Metabolic Profiling of Urine Reveals Distinct Phenotypic Changes in Patients Undergoing Colorectal Excision

Sunreet Randhawa, Reza Mirnezami, Beatriz Jimenez, James Kinross, Olaf Beckonert, Claire Merrifield, Elaine Holmes, Jeremy Nicholson, Ara Darzi. Section of Biomolecular Medicine and Section of Biosurgery and Surgical Technology, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, London SW7 2AZ, UK.

INTRODUCTION

Systems biology has yet to be widely adopted for surgical use however this technology is highly applicable for the development of individualised surgical therapeutics. Colorectal surgery disrupts the gut microbiota and important symbiotic co-metabolic pathways which may be indirectly measured in urine via this approach. This study aimed to determine if a metabolic phenotype exists for the anatomical resection or the invasiveness of surgical approach which accounts for the role of the gut microbiota in mammalian metabolism.

METHODS AND PROCEDURES

27 colorectal cancer (CRC) patients undergoing open or laparoscopic elective surgery for colorectal cancer (CRC) were recruited prospectively at St Mary’s Hospital, London, UK. Urine sampling was performed preoperatively on the morning of surgery and on the 3rd postoperative day. Samples were frozen at -800C. A metabolic, pharmacological and clinical history was taken and oncological staging and grading data were recorded. Urine samples were analysed by high resolution 1H NMR spectroscopy using a Bruker 600 MHz Spectrometer. Spectra consisting of 32,000 data points were phase and base line corrected using TopSpin (Bruker BioSpin, Germany). Raw data were exported into SIMCA 12 and Matlab for multivariate analysis by principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) models. Water and urea were excluded from the spectra.

RESULTS

The median age was 68 (range 47-84) and the male:female ratio was 16:11.There were 11 laparoscopic and 16 open procedures (3 laparoscopic cases were converted to open). Cases comprised: left hemicolectomy (n=8), right hemicolectomy (n=8), anterior resection (n=10) and abdomino-perineal resection (n=1). Cancers were staged according to UICC criteria as stage 1 (n=6), stage 2 (n=8), stage 3 (n=9) and stage 4 (n=4). Unsupervised multivariate analysis of pre-operative urine samples revealed significant metabolic variability according to the anatomical site of the tumour (R2 = 0.46). Creatinine statistically correlated with rectal cancer while colonic cancers demonstrated higher levels of hippurate and aromatic compounds strongly identified as gut microbial co-metabolites. OPLS-DA was able to strongly predict the pre- and postoperative states (R2y (cum)= 0.65; Q2 (cum)= 0.63). In addition, postoperative analysis identified markedly differing urinary metabolic profiles between patients undergoing right versus left hemicolectomy (R2y (cum)= 0.65; Q2 (cum)= 0.63, see figure) and rectal versus colonic surgery (R2y (cum)= 0.65; Q2 (cum)= 0.63). Furthermore, we identified statistical variability between laparoscopic and open cases (R2y (cum)= 0.717; Q2 (cum)= 0.22).

CONCLUSION

Urinary metabolic profiling provides a systems overview of the complex response to colorectal surgery and serves as an important scientific platform for the development of personalised surgical therapies. It is able to achieve this in part by non-invasively measuring gut microbial co-metabolic pathways and systems biology is of benefit in determining the beneficial metabolic affects of minimally invasive surgery.