16. Inflammation

Acute inflammation is a stereotyped vascular and cellular program triggered within minutes of tissue injury or microbial invasion, and every classical sign — rubor, tumor, calor, dolor, functio laesa — has a direct correlate in the attenuation and enhancement behavior that CT records. The initiating events are vasodilation and a sharp rise in microvascular permeability, driven by histamine, bradykinin, prostaglandins (notably $\text{PGE}_2$ and prostacyclin), nitric oxide, and the complement anaphylatoxins $\text{C3a}$ and $\text{C5a}$. Endothelial cell contraction widens interendothelial gaps in postcapillary venules, and the consequent Starling imbalance — a rise in intravascular hydrostatic pressure from vasodilation together with a rise in interstitial oncotic pressure as protein leaks out across the permeable wall — produces a protein-rich exudate rather than the protein-poor transudate of purely hydrostatic edema. This distinction is quantitatively legible on CT: simple transudative fluid measures near $0$–$15\,\text{HU}$, whereas a proteinaceous or cellular exudate typically reads $15$–$30\,\text{HU}$, and frank pus, dense with neutrophils, debris, and proteinaceous material, commonly measures $20$–$45\,\text{HU}$ with internal heterogeneity. The earliest and most sensitive parenchymal sign is fat stranding — the infiltration of normally lucent fat (near $-100\,\text{HU}$) by exudate and edema, raising its attenuation by tens of Hounsfield units to produce reticular, hazy, or 'dirty-fat' opacification adjacent to the inflamed organ. Because fat surrounds most viscera, stranding localizes the epicenter of disease with a precision that the organ itself often cannot.

The vascular phase also explains hyperemic enhancement. Vasodilation and recruitment of capillary beds increase regional blood volume and flow, so inflamed walls and tissues enhance more avidly and earlier than normal; an inflamed appendix, gallbladder, or bowel segment shows mural hyperenhancement, and the engorged feeding vessels become conspicuous. On CT perfusion this is captured as elevated blood flow ($\text{BF}$), blood volume ($\text{BV}$), and — reflecting the leaky endothelium — an increased permeability-surface area product ($\text{PS}$); the same capillary leak that raises $\text{PS}$ produces delayed, persistent interstitial enhancement on equilibrium-phase imaging. Neutrophils, the cellular hallmark of acute inflammation, marginate along activated endothelium expressing E- and P-selectins, then firmly adhere via $\beta_2$-integrins binding ICAM-1 and transmigrate along chemotactic gradients of $\text{C5a}$, leukotriene $\text{B}_4$, and CXC chemokines such as $\text{IL-8}$. Their accumulation, with liquefactive enzymatic digestion of tissue, converts a solid inflammatory mass into an abscess: the cardinal CT signature is a low-attenuation ($0$–$30\,\text{HU}$) collection with a thick, intensely enhancing rim of granulation tissue and surrounding edema, frequently with locules of gas from gas-forming organisms or prior intervention. The temporal arc matters for management — a phlegmon (ill-defined, enhancing, non-drainable inflammatory tissue) typically antedates organized pus by 24–72 hours, and the recognition of a drainable rim-enhancing cavity versus a solid phlegmon is the pivotal decision separating percutaneous drainage from antibiotics. Expert reasoning here is Bayesian: in the right lower quadrant, an enhancing tubular structure $>6\,\text{mm}$ with periappendiceal stranding and an appendicolith makes appendicitis overwhelmingly likely, but the same stranding and rim-enhancing collection demand active exclusion of a perforated cecal carcinoma in the older patient. The dominant failure modes are technical — contrast mistiming that flattens the hyperemic gradient, and beam-hardening or partial-volume artifact that simulates wall thickening — and cognitive: satisfaction of search after finding stranding, and anchoring on 'inflammation' when a necrotic neoplasm produces an identical rim-enhancing cavity.

When the inciting stimulus is not cleared — by persistent infection, indigestible foreign material, autoimmunity, or repeated injury — the acute neutrophilic response gives way to chronic inflammation, a process defined histologically by a mononuclear infiltrate (macrophages, lymphocytes, plasma cells), by simultaneous tissue destruction and attempted repair, and by angiogenesis and fibrosis proceeding in parallel. The macrophage is the orchestrator: activated by $\text{IFN-}\gamma$ from $\text{T}_\text{H}1$ lymphocytes (classical, M1 activation) or by $\text{IL-4}$ and $\text{IL-13}$ (alternative, M2 activation), it secretes a cytokine repertoire — $\text{TNF}$, $\text{IL-1}$, $\text{IL-6}$, $\text{PDGF}$, $\text{TGF-}\beta$, $\text{FGF}$, and $\text{VEGF}$ — that drives fibroblast proliferation, collagen deposition, and neovascularization. The CT phenotype therefore differs fundamentally from acute disease: instead of a sharply hyperemic wall and clean exudate, chronic inflammation produces mural thickening with altered stratification, fibrofatty proliferation, increased regional vascularity, and ultimately fibrosis with stricture and architectural distortion. Crohn disease is the prototypical model and the one most legible on cross-sectional enterography. In active chronic disease the bowel wall thickens (commonly to $>3$–$4\,\text{mm}$ when distended) and retains or exaggerates stratification: a hyperenhancing mucosa, a low-attenuation submucosal band of edema or fat (the 'halo' or 'target' sign), and an intermediate muscularis. Mural attenuation correlates with histologic and endoscopic activity — Bodily and colleagues demonstrated that contrast-enhanced CT enterography mural attenuation and thickness track inflammatory severity — making enhancement a noninvasive surrogate for disease activity that guides escalation of immunosuppressive or biologic therapy.

The mesenteric response is equally informative. Chronic transmural inflammation recruits and dilates the vasa recta supplying the affected segment, producing the comb sign — the original description by Meyers and McGuire termed it 'vascular jejunization of the ileum,' the engorged straight vessels lined up like the teeth of a comb perpendicular to the antimesenteric border — a marker of active hyperemia. Adjacent fibrofatty proliferation ('creeping fat'), reactive mesenteric nodes, and a 'misty' mesentery complete the picture of active chronic inflammation. As the process matures and $\text{TGF-}\beta$-driven fibrogenesis dominates, the signature shifts again: stratification is lost, the wall enhances more homogeneously and less avidly, and a fixed, fibrotic stricture develops with upstream dilatation. This transition is the crux of a clinically vital and difficult distinction — inflammatory (potentially reversible, steroid- and biologic-responsive) versus fibrotic (fixed, surgical) stricture — because they look superficially similar yet demand opposite management. Delayed equilibrium-phase enhancement, the degree of stratification, and CT/MR perfusion parameters all contribute, but the overlap is real and is a recognized diagnostic limitation. The Bayesian differential for a thickened, enhancing bowel segment must weigh chronic inflammatory bowel disease against infectious enteritis, ischemia (where stratification also occurs but with a vascular territory and a different clinical tempo), and — critically — neoplasm, since a short, asymmetric segment with loss of stratification, homogeneous enhancement, and shouldered margins favors malignancy over benign inflammation. The dominant misses are the under-distended loop whose 'pseudothickening' masquerades as disease, the fibrotic stricture mistaken for active inflammation (or vice versa), and anchoring on known Crohn disease while overlooking a superimposed adenocarcinoma or lymphoma arising in chronically inflamed bowel.

Granulomatous inflammation is a distinct, $\text{T}_\text{H}1$-polarized variant of chronic inflammation in which macrophages, frustrated in their attempt to eliminate a persistent or poorly degradable agent, transform into epithelioid histiocytes — cells with abundant eosinophilic cytoplasm and interdigitating membranes that wall off the offending material. The reaction is cytokine-scripted: antigen-presenting cells secrete $\text{IL-12}$, $\text{T}_\text{H}1$ lymphocytes respond with $\text{IFN-}\gamma$ that maximally activates macrophages, and $\text{TNF}$ sustains the aggregate — which is why $\text{TNF}$ antagonists notoriously reactivate latent tuberculosis by dissolving granulomas. Epithelioid cells fuse into multinucleated Langhans giant cells, and a cuff of lymphocytes and a peripheral rim of fibroblasts complete the structure. The single most important CT-relevant branch point is whether the granuloma undergoes central caseous necrosis. Caseation — the cheese-like coagulative necrosis characteristic of Mycobacterium tuberculosis and many fungi — reflects a hypoxic, lipid-rich, acellular core, and it is the substrate for two cardinal CT findings. First, when this necrotic center liquefies and communicates with an airway or duct it cavitates, producing a gas-containing cavity whose wall thickness and nodularity inform the differential. Second, healed caseous foci calcify, depositing dystrophic calcium in the dead tissue to yield the dense ($>100$–several hundred $\text{HU}$) nodules, calcified lymph nodes, and Ghon/Ranke complexes that betray prior granulomatous infection. Non-caseating granulomas — the hallmark of sarcoidosis — by contrast remain solid and tend not to cavitate, though they too may calcify in chronic, fibrotic disease.

These cellular facts dictate organ-specific imaging phenotypes and the reasoning around them. Pulmonary tuberculosis is suspected when CT shows upper-lobe and superior-segment predominance, tree-in-bud opacities (the CT signature of endobronchial spread, caseous material impacting terminal and respiratory bronchioles), thick-walled cavities, and necrotic, rim-enhancing mediastinal nodes with low-attenuation centers; miliary disease produces innumerable $1$–$3\,\text{mm}$ randomly distributed nodules reflecting hematogenous dissemination. Sarcoidosis, by contrast, distributes its non-caseating granulomas along the lymphatics — a perilymphatic nodularity that clusters along bronchovascular bundles, interlobular septa, fissures, and the subpleural surface — accompanied by the classic symmetric bilateral hilar and right paratracheal lymphadenopathy that, when it calcifies in an 'eggshell' or amorphous pattern, mimics old granulomatous infection or silicosis. In the central nervous system a caseating tuberculoma appears as a ring-enhancing lesion whose necrotic core may show a 'target' configuration, an important and often-missed mimic of metastasis, pyogenic abscess, and primary neoplasm. The expert differential is explicitly Bayesian and epidemiologic — tuberculous lymphadenopathy and cavitation rise sharply in prior probability with endemic exposure or immunosuppression, whereas symmetric perilymphatic nodularity with bilateral hilar adenopathy in a young adult shifts the posterior toward sarcoidosis — and tissue confirmation is frequently required because the morphologic overlap among tuberculosis, fungal disease, sarcoidosis, berylliosis, granulomatosis with polyangiitis (which favors cavitating nodules), and necrotic malignancy is substantial. The characteristic misses are dismissing a calcified granuloma's neighboring soft-tissue activity, mistaking a solitary tuberculoma or a calcified granuloma for or against malignancy, and failing to recognize tree-in-bud as active, often transmissible, endobronchial disease.

The preceding mechanisms converge on a finite vocabulary of CT signs, and expert interpretation is the disciplined translation of each sign back to its tissue substrate, its temporal stage, and its differential weight. Fluid attenuation is the most quantitative starting point: a transudate (hydrostatic, protein-poor) sits near $0$–$15\,\text{HU}$, an inflammatory exudate at $15$–$30\,\text{HU}$, pus at $20$–$45\,\text{HU}$, and hemorrhage higher still ($>45\,\text{HU}$ acutely), so a single region-of-interest measurement materially reorders the differential before any morphologic reasoning. Enhancement adds the temporal dimension. Acute hyperemia produces early, avid mural enhancement; capillary leak produces delayed, persistent interstitial enhancement; an organized abscess shows a thick, intensely enhancing rim around a non-enhancing core; and fibrosis enhances late and progressively. Where available, CT perfusion quantifies these states — active inflammation and tumor neoangiogenesis both elevate $\text{BF}$, $\text{BV}$, and $\text{PS}$, which is simultaneously the parameter's power (grading activity, monitoring therapy) and its central pitfall, since inflammation and malignancy overlap substantially in every perfusion metric and neither morphology nor perfusion alone reliably separates a phlegmon from a necrotic tumor or an inflammatory from a neoplastic stricture.

The table below summarizes the principal inflammatory CT signs, their mechanistic basis, and their quantitative signature.

Beyond morphology, the radiologist must interrogate the failure modes that simulate or mask inflammation. Pseudoenhancement artificially raises the apparent attenuation of structures adjacent to densely enhancing vessels or organs, mimicking mural hyperemia; beam hardening and partial-volume averaging fabricate wall thickening and blur small cavities; suboptimal contrast timing flattens the very enhancement gradients on which the diagnosis depends; and an under-distended hollow viscus produces 'pseudothickening' indistinguishable, at a glance, from disease — the corrective being adequate luminal distension, correct phase selection, and thin-section multiplanar review. Cognitive failure modes are at least as consequential: satisfaction of search, in which one inflammatory focus terminates the hunt for a second; anchoring on a benign inflammatory label when a necrotic neoplasm presents identical rim enhancement; and confirmation bias in a patient with known inflammatory disease that obscures a superimposed malignancy or a fibrotic, surgical stricture. The prognostic and management payload is direct — a rim-enhancing drainable abscess versus a non-drainable phlegmon, an active enhancing Crohn segment versus a fixed fibrotic stricture, a cavitating tree-in-bud tuberculosis demanding isolation versus an indolent calcified granuloma, organizing pneumonia that will respond to corticosteroids versus progressive fibrosis — so that the accurate reading of an attenuation value, an enhancement pattern, and a distribution is not descriptive nicety but the determinant of intervention.