VOCPAC
Research type
Research Study
Full title
VOCPAC Investigation of Volatile Organic Compounds to diagnose and stratify men with ProstAte Cancer
IRAS ID
288678
Contact name
Chris Probert
Contact email
Sponsor organisation
University of Liverpool
Duration of Study in the UK
3 years, 2 months, 31 days
Research summary
"Research Summary"
Prostate cancer is common. Symptoms of prostate cancer are common in older men: a lot of them worry about prostate cancer: many request a PSA blood test. Alone, this test can be misleading. There is no screening test for prostate cancer.Volatile organic compounds (VOCs) are emitted from urine. Studies have suggested that VOCs can be used as biomarkers for prostate cancer. Our pilot data supports this: we could define men with prostate cancer with sensitivity of 83% and specificity of 100%: in the same patients, PSA had an accuracy of 62%.
We will translate this finding into a test to aid the diagnosis of prostate cancer, and stratify patients.
We plan to study 350 men with prostate cancer with 350 men with symptoms, in whom no prostate cancer was identified. 70% of samples will be used to create a diagnostic model and the remaining 30% will be used to test the model.
VOCs from urine will be analysed using two devices: a standard gas-chromatography / mass spectrometry (GCMS) system (using 3 different chemical approaches: urine that has been freeze-dried or acidified or alkalinised) and our novel, gas-chromatography-sensor device (GCSD) (acidified and alkalinised samples). We will compare these lab approaches to determine which is the most accurate.
Data from GCMS will be analysed using statistical tools and the current reference library. Potential biomarker compounds will be further investigated using reference standards as comparators.
Data from GCSD will be processed by our published pathways.
Clinical data (age, stage and grade of disease etc) will be used in the modelling. Tools will be developed to determine the specificity and sensitivity, positive and negative predictive value of each of the techniques.
We plan to commercialize the best technique in due course to aid the diagnosis of prostate cancer, as well as which cancers need treatment and which can be monitored.
"Summary of results"
Background Prostate cancer is the most common cancer in men in the United Kingdom. The diagnostic pathway involves prostate examination and prostate specific antigen (PSA) blood test. PSA can be raised in other prostatic diseases; only 25% of men with a raised PSA have prostate cancer. MRI scanning has improved the rate of detection, but there are issues with availability in many countries. Prostate biopsy is invasive, painful, and risks sepsis and bleeding. An improved biomarker would reduce the burden of unnecessary biopsies on men with raised PSA not due to cancer or those with low-grade disease that would not require treatment.Volatile organic compounds (VOCs) are small molecule products of metabolism found in all bodily fluids. Changes in VOCs may be found due to changes in cancer cell metabolism. There is a possibility that these could be used as biomarkers that are more specific for prostate cancer than PSA and/or MRI.
Aims and Objectives
1. Investigate differences in VOC profiles in urine of patients with and without biopsy-proven prostate cancer.
2. Evaluate VOCs as biomarkers for prostate cancer.
3. Investigate whether patients with high-grade prostate cancer have VOC profiles different to those with benign/low-grade biopsy results.Methods
Participants were recruited from the biopsy clinic. Urine is routinely taken to rule out infection and permission was obtained to store this sample at -20oC. Patients previously treated for prostate cancer or those with urinary tract infection were excluded. Samples were split into two aliquots of 1ml, to which sulfuric acid or sodium hydroxide was added (changing the acidity level alters which VOCs are released from the urine). Samples were analysed using gas chromatography-mass spectrometry. A subset of samples were analysed using a prototype gas chromatography-sensor system which can detect smaller VOC abundances. Comparisons were made between subgroups depending on biopsy results (all patients with prostate cancer vs benign, then all patients with high-grade cancer vs benign and low-grade). VOCs that were different between groups were used to train a model on 70% of the participants; this was tested on the remaining participants and compared to other clinical information.Results
949 men were included. Using the routine clinical data, we generated a local risk prediction model that could identify high-grade prostate cancer in 95% of participants, while preventing the need for unnecessary biopsy in 28% of those without high-grade disease.Using GC-MS, we have demonstrated significant differences in the presence of 25 and 9 VOCs in acidified and alkalinized urine, respectively, when patients with prostate cancer were compared to those with benign biopsies.
When the concentrations of compounds were compared, the acidified samples showed higher levels of 5-[(E)-but-2-en-2-yl]-2,2-dimethyloxolane and 4-methylpent-3-enoic acid, and lower levels of (2,4-ditert-butylphenyl)5-hydroxypentanoate in patients with prostate cancer.
Methylsulfonylmethane, 2-ethylhexan-1-ol and 2-methyl-5-prop-1-en-2-ylcyclohex-2-en-1-one (carvone) were found in lower concentrations in alkalinised samples in the group with prostate cancer.
Neither GC-MS or the GC-sensor analysis could identify a reliable urine biomarker for either prostate cancer or high-grade prostate cancer either alone or in combination with the clinical data model.
REC name
Wales REC 7
REC reference
20/WA/0297
Date of REC Opinion
28 Oct 2020
REC opinion
Favourable Opinion