Interstitial Lung Disease Patterns
- Interstitial lung disease (ILD) is scarring or inflammation of the lung's "scaffolding" — the connective-tissue mesh between the air sacs, not the air sacs themselves.
- The chest X-ray usually whispers (faint, hard-to-pin-down haze and small lines); high-resolution CT (HRCT) is where ILD finally speaks up.
- On CT you're sorting a handful of patterns: ground-glass (hazy), reticular (fine lines), honeycombing (clustered cysts), and nodules.
- Honeycombing plus a lower-zone, peripheral, bottom-up gradient points toward the fibrotic end of the spectrum — and it's the one finding that changes everything.
- Where the disease lives matters as much as what it looks like: upper vs lower, central vs peripheral. Geography is half the diagnosis.
Picture the lung as a sponge. Most chest disease soaks the holes of the sponge — pus in pneumonia, water in pulmonary edema. Interstitial lung disease is different: it goes after the rubber of the sponge itself. The walls between the holes get thickened, inflamed, or scarred stiff. The air spaces are innocent bystanders; the scaffolding is the crime scene.
"Interstitium" is just the radiologist's fancy word for that scaffolding — the thin connective-tissue mesh that wraps every air sac and threads alongside every vessel. When it thickens, the lung loses its stretch, like a sponge slowly turning into a kitchen scrub pad.
Why the X-ray is so coy
Here's the frustrating part: on a plain chest radiograph, ILD is a master of mumbling. Early on it can look completely normal. As it progresses you get vague haziness and faint web-like lines, usually worst at the lung bases, but it's the kind of finding where three radiologists squint and two of them say "maybe?" That ambiguity isn't you being bad at this — it's genuinely hard, because the changes are small and the X-ray flattens everything into one shadow.
So the moment ILD is suspected, the workhorse becomes high-resolution CT (HRCT) — thin slices that resolve the lung's fine architecture. Suddenly the mumble becomes a sentence. (If the CT itself feels unfamiliar, the approach to chest CT page is a good warm-up.)
The four patterns you're sorting into
Almost all of ILD reading is pattern recognition. You're putting what you see into one of a few buckets:
| Pattern | What it looks like | The tangible version |
|---|---|---|
| Ground-glass | Hazy gray that you can still see vessels through | Breathing on a cold window — fog, not paint |
| Reticular | A fine net of crisscrossing lines | A fishnet stocking laid over the lung |
| Honeycombing | Clustered small cysts stacked in rows | Bubble wrap, or an actual honeycomb |
| Nodular | Tiny dots scattered through the lung | Someone flicked a wet paintbrush at it |
The key trick with ground-glass: vessels stay visible through the haze. That's how you tell it from true consolidation, where the gray is dense enough to swallow the vessels whole — the same logic as the air–soft tissue contrast you use everywhere else.
Ground-glass is the wishy-washy one of the bunch. It can mean active inflammation (potentially reversible) or very fine fibrosis (not reversible). On its own it doesn't tell you which — you read it alongside the other findings and the clinical story.
Honeycombing: the finding that changes the conversation
If one pattern earns a spotlight, it's honeycombing — clusters of small cysts sharing walls, stacked in layers like the namesake. It's the radiographic signature of established, end-stage fibrosis. The lung has been remodeled so thoroughly that little air pockets form where alveoli used to be.
It often travels with traction bronchiectasis, where scar tissue contracts and yanks the airways permanently open, like guy-wires pulling a tent pole crooked. Together, honeycombing and traction bronchiectasis say this is scar, and scar doesn't go back.
Honeycombing that is peripheral (hugging the pleura) and worst at the lung bases is the textbook fibrotic, bottom-up pattern. Recognizing that specific geography is one of the highest-yield skills in chest imaging.
Geography: where the disease lives
The same pattern means different things depending on where it sits. Two axes do most of the work:
- Up vs down. Some processes prefer the upper zones, others gravitate to the bases. A bottom-heavy, peripheral, fibrotic pattern is a very different beast from an upper-zone-predominant one.
- Center vs edge. Disease hugging the pleural surface points one way; disease clustered around the central airways and vessels points another.
You don't need to memorize a giant table of causes today. You need to describe the case in those terms — pattern, zone, and central-versus-peripheral. That description is what narrows fourteen possible causes down to two or three.
Dependent atelectasis — the lung gently sagging and compressing against the table while the patient lies still — can fake basal ground-glass or fine lines. The tell: it sits in the most dependent part of the lung and tends to vanish on prone imaging, whereas true ILD stays put. Always ask whether the "finding" is just gravity.
Putting it together
ILD isn't one disease — it's a whole family that happens to share the same neighborhood (the interstitium) and the same handful of CT patterns. Your job at this level isn't to name the exact entity from the scanner; it's to recognize that the scaffolding, not the air spaces, is the problem, and to describe the case cleanly: which pattern, which zone, central or peripheral, and is there honeycombing.
Nail that description and you've done the genuinely hard part. Matching it to a specific diagnosis is usually a team sport — radiologist, pulmonologist, and pathologist in the same room — which is exactly the journey the pattern-based ILD deep dive picks up.
If you remember one thing: in ILD, the air sacs are bystanders. Hunt the scaffolding.