Measuring changes in breast tumour tissue from imaging data acquired during chemotherapy infusion is now one step closer to clinical implementation.

A team of engineers and physicians has developed and validated a portable, nine-wavelength near-infrared spectral tomography (NIRST) system that quantifies changes in total haemoglobin (HbT), oxygen saturation, lipid-content scattering amplitude, and scattering properties in the breast (Biomed. Opt. Express 7 2186).

The ability to quantify changes in breast cancer during treatment could help determine whether a specific chemotherapy regimen is effective.

This can be performed using MRI and 18FDG PET – but use of these advanced imaging technologies is often clinically impractical, time consuming, expensive and requires contrast agent injection.

NIRST, on the other hand, is an emerging functional technology that's non-invasive, portable and much less costly.

NIRST has the ability to estimate the intrinsic biophysical composition of tissue, and pilot studies have shown that NIRST can yield prognostic information when used to monitor tumour response during treatment.

Principal investigator Shudong Jiang, associate professor of engineering at the Thayer School of Engineering of Dartmouth College, and colleagues designed a prototype system that integrates frequency-domain (FD) and continuous-wave (CW) data acquisition.

Simultaneous acquisition of FD and CW channels using a photomultiplier and photodiode detection module makes it feasible to dynamically acquire data while a patient undergoes chemotherapy infusion.

Measurements from three FD and six CW channels, for all nine wavelengths spanning a range from 661 to 1064 nm, took just three minutes.

NIRST system design

The NIRST system incorporates FD and CW source modules consisting of laser diodes channelled into two single-source fibres.

Fifteen pairs of photomultiplier and photodiode detectors and one pair of alignment lenses were mounted on a custom programmable rotary switch.

The 16 bifurcated optical bundles, mounted on another plate fixed on top of the rotating circular plate, enable simultaneous acquisition of both FD and CW data.

The other ends of bundles are attached to the breast using one of two adjustable interfaces, one with a flattened curve design and the other with a deeper curve.

After the interfaces are positioned on a breast, optical measurement data are transferred through the fibre-optic interface to the portable NIRST system as a patient sits in an infusion chair.

The researchers used the system to image phantoms, 10 healthy volunteers and one breast cancer patient.

Initial findings

The researchers validated the FD and CW data acquisition using gelatin phantoms.

They were able to recover HbT values that were slightly lower than the actual value, as well as background values for oxygen saturation (StO2), water and lipid using both breast interfaces.

Recovered lipid content of the 10 healthy volunteers varied substantially.

Significantly higher HbT and water were estimated in high-density breast tissue than in low-density tissue.

The recovered physiologically relevant values all fell into reasonable ranges when compared with previous publications.

To assess the system's ability to accurately monitor patient response in an infusion setting, the researchers investigated temporal variations due to patient breathing patterns and/or movement.

Eight continuous 30 minute measurements exhibited less than 5% standard deviation, suggesting that the data were stable.

No significant differences were identified between the left and right sides of the breast for all optical parameters.

This finding suggests that a patient's contralateral breast could be imaged before treatment of the diseased breast to highlight the tumour/background contrast relative to surrounding tissue over the course of therapy.

Relative HbT, an indicator of tissue malignancy, is widely used to assess changes in tumour physiology during neoadjuvant chemotherapy.

In view of the fact that hypoxic tumours have been found to be more resistant to chemotherapy, StO2 may also be an important response predictor.

The ability to compare data for water, lipid and scatter components could also contribute to identifying tumour changes and response.

The researchers also validated system performance with the sole breast cancer participant.

The recovered tumour position corresponded with that highlighted in MR images.

The tumour tissue was differentiated from healthy tissue by significant increases in HbT (a contrast ratio of 1.4) and water (a contrast ratio of 1.2), and decreases in the lipid region.

These findings are consistent with reported results for malignant tumours.

Future research

Jiang told medicalphysicsweb that the team is now working to improve the system design and performance.

The system currently sits in a cart, but they hope to make it more compact.

A study involving a cohort of breast cancer patients is ongoing.

"We are continuing to validate that this system can quantify more accurately changes in the breast. We also want to find out whether the sub-type of breast cancers will have different response during neoadjuvant chemotherapy," Jiang said. "We currently do not have a large enough sample size of patients to determine if sub-types of breast cancers will have different responses; this is one of our future research directions."

Jiang noted that development of a commercial system for use in the clinical oncology infusion suite is a hope, but that much more work on prototype systems needs to be done.

"We want to continue to demonstrate the potential of efficient characterisation of breast cancer by a portable NIRST system," she concluded.

Source: Medical Physics Web