Detecting early signs of cancer with Raman microscopy

Prof Fiona Lyng

Prof Fiona Lyng

Identifying the first signs of cancer as early as possible is key to ensuring the maximum chances of survival. This is true of all cancers, and especially true of cervical cancer, the 4th most common cause of cancer death in women globally. Cervical cancer is largely preventable, and early detection of precancerous signs through routine screening tests like Papanicolaou (Pap) smears or Human papillomavirus (HPV) testing can allow for timely intervention.

While the emergence of the HPV vaccine has been overwhelmingly successful in reducing the incidence of cervical cancers, current gold standard screening strategies are imperfect. Pap smears, for example, have long been cemented as the routine cytology test for identifying cervical abnormalities, but the technique is limited by poor sensitivity. As a result, Pap tests cannot always reliably detect abnormal cell morphologies or precancerous changes, leading to potential false-negative results.

Professor Fiona Lyng, Ph.D, Manager of the Radiation and Environmental Science Centre, Technological University Dublin, has been a leading voice in the field of early cancer detection using optical spectroscopy-based techniques for more than a decade. Prof. Lyng’s research has made transformative contributions to the field, including the development of a novel Raman spectroscopy-based protocol for the detection of cervical precancer, which has shown clinical utility. We recently caught up with Prof. Lyng about her pioneering research, and how HORIBA Raman microscopes have supported her throughout her research career.

The importance of precancer detection

“For any cancer, not just cervical, early detection is key. Being able to pick up precancer is really important before you get down the road to cancer, because at that point, it's already infiltrated and it could be spreading.

Detecting precancerous changes in cervical cells gives a critical window for intervention before cervical cancer can take hold.”

Prof. Lyng explained the limitations of conventional cervical cancer screening methods, and described how this was a motivating factor behind her developing a novel Raman microscopy-based test.

“Pap testing has been around for a long time, and remains the gold standard cytology test for cervical cancer. The problem with the Pap cytology test lies in its subjectivity – the test relies upon clinicians visually inspecting the samples to assess whether the cell morphology is abnormal. Reliable testing leans heavily on the skills and experience of the clinician. Even then, cervical cells can appear morphologically normal, but actually be biochemically abnormal, and that's what we were trying to overcome using Raman spectroscopy.”

In recent years, screening programs have leveraged PCR-based HPV testing, but as Prof. Lyng explained, these too can bring limitations:

“Many of the cervical screening programs today use HPV testing for screening rather than cytology, usually implementing a cytology triage if a positive HPV test is returned. HPV tests have excellent sensitivity in detecting HPV infection, but a low specificity as to whether it’s likely to induce cervical cancer. A woman can have an HPV positive test, but it could be a transient infection with no risk of developing into cancer. Follow up cytology tests then may or may not detect an abnormality due to its own inherent limitations.”

Given that cervical cells can show biochemical signs of precancer long before morphological changes appear, developing testing methods that can detect these early biochemical changes in cervical cells could be a game-changer in the fight against cervical cancer.

Discovering the power of Raman spectroscopy

Setting out to find more reliable methods of detecting precancer, Prof. Lyng and team started to consider optical spectroscopy-based techniques.

“In my postdoctoral research at Daresbury Laboratory, UK, I had been working on confocal fluorescence imaging but my colleague, Dr Mark Tobin, was working on using synchrotron radiation infrared spectroscopy for cancer diagnosis, so when I came to my current institution in Dublin, I was excited to find that they had a HORIBA Raman spectrometer – the only Raman microscope in Ireland at that time in 2000. It was then that I started getting really interested in using spectroscopy for cancer diagnosis, so together with my colleague Prof Hugh Byrne, we started investigating the capabilities of the Raman method.”

In collaboration with Dr Peter Kelehan at Dublin’s National Maternity Hospital, Prof. Lyng and team were given access to clinical cervical tissue samples, and began to find encouraging results.

By probing the tissue with Raman spectroscopy, it’s possible to determine the biomolecular fingerprint of cells, and the technique was found to be well-suited for identifying minute changes associated with precancer.

Research soon moved from tissue samples to clinical cytology samples used in screening thanks to a collaboration with Prof John O’Leary and Dr Cara Martin at the Coombe Women and Infants University Hospital  – Prof. Lyng was keen for Raman testing to be as minimally invasive as possible, while minimizing disruption to the normal screening process.

 

Figure 1: Illustration of the biochemical fingerprint of a cell showing Raman spectrum of a cell together with spectra of DNA and RNA (nucleic acids), phosphatidyl-inositol (phospholipid) and actin (protein)

Figure 1: Illustration of the biochemical fingerprint of a cell showing Raman spectrum of a cell together with spectra of DNA and RNA (nucleic acids), phosphatidyl-inositol (phospholipid) and actin (protein)

“Developing our protocol took a lot of work, and we spent a long time figuring out how best to approach it. For spectroscopy applications, most people use calcium fluoride or other special substrates to mount their samples to optimize conditions, but that wasn’t an option for us, as it would have meant disrupting the whole clinical workflow. If you want something to be adopted easily within a clinical environment, it has to fit straight into the current workflow, so we were adamant about using glass slides, although it caused some complications.”

“A major issue we had was contamination from minute traces of blood in the samples. Since we wanted to use glass slides as these are used in the current clinical workflow and are also much cheaper than other specialist materials, but this limited the excitation wavelengths we could use. Most people use 785 nm for Raman spectroscopy, but that was no good for us using glass, because you get a lot of fluorescence using that wavelength. So that took us down the other end of the spectrum to green light at 532 nm. At this wavelength however, hemoglobin is massively resonant, which disrupted results. But, with great perseverance, trial and error, we came up with a successful procedure.”

One of the most elegant facets to Prof. Lyng’s protocol, published in Nature Protocols [1], is its simplicity and non-destructive nature. This enables the Raman protocol to be used as an add-on to augment the current cervical cancer screening workflow, rather than replace it.

“Our protocol involves minimal sample prep so that we cause very little disruption to the sample. Since Raman spectroscopy is non-destructive, the sample can then be analysed via additional pathways. It could be then stained for cytological examination, or used for molecular biomarker analysis. This way, our protocol can sit on top of the current testing programs without the need for additional sampling.”

Following the successful development of the Raman protocol, Prof. Lyng’s team recently put it to the test, with a large-scale clinical utility study in collaboration with the Coombe Women and Infants University Hospital and funded by the Health Research Board in Ireland [2].

“In 2022, we published our clinical utility study in Raman spectroscopy and cervical precancer, in which we analyzed samples from 662 individuals. This was the largest-scale trial of Raman microscopy for the detection of cervical precancer to date, and we were really pleased with the outcome. An independent blinded test set of samples (a mixture of negative samples and cervical intraepithelial neoplasia (CIN) samples) validated the classification accuracy to be 91.3%.”

Figure 2: Raman spectra and Principal Components Analysis plot showing discrimination between superficial (blue), intermediate (green) and parabasal (red) cervical cells.

Figure 2: Raman spectra and Principal Components Analysis plot showing discrimination between superficial (blue), intermediate (green) and parabasal (red) cervical cells.

A workhorse microscope supporting cutting-edge research

Technological University Dublin upgraded the original HORIBA Raman microscope LabRam system to two LabRam HR800 systems, and added a pair of HORIBA XploRA™ systems in 2009. The XploRAs have been a critical part of Prof. Lyng’s research ever since, and have been the mainstays throughout the development of the Raman-based cervical precancer detection protocol.

“The XploRA systems have helped us enormously in the development of our protocol. It's a system that could readily be used in a hospital lab, because unlike our older Raman microscope system, it fits onto a benchtop and is very straightforward to operate by non-specialists. We’ve had numerous researchers working on our projects from various backgrounds, and we’ve been able to rapidly get them up to speed and using the XploRA with confidence. They're real workhorses – this year marks 15 years since we got our XploRA systems, and they're still going strong and delivering the robust data we need to support our research.”

Future research plans

Since HPV testing is increasingly used in cervical screening programs, Prof. Lyng and her team have been investigating the possibility of using Raman spectroscopy to augment this testing to stratify women who are HPV-positive on screening. They propose that Raman could be leveraged as a triage test to discriminate between transient and transforming HPV infections. Preliminary work has been published [3], showing promising results to meet this unmet clinical need.

“Although we've been heavily focused on cytology up until recently, things really have changed over the last few years in the way cervical cancer is screened. It’s now all about HPV testing – so we have to move with the times. Our pilot study found Raman spectroscopy to be able to classify transforming HPV infections from transient infections with a sensitivity of 85% and specificity of 92%, which is really encouraging. Better HPV triage would allow women with high-risk persistent HPV infection to be referred for immediate treatment, while women with a low-risk transient infection could avoid unnecessary overtreatment.”

In addition to their work in augmenting HPV triage, Prof. Lyng is also interested in broadening the utility of Raman spectroscopy to detect precancerous changes in cancer types beyond cervical. 

“I’m interested in using Raman spectroscopy for the detection of alternate types of precancer. We’ve previously done work on lung and thyroid samples, but we’re particularly excited about our work in the oral precancer space. Oral samples are ideal because collection is minimally invasive – you just open the mouth and use a brush to collect the cells, and then process them the same way that we do for cervical cancer samples.”

“Our initial studies with the Dublin Dental University Hospital have found Raman spectroscopy to distinguish between the different grades of oral precancer or oral dysplasia, and that has worked really well [4]. We have a new project starting this year funded by Science Foundation Ireland investigating whether we could use Raman spectroscopy and minimally invasive saliva and cell samples for improved management of oral potentially malignant lesions. For all of our ongoing studies into precancer detection, our XploRA systems will continue to prove invaluable.”

References

  1. Traynor, D., Behl, I., O’Dea, D. et al. (2021). Raman spectral cytopathology for cancer diagnostic applications. Nature Protocols, 16, 3716–3735. https://doi.org/10.1038/s41596-021-00559-5
  2. Traynor, D., Duraipandian, S., Bhatia, R. et al. (2022). Development and Validation of a Raman Spectroscopic Classification Model for Cervical Intraepithelial Neoplasia (CIN). Cancers, 14(7), 1836. https://doi.org/10.3390/cancers14071836
  3. Traynor, D., Martin, C. M., White, C., et al. (2021). Raman Spectroscopy of Liquid-Based Cervical Smear Samples as a Triage to Stratify Women Who Are HPV-Positive on Screening. Cancers 2021;13(9):2008. doi: 10.3390/cancers13092008
  4. Behl I, Calado G, Vishwakarma A, Traynor D, Flint S, Galvin S, Healy CM, Pimentel ML, Malkin A, Byrne HJ, Lyng FM. Classification of cytological samples from oral potentially malignant lesions through Raman spectroscopy: A pilot study. Spectrochim Acta A Mol Biomol Spectrosc. 2022;266:120437. doi: 10.1016/j.saa.2021.120437

Follow Us 

    

留言咨询

如您有任何疑问,请在此留下详细需求信息,我们将竭诚为您服务。

* 这些字段为必填项。

留言咨询

如您有任何疑问,请在此留下详细需求信息,我们将竭诚为您服务。

* 这些字段为必填项。

Corporate