Peritoneal Adhesions Show Distinct Tissue States

Peritoneal adhesions are bands of tissue that can form after abdominal surgery. Their visible forms range from thin strands to dense attachments, but appearance alone does not explain whether those forms represent different biology. Two independent studies examine opposite ends of the process: a new analysis classifies mature human adhesions by their tissue structure, while an earlier mechanistic study tests how inflammatory DNA scaffolds contribute to adhesion formation.

Evidence

The new study examined adhesion tissue from 77 patients. Researchers used Masson's Trichrome staining to visualize extracellular matrix, the protein-rich framework that gives tissue much of its structure. A qualitative histological analysis separated the specimens into five clusters: stringy, dense, loose, fatty, and mixed. This was not simply a relabeling of surgical appearance; the groups differed in measurable microscopic features.

Stringy and dense adhesions had the highest extracellular-matrix fraction and vessel area. Loose and fatty adhesions had lower matrix density and vascularity. Immunohistochemical staining for alpha-smooth-muscle actin, a marker used to identify myofibroblasts, and PGP9.5, used to visualize nerve fibers, also showed cluster-specific differences in cellular composition and innervation. Cluster distribution was significantly associated with clinical features including overall adhesion load and previous peritonitis. The study therefore supplies a five-state tissue framework, but its abstract does not report an externally validated classifier or show that a cluster predicts symptoms, recurrence, or surgical difficulty.

The independent iScience study investigated an earlier step. Researchers induced peritoneal injury in mouse models, altered pathways that produce or clear neutrophil extracellular traps, or NETs, and examined human surgical samples. NETs are webs of extracellular DNA decorated with neutrophil proteins. In wild-type mice, extracellular DNA at the injury site peaked at 72 hours, while cumulative adhesion formation reached its maximum at 21 days.

Removing the DNA-clearing enzymes DNASE1 or DNASE1L3 increased experimental adhesions, with the stronger phenotype in DNASE1L3-deficient mice. By contrast, mice lacking PAD4, which limits aggregated NET formation, had reduced adhesions. Topical treatment with the DNASE1L3 analog NTR-10 reduced adhesion scores at day 21, and treatment with dornase alfa or NTR-10 reduced peritoneal thickness. In additional mouse models involving intestinal anastomosis, surface injury, and thermal injury, dornase alfa reduced adhesions without a detected increase in wound infection, incisional hernia, or anastomotic failure.

Human samples provided observational support for the mechanism. Adhesions contained the NET-associated proteins neutrophil elastase and myeloperoxidase, as well as citrullinated histone H3 and extracellular DNA. High-resolution mass spectrometry also detected several NET-borne proteins. The investigators found modified fibrin in adhesion tissue, including oxidation and citrullination near sites normally cut during fibrin breakdown. These human findings show that NET-related material is present in adhesions; they do not establish that NETs caused any individual patient's adhesion.

Analysis — From early scaffold to mature states

The cross-study pattern suggests a staged model, and this is analysis rather than a jointly tested conclusion. The mechanistic study places extracellular DNA, neutrophil proteins, fibrin, and DNA-clearing enzymes early in adhesion formation. The newer human study shows that mature adhesions do not settle into one uniform tissue state: matrix abundance, vascularity, myofibroblast staining, and innervation vary across five categories. One plausible connection is that differences in the intensity, duration, or clearance of the early inflammatory scaffold help shape later tissue organization. Prior peritonitis being associated with cluster distribution is consistent with an inflammatory contribution, but it does not prove that NET activity created a particular cluster. The studies become more informative together because they separate initiation from mature architecture. A decisive test would follow human tissue or matched experimental samples over time, measure NET and fibrin markers early, and determine whether those measurements predict a prespecified histological state later. Until that is done, the proposed sequence remains an emerging model, not a validated clinical classification or prevention strategy.

Limitations

The 77-patient classification is available in the ingested pack only as a PubMed abstract. Its five groups arose from qualitative histological assessment, and the available text does not provide inter-rater reliability, detailed patient selection, cluster sizes, adjusted effect estimates, or prospective validation. Associations with adhesion load and prior peritonitis cannot establish direction or causation. Larger cohorts would be needed to test whether the framework is reproducible or clinically informative.

The NET/DNase evidence is predominantly preclinical. Human tissue demonstrated co-location and protein signatures, while causal perturbations and DNase treatment occurred in mice. The study focused on early formation and did not establish long-term recurrence, patient outcomes, human dosing, treatment duration, or adverse effects. Its authors also disclosed advisory, ownership, or research-funding relationships with a company developing NET-directed therapies. Finally, the two studies did not analyze the same specimens: neither tested whether NET abundance differs among the five new histological clusters. The synthesis therefore identifies a testable bridge between inflammation and tissue structure, not evidence for medical decisions.