An email landed in my inbox this morning from Ben Lipsky — a long-time friend, mentor, and research colleague — drawing the working group’s attention to a new study that, in his words, may be “some of the best data yet to help differentiate IDSA (and WIfI) moderate and severe infections.”
The question at the heart of Ben’s note is not new, but it has become unusually urgent: should “moderate” and “severe” diabetic foot infections (DFI) continue to live in the same bucket for the purposes of classification, antibiotic selection, and trial design — or do they behave like fundamentally different diseases?
What the new study shows
In Pathogen profile and inflammatory biomarkers in diabetic foot infections: correlation with disease severity, just published online in BMC Infectious Diseases by Nuermaimaiti and colleagues, the investigators retrospectively compared 80 people with diabetic foot infection to 80 matched diabetic foot patients without concurrent infection. They cultured the wounds and measured five serum inflammatory biomarkers: IL-17, IL-12p70, HMGB1, IL-8, and CRP.
A few things stood out:
- Among 131 isolates, Gram-positive (~49%) and Gram-negative (~47%) organisms came up at near-equal frequency, with a small fungal tail (~4%) — reinforcing that empirical regimens covering both classes remain reasonable.
- All five biomarkers were significantly higher in infected vs. non-infected feet (P < 0.05).
- Most importantly for Ben’s question, severity correlated strongly and monotonically with each marker on Spearman analysis — r values between 0.59 and 0.73, all P < 0.001. HMGB1 (r = 0.722) and CRP (r = 0.726) led the pack.
- Levels fell significantly after standard treatment, suggesting these markers track resolution as well as severity.
That last point matters. It is one thing to find a biomarker that distinguishes “infected” from “not infected.” It is quite another to show a biological signal that scales with the clinical severity gradient and then walks back down with successful therapy. That is the behavior of a tracking marker, not just a yes/no test.
Why this is relevant to the lumping-vs-splitting debate
For years, some in the field — including a number of us — have argued that the IDSA “moderate” category is a noisy bin holding patients who really belong in two clinically distinct populations. Ben, Warren Joseph, and Michael Silverman proposed an update to the IDSA classification back in 2017, motivated in part by the failure of a topical-antibiotic trial in mild infections — a result that suggested the group we were calling “mild” was probably itself a mix of infected and not-quite-infected wounds.
More recently, we have argued for borrowing from oncology’s Gleason-style thinking when classifying mildly infected DFUs — recognizing that even within a single severity tier, biological behavior may vary substantially. The same logic applies upward, at the moderate/severe boundary.
What this new biomarker dataset adds is a quantitative, tissue-level fingerprint that does not collapse moderate and severe into one. If serum HMGB1 and CRP scale with severity — and rise above the noise of “moderate” into a clearly different range at “severe” — that is mechanistic support for what the bedside has been telling us for decades: a person with a moderate cellulitis around a forefoot ulcer and a person with a fulminant deep-space infection threatening the limb are not the same patient. They should probably not sit in the same trial arm or stewardship pathway.
None of this displaces good clinical judgment, of course. The WIfI framework already grades foot infection on a 0–3 scale, and the recent CID review by Cortes-Penfield and colleagues reminds us that severity assessment is part of a larger, multidisciplinary loop. But biomarker data of this kind put quantitative meat on the bone.
Where this fits in the broader research community
Ben, Larry, Dane, and I have worked together for years- see this research constellation map. This kind of incremental, careful work — small studies that tighten the screws on classification — is exactly what that web is for. It is rarely flashy. It is often the most useful thing a field can do.
Citation: Nuermaimaiti Z, Abudurexiti G, Simayi M, Yibolayin A. Pathogen profile and inflammatory biomarkers in diabetic foot infections: correlation with disease severity. BMC Infect Dis. 2026. doi:10.1186/s12879-026-13386-4
#DiabeticFoot #DiabeticFootInfection #IDSA #WIfI #Biomarkers #Inflammation #HMGB1 #CRP #IL17 #Lipsky #ActAgainstAmputation @ALPSlimb @KeckSchool_USC @USC_vascular @ResearchatUSC
Leave a Reply