Max Planck Institute spots spectral fingerprints of cancer

15 Apr 2025

Laser-based molecular spectroscopy could assist diagnosis under real-world conditions.

A project group including Max Planck Institute of Quantum Optics (MPQ) and Ludwig Maximilian University of Munich (LMU) has developed an electric-field molecular fingerprinting method to help spot evidence of cancer.

Published in ACS Central Science, the technique offers a robust technological framework for disease phenotyping under real-world conditions, according to the team.

"Sensitive and specific analytical methods have led to the discovery of numerous molecular biomarker candidates," noted the project in its ACS paper.

"However, current techniques are often still limited in the range of molecular species that they can probe at once."

The team's solution involves infrared molecular fingerprinting, whereby phenotype detection is based on patterns of change "across the entire molecular landscape."

Fourier transform infrared spectroscopy has been regularly deployed as a means of detecting specific molecular fingerprints in the blood, but FTIR's sensitivity can be limited by strong background signals from thermal excitation.

Electric-field molecular fingerprinting (EMF) can overcome this limitation, exciting a sample with an ultrashort IR pulse lasting tens of femtoseconds rather than continuous irradiation. This makes the molecules emit their subsequent resonant vibrational response for periods extending over hundreds of femtoseconds or several picoseconds.

The responses can then be time-resolved into individual infrared electric-field molecular fingerprints, as a route to profiling complex molecular mixtures.

The new MPQ project is "the first proof-of-concept biomedical application of EMF, demonstrating that the measured fingerprint patterns robustly acquire disease-specific information from liquid blood plasma," noted the team.

Translation into clinical practice

In trials the project applied its EMF method to blood plasma from 2500 participants, including people with lung, prostate, breast or bladder cancer and those without cancer.

An initial training stage used responses from individuals with and without cancer to teach a machine learning model to identify molecular signatures associated with the four types of cancer. This exploits the way that different cancers have different effects on the blood stream: lung tumors may release more metabolic and cellular products into the bloodstream, for example, given the closer proximity and exchange with the circulatory system.

After this training the technique was tasked with detecting lung cancer-specific infrared signatures and differentiating them from control samples obtained from individuals without cancer, doing so with an accuracy of up to 81 percent. Performance for the other three cancers was not as good, but the team aims to improve the approach and expand it to test additional cancer types and other health conditions.

Although the current EMF experiment covered only a small fraction of the potential molecular fingerprinting spectral region, the stability and reproducibility of the results over the trial's extended experimental period is encouraging for ultimate clinical viability, said the team.

"Laser-based infrared molecular fingerprinting detects cancer, demonstrating its potential for clinical diagnostics," commented Mihaela Žigman from MPQ. "With further technological developments and independent validation in sufficiently powered clinical studies, it could establish generalizable applications and translate into clinical practice, advancing the way we diagnose and screen for cancer today."