As already stated, there is a diffuse resistance in using QoL assessment in surgical practice. Most surgeons believe they can be useful only for research, but even in this field QoL is an add-on rather than an internal part of most trials. There is a common perception among surgical oncology specialists that this parameter is useful, especially in the comparison of different therapies for advance stage disease where there are low chances of improving survival. In other words, the concept of "QoL" is strictly linked to palliative care.
Nowadays, evidence is accumulating to suggest that QoL per se plays a role as a prognostic factor. Baseline QoL predicts survival in different types of cancer, such as myeloma , head and neck cancer , breast cancer  and oesophageal carcinoma . Also, two large cohort studies [32, 33] reached the same results by analysing different malignancies.
The first studies on rectal cancer obtained the same results. Earlam  demonstrated that a better QoL (measured by the Rotterdam Symptom Checklist Score) was associated with improved survival in patients with colorectal liver metastasis. Maisey et al.  retrospectively analysed patients with advanced colorectal cancer enrolled in 4 different clinical trials for testing different chemotherapy regimens. They found that baseline QoL was an independent prognostic indicator in all the patients involved in these four phase III clinical trials. Both previous studies take account of patients with advanced stage disease (inoperable cancer) and the QoL assessment was done only before therapeutic manoeuvres.
In 2001, Camilleri-Brennan et al.  published an interesting study where they analysed the change of QoL score among 65 patients undergoing curative surgical resection. They found two important far-reaching conclusions:
1. The QoL scores are dynamic and should be checked over time. A worsening in specific items could suggest the presence of specific problems (for example: loss in appetite can be associated with early bowel obstruction because of recurrence).
2. In their work, the combination of sociodemo-graphic and QoL scores could predict 1-year survival with an accuracy of 76.8%.
Even if larger trials are needed to confirm these results, these conclusions are very important. Collecting information by routine use of QoL scores can help physicians to follow patients over time and plan treatment on specific areas affected by the disease. In our experience (unpublished data), weight loss and gastrointestinal symptoms (such as nausea and vomiting) are early predictors of tumour recurrence and, sometimes, they start 3-4 months before radiological finding or increase in neoplastic markers.
Was this article helpful?