While I congratulate the authors on a very thoughtful and relevant review, I am going to respectfully challenge some of the nuances in this otherwise excellent paper.

Opioid Responsiveness and Selection

The issue of interindividual variability of opioid responsiveness is one of the most fascinating aspects of opioid pharmacology. There are few other pharmacologic agents with a 1,000-fold variability in effective clinical doses. While the pharmacogenomic studies described are interesting, it remains premature to draw definitive conclusions. Hopefully the biobank being developed by Prof. Kassa and his international colleagues from the European Palliative Care Research Collaborative will help add clarity to this picture.

The discussion on opioid selection is well presented. Despite the concerns raised by the authors, I think it is important to indicate that the development of liver failure in a patient who has been previously stable on an opioid does not necessitate changing the opioid or its dose unless the patient develops new dose-limiting adverse effects. There are important differences between liver failure caused by tumor infiltration and that caused by global hepatocellular disease. In the setting of metastases, there are usually well preserved areas of healthy hepatocytes that continue to function well until a very advanced stage of liver infiltration. Indeed, it is not uncommon to observe good preservation of synthetic and metabolic liver function in patients with relatively advanced liver metastases; this function is maintained by the residual healthy parenchyma.

Regarding the use of fentanyl(Drug information on fentanyl), I would challenge some of the assertions made. While I deplore decisions made on the basis of a seductive technology, there is a rationale for many patients to consider fentanyl as a viable and appropriate first-line option. This is particularly true for patients with an already heavy burden of oral medications and those for whom there may be problems of compliance. In many countries, a generic form of transdermal fentanyl is now available and the cost of this therapeutic option is falling. For many insured patients, this consideration may even be irrelevant. There are no data to indicate that titration with transdermal fentanyl need be slower than the titration of controlled-release morphine(Drug information on morphine). Indeed, all of the comparative studies of transdermal fentanyl and oral morphine showed no difference in level of pain control. Though the use of transdermal fentanyl is not generally recommended as first-line therapy, it is an overstatement to maintain that it is an "inappropriate" option.

The concern regarding dose selection when switching to or from transdermal fentanyl is also overstated. That equianalgesic conversion tables need to be used with caution when making these calculations as part of an opioid switch holds true for all opioids. This admonition is not specific to fentanyl.

Pain Assessment

In their discussion on the assessment of pain, the authors state that a realistic goal is a 33% to 50% reduction in pain intensity, or a 2-point decrease in the 10-point numerical rating of pain (as well as maintained or improved physical function, reduced pain interference, and minimal opioid side effects). These assertions are based on acute pain studies . Although such parameters may be meaningful in the context of chronic pain, the notion of satisfactory relief to a tolerable level remains a gold standard. Although this is variable, in practice, pain rated less than 5 on a 10-point scale is usually associated with low levels of interference with function.[1]

Adverse Effects

Some adverse effects appear to transiently and spontaneously abate after the initiation of an opioid or after dose escalation. This phenomenon has been well demonstrated in a prospective study on the effect of morphine dose escalation on cognitive performance.[2] The study demonstrated that cognitive impairment commonly improved after 7 days. This phenomenon, although often described, has not been formally studied with regard to other adverse effects. Indeed, the oft-stated assertion that nausea generally will resolve over several days is not supported by data.