Chest x-ray eponyms in pulmonary embolism were never separate diseases
The classic chest radiograph signs of pulmonary embolism are still taught as if they sit in neat diagnostic boxes. They do not. Westermark, Hampton, Fleischner, Chang and Palla describe different radiographic expressions of the same pathophysiology: proximal arterial obstruction, distal vascular pruning, infarction and secondary atelectasis. Once those signs are read together, they form a coherent vascular story rather than a set of unrelated curiosities.
A central theme runs through the historical literature: the meanings of these eponyms changed over time. Later teaching compressed broad physiological observations into simplified labels, and in several cases the modern shorthand no longer matches the original description. Fleischner’s contribution, for example, was not a single static enlarged artery discovered in isolation. His work tied pulmonary arterial calibre to haemodynamic stress and serial change. Westermark’s original observation was not just “hyperlucency”; it was regional vascular depletion. Hampton’s lesion was not a tidy wedge in every case, but a pleural-based infarct with variable rounded and truncated forms.
The practical consequence is straightforward.
A chest radiograph can suggest embolism, but it cannot safely rule it in or out on its own.
Westermark sign reflected distal vascular depletion, not simple hyperlucency
Nils Westermark’s 1938 series remains one of the clearest early radiographic-pathological correlations in this field. He reviewed 26 patients with pulmonary embolism confirmed at autopsy and followed another 18 clinically during evolution and recovery. In embolism without infarction, the affected region could appear relatively more lucent because peripheral vascular markings were diminished or absent. Central vessels were still visible, sometimes terminating abruptly at the edge of the oligemic territory. The sign therefore represents a perfusion defect visible on plain film, a rare window into haemodynamics before the era of CT pulmonary angiography.
The modern habit of equating Westermark sign with hyperlucency alone misses the anatomy. Hyperlucency is a surface appearance; vascular depletion is the mechanism. The distinction matters because distal pruning, not brightness itself, is what makes the pattern biologically meaningful. It also explains why the sign is uncommon and unreliable as a standalone diagnostic tool: regional perfusion must fall enough to be seen, yet not be masked by infarction, effusion, or competing parenchymal change.
Hampton hump emerged from infarction morphology, not a single wedge template
Aubrey Hampton and Benjamin Castleman published their autopsy-radiograph correlation in 1940 and placed pulmonary infarction on much firmer radiologic ground. Their work described pleural-based infarcts with a convex inward margin, often hemispherical or truncated-cone in shape. The familiar “hump” is the contour of that lesion rather than a fixed triangular wedge. This is more than semantic housekeeping. Infarction in pulmonary embolism evolves from pleural-based ischemic injury, haemorrhage and tissue necrosis, so the shadow can vary depending on timing, extent and surrounding collapse.
One sentence captures the bedside relevance: the opacity is not the clot, but the lung response to the clot.
Hampton and Castleman also helped establish why these findings often overlap. Infarction can trigger adjacent atelectasis and pleural reaction, while the embolic territory itself may show mixed density patterns rather than a single canonical silhouette. That is why Hampton hump remains a useful clue but a poor gatekeeper. It is a sign of parenchymal consequence, not a direct map of arterial occlusion.
Fleischner sign and Fleischner lines describe different lung responses
Felix George Fleischner’s work straddled two different manifestations that later became entangled in teaching. On one side was proximal or central pulmonary arterial enlargement, the modern Fleischner sign. On the other was the linear shadowing later called Fleischner lines, which Baron clarified as an indirect, non-specific manifestation caused by focal bronchial obstruction and atelectasis rather than direct visualization of vascular blockage. Conflating the two blurs a crucial physiological difference: one reflects altered arterial calibre under pressure, the other reflects secondary collapse and healing.
Fleischner’s serial observations showed dynamic changes in arterial size, including a progression from 11 to 14 to 22 and then 15 mm in one described case. This is exactly the kind of haemodynamic behaviour one would expect in acute obstructive pulmonary vascular disease, where pressure loading and flow redistribution can enlarge the proximal artery before later partial normalization. The finding is elegant, but it is not a standalone diagnostic test.
Chang and Palla were related, but they were not the same sign
Chang’s 1962 work established normal measurements for the right descending pulmonary artery, creating a baseline for later interpretation. In 1965, Chang and Davis reported enlargement of that vessel in 23 consecutive patients with pulmonary infarction, with diameters ranging from 17 to 22 mm in 19 right-sided infarcts and 17 to 26 mm in four left-sided infarcts. The key point is transient calibre enlargement. Chang sign is not vessel amputation, and it is not defined by an abrupt cutoff.
Palla’s 1983 analysis used quantitative comparison in 374 consecutive patients with suspected embolism and focused on the same right descending pulmonary artery, but with a different morphologic emphasis: enlargement, density and a characteristic sausage-like tapering. Later teaching blurred Chang and Palla into near-synonyms, yet they emerged from distinct observations. Chang described measurement and transient enlargement in infarction. Palla described a distorted vessel contour with abrupt distal tapering in acute embolic disease.
Prospective studies showed why CXR clues lost ground in the CTPA era
The prospective literature was blunt about the limits of these signs. Greenspan’s 1982 blinded-reader study of 152 patients suspected of PE found poor overall diagnostic accuracy for chest radiography, with no meaningful relationship between reader experience and performance. The PIOPED data later reinforced the same message. In 1,063 patients with suspected pulmonary embolism, 383 had angiographically confirmed PE and 680 did not, yet Westermark sign, prominent central pulmonary artery and Hampton hump all performed poorly as independent predictors.
This is the point at which chest radiography changed role. It stopped being a primary diagnostic instrument for embolism and became a triage tool for alternatives, complications and occasional clues. A radiograph may still show regional oligemia, pleural-based infarction, proximal arterial enlargement or plate-like atelectasis, but modern practice asks a different question: does the film justify urgent definitive imaging, and what else could explain the patient’s symptoms?
Historical data from the International Cooperative Pulmonary Embolism Registry also showed that pulmonary artery enlargement was reported in only a minority of cases, while cardiomegaly and pleural abnormalities were more common. Even then, the radiographic signal was incomplete. The broader lesson is anatomical rather than nostalgic. These eponyms are still useful shorthand, but they are best understood as related consequences of vascular obstruction, not as isolated diagnostic trophies.
Historical names are still useful, but only if the physiology stays intact
The original papers show a tighter relationship among these radiographic signs than modern teaching usually allows. Westermark’s depletion, Hampton’s infarct morphology, Fleischner’s haemodynamic enlargement, Chang’s calibre measurement and Palla’s sausage-shaped right descending artery all sit on the same pathophysiological chain. Once that chain is remembered, the eponyms regain value without being overinterpreted.
The unresolved problem is not whether the names are familiar. It is whether clinicians remember what they actually described when the film is sitting on the screen in front of them.