lymph-node test shows clinical potential
07 feb 2011
a revolutionary method for identifying breast cancer in lymph nodes using raman spectroscopy is being tested by researchers and surgeons at a uk hospital.
surgical management of breast cancer has come a long way since the days of halsted’s radical mastectomy. but even so, for some patients two or more operations are still required before a surgeon can be assured of removing all cancerous tissue.
the first operation removes the visible breast cancer and samples the axillary lymph nodes for signs of any spread of the disease. if post-operative tests confirm that spread, then the patient must undergo an additional procedure to remove all the nodes in the affected area. this is not just a very real cause of psychological stress and anxiety; it also requires the patient to undergo the risks of additional surgery and general anaesthesia.
a preferable approach - not just for the patient, but also the surgeon and the health care system as a whole - would be some form of intra-operative assessment, allowing the immediate excision of all of the axillary lymph nodes, if required.
raman spectroscopy may offer a solution: the light-scattering analytical technique can detect differences in tissue composition, and its efficacy has already been demonstrated in a wide range of cancer pathologies.
in addition, research using raman spectroscopy and principal component analysis （pca） to assess axillary lymph nodes has now shown that the technique can match the sensitivities and specificities of other techniques that are used currently. however, the mapping techniques required for analysis took many hours to produce for each lymph node tested - clearly inappropriate for intra-operative use.
“for ‘one-stop’ axillary surgery to become a reality for all patients, a technique that analyzes the node quickly, accuray, simply and economically needs to be developed,” said jonathan horsnell, a researcher at the uk’s gloucestershire royal hospital. “the classical histopathological techniques of frozen section analysis and touch imprint cytology have been used in this context, but the poor sensitivity and the need for an experienced histopathologist have prevented their universal availability.”
raman spectrometer
probe-based spectrometer
it was with this in mind that horsnell and colleagues analyzed the existing data further, to mimic the use of a probe-based raman spectrometer by selecting equally spread points across the maps. the results suggested the potential for using a raman probe, with specificities remaining above 90%. although the sensitivity fell as smaller metastatic deposits were missed, it remained at a level equivalent to that reported for frozen section and touch imprint cytology.
and in comparison to a microscope, the larger spot size of a probe has the potential to overcome this reduced ability to detect small metastatic deposits, by collecting light scattered from a greater volume of tissue per spectrum. after an extensive review of compact fiber-optic raman spectrometers, the gloucestershire royal hospital team chose to use the miniram ii manufactured by the us company b&w tek to demonstrate this effect, largely because it was the only system in its class to use both a thermoelectrically cooled ccd array and a volume bragg grating （vbg） stabilized laser source.
this spectrometer can collect scattered light from depths of up to 4 mm with an acquisition time of 16 seconds per spectra. thus, by sampling a far greater volume of tissue per spectrum （rather than collecting many hundreds of spectra from small points across a tissue field）, the researchers have been able to reduce the time required for tissue assessment to a level that would be acceptable in a clinical scenario.
clinical efficacy
their work, recently published in the journal analyst, demonstrated the clinical efficacy of this volume sampling technique. in the study, they assessed 38 lymph nodes from 20 patients undergoing surgery for newly diagnosed breast cancer using the miniram ii, and compared the results with a standard histopathological assessment of each node.
as anticipated, the results improved in comparison to the probe models, with specificities of up to 100% and sensitivities of up to 92% when differentiating between normal and metastatic nodes. these results were achieved using a “leave one node out”, cross-validated assessment of five random spectra from each node. as a consequence, the time required to collect five spectra was less than 90 seconds - significantly quicker than any intra operative technique currently in use. though promising, these results remain experimental and were achieved using equipment within a laser laboratory, on tissue that had been “snap” frozen at the time of surgery.
but in ongoing research, the equipment is now inside an operating theater, where it can assess the node as soon as it is removed from the patient. since the 785 nm laser in the miniram ii spectrometer does not damage the tissue under investigation, a direct comparison between the raman results and those achieved using standard histopathological tests can be made. it is hoped that the results of these studies will be published in spring 2011 and, if successful, will signal the start of a major clinical trial.
on top of the obvious benefits of undergoing one operation rather than two, it is estimated that savings of up to $2,000 for every patient required to have immediate axillary surgery could be achieved. this, of course, also makes the approach an attractive one for health care providers.
“the advantages remain clear in our minds,” said a surgeon involved in the study. “it’s quick, independent of immediate pathology review, non-damaging （allowing post operative confirmation） and, with few ongoing consumable requirements, very cost-effective.”
“if its efficacy can be demonstrated in real time, then this could become the standard assessment method of choice in the years to come.”