Chemoradiotherapy for Cervical Cancer — Do Questions Remain?

The introduction of chemoradiotherapy into routine clinical practice for locally advanced carcinoma of the cervix is one of the success stories of the National Health Service (NHS). In February 1999, the US National Cancer Institute issued a rare clinical alert [1] stating that ‘strong consideration should be given to adding chemotherapy to radiation therapy in the treatment of cervical cancer’. This was in advance of the publication of five randomised phase III trials showing that platinum-based chemotherapy, when given concurrently with radiotherapy, significantly improves survival. A questionnaire survey of UK practice is a noun, practise is a verb in 2000 [2] showed that 96% of respondents (46 of 48 oncologists representing 46 UK centres) had begun using chemoradiotherapy for the treatment of cervical cancer less than 1 year after the issue of the National Cancer Institute alert. This encouraging finding suggests that rapid implementation of effective treatments is possible in the NHS, despite the resource constraints.

The Royal College of Radiologists (RCR) has conducted a retrospective audit of cervix cancer patients treated with radical radiotherapy (with or without surgery and/or chemotherapy) in UK cancer centres in 2001 and 2002. The results are published in this issue of Clinical Oncology [3]. Data were available on 1075 patients from 42 centres in the UK, making this one of the largest reported series of cervix cancer patients in the chemoradiotherapy era. The audit showed an 11% increase in 5-year overall survival for patients treated with chemoradiotherapy (55%) compared with those treated with radiotherapy alone (44%). There was also a 9% increase in overall survival for the entire cohort of patients treated with radical radiotherapy (56%) compared with a previous RCR audit of patients treated in 1993 (47%) [4].

The RCR audit represents a commendable undertaking to collect large-scale national outcome data for a relatively rare tumour. However, such audits are difficult to conduct and there are some omissions in the data collection, which the authors acknowledge. Although the rates of local recurrence were collected, the dates of recurrence were not collected and it was therefore not possible to calculate Kaplan–Meier estimates for local recurrence (or distant metastases). This is an important omission, as radiotherapy is primarily a local treatment and may account for the incongruity of some of the results of the analysis. Chemoradiotherapy did not seem to have an effect on local recurrence or distant metastases, despite the improvement in overall survival. The crude rates of local recurrence and distant metastases in the chemoradiotherapy group were higher at 22% and 27%, respectively, compared with 19% and 19% for the radiotherapy alone group. Moreover, although 1075 patients were available for survival analysis, 244 patients (23%) had to be excluded from the Cox analysis due to missing values for one or more variables included in the model.

The audit reported no difference in the crude rates of serious late toxicity for the patients treated with chemoradiotherapy (10%) compared with radiotherapy alone (8%). However, there were six treatment-related deaths in the chemoradiotherapy group compared with none in the radiotherapy group. It is generally accepted that retrospective collection of toxicity data tends to underestimate the true incidence. Contributors were also required to score toxicity using the Franco-Italian glossary, which is not commonly used in UK clinical practice. There are, therefore, caveats to the results of the toxicity analysis as the authors have pointed out. Radiation-induced toxicity is a major cause of morbidity in cervical cancer patients, as is evident from the accompanying editorial in this journal from a group of Patient Research Partners [5]. Even those side-effects graded clinically as ‘mild’ or ‘moderate’ can have a detrimental effect on quality of life for cervical cancer survivors. It is incumbent on us in the profession to collect high-quality toxicity data to help our patients make informed choices about chemoradiotherapy for cervical cancer, particularly as a recent meta-analysis [6] has shown that the survival benefits are smaller than originally predicted.

Perhaps a message from the RCR audit is that retrospective data collection can have significant limitations and a national cervix cancer database for prospective data collection is urgently needed. Such databases have been successfully established under the auspices of the NHS Information Centre for other tumour sites, including bowel (NBOCAP), head and neck (DAHNO), lung (LUCADA), breast (mastectomy and reconstruction), and oesophago-gastric cancers. The cervix cancer community is currently facing the exciting prospect of another intervention with a major effect on local control and toxicity. Magnetic resonance imaging-based image-guided brachytherapy has been shown to improve local control by 20% while reducing serious late toxicity by 10%. At a national meeting on ‘Implementation of image-guided brachytherapy for cervix cancer in the UK’ held in September 2008 and attended by 55 delegates from 29 centres, there was overwhelming support for the establishment of a national cervix cancer database for prospective data collection to evaluate the effect of introducing this new technique.

Over the last two decades there have been unprecedented developments and advances in radiotherapy equipment and techniques. However, there is regrettably a paucity of evidence for the clinical benefits of such technology, which has severely hampered its uptake in the UK. Unlike chemotherapy, radiotherapy does not lend itself to evaluation within randomised clinical trials, as techniques vary considerably with equipment and it is rare that any centre will have access to more than one type of equipment. Moreover, there are ethical concerns about withholding advanced radiotherapy techniques with documented dosimetric advantages from patients, even if the clinical benefits have not been ascertained. On the other hand, the need for objective evidence of clinical benefit is greater than ever in the current financial climate if radiotherapy departments are to persuade trust boards and healthcare commissioners to invest in highly expensive equipment. The pioneers of image-guided brachytherapy for cervix cancer have shown that it is possible to produce incontrovertible evidence of clinical benefit without the need for clinical trials by meticulous collection and reporting of clinical and dosimetric data.

The RCR audit of chemoradiotherapy for cervix cancer is a testimony to the desire and commitment of the clinical oncology community to evaluate the effect of introducing new interventions and treatments within the NHS. However, it also serves to illustrate the limitations of such endeavours if they are not supported by appropriate resources for data collection and information technology. In setting out his vision for the NHS [7], the new Secretary of State for Health, Andrew Lansley, has stated that ‘The NHS must make sure that patient outcomes and clinical evidence are at the heart of any changes to health services’. Now is therefore an opportune time to establish a national UK cervix cancer database for prospective data collection to continue to monitor the effect of chemoradiotherapy and other future interventions. It could also be the basis for an alternative model to investigate the clinical benefits of advanced radiotherapy techniques and other new technologies, particularly for rare tumours.