Perspectives in Circulation Today

Osteomyelitis

The osteomyelitis in the first toe and metatarsal head of this 47 year old lady, a diabetic since age 15, had failed to heal in spite of two hospitalizations in California for debridements and parenteral antibiotic therapy. She came to Bryn Mawr to be treated as an outpatient. How would you have treated her and what success would you expect?

General Considerations and Imaging Procedures: The development of osteomyelitis adds to the morbidity  and costs of foot infections. Longer and more aggressive antibiotic therapy and more surgery are likely. While serial x-rays of the foot over several weeks, taken together with obvious clinical findings such as abscesses, draining fistula, cellulitis etc, may show definitive evidence of osteomyelitis, early diagnosis or diagnosis in the absence of such findings may require other visualization techniques. Most studies with the techniques listed below involve relatively small numbers of patients making statistical evaluation of the techniques difficult. Measures of precision and confidence limits are commonly not included in such studies. In evaluating the test results, the clinician is best reminded the number of patient results fall into four possibilities: (1) The positive tests do indeed represent  true positives, (T)rue (P)ositive; (2) The positive tests are false positives, (F)alse (P)ositive; (3) The negative tests are true negative results, (T)rue (N)egative; (4) The negative results are (F)alse (N)egative… or TP, FP, TN and FN. The test results  must be compared to some kind of "gold standard". The sensitivity of a test is the probability that it will produce a true positive result when compared to the standard and is calculated as TP/(TP + FN) . Again, compared to a "gold standard", the specificity of a test represents the probability the test will produce a true negative result when used on a healthy population and is calculated as TN/(TN+FP). The precision of the sensitivity and specificity estimates or likelihood ratios should be stated  In contrast, the specific values given for diagnostic accuracy are merely estimates. Other calculations that may be helpful include the Positive Predictive Value (TP/(TP+FP) and the Negative Predictive Value (TN/(TN+FN). The positive predictive value is the probability that a person is indeed infected when a positive test result is reported; the test should only be calculated from studies that legitimately reflect the number of people in the population under study who are infected with the disease of interest at that time. The negative predictive value is the probability that a person is not infected when a negative test result is reported; it should only be used if prevalence is available from the data. For the sake of completion, one might consider the Positive Diagnostic Likelihood Ratio {[TP/(TP+FN)]/[FP/(FP+TN)]} and the  Negative Diagnostic Likelihood Ratio {[FN/(TP+FN)]/[TN/(FP+TN)] or (false negative rate)/(true negative rate)}. The latter two tests are not dependent upon the prevalence of disease. The positive DLR represents the odds ratio that a positive test result will be reported in an infected population compared to the odds that the same result will be reported in a noninfected population (High numbers signify a good test. Thus, a ratio of 10 means that for every 1% of a control population that tests positive, 10% of the infected population tests positive.) The negative DLR represents the odds ratio that a negative test result will be observed in an infected population compared to the odds that the same result will be observed among a noninfected population. (In this case, helpful tests will have negative DLRs close to 0, and less useful tests will have higher negative DLRs. Thus, for example,  for a  ratio equal to 10,  for every one false negative,  10 true negatives are reported.

            Data can be collected in patients most likely to have osteomyelitis from clinical findings, in patients suspected of having osteomyelitis, in those with ulcers and no other suggestion of osteomyelitis, and in those with findings suggesting osteomyelitis is unlikely. The success of the various tests may vary in each group. See  our library for abstracts of such studies (www.circulatorboot.com/literature/cellulit.html).  Thus, Croll et al found the MRI to be superior to other tests in documenting osteomyelitis in patients with obvious foot infections while Newman et al found the indium leukocyte scan to be superior to the MRI in patients with unsuspected osteomyelitis..

Tests Used in the Diagnosis of Osteomyelitis in the Diabetic Foot

Tests                                                    Specificity                                 Sensitivity         +PV  -PV                   

MRI                                                     75-100%                                 82-100             50       58

Technetium Scan  22                            50                                            50-100                                    

Indium Scan  19                                   67-90                                       33-90               70      100

Plain x-ray                                            75-83                                      22-75

Gallium                                                 67-70

CT scan                                               65-70                                        65-70

Probing to bone                                    85                                            66                     89        56

Staphylococcus

  Staphylococcus is especially common in diabetic bone infections. It spreads from superficial ulcers to the adjacent bone. The organism may have receptors or adhesions for bone matrix and cartilage and a fibronectin adhesin allowing it to bind to implanted devices.  The organism can be internalized by cultured osteoblasts and survive intracellularly possibly explaining the persistence of some infections. Signs of infection for more than ten days suggest the development of necrotic bone and chronic osteomyelitis.

Diagnosis of osteomyelitis and its pathophysiology.

First, the diagnosis must be suspected. The presence of exposed pus-covered bone represents most certainly osteomyelitis. Bone that may be probed through fissures is most likely infected. Baseline x-rays may not show abnormalities while later films do; bacteria do not consume calcium and are not responsible for the changes seen on x-ray. The immune response of the patient with the resultant dissolution of dead bone and the deposition of new bone are the processes revealed by the x-ray. The results of scans and x-rays depend on the stage of the osteomyelitic process. The latter usually begins as an extension of a cellulitic process in a diabetic. The periosteum may be penetrated and eventually separated from the bone. The superficial osteoblasts may be killed and the bacteria enter the bone along the haversian canals and lacunar system, killing the osteocytes and disrupting the membrane normally separating the bony matrix and minerals from the extracellular fluid. The denuded matrix becomes a foreign body or sequestrum that invites attack and resorption by the scavenger cells in the body. An adequate blood supply is necessary for the resorption process and together with control of the infection for the preservation of the periosteum.  The latter retains osteogenic potential and may surround the inflamed bone with an involucrum, which may be remodeled with time to produce a normal bone outline. Hence, scans early in the process showing an increase in blood flow or the concentration of leukocytes may be positive. A technetium scan is a nonspecific test with a specificity too low to confirm the diagnosis of osteomyelitis in many clinical situations. A leukocyte scan is more specific. Thus, Newman et al found osteomyelitis to underlie 68% of ulcers as determined by bone biopsy and culture and detected these by leukocyte scanning with Indium in 111 oxyquinoline with an 89% sensitivity. Sixty-seven percent had no evidence of inflammation. The image intensity decreased by 16-34 days of antibiotic therapy and normalized by 36-54 days. In another report they noted that they detected 100% of biopsy-proven osteomyelitis with their 111 In scans while with magnetic resonance imaging they detected but 29%. Weinstein et al might not contest the superiority of leukocyte cans over MRI's, but they do report that MRI is significantly more sensitive than plain x-rays and technesium and gallium scans.  As the scans and sedimentation rate become normal, the x-rays may show bone destruction and resorption. Over the subsequent year the bone may remodel and gain a sharp well-mineralized outline that may or may not conform to its original shape. Sequential changes on x-rays do not necessarily indicate advancing infection and in the presence of improvement in the inflammatory process, closing of the skin and normalization of the white count and sedimentation rate, may be a normal part of the healing process. Such bony changes are likewise seen with healing traumatic fractures when they commonly are allowed to heal without surgical intervention. The need for bone scans and biopsy is questionable in clinical practice. Bone biopsy adds additional injury to the wound. While bone biopsy cultures are more definitive than cultures of swabs of the ulcer or drainage fluid, the latter are not traumatic and repeated as necessary over time will include the organism found on the biopsy. The scans are expensive and rarely alter treatment; the presence of osteomyelitis slows the healing of ulcers and cellulitis thus prolonging their treatment to a time when the plain x-ray will be abnormal.

Problems with standard antibiotic therapy.

Effective antibiotic therapy may be difficult especially in the elderly. The latter have an increased incidence of drug allergies, gastrointestinal problems affecting proper nutrition and drug tolerance, cardiac problems affecting volume and solute overload and nephrotoxicity associated with diabetes and use of diuretics. In all patients, an effective antibiotic must have activity against the invading bacteria and must reach the infected tissues in sufficient concentration. The neuropathic disease described in our last Newsletter may decrease the delivery of antibiotics and, once sufficient vascular occlusions exist, tissue destruction is frequently found to progress in spite of adequate serum levels of appropriate antibiotics. Prolonged antibiotic therapy is commonly prescribed for patients with osteomyelitis and at considerable risk and costs both in the hospital and the outpatient clinic. The real costs of the procedure include the costs of the antibiotics, the infusion equipment, the visiting nurse and any complications. The visiting nurse commonly is reimbursed more than a boot physician who provides both antibiotic local injections and boot therapy in his office! The indwelling intravenous catheter has become more important today as HMO's force patients to be discharged from the hospital to receive long-term antibiotics at home. Central venous catheters account for an estimated 90% of all nosocomial bloodstream infections with multiple lumen catheters especially having a high risk. Such infection may be reduced somewhat with antibiotic-coated catheters. Other reported complications include sterile phlebitis in 2-10%, large vein thrombosis soon after or many months after catheter placement, pulmonary emboli, superior vena cava syndrome, air embolism (potentially fatal), catheter fragment embolization, catheter tip migration to the right atrium or the jugular vein, catheter erosion through a vein or the right atrium (producing pericardial tamponade), intracatheter clots, fluid leaks through small holes in the catheter causing fluid extravasation or contiguous mass formation, rare idiosyncratic hypersensitivity reactions to the catheter substance, and exit-site infections, tunnel infections and catheter-related bloodstream infections. Finally, a rare form of infective endocarditis may occur when a malpositioned catheter traumatizes the tricuspid valve resulting in platelet-fibrin thrombi that become infected. (Ref: www.circulatorboot.com/literature/cellulit.html.)

Morbidity of Osteomyelitis as Commonly Treated

Surgical debridements or amputations are commonly deemed necessary when viualization procedures are read as showing necrotic or diseased bone. Bamberger described a significant amputation rate in their study of osteomyelitis. Of their 51 diabetic patients, 15 patients had a BK amputation and 9 a toe amputation. The absence of necrosis and/or gangrene, the presence of swelling and the use of antimicrobial therapy active against the isolated pathogens for at least 4 weeks intravenously or combined orally-intravenously for ten weeks predicted a good outcome. Fejfarova et al found 26% of their 191 diabetic foot ulcer patients underwent amputations (88% minor and 12% major). Compared to their patients not having amputations, those having amputations were significantly more likely to have resistant organisms, especially resistant Staphylococcus, in their foot cultures (43% vs 14%) and to have osteomyelitis (69% vs 13%) and peripheral vascular disease (79% vs 60%). Venkatesan et al prescribed oral antibiotics over a range of 5-72 weeks (median 12 weeks) to 22 diabetic patients with overt osteomyelitis with resolution of the osteomyelitis in 17 over 5-73 months (median 27); four patients did not respond and had amputations while one patient had a recurrence of osteomyelitis at the same site six years later. Weinstein et al: maintain that early follow-up and complete resection of abnormal bone on MRI correlated with clinical healing.

Our Patient at Four year follow-up

For our method of treatment see http://www.circulatorboot.com/literature/angiology2.html. For multiple other cases so treated and spared surgery see http://www.circulatorboot.com/casehistory/csemenu1.html.

Perspectives in Circulation Today

Volume 1, Number 6