KL-6 separates faster from slower lung decline in SSc-ILD
Circulating biomarkers in systemic sclerosis-associated interstitial lung disease may be moving closer to bedside utility. In the SENSCIS biomarker analysis, baseline KL-6 emerged as the clearest signal for progressive lung disease, while treatment with nintedanib shifted markers linked to epithelial injury and collagen synthesis. The findings sit at the intersection of prognosis and pharmacology: one set of proteins helped identify patients headed toward steeper functional loss, and another tracked biologic response to antifibrotic therapy.
A baseline KL-6 concentration above 1000 U/mL separated a group with a more aggressive course. Patients above that threshold had a greater rate of decline in forced vital capacity over 52 weeks than those at or below 1000 U/mL, with estimates of -132.5 mL versus -56.4 mL and a P value of .009. In practical terms, this is the kind of difference that reflects ongoing alveolar epithelial damage and failed repair, the biological terrain in which fibrotic interstitial lung disease advances.
The study focused on patients with systemic sclerosis diagnosed within 7 years of their first non-Raynaud symptom, a population in which lung fibrosis often drives disability and survival risk. Serum and plasma markers covered inflammation, epithelial dysfunction, extracellular matrix synthesis, and extracellular matrix turnover. That design allowed the investigators to ask two separate questions at once: which markers forecast worsening lung function or skin disease, and which biomarkers change when antifibrotic treatment is introduced.
Baseline biomarker patterns pointed to distinct fibrotic pathways
Two markers linked to lung function decline stood out in uncorrected analyses: KL-6 and citrullinated vimentin (VICM) degraded by MMP-2/8. Both sit within the broader biology of tissue injury and remodeling, where epithelial stress, matrix turnover, and protease activity fuel progressive scarring. Their association with FVC decline fits the clinical phenotype of SSc-ILD, in which damaged epithelium and dysregulated repair continually remodel the lung interstitium.
The skin data told a different story. Higher baseline N-terminal propeptide of type III collagen (Pro-C3), chemokine (C-C motif) ligand 2, and C-reactive protein were associated with less improvement in modified Rodnan Skin Score at week 52 after correction for multiple comparisons. Skin thickening in systemic sclerosis is not a simple cosmetic feature; it is a readout of immune activation, fibroblast stimulation, and collagen deposition. These markers point toward a patient subgroup in whom the fibrotic program is more entrenched and less likely to soften over a year.
One sentence captures the clinical tension here: the biomarker profile separated biology before the bedside exam could.
Nintedanib shifted epithelial injury and collagen synthesis markers
Treatment effects were more selective than the prognostic signals. Over 52 weeks, patients receiving nintedanib showed decreases in CA-125 and pro-C6 compared with placebo. Both markers are relevant to the fibrotic process, with CA-125 serving as an epithelial dysfunction signal and pro-C6 reflecting type VI collagen synthesis. The direction of change is consistent with an antifibrotic mechanism that dampens tissue injury responses and slows extracellular matrix production.
The most striking mechanistic result came from mediation analysis. Nearly half of nintedanib’s effect on FVC change at week 52—48.0%—was attributed to the treatment-related decrease in CA-125 at week 24. That is an unusually concrete bridge between a circulating marker and a functional endpoint. It suggests that epithelial injury, rather than collagen turnover alone, may be one of the main biological routes through which the drug preserves lung capacity in this population.
Nintedanib does not reverse established fibrosis, but it appears to interrupt the signals that keep fibrosis active. By suppressing pathways tied to fibroblast activation and matrix deposition, the drug can slow the pace at which lung function erodes. The biomarker pattern in this analysis fits that therapeutic logic.
Why the KL-6 threshold matters at the bedside
A dichotomized KL-6 threshold is clinically appealing because it converts a continuous biomarker into a simpler risk signal. In a disease where treatment decisions often hinge on how quickly lung function is slipping, a baseline value above 1000 U/mL could help identify patients more likely to follow a progressive trajectory. The data do not make KL-6 a stand-alone decision tool, but they do support its use as part of a broader risk assessment that includes spirometry, imaging, and clinical context.
The value of that approach is not merely statistical. Patients with SSc-ILD often have overlapping drivers of decline, and the tempo of disease can be difficult to judge from symptoms alone. A marker that tracks epithelial injury may flag risk before a major functional drop appears on routine testing. In a fibrosing lung disease, that lead time matters because it can influence when antifibrotic therapy is started and how closely lung function is followed.
What the biomarker panel can and cannot do
This analysis strengthens the case for biomarker-informed phenotyping in SSc-ILD, but it also leaves several important limits in view. The biomarker associations were not uniformly significant across all analyses, and the study does not establish any single circulating protein as a replacement for clinical assessment or serial pulmonary function testing. Some signals held up only in uncorrected analyses, while others persisted after correction for multiple comparisons, a reminder that biomarker discovery is more fragile than biomarker adoption.
Even so, the biologic story is coherent. KL-6 and VICM point toward epithelial injury and matrix remodeling as markers of progression. Pro-C3, CCL2, and CRP connect systemic inflammation with limited skin improvement. CA-125 and pro-C6 move in the direction expected under antifibrotic treatment, and the mediation analysis links CA-125 to the functional preservation of FVC. The study therefore captures both prognosis and treatment response in the same disease state, using the blood as a window into the lung.
What remains unresolved is how best to integrate these measurements into routine care. The SENSCIS biomarker findings support risk stratification, but they do not yet define an action threshold beyond the KL-6 cut point tested here. They also do not show whether biomarker-guided treatment changes outcomes more effectively than standard clinical monitoring. For now, the practical message is narrower: in SSc-ILD, KL-6 identifies a more progressive lung phenotype, and nintedanib leaves a measurable imprint on markers of epithelial dysfunction and collagen synthesis.