Novel MRI Technique Distinguishes Lung Cancer Types in Spinal Metastases Without Biopsy

When lung cancer spreads to the spine, the tumor type driving that spread matters enormously. Lung adenocarcinoma and lung squamous cell carcinoma are the two most common forms of non-small-cell lung cancer, and they respond to different treatments and carry different prognoses. Yet for many patients with spinal metastases, identifying which subtype is responsible requires a biopsy, an invasive procedure that can be difficult, risky, and sometimes impossible in fragile patients. A new prospective study published in BMC Medical Imaging suggests that a cutting-edge magnetic resonance imaging technique may be able to make that distinction noninvasively, simply by reading the chemical signature of the tumor tissue itself.

The research, led by Yuning Li, Dongyang Wu, and Xiaozhong Li of the Affiliated Hospital of Gansu University of Chinese Medicine in Lanzhou, China, focused on a method called amide proton transfer-weighted imaging, or APTw. Unlike conventional MRI, which primarily maps the abundance of water and fat in tissue, APTw detects exchangeable protons bound to amide groups within mobile proteins and peptides. Because malignant tissues typically overexpress intracellular proteins and peptides, APTw signal can serve as a sensitive proxy for cellular composition and metabolism. In effect, the technique offers a molecular contrast mechanism layered on top of the anatomical detail that MRI already provides.

To test whether this molecular contrast could separate the two lung cancer subtypes, the team enrolled 49 patients with spinal metastases from pathologically confirmed lung adenocarcinoma or lung squamous cell carcinoma between May 2025 and April 2026. Thirty-four patients had adenocarcinoma and fifteen had squamous cell carcinoma. Before any treatment began, every patient underwent conventional spinal MRI, APTw imaging, and diffusion-weighted imaging. The researchers then measured APTw values and apparent diffusion coefficient values, or ADC values, within the metastatic lesions themselves, comparing the two groups statistically and evaluating diagnostic performance with receiver operating characteristic analysis.

The results were striking. Spinal metastases from lung adenocarcinoma showed significantly higher APTw values than those from squamous cell carcinoma, averaging 5.26 percent versus 2.98 percent, a difference that was highly statistically significant. This gap reflects underlying biological differences between the two tumor types. Adenocarcinomas are gland-forming tumors rich in secretory proteins and mucin, which translate into a denser population of exchangeable amide protons and therefore a stronger APTw signal. Squamous cell carcinomas, by contrast, tend to be more keratinizing and less protein-dense in ways that produce a weaker signal on this imaging modality.

ADC values, which quantify how freely water molecules diffuse through tissue, also differed between the groups, though less dramatically. Adenocarcinoma metastases averaged 1.18 times ten to the minus three square millimeters per second, compared with 1.09 for squamous cell carcinoma, a smaller but statistically significant difference. ADC generally falls in tissues with high cellularity, where crowded cells restrict water movement, so the modest gap between the two subtypes suggests that cell density alone does not fully capture the biological distinctions that APTw reveals.

When the researchers evaluated how well each measure distinguished the two cancer types, APTw clearly outperformed ADC. The area under the receiver operating characteristic curve, a standard measure of diagnostic accuracy, reached 0.87 for APTw, while ADC achieved only 0.68. Combining both measures into a single model pushed the AUC to 0.89, but a DeLong test, which formally compares the accuracy of competing diagnostic approaches, showed that the combined model was not significantly better than APTw alone. In other words, adding diffusion information to the molecular imaging signal provided little incremental benefit in this cohort.

That finding carries practical weight. If APTw alone can achieve an AUC of 0.87, clinicians may be able to rely on a single advanced imaging sequence rather than a multi-parametric workup when trying to characterize a spinal lesion. The authors conclude that APTw may serve as a potential noninvasive imaging marker for differentiating lung adenocarcinoma from lung squamous cell carcinoma in patients with spinal metastases, while ADC provided limited additional diagnostic improvement beyond APTw in this cohort.

The clinical context makes this advance particularly meaningful. Spinal metastases are among the most common complications of advanced lung cancer, and they can cause severe pain, vertebral collapse, and compression of the spinal cord. Treatment decisions, from targeted therapies and immunotherapy to surgical decompression and radiation, often hinge on knowing the tumor’s histological subtype. For patients whose primary lung tumor was never biopsied, or whose metastatic lesion is in a location where needle access is dangerous, a noninvasive imaging marker could spare them an invasive procedure while still guiding therapy.

APTw itself is part of a broader movement in radiology toward molecular imaging, in which MRI sequences are tuned to detect specific chemical exchange processes rather than gross anatomical changes. The technique exploits the magnetization transfer between free water protons and the amide protons of endogenous mobile proteins, producing contrast that scales with protein concentration and pH. Because it requires no exogenous contrast agent, APTw avoids the risks associated with gadolinium injection and can be integrated into standard clinical MRI protocols with appropriate hardware and sequence tuning.

The study has limitations typical of an early prospective investigation. With 49 patients, the squamous cell carcinoma group was relatively small, and the findings will need validation in larger, multicenter cohorts before APTw can be adopted as a routine diagnostic tool. The authors also note that the combined APTw and ADC model did not significantly outperform APTw alone, suggesting that future work should focus on refining the APTw measurement itself rather than simply stacking additional imaging parameters. Still, the results mark an important step toward a future where the molecular identity of a cancer spread to bone can be read directly from an MRI scan, without a single needle touching the patient.

Subject of Research: Noninvasive MRI differentiation of lung adenocarcinoma and squamous cell carcinoma in spinal metastases

Article Title: Amide proton transfer-weighted imaging and apparent diffusion coefficient for differentiating lung adenocarcinoma from lung squamous cell carcinoma in patients with spinal metastases: a prospective study

Article References: Li, Y., Wu, D., & Li, X. (2026). Amide proton transfer-weighted imaging and apparent diffusion coefficient for differentiating lung adenocarcinoma from lung squamous cell carcinoma in patients with spinal metastases: a prospective study. BMC Medical Imaging. https://doi.org/10.1186/s12880-026-02760-y

Image Credits: AI Generated

DOI: 10.1186/s12880-026-02760-y

Keywords: amide proton transfer-weighted imaging, apparent diffusion coefficient, lung adenocarcinoma, lung squamous cell carcinoma, spinal metastases, molecular MRI, diffusion-weighted imaging, non-small-cell lung cancer, cancer imaging, diagnostic accuracy, receiver operating characteristic, noninvasive diagnosis