Comment on: “FGF-23 and Long-Term Outcome in Peripheral Artery Disease”

To the Editor,

We read with great interest the article by Zierfuss et al ¹ investigating the prognostic value of Fibroblast Growth Factor-23 (FGF-23) in a cohort of 298 patients with peripheral artery disease (PAD). While the authors present a valuable 10-year follow-up, a critical appraisal of their methodology and data reveals profound contradictions that challenge their primary pathophysiological conclusions.

First, there is a glaring disconnect between the authors’ mechanistic hypothesis and their statistical reality. Throughout the discussion, Zierfuss et al ¹ heavily attribute the detrimental effects of elevated FGF-23 to progressive vascular calcification and medial sclerosis. However, according to their own fully adjusted multivariable analysis (Table 4, Model 3), FGF-23 completely fails to independently predict cardiovascular (CV) mortality (Hazard Ratio 1.19, 95% CI 0.85-1.67). It is pathophysiologically contradictory to claim that a biomarker drives fatal outcomes primarily via vascular calcification when it cannot statistically predict CV death. This strongly suggests that the observed all-cause mortality signal is driven by non-CV factors—such as frailty, recurrent infections, or unmeasured comorbidities—rather than the proposed mineral homeostasis dysregulation. Recent robust evidence supports this alternative explanation: elevated FGF-23 has been independently validated as a strong predictor of non-CV postoperative mortality in highly frail cohorts and is strongly associated with acute infection-related mortality.^2,3

Second, the authors’ reliance on a C-terminal FGF-23 (cFGF-23) assay severely confounds their findings. Recent evidence has fundamentally shifted our understanding of FGF-23 metabolism: cFGF-23 assays capture both intact, biologically active FGF-23 and its inactive cleaved fragments.^4,5 Crucially, the cleavage of intact FGF-23 into cFGF-23 is markedly upregulated by systemic inflammation and functional iron deficiency.^4,5 Since chronic inflammation and unrecognized iron deficiency anemia are highly prevalent in PAD populations,^4,5 the elevated cFGF-23 levels in the deceased cohort likely reflect a state of inflammatory or hematological distress rather than a primary disruption in the calcification axis. Without adjusting for iron status (e.g., ferritin, transferrin saturation) or measuring intact FGF-23, claiming that these results reflect early “chronic kidney disease-mineral and bone disorder (CKD-MBD)” is unsubstantiated.

Third, for a study focused exclusively on a PAD cohort, the complete omission of limb-specific clinical endpoints is highly problematic. Recent molecular analyses demonstrated that FGF-23 acts locally within peripheral arteries to promote athero-inflammation, specifically by upregulating tumor necrosis factor (TNF)α in vascular tissue. ⁶ Furthermore, Freise et al ⁷ showed that FGF-23 actively remodels the vascular extracellular matrix via glycosaminoglycan induction. Given these direct local athero-inflammatory effects, evaluating Major Adverse Limb Events (MALE)—such as major amputation or critical limb ischemia—is mandatory. By only reporting fatal events, Zierfuss et al ¹ omitted the most clinically relevant outcome for limb preservation in PAD patients.

In summary, while Zierfuss et al ¹ propose FGF-23 as a novel risk stratification tool in PAD, their own data undermine the CV calcification hypothesis due to the loss of CV-mortality significance in the fully adjusted model. Compounded by the critical limitations of measuring C-terminal rather than intact FGF-23 without adjusting for iron deficiency, and the absence of limb-specific outcomes, the pathophysiological and clinical utility of FGF-23 in this context remains highly questionable. We hope to describe more specific details regarding CV and limb-related outcomes, and first address key confounding factors such as iron metabolism and inflammation. Only after that will we recommend FGF-23 as a standard indicator for clinical practice.