Rapid diagnosis of pedal osteomyelitis in diabetics with a technetium-99m-labeled monoclonal antigranulocyte antibody☆☆☆
Article Outline
Abstract
An in vivo method of labeling white cells that diagnoses diabetic pedal osteomyelitis safely, rapidly, and accurately is desirable. The objectives of this investigation were to evaluate a technetium-99m-labeled monoclonal antigranulocyte antibody for diagnosing diabetic pedal osteomyelitis, compared with indium-111-labeled leukocyte and 3-phase bone imaging for this purpose. Twenty-five diabetic patients with pedal ulcers, 22 in the forefoot and 3 in the midfoot, underwent antibody, indium-111-labeled leukocyte, and technetium-99m methylene diphosphonate 3-phase bone imaging. The 1-hour antibody, 24-hour labeled leukocyte, and 3-phase bone images were interpreted separately for the presence of osteomyelitis. The antibody and labeled leukocyte images also were interpreted together with the bone images to determine if the combined study was more accurate than each individual study. There were 10 cases of osteomyelitis among the 25 patients. The sensitivity, specificity, and accuracy of the antibody were .90, .67, and .76, respectively. These results were not significantly different from those obtained with labeled leukocyte imaging: .80, .67, and .72, respectively. The antibody was significantly more specific (P = .004) than 3-phase bone imaging (.27). Interpreting the antibody together with the bone scan did not alter the results. When interpreted with the bone images, the accuracy of labeled leukocyte imaging improved from .72 to .80. This was not significantly more accurate than either the antibody or labeled leukocyte imaging alone. The data suggest that the monoclonal antigranulocyte antibody is comparable with in vitro labeled leukocyte imaging for diagnosing pedal osteomyelitis in diabetic patients, and warrants further investigation in a larger population. (The Journal of Foot & Ankle Surgery 42(1):2-8, 2003)
Keywords: diabetic foot infection, osteomyelitis, labeled leukocyte imaging, antigranulocyte antibody
Nuclear medicine imaging often is used to diagnose osteomyelitis in diabetic patients with pedal ulcers. Although 3-phase bone scintigraphy has an accuracy of 90% or greater for diagnosing osteomyelitis in otherwise normal bone, the specificity of the test decreases in the setting of pre-existing conditions such as the neuropathic joint, which is often present in the diabetic patient's foot (1). Labeled leukocyte scintigraphy is used in addition to, or instead of, bone imaging in this population because of its greater specificity. However, there are disadvantages to this technique. The procedure is labor intensive, and requires in vitro handling of blood products. An in vivo method of labeling white cells that would diagnose diabetic pedal osteomyelitis safely, rapidly, and accurately is very desirable.
One such agent that has been used for white cell labeling is a monoclonal antigranulocyte antibody that binds in vivo to leukocytes. It is a murine monoclonal immunoglobulin M antibody that binds to the CD15 antigen present on human polymorphonuclear leukocytes. Intravenously administered radioactivity concentrates in areas of infection and inflammation where leukocytes have accumulated, and there are data that confirm that the agent is useful for imaging infection in humans 2, 3, 4, 5.
The objectives of this investigation were to evaluate this technetium-99m-labeled monoclonal antigranulocyte antibody (Moab) (LeuTech; Palatin Technologies, Inc., Princeton, NJ) for diagnosing diabetic pedal osteomyelitis, and to compare it with indium-111-labeled leukocyte (WBC) and 3-phase bone imaging for this purpose.
Materials and methods
The institutional review boards of the participating institutions approved this study. Written informed consent was obtained after the nature of the investigation had been explained fully to the participant. Eligibility criteria for entry into the study included diabetic patients older than 18 years of age with a peripheral leukocyte count of at least 2,500/mm3, who were suspected of having osteomyelitis underlying a pedal ulcer based on the presence of one or more of the following: localized pain, fever greater than 100°F for at least 3 days, elevated peripheral leukocyte count, elevated erythrocyte sedimentation rate, radiographic findings suggestive of osteomyelitis, or positive blood or wound cultures. Patients with granulating surgical incisions or who had received 7 or more days of antibiotic therapy at the time of enrollment were excluded. Patients were required to undergo WBC and 3-phase bone imaging within 1 week of the Moab. Twenty-five patients, 17 men and 8 women, have been enrolled to date. In 22 patients, the ulcer was in the forefoot, and in 3 it was in the midfoot. Final diagnoses were based on bone biopsy examination and culture in 20 patients. In the remaining 5 patients, the final diagnosis was made by an experienced clinician (R.C.) based on all available data (other than the results of the investigational agent).
The Moab was supplied as a lyophilized, sterile formulation, containing 250 μg of antibody. At the time of use, 0.2-0.35 mL of 99mTcO 4−, containing 740-1,480 MBq of 99mTc, was added to the kit, and the mixture was then incubated at 37°C for 30 minutes. After incubation, a sufficient volume of 500 mg/mL ascorbic acid injection was added to the vial to bring the final preparation volume to 1 mL. Quality control was performed using instant thin layer chromatography. All kits used for patient studies had less than 10% free pertechnetate. The dosage for patient administration, drawn immediately before use, was administered within 6 hours after reconstitution. Patients were injected with 370-740 MBq (75-125 μg) 99mTc-labeled antibody.
Imaging was performed on a large field-of-view gamma camera equipped with a low-energy, high-resolution, parallel hole collimator. Energy discrimination was accomplished using a 20% window centered on the 140 keV photopeak of 99mTc. Five-minute dorsal, plantar, medial, and lateral images were acquired using a 256 × 256 × 16 matrix. All patients were imaged within 1 hour after injection of the tracer; some patients had additional imaging at later time points. Only images acquired during the first hour were analyzed subsequently.
For WBC, 40 mL of whole blood was withdrawn for labeling with 111In-oxine, according to the method of Thakur et al. (6). Approximately 18.5 MBq of 111In-labeled autologous leukocytes were injected and imaging was performed 18-30 hours later. Images were acquired on a large field-of-view gamma camera equipped with a medium-energy, parallel hole collimator. Energy discrimination was accomplished using a 15% window centered on the 174 keV photopeak and a 20% window centered on the 247 keV photopeak of 111In. Images were acquired for 10-15 minutes per view, using a 128 × 128 × 16 matrix. Views obtained were the same as those obtained for the Moab study.
Three-phase bone scintigraphy was performed with 740 MBq of 99mTc-methylene diphosphonate. Imaging was performed on a large field-of-view gamma camera, equipped with a low-energy, high-resolution, parallel hole collimator. Energy discrimination was accomplished using a 20% window centered on the 140 keV photopeak of 99mTc. Dynamic acquisitions were performed using a 64 × 64 × 16 matrix. Static images were acquired using a 256 × 256 × 16 matrix. For the dynamic and blood pool images, the plantar view was used. For delayed images, the views acquired were identical to those obtained for the Moab and WBC studies.
The Moab, WBC, and 3-phase bone studies initially were interpreted separately. Subsequently, the Moab and WBC were each interpreted in conjunction with the bone scan (Moab/Bone, WBC/Bone). A single experienced nuclear physician (C.J.P.), who had no knowledge of the results of other studies or the final diagnoses, reviewed the studies randomly and interpreted them according to the following criteria: 1) Moab: images were interpreted as positive for osteomyelitis when focal activity, felt to be bony, was increased relative to adjacent activity; 2) WBC: images were classified as positive for osteomyelitis when focally increased activity, equally well seen on the dorsal and plantar views, was present (7); 3) 3-phase bone scan: focal hyperperfusion, focal hyperemia, and focally increased bony uptake on delayed images was interpreted as positive for osteomyelitis; 4) Moab/Bone: the combined study was interpreted as positive for osteomyelitis if there was abnormal uptake in the same region on both studies; 5) WBC/Bone: the combined study was interpreted as positive for osteomyelitis if there was abnormal uptake in the same region on both studies (8).
Sensitivity, specificity, and accuracy were calculated for the Moab, WBC, and 3-phase bone studies individually, and for Moab/Bone, and WBC/Bone. The significance of differences in the sensitivity, specificity, and accuracy of each of these 5 tests was determined with the Student t test. A P value of less than .05 was considered statistically significant.
Results
There were 10 cases of osteomyelitis among the patients studied. Histopathologic confirmation was available for 9 patients. In the tenth patient the final diagnosis of osteomyelitis was made on the basis of the clinical data. Fifteen patients did not have osteomyelitis. This was confirmed by bone biopsy examination and culture in 11 patients and clinically in 4 patients. In 1 of these 4 patients the pedal ulcer healed spontaneously and the biopsy procedure was cancelled. The remaining 3 patients responded to therapy for the pedal ulcer and osteomyelitis therefore was deemed unlikely (Table 1).
Table 1. Patient data
| Patient | Age | Sex | Site | Moab | WBC | 3-Phase Bone | Moab/Bone | WBC/Bone | Proof |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 31 | M | R 2nd MT | TN | TN | FP | TN | TN | C |
| 2 | 77 | M | L 3rd MT | TN | TN | FP | TN | TN | H |
| 3 | 62 | F | R 4th MT | TP | FN | TP | TP | FN | H |
| 4 | 47 | M | L 1st Toe | TN | TN | FP | TN | TN | C |
| 5 | 84 | F | R 4th Toe | TN | FP | TN | TN | TN | H |
| 6 | 74 | M | L 1st MT | TN | FP | TN | TN | TN | H |
| 7 | 63 | M | R 2nd Toe | FP | TN | FP | FP | TN | H |
| 8 | 47 | M | R Hind-foot | TP | TP | TP | TP | TP | H |
| 9 | 54 | F | L 4th Toe | TN | TN | TN | TN | TN | H |
| 10 | 77 | F | R 1st Toe | TP | TP | TP | TP | TP | H |
| 11 | 40 | M | L 3rd Toe | TN | TN | FP | TN | TN | H |
| 12 | 60 | M | L Mid-foot | FP | TN | FP | FP | TN | H |
| 13 | 52 | M | L 1st Toe | TN | TN | FP | TN | TN | C |
| 14 | 53 | M | R 1st Toe | FP | FP | FP | FP | FP | H |
| 15 | 54 | F | R 1st Toe | TP | TP | TP | TP | TP | H |
| 16 | 55 | M | R 1st Toe | FN | FN | FN | FN | FN | H |
| 17 | 47 | M | L 1st Toe | TN | TN | FP | TN | TN | C |
| 18 | 82 | F | R 1st Toe | TP | TP | TP | TP | TP | H |
| 19 | 59 | M | R Mid-foot | TP | TP | TP | TP | TP | H |
| 20 | 75 | M | R 1st MT | TP | TP | TP | TP | TP | H |
| 21 | 83 | M | L 2nd MT | FP | FP | FP | FP | FP | H |
| 22 | 54 | F | R 4th Toe | TP | TP | TP | TP | TP | C |
| 23 | 72 | F | R 2nd MT | TN | TN | TN | TN | TN | H |
| 24 | 62 | M | R 1st Toe | TP | TP | TP | TP | TP | H |
| 25 | 66 | M | R 4th MT | FP | FP | FP | FP | FP | H |
The Moab was very sensitive (.90), and moderately specific (.67). False-positive results were related to adjacent soft-tissue infection. The Moab was slightly but not significantly more sensitive than WBC, and as sensitive as 3-phase bone imaging. The specificity of the Moab was identical to that of WBC and significantly higher than that of 3-phase bone imaging (P = .004). The results of the Moab and WBC were concordant in 20 patients: 8 true positives, 8 true negatives, 3 false positives, and 1 false negative (Figs. 1 and 2).
Fig. 1.
Both the Moab and WBC images show focally increased activity in the right great toe of a 54-year-old woman with biopsy examination-proven osteomyelitis of this digit.
Fig. 2.
All 3 studies are false positive for osteomyelitis in a 53-year-old man with a pedal ulcer underlying the right great toe. Bone biopsy examination revealed no evidence of osteomyelitis. Interpreting the bone images with either the Moab or WBC does not change their interpretation.
Fig. 3.
The Moab study is true negative, whereas the WBC study is false positive in a 74-year-old man with a pedal ulcer underlying the left first metatarsal bone. Left lower extremity amputation revealed gangrene but there was no evidence of osteomyelitis. When the WBC images were read in conjunction with the bone scan, the study was classified correctly as negative. The Moab results are unchanged.
Fig. 4.
The Moab is true positive and the WBC is false negative in a 62-year-old woman with biopsy examination-proven osteomyelitis of the fourth metatarsal of the right foot. The addition of the bone images does not alter the interpretation of either the Moab or WBC.
The results of Moab/Bone were identical to those of the Moab alone. WBC/Bone was superior to WBC alone. Although WBC/Bone was the most accurate study in this investigation, it was not significantly more accurate than Moab, Moab/Bone, or WBC (Tables 1 and 2).
Table 2. Imaging results
| Moab | WBC | 3-Phase Bone | Moab/Bone | WBC/Bone | |
|---|---|---|---|---|---|
| Sensitivity | 9/10 (.90) | 8/10 (.80) | 9/10 (.90) | 9/10 (.90) | 8/10 (.80) |
| Specificity | 10/15 (.67) | 10/15 (.67) | 4/15 (.27) | 10/15 (.67) | 12/15 (.75) |
| Accuracy | 19/25 (.76) | 18/25 (.72) | 13/25 (.52) | 19/25 (.76) | 20/25 (.80) |
Discussion
The most commonly encountered limb-threatening foot problem in the diabetic patient is the mal perforans or pedal trophic ulcer. This lesion accounts for more than 90% of all cases of diabetic pedal osteomyelitis (9). Although the diagnosis of osteomyelitis can be made with certainty when bone is visible to the naked eye or a blunt probe can be passed to the bone, the majority of patients with pedal osteomyelitis lack obvious clinical signs and symptoms (other than the ulcer) and the diagnosis frequently is not appreciated clinically 7, 10. As a result, imaging studies are an important component of the evaluation of these patients. The most commonly performed radionuclide studies are 3-phase bone and labeled leukocyte imaging. The 3-phase bone scan, as the data in this study show, is sensitive but not specific, and its role in the diagnosis of diabetic pedal osteomyelitis usually is limited to that of a screening test (11). Labeled leukocyte scintigraphy is the current radionuclide gold standard for diagnosing this entity. Initially, white cells were labeled with 111In 7, 8. More recently, 99mTc has been used 12, 13. Regardless of which radionuclide is used, however, cell labeling must be performed in vitro. This procedure is labor intensive, has limited availability, and requires direct contact with blood products. In the mid- and hindfoot, normal bone marrow activity can confound image interpretation, and complementary marrow imaging may be needed to clarify the findings (14). When using 111In as the radiolabel, images are performed routinely the day after injection of labeled cells. Image resolution is poor and it may not always be possible to separate soft tissue from bony infection. The use of 99mTc as the radiolabel overcomes some, but not all, of these problems. Results, at least in some conditions, may be available within 3-4 hours after injection, and image quality is superior to that obtained with 111In. The need to perform dual tracer imaging introduces its own logistic problems. However, the use of 99mTc does not obviate what is currently the single greatest impediment to white cell imaging: the need to perform the labeling in vitro. An agent that is comparable in accuracy, but without the disadvantages of in vitro labeled leukocyte imaging, represents a significant improvement over current methodology.
The results obtained with the Moab in this population are encouraging. The accuracy of the Moab was comparable with that of WBC. The results of both studies were concordant in 20 of 25 patients, and this suggests that the behavior of leukocytes labeled in vivo with the Moab is similar to that of leukocytes labeled in vitro with 111In.
The ideal radionuclide method for diagnosing diabetic pedal osteomyelitis would entail only one study. WBC/Bone was slightly, but not significantly, more accurate than either WBC or Moab in this series. This suggests that perhaps the Moab alone will be sufficient.
Although the results obtained are encouraging, there were limitations to the study. In 22 of 25 patients (88%), the area of concern involved the forefoot. Consequently, these results cannot automatically be extrapolated to the mid- and hindfoot, where the Charcot joint is encountered more frequently. A separate investigation is needed to determine the value of this agent in the patient with a neuropathic joint. Patients with open, granulating, surgical incisions were excluded from this investigation and, as in the case of the neuropathic joint, the role of the Moab in this population remains to be determined. Finally, only patients receiving antibiotic therapy for less than 7 days were eligible for entry into the study, and consequently the effect of a more prolonged course of treatment on the accuracy of the Moab is unknown.
Summary
Our data suggest that the monoclonal antigranulocyte antibody under investigation, which was comparable with in vitro labeled leukocyte imaging for diagnosing pedal osteomyelitis in diabetic patients, warrants further investigation in a larger population.
References
- . The scintigraphic diagnosis of osteomyelitis. AJR Am J Roentgenol. 1992;158:9–18
- . Imaging inflammatory diseases with neutrophil-specific technetium-99m-labeled monoclonal antibody anti-SSEA-1. J Nucl Med. 1996;37:1789–1795
- . Effects of a 99mTc-labeled murine immunoglobulin M antibody to CD 15 antigens on human granulocyte membranes in healthy volunteers. J Nucl Med. 1999;40:2107–2114
- . Intraindividual comparison of 99mTc-labelled anti-SSEA-1 antigranulocyte antibody and 99mTc-HMPAO labelled white blood cells for the imaging of infection. Eur J Nucl Med. 1998;25:386–393
- . Neutrophil-specific Tc-99m-labeled anti-CD 15 monoclonal antibody imaging for diagnosis of equivocal appendicitis: clinical evaluation of safety, efficacy and time performance characteristics. J Nucl Med. 2000;41:449–455
- . Indium-111-labeled autologous leukocytes in man. J Nucl Med. 1977;18:1014–1021
- . Unsuspected osteomyelitis in diabetic foot ulcers: diagnosing and monitoring by leukocyte scanning with In-111 oxyquinoline. JAMA. 1991;266:1246–1251
- . Diagnosis of osteomyelitis of the foot in diabetic patients: value of 111In-leukocyte scintigraphy. AJR Am J Roentgenol. 1991;157:527–531
- . Osteomyelitis in the feet of diabetic patients. Long term results, prognostic factors, and the role of antimicrobial and surgical therapy. Am J Med. 1987;83:653–660
- . Probing to bone in infected pedal ulcers. A clinical sign of underlying osteomyelitis in diabetic patients. JAMA. 1995;273:721–723
- . Scintigraphic evaluation of the diabetic foot. In: Freeman LM editors. Nuclear Medicine Annual. Philadelphia: Williams and Wilkins,; 2000;p. 143–172
- . Contribution of technetium-99m hexamethylpropylene amine oxime labelled leucocyte scintigraphy to the diagnosis of diabetic foot infection. Eur J Nucl Med. 1998;25:132–138
- . Diagnosis of pedal osteomyelitis with Tc-99m HMPAO labeled leukocytes. J Foot Ankle Surg. 1997;36:120–126
- . Marrow versus infection in the Charcot joint: indium-111 leukocyte and technetium-99m sulfur colloid scintigraphy. J Nucl Med. 1998;39:346–350
☆ Address correspondence to: Christopher J. Palestro, MD, Division of Nuclear Medicine, Long Island Jewish Medical Center, 270-05 76th Ave, New Hyde Park, NY 11040. e-mail: palestro@lij.edu.
☆☆ Supported by a grant from Palatin Technologies, Inc., Princeton, NJ; research grants from Palatin Technologies (C.J.P., S.L.K., F.L.W.); honoraria from Palatin Technologies (C.J.P., S.L.K.). S.L.K. is a stockholder in Palatin Technologies.
PII: S1067-2516(03)70045-3
doi:10.1053/jfas.2003.0420001
© 2003 American College of Foot and Ankle Surgeons. Published by Elsevier Inc. All rights reserved.
