Why don’t painkillers work for everyone? A guide to pharmacogenetics-based chronic pain treatment

Chronic pain is notoriously hard to treat. People can live with pain for months or years while trying one remedy after another, and many get only partial relief even with good treatment. One reason is that medications are not matched to each person's genetic makeup, which affects how they are broken down. Pharmacogenetic tests give prescribers a way to personalize chronic pain treatment and get better results. This article reviews the main treatment options for different types of pain and how pharmacogenetic tests like Pillcheck can help.

Pain medications include opioids, nonsteroidal anti-inflammatory drugs (NSAIDs), antiepileptics, tricyclic antidepressants (TCAs) and serotonin-norepinephrine reuptake inhibitors (SNRIs). Physical rehabilitation, physiotherapy, cognitive behavioural therapy (CBT) and acupuncture can also improve quality of life for people with chronic pain. Medical cannabis is being studied as another option, but the evidence for cannabinoids in chronic pain is still limited.

Each type of treatment has its own side effects, summarized in the chart below.

How do genes affect opioids?

Common opioids include codeine (Tylenol No. 3 and No. 4), tramadol, hydrocodone, oxycodone, morphine, hydromorphone and fentanyl. Opioids are used for short-term pain around surgery and for cancer and palliative care, but for chronic pain they should be limited to pain that does not respond to other treatments. Long-term use can increase sensitivity to pain and lead to higher doses, with a greater risk of overdose and addiction. In the SPACE randomized trial of people with chronic back pain or hip or knee osteoarthritis, opioids were no better than non-opioid medications for pain-related function over 12 months, while pain intensity was better and side effects were fewer with non-opioid medications.

Opioids fall into two groups:

1. Prodrugs, such as codeine, tramadol and, in part, hydrocodone, which must be converted by the liver enzyme CYP2D6 into stronger forms to relieve pain.
2. Directly acting opioids, such as morphine, hydromorphone, oxycodone and fentanyl, which do not need CYP2D6 activation. Oxycodone is partly broken down by CYP2D6, but it works mostly as it is, and current guidelines do not recommend gene-based dosing for it.

Both groups come in fast-release forms for breakthrough pain and slow-release forms for longer relief.

Genes and other medications both affect the response to opioids. According to the 2021 CPIC opioid guideline, people with little or no CYP2D6 activity (poor metabolizers) get little relief from codeine and tramadol, while ultrarapid metabolizers convert them too quickly and are at risk of serious toxicity. Both groups should avoid codeine and tramadol. Medications that block CYP2D6, such as some antidepressants, can have the same effect as poor metabolism. If such a medication is stopped while someone is taking codeine or tramadol, opioid levels can rise suddenly. In a 2019 clinical trial, people with chronic pain who were CYP2D6 intermediate or poor metabolizers had better pain control when their medications were chosen using their CYP2D6 results. Learn more in Pain and depression drugs not working? and Tylenol 3 Side Effects: Why Codeine Doesn't Work for Everyone.

Pillcheck also reports the opioid receptor gene OPRM1. Opioids need extra caution in people with a history of mental health or substance use problems. In older adults, opioids can cause confusion (delirium) and constipation, and NSAIDs and tricyclic antidepressants also need caution, so pharmacogenetic results can help guide safer choices.

How do genes affect NSAIDs such as ibuprofen?

Nonsteroidal anti-inflammatory drugs, such as ibuprofen, naproxen, celecoxib, diclofenac and meloxicam, reduce pain, fever and inflammation and are a first choice for inflammatory pain. Long-term use increases the risk of stomach bleeding, high blood pressure, heart attack, stroke and kidney problems.

Many NSAIDs are broken down by the CYP2C9 enzyme. The 2020 CPIC guideline for NSAIDs recommends:

• Celecoxib, ibuprofen, flurbiprofen and lornoxicam: usual starting dose for normal metabolizers. Some intermediate metabolizers should start at the lowest dose. Poor metabolizers should start at 25 to 50% of the lowest dose or use another medication.
• Meloxicam: some intermediate metabolizers should start at half the lowest dose or use another medication, and poor metabolizers should use another medication.
• Piroxicam and tenoxicam: these stay in the body a very long time, so intermediate and poor metabolizers should use another medication.

Naproxen and diclofenac are not significantly affected by CYP2C9 genes. Learn more in Are you a Tylenol or Advil person?

Do genes affect gabapentin, pregabalin and carbamazepine?

Gabapentin (Neurontin) and pregabalin (Lyrica) are used for nerve pain, such as after shingles or from diabetes, but they help only some people, and less so for other types of nerve pain. Both are cleared by the kidneys without being broken down by the liver, so liver enzyme genes do not affect them. Their dose depends on kidney function.

Other antiepileptics, such as carbamazepine, are used for some nerve pain conditions like trigeminal neuralgia. Carbamazepine can rarely cause a severe skin reaction called Stevens-Johnson syndrome. The risk is linked to immune genes: HLA-B*15:02, more common in people of Asian ancestry, and HLA-A*31:01, found in many populations including people of Japanese and European ancestry. Pillcheck does not report HLA genes (see the FAQ below).

How do genes affect tricyclic antidepressants?

Tricyclic antidepressants (TCAs) include amitriptyline (Elavil), nortriptyline, desipramine, imipramine, doxepin and others. Amitriptyline and other TCAs can help with both depression and chronic pain, but they often cause side effects such as dry mouth, constipation, urinary problems, drowsiness and dizziness. TCAs are broken down by CYP2C19 and CYP2D6. At the low doses used for pain, the CPIC guideline for TCAs recommends no dose change for poor or intermediate metabolizers, with close monitoring for side effects, while CYP2D6 ultrarapid metabolizers may get too little effect and should consider another medication. Dose adjustments based on genes matter more at the higher doses used for depression.

How do genes affect SNRIs such as duloxetine and venlafaxine?

SNRIs such as duloxetine (Cymbalta), venlafaxine (Effexor XR), desvenlafaxine (Pristiq) and levomilnacipran (Fetzima) are antidepressants that can also help with chronic pain, including fibromyalgia and other pain linked to the central nervous system, especially in people who also have depression. According to the 2023 CPIC guideline:

• Venlafaxine depends on CYP2D6. For poor metabolizers, CPIC recommends considering another medication, such as desvenlafaxine, which is broken down by CYP3A4 rather than CYP2D6.
• Duloxetine is broken down by CYP1A2 and CYP2D6. CYP2D6 poor or ultrarapid metabolizers may have more side effects or a weaker response, but current guidelines do not give specific dose adjustments.
• Levomilnacipran is broken down mainly by CYP3A4.

Antibiotics such as clarithromycin, some antifungals and grapefruit juice can block CYP3A4 and raise levels of medications it breaks down, while phenobarbital, phenytoin, rifampin, St. John's wort and glucocorticoids can speed it up and reduce their effect. Learn more in Most effective treatments for fibromyalgia.

Can cannabidiol (CBD) be combined with pain medications?

Cannabidiol (CBD) is approved for some rare forms of epilepsy and is being studied for chronic pain. CBD is generally well tolerated, while products containing THC can cause psychoactive side effects. CBD is broken down by several liver enzymes and can block others, so it interacts with many medications. Talk to an expert pharmacist before adding medical cannabis to other medications.

How can Pillcheck help?

Pharmacogenetic tests such as Pillcheck show how your body processes many pain medications. Pillcheck includes a medication review by clinical pharmacists trained in pharmacogenetics, who recommend how to optimize the treatment plan. Pillcheck can be reimbursed through some workplace benefits plans and may be eligible for coverage through disability claims services.

Pillcheck testing is performed in CAP- and COLA-accredited clinical genetic laboratories located in Canada. Pillcheck reports on 24 genes, increasing coverage with new guidelines and drug labels (see the genes tab), and covers 355 medications, which account for over 70% of dispensed prescriptions. The Pillcheck test panel ensures high accuracy for people of non-European and admixed ancestries. Your results last a lifetime, and report updates are free of charge.

Summary

• For chronic pain, non-opioid medications and non-drug treatments should come first, and opioids should be a last option.
• How well pain medications work, and how well they are tolerated, depends on genes, other medications and health conditions.
• Pharmacogenetic testing helps prescribers choose the pain medications most likely to be effective and safe for each person.

Frequently asked questions

Why don't painkillers work the same for everyone?

Genes affect how quickly the liver breaks down many pain medications. Other medications, age and health conditions also play a role.

Which painkillers are affected by genes?

Codeine and tramadol are strongly affected by CYP2D6, and several NSAIDs, such as celecoxib, ibuprofen, meloxicam and piroxicam, by CYP2C9. Tricyclic antidepressants and venlafaxine are also affected. Gabapentin and pregabalin are not.

Are opioids a good choice for chronic pain?

Usually not as a first choice. In a randomized trial, opioids were no better than non-opioid medications for pain-related function in chronic back pain or hip or knee arthritis, and they caused more side effects.

Can a pharmacogenetic test help with chronic pain?

Yes. In a clinical trial, people who were CYP2D6 intermediate or poor metabolizers had better pain control when their medications were chosen using their test results.

Does Pillcheck report HLA genes?

No. HLA genes are linked to rare but serious allergic reactions to some medications, such as severe skin reactions (Stevens-Johnson syndrome, or SJS) and DRESS. Clinical guidelines (CPIC) recommend avoiding certain drugs in people who carry specific HLA variants. Examples are abacavir (HLA-B*57:01), carbamazepine, oxcarbazepine and phenytoin (HLA-B*15:02), carbamazepine (HLA-A*31:01) and allopurinol (HLA-B*58:01).

We don't include HLA results in Pillcheck reports because these results can easily be misread:

• Carrying a risk variant doesn't mean you will react. Most people who carry one take the medication without problems.
• Not carrying a known variant doesn't mean you're safe. People without any known risk variant can still have a severe reaction. A 2025 US study found that more than one-third of people with severe allopurinol reactions did not carry HLA-B*58:01, the variant usually tested for. Among Black patients in the study, the figure was 45%. Many risk variants haven't been discovered yet, especially in populations that have been studied less.
• Most tests check only a few variants. HLA genes are among the most variable in the body, with thousands of known versions. A test for one variant can miss closely related ones that also carry risk.

A negative HLA result could therefore give false reassurance. Whatever your genetic results, watch for early warning signs after starting a new medication, such as a rash, fever, blistering or sores in the mouth. Contact a healthcare provider right away if they appear.

If your doctor recommends HLA testing before you start a specific medication, they can order it through a clinical lab. If you've had a reaction before, an allergist or dermatologist can advise whether tests such as patch testing could help identify the cause.

Related

• Learn more about Pillcheck for chronic pain
• Pillcheck in your practice

References

1. Crews KR, Monte AA, Huddart R, et al. Clinical Pharmacogenetics Implementation Consortium Guideline for CYP2D6, OPRM1, and COMT Genotypes and Select Opioid Therapy. Clin Pharmacol Ther. 2021;110(4):888-896.
2. Theken KN, Lee CR, Gong L, et al. Clinical Pharmacogenetics Implementation Consortium Guideline (CPIC) for CYP2C9 and Nonsteroidal Anti-Inflammatory Drugs. Clin Pharmacol Ther. 2020;108(2):191-200.
3. Hicks JK, Sangkuhl K, Swen JJ, et al. Clinical Pharmacogenetics Implementation Consortium guideline (CPIC) for CYP2D6 and CYP2C19 genotypes and dosing of tricyclic antidepressants: 2016 update. Clin Pharmacol Ther. 2017;102(1):37-44.
4. Bousman CA, Stevenson JM, Ramsey LB, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) Guideline for CYP2D6, CYP2C19, CYP2B6, SLC6A4, and HTR2A Genotypes and Serotonin Reuptake Inhibitor Antidepressants. Clin Pharmacol Ther. 2023;114(1):51-68.
5. Krebs EE, Gravely A, Nugent S, et al. Effect of Opioid vs Nonopioid Medications on Pain-Related Function in Patients With Chronic Back Pain or Hip or Knee Osteoarthritis Pain: The SPACE Randomized Clinical Trial. JAMA. 2018;319(9):872-882.
6. Smith DM, Weitzel KW, Elsey AR, et al. CYP2D6-guided opioid therapy improves pain control in CYP2D6 intermediate and poor metabolizers: a pragmatic clinical trial. Genet Med. 2019;21(8):1842-1850.
7. Sostres C, Lanas A. Appropriate prescription, adherence and safety of non-steroidal anti-inflammatory drugs. Med Clin (Barc). 2016;146(6):267-272.