Wedge-Shaped Lung Scars Offer Clue to Hidden Clots in Cancer Scans

When a cancer patient arrives at the hospital struggling to breathe, one of the most urgent questions facing clinicians is whether a pulmonary embolism — a blood clot lodged in the arteries of the lungs — is to blame. For oncology patients, that question is surprisingly hard to answer, because their lungs are rarely radiologically quiet. A new study published in Frontiers in Radiology has now mapped, in unusual detail, the tangle of lung and pleural abnormalities that show up on CT pulmonary angiography (CTPA) scans in cancer patients, and the results reveal both a tantalizing clue and a sobering caution about how much these scans can really tell us.

The research team, led by investigators at Al-Quds University in Jerusalem together with colleagues at St. Vincent’s University Hospital in Dublin, conducted a retrospective observational study of 113 consecutive cancer patients who underwent CTPA for suspected acute pulmonary embolism between January 2022 and January 2026. Their goal was straightforward but clinically important: to characterize the pleural and lung parenchymal abnormalities visible on these scans and determine which, if any, actually distinguish patients with confirmed embolism from those without. Pulmonary embolism remains one of the top three causes of cardiovascular death worldwide, and malignancy is among the most potent risk factors for developing one, so the stakes of accurate interpretation in this population are high.

What the team found first was sheer complexity. Pleural effusion — fluid around the lungs — was the single most common abnormality, present in 58.4 percent of all patients. Pulmonary nodules or masses appeared in 46.0 percent, atelectatic or collapse changes in 45.1 percent, consolidation in 38.1 percent, and ground-glass opacity in 30.1 percent. Interstitial or septal changes, infectious or inflammatory patterns, emphysema, bronchiectatic tree-in-bud changes, cavitary lesions, and infarct-like wedge-shaped opacities rounded out a long list of coexisting findings. In other words, the thorax of a cancer patient undergoing CTPA is almost never a clean canvas: it is a layered superimposition of tumor effects, treatment-related lung injury, infection, and chronic structural change, all of which can mimic or obscure the signature of an embolus.

Embolism itself was confirmed in 32 of the 113 patients, or 28.3 percent — a substantial yield that underscores the genuine thrombotic burden carried by oncology populations. The prevalence was highest among patients with gastrointestinal, gynecologic, and thoracic malignancies. The median age of the cohort was 59 years, and 42.5 percent were female. Notably, when the researchers examined where the clots actually sat, more than 71 percent of PE-positive patients had bilateral emboli, distributed across central, segmental, and mixed peripheral vascular territories. In this cohort, embolism was frequently not a small, isolated finding but an anatomically extensive thromboembolic process, a pattern that carries implications for how thoroughly radiologists must scrutinize the entire pulmonary arterial tree.

Against that noisy background, one finding stood out. Infarct-like wedge-shaped opacity — a peripheral, cone-shaped density classically associated with pulmonary infarction downstream of an occluded vessel — appeared in 12.5 percent of PE-positive patients compared with just 1.2 percent of those without embolism. That corresponds to an unadjusted odds ratio of 11.43, and in a deliberately parsimonious multivariable model restricted to three imaging predictors, the association held with an adjusted odds ratio of 11.72. From a pathophysiologic standpoint, the signal is biologically plausible: when a clot blocks a segmental pulmonary artery, the distal lung tissue can infarct, producing exactly this focal, peripheral wedge of damaged parenchyma on the scan.

But here is where the study’s statistical rigor matters. The researchers applied Benjamini–Hochberg false-discovery-rate correction across twelve imaging comparisons, and none of the associations — including the wedge-shaped opacity signal — survived that correction (q = 0.133 for the wedge finding). The confidence intervals were strikingly wide, stretching from 1.23 to 106.62, because only five patients in the entire cohort had this finding. The authors are appropriately candid: these results should be considered hypothesis-generating rather than confirmatory, and the wedge-shaped opacity should be viewed as a clue that heightens suspicion of embolic infarction when it appears, not as a standalone diagnostic predictor.

Equally interesting was what the data suggested in the opposite direction. Ground-glass opacity and interstitial or septal change were both less frequent in PE-positive patients — appearing in 15.6 percent versus 35.8 percent, and 3.1 percent versus 25.9 percent, respectively. The explanation likely lies in radiologic specificity. Acute pulmonary embolism produces focal, peripheral change reflecting segmental vascular occlusion, whereas ground-glass and interstitial abnormalities reflect diffuse processes affecting the alveolar-interstitial compartment: infection, pulmonary edema, lymphangitic tumor spread, drug-related pneumonitis, or radiation injury. Their relative predominance in PE-negative examinations suggests they should prompt careful consideration of non-embolic explanations for a patient’s breathlessness — though the authors caution that this inverse pattern may partly reflect referral behavior, and embolism must still be assessed on direct vascular findings regardless of the surrounding parenchyma.

Methodologically, the study has notable strengths. The team used consecutive sampling to minimize selection bias, extracted data from formal radiology reports produced by a consultant thoracic radiologist who had no knowledge of the research question, classified imaging terminology according to the Fleischner Society glossary, and followed STROBE reporting guidelines. The exploratory regression model was deliberately limited to three predictors to avoid overfitting given only 32 PE events, and it was internally validated with 2,000 bootstrap resamples, yielding an optimism-adjusted AUROC of 0.699 — only marginally below the apparent value of 0.707 — with acceptable calibration. Still, the authors acknowledge real limitations: a single-center retrospective design, modest event counts, missing data on chemotherapy, radiotherapy, anticoagulation, and infection, and the inability to definitively attribute individual findings to embolism rather than overlapping disease processes.

The practical message for clinicians is a dual-purpose reading of CTPA in oncology patients: first as a vascular study hunting for intravascular filling defects, and second as a comprehensive thoracic examination for alternative and coexisting causes of symptoms. The authors advocate a structured reporting approach that clearly separates direct vascular evidence of embolism, parenchymal findings plausibly related to embolic injury, and ancillary abnormalities pointing toward non-embolic disease. That kind of clarity, they argue, could reduce the over-attribution of nonspecific lung findings to embolism and improve communication between radiologists and treating oncologists.

Looking ahead, the team calls for larger, multicenter studies with prospective designs, blinded radiologic re-review, formal interobserver agreement analysis, and standardized severity scoring to confirm whether infarct-like wedge-shaped opacity truly is the most embolism-suggestive pattern in cancer patients. They also propose examining how cancer subtype, metastatic burden, recent chemotherapy or radiotherapy, and infectious status modify these imaging relationships — and correlating CT patterns with laboratory data, clot burden, and clinical outcomes to determine whether these findings are prognostically as well as diagnostically relevant. For now, the study delivers a nuanced verdict: in the imaging-dense world of cancer care, lung findings abound, most are nonspecific, and only careful, contextual interpretation — anchored in the direct visualization of the pulmonary arteries — can reliably separate clot from the many other culprits competing for attention on the scan.

Subject of Research: Pleuroparenchymal CT findings and their association with pulmonary embolism in cancer patients

Article Title: Pleuroparenchymal abnormalities in cancer patients undergoing CTPA for suspected pulmonary embolism

Article References: Allahham, M. F., Mustafa, H., Faroun, J., Amro, A. M., Allahham, F. M., Shrateh, O. N., & Murphy, D. (2026). Pleuroparenchymal abnormalities in cancer patients undergoing CTPA for suspected pulmonary embolism. Frontiers in Radiology, 6, Article 1901919. https://doi.org/10.3389/fradi.2026.1901919

Image Credits: AI Generated

DOI: 10.3389/fradi.2026.1901919

Keywords: pulmonary embolism, cancer, CT pulmonary angiography, radiology, pleuroparenchymal abnormalities, pulmonary infarction, ground-glass opacity, venous thromboembolism, oncology imaging, thoracic imaging, pleural effusion, diagnostic imaging