Pain is a main concern for most of patients with cancer: pain is limiting for their quality of life and also has a huge impact on their outcome [79
]. Opioids are the most frequently prescribed drugs for cancer pain management. Despite their wide adoption, concerns still remain about both their efficacy and safety. Some patients still suffer for incomplete pain relief and side effects of are an issue in the short- and long-term.
One of the major contributors in determining pharmacologic response (including opioids) is the genetic background, which is different among patients and can influence several aspects of drug response (pharmacokinetics) and sensitivity (pharmacodynamics). Theoretically, mapping the inter-individual variability in genes involved in opioid response is the “Holy Grail” for pain physicians. Knowing how a patient may react to a specific drug may actually be the main step towards a more effective “personalized” therapy. Pharmacogenetics has the specific endpoint of understanding those genetic variations, in order to tailor specific treatments to each and every patient. This is even more important in cancer patients: they are usually multi-pharmacologic, more prone to drug–drug interactions and potential adverse events. The “trial-and-error” approach is ineffective, dangerous, costly, and causes delays in effective care: pharmacogenetics and personalized medicine may help in overcoming this obsolete approach, using genetic tests to assess each patient’s risk for adverse events and the likelihood to positively respond to a given drug, thereby allowing an “informed” drug selection.
In such a scenario, different genetic variants have been investigated for many aspects of opioids, including both PK and PD features. In a clinical purpose, it is even more important for prescribers to understand how genetic variants affect efficacy and safety, i.e., how each variant and their combinations affect pain relief and the development of side effects. Testing a patient for this purpose may help changing prescriptions to improve therapeutic profile, determining the most effective drug at the safest dose.
In the field of PK (which basically explains how the drug is metabolized by the body) the main actor is the cytochrome P 450 family, which is involved in liver metabolism of many drugs (including opioids). Different isoforms are active in opioid metabolism, but CYP2D6, 3A4, and 3A5 are the main (or at least those who have been mostly investigated).
Testing CYP2D6 in the clinical setting provided interesting data; although less of them come from the specific scenario of cancer pain management, they suggest that CYP family testing may help better opioid prescribing. It’s probably too early to tell if we can screen patients to specifically identify those who will respond better and with lower doses, although studies to genotype patients and guide oxycodone prescribing are in place [18
], and a recent trial showed that CYP2D6
-guided opioid therapy improves pain control in CYP2D6
intermediate and poor metabolizers [20
]. In any case, CYP testing may help in screen patients with extreme phenotypes, i.e., those prone to treatment inefficacy or serious adverse events [80
A recent paper has even proposed an alternative to classic nomenclature that is more practical for clinician [7
]. This system classifies 3 different phenotypes for the CYP2D6
gene (functional, sub-normal, dysfunctional). With this approach, patients with dysfunctional alleles should be treated with caution: CYP2D6 substrate drugs (including opioids) should probably be avoided since they may result in poor efficacy and/or increased adverse events (see Table 1
). Specific indications already exist for some opioids. The Clinical Pharmacogenetic Implementation Consortium (CPIC) has published clinical recommendations for codeine dosing based on CYP2D6
]. Both CYP2D6
UMs and PMs should avoid codeine: in the former group a rise in morphine formation will increase toxicity, while in the latter no analgesic effect is weaker (due to reduced morphine formation). Literature provides less evidence for other opioids; caution should especially be taken with the use of tramadol, oxycodone, and hydrocodone.
The Dutch Pharmacogenetics Working Group (DPWG) has produced other relevant guidelines [82
]. Even though (in general) some difference is retrievable between guidelines, efforts have been made to analyze and harmonize them [83
]. In the specific case of opioids (namely codeine) the CPIC and DPWG recommendation fit well and complement each other; recently, they were the bases for the development of the pharmacogenetics-guided passport, i.e., a panel of actionable germline genetic variants for pre-emptive pharmacogenetics testing [84
]. Furthermore, a recent study by Gammal et al. has demonstrated the successful application of this genetics-driven approach in patients. The authors describe the implementation of pharmacogenetics-based codeine prescribing that accounts for CYP2D6 metabolizer status. Clinical decision support was implemented to guide codeine prescribing in CYP2D6 UM and PM (recommending against codeine for patients with these high-risk CYP2D6 status). None of them were prescribed with codeine. Using genetics to tailor treatments had an important effect by limiting codeine use to patients who could safely receive and benefit from it [80
]. This is another proof-of-concept for safer and effective use of opioids when prescriptions are tailored on genetic background.
CYP2D6 status is also important when other medications affecting the same/other pharmacokinetic pathways are concomitantly administered. Several drugs have common metabolic pathways and combining them can alter drug kinetics and lead to lethal consequences in patients with existing comorbidities or a pre-existing UM status.
Further, since CYP450 family affects many drugs and their metabolism, whenever patients display abnormal reactions to these drugs, a genetic contribution should be always considered. Once investigated, such a contribution may unravel an unexpected abnormal response to opioids.
In these cases, or whenever drug–drug interaction may be suspected, molecules with different metabolic pathways may be a better choice. Since the available data on UGT
genetic variability failed to demonstrate any important association with opioid response [66
], morphine and hydromorphone (that are metabolized by UGT and not by CYP family members) are a good choice to reduce variability in opioid efficacy and the development of side effects, and may not predispose to interference with other compounds with different metabolic process. [85
], as well, is a good alternative. Although it is similar to tramadol, Tapentadol is not a pro-drug and does not rely on metabolism to produce an analgesic effect (no CYP interference) [85
]; this makes it a useful molecule for patients who do not respond to more commonly used opioids due to their genetic disposition. Tapentadol is an opioid and a noradrenaline re-uptake inhibitor (similar to tramadol): its pharmacogenetics may be primarily deduced from known associations of OPRM1
and COMT gene variants [86
Pharmacogenetics is a major tool for clinicians when approaching to PD of opioids. In this context, more data are available than for PK, especially for OPRM1 and COMT genes. Many open issues remain that need to be fixed in the future, but it is honest to affirm that identifying patients with some specific variants of these two genes may help in better opioid prescription.
Patients with OPRM1
A/A genotype and patients with COMT
Met/Met genotype have been invariably associated with a more favorable phenotype, also in specific cancer patients’ populations: less pain sensitivity, better response to opioids with lower consumption, less side effects [35
]. These results have been mainly shown in patients prescribed with morphine, but no data seem to exclude that they can be replicated with other types of opioids (Table 2
Further, since opioids are immunosuppressive and affect the endocrine system [3
], they might directly and/or indirectly influence long-term cancer recurrence. MOR is a major element underlying (and connecting) pain, endocrine system, and immune function: OPRM1 might be used in the near future to customize opioid therapy, avoiding not only opioid side effects but also disease progression [6
As for CYP testing, pharmacogenetics may help screening patients prone to ineffective pain relief and adverse events. The same authors proposed a simple classification (as for CYP2D6
) for clinical management of patients with OPRM1
variants based on 3 allelic variants: Normal, sub-functional, dysfunctional [7
]. When OPRM1
is subnormal or dysfunctional, the patient is less likely to benefit from opioids, and more likely to experience side effects and dependence [7
]. This approach would also change the prescribing approach whenever a patient does not get effective pain relief: escalating opioids, as well as adding other opioids may not increase analgesia if the patient has a dysfunctional OPRM1
status. Relying on non-opioid drugs or alternative therapies may be better, in this case.
Of course, combining data on pharmacokinetics and pharmacodynamics may further help in stratifying patients more prone to receive opioids. Of note, the list of gene variants affecting opioid metabolism is not restricted to OPRM1
: several other genes have been identified that code for the cell’s channels and transporters involved in opioid signaling that finally reflect in different patient’s response to a given drug (see Table 2
). These other genes act alone or in combination and may lead to different opioid consumption and efficacy, as well as to a different side effects profile. When combined, all of these variants may configure some panels for genetic testing in patients prescribed with opioids.
Since different genes regulate pain and analgesia, approaching pharmacogenetics with a multiple SNPs approach may be promising [53
]. To the same extent, genome-wide association studies (GWAS) would be required to identify all genetic factors involved in pain and opioid response, being helpful for better prescriptions. GWAS found new SNPs, whose mechanisms might not seem clear, but associated with opioid response. Recently a GWAS on more than 1000 cancer patients identified eight SNPs significantly associated (p
< 1.0 × 10−3
) with pain relief. However, rs12948783 of the RHBDF2
gene showed the best statistical association [87
]. This genetic variant rs7104613 is located on a gene encoding for a cell-adhesion protein that promotes the attachment of spinal cord and sensory neuron cells and the outgrowth of neurites in vitro. This result underlines the involvement of genes related to transmission of nerve impulse, and not opioid pharmacology, in mediating the main modulatory effect of genetic variability on the perception of pain.
A recent analysis on 4 different populations with thousands of patients identified a variant on chromosome 15, rs12442183, near RGMA
, associated with opioid dependence (p
= 1.3 × 10−8
encodes repulsive guidance molecule A, which is a central nervous system axon guidance protein. GWAS (Genome—Wide Association Studies) unraveled new lead into our understanding of pain pathophysiology [89
], confuting the traditional attention paid to genes involved in opioid pharmacology only. Other GWAS data suggest that opioid therapy outcome more likely results from interaction of both pathways [90
], also because some of these “new” genes function is still unknown or uncertain. However, GWAS studies provide promising targets that might contribute to personalized medicine.
Better prescribing and better patients’ management may also have a pharmaco-economic impact. For example, 70% of cardiac drugs showed a different response or a different risk of adverse effects based on genetics, supporting the idea that having this information before prescribing is warranted [91
]. The same was also proven for warfarin treatment [92
]. Advances in pharmacogenetics may optimize patient outcomes while reducing costs associated with inappropriate or unsafe prescribing [93
]. Even if a few studies are available, there is increasing evidence that pharmacogenetics may help reducing healthcare costs. Direct costs may be influenced by better drug’s biological activity, while indirect costs may be lowered by the prevention of side effects (especially addiction in non-cancer patients [94
]). Finally, pharmacogenetics could be useful to predict drug–drug interactions [95
], really common in cancer patients because of the concomitant intake of several different drugs (especially in slow or ultra-rapid CYP4502D6 metabolizers).
On the other hand, costs for obtaining and interpreting samples, as well as counseling patients should be accounted. However, broader data and more knowledge, combined with refined and cheaper testing, may help overcoming the obsolete “trial-and-test” philosophy, improving patients’ outcome and impacting on healthcare costs [97
]. Some proof already exists: a recent trial including pharmacogenomics information in the healthcare pathway of inpatients with high risk profile, reduced readmissions by 52%, emergency department visits by 42% and mortality by 85%, with an estimated cost savings per patient of $
4382 in just 60 days [98
The same concept may be applied to pharmacologic development in industries [99
], which may save lot of money during drug development. It is currently difficult to estimate the actual effect of introducing pharmacogenetics in the clinical day-case scenario and the most relevant tests to be used, but genetic testing may be extremely useful in overcoming the important rate-limiting step of costs in providing adequate therapies.
In the meantime, when a genomic-oriented medical approach is pursued, ethical concerns rise (as well as with any “omics” approach), especially in the field of pain medicine. Genomics (and other “omics”) technologies produced not only a deep change in biomedical research, but the potential to discover large amounts of possibly important incidental information is an issue, as well as physicians’ obligations arising from this type of the research, protection of the included individuals and personal data, and communication of such research to patients and the community [100
In the specific context of opioids, pharmacogenetics may exacerbate inequalities in the delivery of healthcare [101
]. Genetic testing may identify patients more prone to effective analgesic response, but also those at high risk of opioid adverse events, such as abuse and addiction.
This information may lead to different treatments covered by health insurance companies, since patients more prone to have “poor profile of response” would be unraveled before treatment [101
Hence pharmacogenetics could raise the following questions: who decides to perform the test: patients? Care providers? Insurance company? Once a genetic-driven prescription is pursued, how the needs of insurance companies will fit with the variability of the clinical scenario?
Finally, pharmacogenetics information can vary according to racial or ethnic origin. Discovering a genetic variant that influences a medicine’s response in a particular ethnic group, may lead to the design of medications for a specific ethnic group or for an ethnic group most likely to be able to afford the medications [102
]. This would be a concern if the ethnic group more prone to a “good response” to a drug is not equally able to afford it.
Once the concept of pharmacogenetics becomes mainstream and knowledge will increase in patients as in care physicians, one may lie or forge his profile to obtain specific drugs (and abuse of them) [103
]. This would be the nemesis of pharmacogenetics: born to improve outcome, it would be the cause of abuse and diversion.
Of course, these are potential scenarios that may configure or not, but they should be kept in mind when evaluating the potential implications of genetic profiling for opioid prescriptions, especially in cancer patients, in which ethical concerns are strictly related to life expectations.