EPA (omega 3 fatty acid from fish oil) has been studied as part of a nutritional supplement in patients with cachexia anorexia from pancreatic cancer. Despite promising results in several studies, a systematic review has not shown it to be better at improving weight or quality of life than a nutritional supplement without EPA. 
A number of studies examining the benefits of dietary supplements, such as minerals, vitamins and proteins have been undertaken in cancer patients, yet no systematic reviews have established positive recommendations. 
Weight loss due to primary cachexia anorexia syndrome from progressive malignancy is not reversed by supplying additional nutrition. 
Few studies have examined the impact of medically assisted nutrition on survival or quality of life in palliative care patients [9, 10] or those with dementia.  Clinical guidelines for use of artificial nutrition in the palliative care setting are evolving.  The proposed guidelines suggest criteria for identifying who might benefit, including nutritional criteria, life expectancy, quality of life, and functional capability. Patients with secondary cachexia (due to obstruction or swallowing difficulties, without the inflammatory processes of the cancer cachexia syndrome) may benefit from a trial of enteral or parenteral nutrition. It should be recognised that artificial feeding has significant risks, complications and burdens. 
In patients with advanced dementia and cachexia, artificial feeding (by tube) does not prolong life or prevent aspiration. [14,15] A recent systematic review compared the outcomes of enteral nutrition in dementia and other conditions, and showed similar mortality for patients with dementia and other advanced conditions receiving enteral nutrition. The data included were of variable quality and the issue is acknowledged as ethically challenging. 
A systematic review revealed that the frequency of use of artificial nutrition in cancer patients in the last week of life may range from 3 – 53%, and artificial hydration may be used in up to 80%, but withholding or actively withdrawing artificial hydration or nutrition is rarely studied.  There was no clear-cut benefit in terms of survival, comfort or quality of life from either treatment. The evidence that it may improve or worsen specific symptoms (thirst, delirium, nausea, fluid overload) is not consistent.
- The optimal timing and strategy for providing nutritional support is not known.
- There is limited evidence about the effects of nutritional counselling on palliative care patients, but non-sustained improvements in intake have been shown. 
- Decisions about artificial nutrition in palliative care patients should include an assessment of both the benefit and burden of all treatments offered.
- The use of enteral and parenteral nutrition in terminally ill cancer patients approaching the end of life is rarely indicated. Dying patients lose their ability to swallow safely in the last few days of life. Studies support the view that dying patients require only minimal amounts of food and fluids to reduce thirst and hunger. [6,9]
- The use of enteral and parenteral nutrition in patients with advanced cancer and a life expectancy of months or more is controversial.  Clinical guidelines for use of parenteral nutrition in the palliative care setting are evolving and the need to develop evidence to support decision making in this topic is recognised. [12,13]
- Artificial feeding in advanced dementia is ethically problematic and does not provide good palliation. In general the focus should be on offering whatever oral intake the patient will accept, and on good mouth care. However families and aged care staff may both require support and education in this area. [2,15]
- Decision-making about artificial nutrition for patients who lack capacity requires a framework that balances treatment risks/benefits /burdens and quality of life considerations, and family and clinician perspectives, with a concern not to prolong suffering. However 'quality of life' is a concept that has varying interpretations, and discussion between clinicians and family needs to aim for a shared understanding.