Volume doubling time (VDT) as a quantitative malignancy predictor — ризик злоякісності вогнища легені за швидкістю росту на серійній КТ
Radiologists have long compared nodule diameters between chest CTs, but diameter is a deceptively crude yardstick. A nodule growing from 8 mm to 10 mm looks like a modest 25% increase in diameter — yet because volume scales with the cube of diameter, that same nodule has nearly doubled in volume (≈95% increase). Volume doubling time (VDT) converts this cubic relationship into a single, clinically interpretable number: the number of days it takes the nodule’s volume to double, assuming exponential (constant relative rate) growth.
The calculation behind VDT:
1. Measure the nodule on scan 1 (diameter d1, or volume V1 from semi-automated volumetric segmentation) 2. Measure the same nodule on scan 2, taken t days later (diameter d2, volume V2) 3. Assuming exponential growth, V(t) = V1 · 2^(t / VDT) 4. Solve for VDT: VDT = t · ln(2) / ln(V2 / V1)
Why volume, not diameter, is preferred: • A sphere’s volume is (4/3)πr³ — proportional to the cube of the radius (or diameter) • A diameter increase from 8→10 mm (+25%) corresponds to a volume increase of (10/8)³ ≈ 1.95×, i.e. almost a full doubling • Diameter-based "growth" thresholds (commonly ≥2 mm change) can miss early, biologically important growth in small nodules, or over-call growth that is really measurement noise • Modern CT workstations increasingly report semi-automated volumetric measurements directly, reducing the need to manually cube diameters
Practical shorthand many radiologists use at the reading station: • A nodule growing detectably in under 1 year (~365 days) is watched closely • A nodule that shows literally zero interval change over ≥2 years is traditionally considered a reasonable (though not absolute) indicator of benignity • Between those extremes, VDT is placed on a risk continuum rather than treated as a hard yes/no cutoff
Malignant cells tend to divide faster and less predictably than the tissue around them, and this shows up directly in serial imaging as a short volume doubling time. A nodule that doubles in volume in a matter of weeks to a few hundred days is behaving like a proliferating tumor, not a scar or a resolving infection — and that growth pattern alone can push the malignancy probability high enough to justify tissue sampling even when the nodule’s size or shape looked unremarkable at baseline.
Rapid growth is a strong but not exclusive signal:
• Primary lung cancers (especially adenocarcinoma, squamous cell carcinoma) commonly show VDT in the range of roughly 20–400 days • Very short VDT (under ~20–30 days) is unusual even for aggressive cancer and should raise consideration of an infectious or inflammatory process (rapidly resolving/evolving pneumonia, organizing pneumonia, abscess) rather than assuming malignancy by default • Metastatic deposits and some high-grade tumors can double even faster than typical primary lung cancers • A short VDT measured on only two data points can be misleading if one measurement is corrupted by an unrelated process (adjacent atelectasis, overlapping vessel) — correlate with the images directly, not just the numbers
Why rapid growth escalates management: • Elevated pretest probability of malignancy changes the risk/benefit calculus of an invasive test • A PET/CT adds functional information (FDG avidity) that correlates with metabolic activity and can support the growth-rate signal • Tissue diagnosis (CT-guided percutaneous biopsy, bronchoscopic biopsy, or navigational bronchoscopy) is typically pursued rather than extending surveillance further, since further watchful waiting on a rapidly growing lesion risks disease progression • Multidisciplinary discussion (radiology, pulmonology, thoracic surgery) is standard once growth rate alone crosses into the concerning range
A nodule that is unchanged, or grows only glacially, over an extended observation window is one of the most reassuring findings in nodule management — long VDTs (well beyond 600 days) or true two-year stability are classically associated with benign etiologies such as granulomas or intrapulmonary lymph nodes. But this reassurance has a well-documented exception: certain ground-glass and part-solid nodules on the adenocarcinoma spectrum can grow extremely slowly — sometimes over many years — while still representing a genuine, if indolent, malignant process.
Two things are true at once:
• For solid nodules, sustained size stability over two full years has long been used as a practical (if imperfect) marker that further surveillance can stop — most benign solid nodules (granulomas, intrapulmonary lymph nodes, scars) simply do not grow • For subsolid nodules — pure ground-glass nodules and part-solid nodules — the adenocarcinoma spectrum (atypical adenomatous hyperplasia → adenocarcinoma in situ → minimally invasive adenocarcinoma → invasive adenocarcinoma) can progress extremely slowly, with reported doubling times often exceeding 1000 days, sometimes several years • A subsolid nodule can therefore appear "stable" across a standard follow-up window and still be a slowly progressing malignancy; the appearance of a new solid component, rather than a change in overall size, is often the more sensitive warning sign for these lesions • Because of this, current guidance (e.g. Fleischner Society recommendations) generally favors longer total surveillance duration for persistent subsolid nodules than for solid nodules before growth can be confidently excluded
Practical implication: • Slow/stable growth still lowers malignancy probability substantially and is the basis for extending or ending surveillance in most cases • Nodule subtype (solid vs. part-solid vs. pure ground-glass) should always be factored in alongside the raw growth-rate number before declaring a subsolid lesion "reassuringly stable"
VDT is only as reliable as the two measurements that feed it, and CT nodule measurement carries real, well-quantified variability. Manual diameter measurements by different readers — or even the same reader on different days — routinely differ by 1–2 mm. Because volume scales with the cube of diameter, that small, ordinary measurement noise can masquerade as meaningful growth or falsely suggest stability, especially for nodules under about 10 mm where a 1 mm difference is a much larger fraction of the total size.
Where measurement variability comes from:
• Manual caliper placement: subjective choice of nodule margin, especially for irregular or spiculated borders and for subsolid nodules with indistinct edges • Reconstruction and slice thickness differences between scans: thicker slices blur small nodules and bias size estimates • Window/level settings: lung windows displayed differently between reads can shift perceived margins • Patient positioning and respiratory phase: partial volume effects change apparent nodule size scan to scan • Semi-automated volumetric segmentation reduces but does not eliminate this variability, and still requires a technically adequate, motion-free scan
How radiology practice minimizes false growth/false stability calls: • Use consistent scan protocol (same slice thickness, reconstruction kernel) across follow-up studies whenever possible • Prefer volumetric or semi-automated measurement over manual single-diameter calipers, particularly near decision thresholds • Require growth to exceed a minimum meaningful change (not just any registered difference) before labeling a nodule as "growing" — small apparent changes within known measurement error are not treated as true growth • When growth is borderline, a short-interval confirmatory follow-up scan (rather than immediate escalation) is often used to see whether the trend is reproducible • Side-by-side comparison with prior images (not just the numeric report) remains an essential sanity check before trusting a VDT calculation
Growth rate does not act alone — it is layered onto baseline nodule characteristics (size, solid vs. subsolid composition, margin, location, patient risk factors such as smoking history) to arrive at an overall malignancy probability. That combined estimate then drives one of three practical pathways: keep watching at the same interval, extend the interval because risk has fallen, or escalate straight to diagnostic workup because risk has risen enough that further watching is not the safer option.
How growth rate is folded into the final decision:
• Growth rate is a dynamic risk factor layered on top of static baseline features (size at detection, solid/part-solid/ground-glass composition, spiculation, upper-lobe location, patient age and smoking pack-years) • A nodule with otherwise low-risk baseline features that shows convincing rapid growth is still escalated — growth rate can override an initially reassuring baseline impression • Conversely, a nodule with some concerning baseline features (e.g. spiculated margin) that remains genuinely stable over a sufficiently long observation period gains substantial reassurance from that stability • Structured frameworks (Fleischner Society follow-up recommendations, Lung-RADS categories) build interval-based surveillance schedules precisely so that growth can be assessed with adequate, standardized time between scans rather than reacting to noisy short-interval comparisons • The observation period itself matters: a "stable" reading after only 1–2 months carries far less weight than the same reading after 12–24 months, which is why measurement confidence and total elapsed surveillance time are tracked alongside the raw growth-rate number
The three practical endpoints: 1. Diagnostic workup (biopsy, PET/CT, or direct surgical consultation) — triggered by rapid growth, or growth rate combined with high baseline risk 2. Continue surveillance at the current interval — growth rate is indeterminate/moderate, or the observation period so far is too short to be confident either way 3. Extend the surveillance interval, or stop surveillance — growth rate is reassuringly slow/absent over a sufficiently long, technically reliable observation period
No single growth-rate number is diagnostic on its own. VDT is a probabilistic input, refined by nodule subtype, baseline morphology, patient risk factors, and — critically — how much confidence the measurement technique and elapsed time actually support. The output is not a verdict but a next step: watch, watch longer, or find out for certain.