Page 200 - Risk quantification and modification in older patients with colorectal cancer

Chapter 11
for outcomes of CRC surgery, have acceptable discrimination but unsatisfactory calibration. As a consequence, the use of poorly calibrated models that overestimate individual risks could lead to worse outcomes compared to not using a prediction model.2 Moreover, many prediction models showed unsatisfactory performance in validation studies (Chapter 2). In addition, prediction models for outcomes of CRC need a periodic update owing to possible changes of the population and certain changes in therapy.3 This applies to surgical CRC prediction models, because of the many efforts made to improve outcomes of colorectal cancer surgery for patients (i.e. auditing, ERAS including laparoscopic surgery, neoadjuvant treatment and wait-and-see for rectal cancer and selective use of defunctioning stoma).4,5 For the GerCRC model from Chapter 6, we initiated validation including evaluation of its calibration, and this will be completed in 2020.
With regard to the field of surgical oncology, we have shown that gender, comorbidity, physical functioning (need for ADL assistance and use of a mobility aid), cognitive functioning (previous delirium) and tumour location are useful predictors for postoperative complications and have incorporated this in the GerCRC model. The GerCRC model underlines the importance of taking geriatric- predictors into account when conducting prognostic research in the field of surgical oncology. It is possible that the discriminatory value of the GerCRC model could be enhanced with other geriatric parameters or physical performance measures such as the need for help with Instrumental Activities of Daily Living (IADL) or physical performance measures such as walking speed or grip strength. A planned validation study will demonstrate whether further improvement of the performance of the GerCRC model is needed. Hence this would require fewer patients (and events) to investigate.6 As highlighted in Chapter 2, the geriatric screening tool G8 alone is not useful as a prognostic tool for complications of CRC surgery.
Body composition research
Previously published cut-off values for radiologically assessed low skeletal muscle mass and density do not apply to older patients. We have shown that physical functioning reflected by the use of a mobility aid has better potential as a predictor for complications and survival then a single CT-measurement of muscle mass or muscle density. The big challenge for body composition research is determining interpersonal variation and pathological loss of muscle mass and quality7 and
198