconstructive techniques may be needed including free flaps. (77
- 80)
PTO and CO are some of the most challenging and difficult orthopaedic conditions.
They are however, unfortunately common in the conditions of modern Africa.
Acknowledgement: I would like to thank Dr. Andrew Howard for reading an initial draft of this Review and making useful suggestions.
Brian Ostrow MD, FRCS(C)
Guelph, Ontario, Canada
Click
here to join the Surgery in Africa Discussion Group
| Clinical cases | |
| Acute Hematogenous Osteomyelitis | |
| Chronic Osteomyelitis | |
| Post-trauma Osteomyelitis | |
| Questions | |
Click here to join the Surgery in Africa Discussion Group