TO THE EDITOR:
I congratulate Skehan et al. (1) on their recent study on the mechanism by which 99mTc-sulesomab (LeukoScan; Immunomedics, Inc.) accumulates in inflammation and infection. 99mTc-Sulesomab is a Fab′ monoclonal antibody directed against the nonspecific cross-reacting antigen of granulocytes (NCA-90), which is overexpressed by activated granulocytes. I waited some months before writing this letter because I wished to discuss the 99mTc-sulesomab kinetic data of Skehan et al. in light of clinical data that emerged from a recent study at our center (2). Interestingly, Skehan et al. compared 2 proteins characterized by a similar molecular weight—99mTc-sulesomab and 99mTc-labeled human serum albumin (HSA)—to investigate 99mTc-sulesomab clearance and uptake by the infectious site. As is known, capillary permeability increases in inflamed tissues. Both HSA and 99mTc-sulesomab are therefore expected to accumulate, at least in part, in the inflammation region by a nonspecific mechanism. The accumulation is expected to be bidirectional in the case of a “neutral” protein, as happens for HSA, whereas the accumulation should be monodirectional in the case of a protein that specifically binds to cells or tissues expressing the NCA-90 antigen and located in the extravascular interstitial space, as should happen for 99mTc-sulesomab.
The study of Skehan et al. (1) elucidated some important points: The monoclonal Fab′ antibody of 99mTc-sulesomab does not significantly bind to circulating quiescent granulocytes (<5% binding). Instead, clearance of 99mTc-sulesomab, as evaluated by Patlak–Rutland analysis, is about 3 times greater than that of HSA. These data agree with an in vitro study conducted by the same authors (1), in which, unlike HSA, primed and activated granulocytes showed a 3–4 times higher affinity for 99mTc-sulesomab than for quiescent granulocytes. Despite these observations, the authors found that the target-to-background ratios of HSA and 99mTc-sulesomab were similar at 60, 180, 240, and 360 min after radiotracer injection. I think that the authors decided to stop their study 6 h after 99mTc-sulesomab administration because some large clinical trials have found that 99mTc-sulesomab makes possible imaging and diagnosis of infection very soon, that is, within a few hours, after radiotracer injection (3,4).
However, on the basis of the study of Skehan et al. (1), the nuclear medicine reader might realize that despite specific binding of 99mTc-sulesomab to primed and activated granulocytes in the infectious site, the prevalent mechanism of 99mTc-sulesomab accumulation is related simply to the nonspecific increase of capillary permeability exactly as for HSA, at least within 6 h after injection. Thus, the question remains open of what, if any, timing should be adequate for evaluating a specific 99mTc-sulesomab binding prevailing on the nonspecific binding.
In a recent prospective study by Rubello et al. (2), the results of 253 consecutive 99mTc-sulesomab examinations of 220 patients with proven or suspected peripheral bone infection were evaluated. The protocol used in our study included both early (4 h) and, at variance with previous protocols, delayed (24 h) acquisition of 99mTc-sulesomab images. Moreover, for interpreting 99mTc-sulesomab findings, we evaluated the early (4 h) uptake pattern versus the delayed (24 h) uptake pattern. Specifically, a pattern of increasing uptake was judged as infection (true-positive result), whereas a pattern of decreasing uptake was judged as nonspecific early accumulation (false-positive result). By adopting these interpretation criteria, we obtained a significantly improved specificity for the 99mTc-sulesomab examination. In details comparing early and delayed imaging, specificity was 75% versus 87.5% in patients with diabetic foot infection and 76.2% versus 85.7% in patients with other peripheral bone infections or prosthetic joints (2). This increase in specificity was related strictly to the identification of some false-positive findings due to nonspecific 99mTc-sulesomab uptake on early images alone, for example, when blood-pool activity was significantly high. Thus, in our experience, delayed 24-h 99mTc-sulesomab imaging was useful in detecting nonspecific early 99mTc-sulesomab uptake and, as a consequence, in identifying cases of specific uptake to granulocytes at the site of infection. Of note, 24 h after 99mTc-sulesomab injection, background activity is nearly negligible in evaluations of peripheral bone, especially considering the blood-pool activity present (2).
It can be concluded that, from a clinical point of view, nuclear medicine physicians should consider reevaluating acquisition protocols and interpretative criteria for 99mTc-sulesomab imaging, particularly delayed 99mTc-sulesomab imaging. Lastly, I encourage Skehan et al. to continue their investigation and, in particular, to evaluate the kinetic characteristics of 99mTc-sulesomab on delayed 20- to 24-h imaging. Further scientific contributions to this field would be extremely useful in better establishing the role of 99mTc-sulesomab imaging in the diagnosis of peripheral bone infection.
REPLY:
Dr. Rubello makes an interesting point that is reminiscent of claims for other speculative specific, infection/inflammation-targeting agents—Infecton (Draximage Inc.), for instance (1)—that imaging at 4 and 24 h yields a higher clinical specificity than imaging at 4 h alone. It would perhaps be helpful, here, to draw a distinction between 2 different definitions of specificity: Clinical specificity is a measure of the number of false-positive results; “radiopharmaceutical” specificity is to do with whether an agent localizes in a lesion through a well-defined physiologic process for which it was designed. There appears to be a widely held view, as also hinted at by Dr. Rubello in his letter, that if accumulation (i.e., target-to-background ratio) continues to increase over 24 h, the agent must be specific by the second of these 2 criteria. Labeled leukocytes are specific on both, but a radiolabeled protein, such as polyclonal IgG, only on the first. That does not mean to say that IgG would not show increasing localization over time. It clearly does (2), and moreover, 99mTc detached from the protein may be retained in tissue (3). Therefore, the fact that an agent, including 99mTc-sulesomab (LeukoScan; Immunomedics, Inc.), gives better images at 24 h throws no light on its mechanism of accumulation, which could still be “nonspecific.”
The purpose of our study was obviously not a clinical comparison between 99mTc-sulesomab and human serum albumin but an attempt to clarify mechanisms of 99mTc-sulesomab accumulation in an inflammatory lesion, especially because the concept that 99mTc-sulesomab binds to circulating granulocytes is clearly erroneous, as shown by negligible cell binding in blood obtained ex vivo (4). Perhaps we should have extended our study to 24 h, although by then, among other problems, there would have been significant detachment of 99mTc from the respective proteins, rendering quantitative studies difficult to conduct or interpret.
If new tracers for inflammation require imaging beyond 4 h, as the evidence seems to support, then perhaps we should be looking for radionuclides more appropriate than 99mTc with which to label them. Alternatively, perhaps we should be imaging at a single time point, 7–8 h, instead of the 4- and 24-h time points that seem to be ingrained in our imaging protocols, and not just those for imaging inflammation.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.