Radiotherapy is the choice of treatment for the locally advanced stages of cervical cancer. Resistance of tumours to radiotherapy, due to intrinsic factors, is a serious hurdle. Now, researchers in India have shown that Raman spectroscopy could be helpful to oncologists hoping to predict how well a particular patient will respond to the therapy. The findings could enable clinicians to adopt a patient-specific treatment regimen.

Cervical cancer is a leading cause of cancer death. In India, for instance, it accounts for almost a quarter of mortality among women, with 370,000 new cases diagnosed every year. Moreover, a quarter of new cases around the world are reported in India. Despite these figures, researchers estimate that 90% of deaths could be prevented with early diagnosis through effective screening followed by treatment.

Chilakapati Murali Krishna of the Center for Laser spectroscopy, Manipal University, Manipal (presently at Cancer Research Institute, ACTREC, Navi Mumbai) explains that malignancy progresses through several stages to carcinoma in situ and finally to invasive cancer. Currently, a combined radiotherapy and chemotherapy regimen is used for locally advanced stages.

Conventionally, immediate tumour response to radiotherapy is clinically assessed at the end of the treatment by measuring the degree of tumour shrinkage. There are no established methods to predict outcomes during the early stages of treatment. Cancers can develop resistance to radiotherapy, just as they can against drugs, which represents a serious obstacle to treatment. Factors such as DNA aneuploidy (chromosome abnormalities), S-phase fraction (a measure of tumour growth rate) and proliferation kinetics (cell division speed), tumour vascularity (blood vessel growth) and hypoxia (oxygen starvation) and glutathione content all influence radiation resistance and interfere with determination of the success of treatment.

Given that a typical radiotherapy regimen lasts for approximately four months, an objective approach to predict tumour response to radiotherapy, at the early stages of treatment, would be very beneficial to the patient as it could be used to create an individualized treatment regimen. This would also reduce the financial burden on patients.

Krishna points out that cervical cancer is a very well-studied form of malignancy and has been investigated keenly with several optical spectroscopic techniques previously, including Fourier transform infrared (FTIR), fluorescence (LIF), and combination of reflectance and fluorescence spectroscopy, as well as Raman. However, all of these studies focused on diagnostic and screening applications. Krishna and clinical colleagues led by M.S. Vidyasagar and colleagues in the Department of Oncology, Shirdi Sai Baba Cancer Hospital and Research Center, Manipal University, Manipal. Vidyasagar and colleagues have, for the first time, demonstrated the feasibility of Raman spectroscopic prediction of immediate tumour response to radiotherapy using tissues that are collected as early as after second fraction of the treatment. Raman spectroscopy, among the spectroscopy methods is more suited to in vivo applications as it is not affected by water bands and employs less harmful near-infrared radiation.

In order to explore whether the technique might make a viable predictive tool, the researchers recorded spectra from malignant tissues collected before and 24 hours after patients had received a second round of radiotherapy. They analysed the resulting data using PCA (Principal Components Analysis) and looked for a correlation between radiotherapy responses of the patients which was available at the end of the treatment.

In their study, the team analysed 106 spectra from 25 malignant and radiotherapy samples. Of those twenty-five patients, eighteen subjects showed complete response, three patients showed partial response, and the remaining four showed no response on clinical examination.

The team found that the mean Raman spectra corresponding to different levels of tumour response, complete, partial, and no response, showed "minute but significant variations". They corroborated their results using supervised classification of the spectra. "Our study suggests that Raman spectroscopic prediction of tumour radiotherapy response in cervical cancer is feasible," Krishna says.

"If outcomes can be predicted sufficiently early, after the second fraction of treatment, say, then suitably modified treatment modalities could be implemented for non-responding subjects," he adds, "With suitable fibre-optic probes there is also the prospect of being able to carry out these examinations under in vivo/in situ conditions, without a biopsy."

Related links:

• Biopolym, 2008, in press
• Krishna Page

Article by David Bradley