Using an Electronic Nose to Predict Gastrointestinal Consequences of Pelvic Radiotherapy (PREDICT)

There have been enormous improvements in outcomes for patients diagnosed with cancer in recent years. Patients survive longer and the number of survivors after cancer therapy increases 3% annually in the UK and 11% in the USA.It is estimated that there will be 4 million cancer survivors in the UK by 2030.Patients after cancer treatments have many unmet needs which are often caused by the very therapies given to cure or control the cancer. Of all the possible chronic physical side effects of cancer treatment, gastrointestinal (GI) symptoms are the most common and not only have the greatest impact on daily activity but also frequently carry serious financial, psychological and social implications.

During treatment for the cancer, GI symptoms are common and their causes are poorly researched despite their impact on patients' quality of life and the fact that when they occur they often require dose reduction or cessation of chemotherapy and/or radiotherapy with potential impact on outcome. In addition, the GI side effect may not settle once anti-cancer treatment is stopped. In patients receiving radiotherapy for a tumor in the pelvis, more severe acute reactions predispose to worse long term side effects.

At least 17,000 British patients are treated annually with radiotherapy for pelvic cancer. Up to 80% of these patients are left with chronic alteration in GI function, 50% state that this affects their daily activity and 30% that this change in function has a moderate or severe effect. While progress has been made in defining optimal management of chronic changes in bowel function after cancer treatment and UK multi-professional guidance to aid clinicians has been published it would be extremely valuable to be able to predict those in advance who might go on to develop serious problems as a result of radiotherapy so as to have the opportunity to provide them with pre-treatment counseling, potentially modify the cancer treatment and introduce toxicity modifying therapies at the earliest opportunity.

It is widely believed that technical advances in the delivery of radiotherapy will abolish GI problems however, recent compelling data demonstrate that 30% of patients treated with "perfect" radiotherapy - meeting every described constraint still develop significant un-predicted problems. While mechanisms by which gastrointestinal symptoms occur after cancer treatment are starting to be understood and personal parameters which change the risk profile for individuals are being identified (body mass index, concomitant chemotherapy, use of a statin or ACE inhibitor, the presence of diabetes mellitus, hypertension, connective tissue disorders or HIV infection) the reasons why some patients remain symptom free whilst others experience severe side effects remains under researched.

It has become clear that development of GI toxicity is not solely related to the dose and way the radiotherapy is delivered, and a phenomenon independent of the radiation called the 'the consequential effect' is a second reason why some people get chronic side effects after radiotherapy. An important driver of the 'consequential effect' is increasingly believed to be the microbiota, the vast numbers of bacteria that live in our guts. The gastrointestinal microbiota are a complex ecosystem of up to 1,000 bacterial species in any one individual. The species vary greatly between individuals but within each individual, the flora composition remains stable for the majority of the species over time. The diversity of the microbiota is high in healthy people and low in people with GI side effects after pelvic radiotherapy, a process which very closely parallels findings in people with other much more intensively researched inflammatory conditions of the bowel.

Normal gut flora produce gases from their metabolites (Volatile Organic Compounds or VOCs). The gold standard for analyzing volatile organic compounds is gas chromatography/mass spectrometry (GCMS).

GCMS analyses gasses to identify all chemical components of that gas. However, this is very expensive and requires a specialized laboratory to process the samples hence is not a viable option in day to day clinical practice.

Two alternative techniques can be applied to analyse those VOCs: electronic sensing (e-nose) or High Field Asymmetric Ion Mobility Spectrometry (FAIMS).

Electronic sensing techniques are applied via an instrument (electronic nose) that attempts to replicate the biological olfactory system, by investigating samples as a whole, instead of identifying specific chemicals within a complex sample. The air above the sample (head space) is drawn into the e-nose and passed across an array of chemical sensors. The size of the array varies, but most are between 6 and 32 sensors. As each sensor is dissimilar, the interaction between the sensor and the sample is unique and an olfactory signature for this complex odor can be created.

Fields Asymmetric Waveform Ion Mobility Spectometry (FAIMS) is a new technology capable of separating gas phase ions at atmospheric pressure and at room temperature. As with the electronic nose, FAIMS can be used for the real time analysis of complex chemical components, looking at the total chemical composition of a sample. Differences in the way the ionised molecules move in high electric fields are used to draw a mobility signature of a complex sample.

The e-nose and FAIMS techniques have been used successfully in small studies to diagnose lung disease, diabetes, bladder cancer, tuberculosis, cyanide poisoning, renal failure and schizophrenia. Focusing on gastrointestinal disorders, consistent changes in the VOCs produced by patients with inflammatory bowel disease, bile acid malabsorption, gastrointestinal cancer, coeliac disease, Clostridium difficile infection can be identified compared to healthy individuals.

In a paper published in 2012 of a retrospective blinded pilot study using the e-nose in 23 patients in whom the investigators analysed stool samples collected immediately before the start of radiotherapy. Of the 23 patients selected, 11 subsequently had minimal or no side effects from radiotherapy and 12 had severe toxicity. The e-nose analysed the samples and astonishingly identified with 100% accuracy two completely distinct groups of patients: those who would develop severe toxicity during radiotherapy and those who would not. The results of this study suggest that the substantial differences in the volatile gases produced by stool samples in the two populations, have a critical impact on toxicity and the reason that different fermentation products occur is either because the microbiological composition of stool differs between the two groups, or because bacterial function is somehow different. Whatever the reason, these findings suggests that the microbiota plays a critical role in the initiation of radiation induced inflammation are completely consistent with findings from other studies using very complex methodologies and with studies investigating the role of the microbiota in other GI conditions .