WHO analgesic ladder & opioid titration for cancer pain — from assessment through strong-opioid dosing and rotation in palliative care
Roughly 60–90% of patients with advanced cancer experience clinically significant pain, yet worldwide surveys repeatedly find that a large fraction receive inadequate analgesia — often because assessment is rushed or skipped entirely. Every subsequent decision on the WHO analgesic ladder — which drug class, which starting dose, which adjuvant — depends on a structured, repeatable assessment that captures intensity, mechanism, temporal pattern, and the patient's own goals of care.
The 0–10 Numeric Rating Scale (NRS) is the most widely used bedside tool: patients rate current pain, average pain over the last 24 hours, and worst pain, since a single snapshot underestimates the burden of breakthrough episodes. Visual Analogue Scales (VAS) and the Wong-Baker FACES scale serve as alternatives for patients with cognitive impairment, low literacy, or pediatric populations.
PQRST structures the history systematically:
• Provocation/Palliation — what worsens or relieves the pain (movement, position, prior analgesics tried and their effect) • Quality — patient's own descriptors: "aching," "throbbing," and "cramping" suggest nociceptive somatic or visceral pain; "burning," "shooting," "electric," or "tingling" suggest a neuropathic component • Region/Radiation — anatomical site and referral pattern; dermatomal radiation strongly suggests nerve root or plexus involvement • Severity — current, average, and worst NRS scores, plus impact on function, sleep, and mood • Timing — constant background pain vs. intermittent breakthrough episodes, and whether episodes are predictable (incident, tied to movement or a procedure) or unpredictable (spontaneous)
Reassessment at every dose change is as important as the initial assessment — the ladder is not a one-time decision tree but an iterative loop of dose, observe, and adjust.
Pain mechanism determines which drug classes are likely to help, independent of where a patient sits on the WHO ladder:
• Nociceptive somatic pain — from bone, soft tissue, or joint nociceptor activation (e.g. vertebral or rib metastases); typically well-localized, aching or throbbing, and reliably opioid- and NSAID-responsive • Nociceptive visceral pain — from stretch, distension, or ischemia of hollow or solid organs (e.g. hepatic capsule distension, bowel obstruction); often poorly localized, deep, cramping, and may be referred to a distant dermatome • Neuropathic pain — from tumor infiltration or treatment-related injury to nerve, plexus, or spinal cord (e.g. brachial plexopathy from apical lung tumor, chemotherapy-induced peripheral neuropathy); burning, shooting, or electric in quality, often with allodynia or hyperalgesia • Mixed pain — the most common real-world presentation in advanced cancer, combining nociceptive and neuropathic components in a single lesion (e.g. vertebral metastasis with cord or root compression)
Neuropathic and mixed pain typically show a partial but incomplete response to opioids alone, which is why adjuvant analgesics — gabapentinoids, tricyclic or SNRI antidepressants, corticosteroids — are introduced in parallel rather than sequentially after opioid failure.
Breakthrough pain is a transient flare of pain that exceeds otherwise well-controlled background pain. Recognizing its subtype changes management:
• Incident pain — predictable, triggered by a specific activity (walking, dressing changes, coughing); can be pre-treated with a rapid-onset opioid dose 20–30 minutes before the trigger • Spontaneous (idiopathic) pain — unpredictable, no identifiable trigger; managed reactively with an as-needed rescue dose • End-of-dose failure — pain recurring predictably before the next scheduled background dose is due; signals that the background regimen's dose or interval needs adjustment, not simply more rescue doses
Validated tools such as the Alberta Breakthrough Pain Assessment Tool structure the history of episode frequency, intensity, onset speed, duration, and precipitating factors, and should be repeated whenever the background regimen changes.
The WHO estimates that in many low- and middle-income countries, opioid consumption remains far below the level required for adequate cancer pain control — a gap driven less by clinical uncertainty than by regulatory restriction, supply-chain limits, and prescriber reluctance ("opiophobia"). Assessment quality closes only half the gap; access to the medicines it recommends closes the rest.
For mild pain (NRS 1–3), the WHO ladder begins not with an opioid but with a non-opioid analgesic — paracetamol (acetaminophen) and/or an NSAID — combined from the outset with an adjuvant matched to the pain mechanism identified during assessment. Step 1 is frequently under-utilized: clinicians move straight to opioids for cancer pain, missing a genuinely effective, low-side-effect option, particularly for the somatic bone-pain component that responds well to NSAIDs.
Paracetamol (acetaminophen) acts centrally, likely via inhibition of prostaglandin synthesis in the CNS and modulation of the endocannabinoid system; its exact mechanism remains incompletely defined despite over a century of clinical use. Standard dosing is 500–1000 mg every 4–6 hours, capped at 4 g/day in adults, reduced to 2–3 g/day in hepatic impairment, low body weight, or chronic alcohol use — relevant in cachectic cancer patients.
NSAIDs (ibuprofen, naproxen, diclofenac, celecoxib) inhibit cyclooxygenase (COX-1/COX-2), reducing prostaglandin-mediated sensitization of peripheral nociceptors — making them particularly effective for bone metastasis pain, where prostaglandin release from osteolytic activity is a key driver. Cancer-specific cautions are significant: thrombocytopenia (chemotherapy-induced) raises bleeding risk with any NSAID; concurrent corticosteroid use (common in palliative care for appetite, nausea, or peritumoral edema) roughly doubles GI bleeding risk; renal impairment, dehydration, and concurrent nephrotoxic chemotherapy (cisplatin) raise acute kidney injury risk. COX-2 selective agents (celecoxib) reduce GI risk but retain renal and cardiovascular concerns.
Adjuvants are introduced at Step 1 rather than reserved for opioid failure, because they treat mechanisms opioids address poorly:
• Corticosteroids (dexamethasone 4–8 mg/day) — reduce peritumoral edema and inflammatory mediator release; first-line for bone pain, nerve or spinal cord compression, raised intracranial pressure, and hepatic capsule distension; benefit must be weighed against hyperglycemia, myopathy, and infection risk with prolonged use • Bone-modifying agents (zoledronic acid, denosumab) — reduce skeletal-related events and bone pain in metastatic disease by inhibiting osteoclast-mediated bone resorption; onset of analgesic benefit takes weeks, so they complement rather than replace acute analgesia • Gabapentinoids (gabapentin, pregabalin) — first-line for neuropathic pain; modulate the α2δ subunit of voltage-gated calcium channels, reducing neuronal hyperexcitability; titrated slowly to limit sedation and dizziness • Tricyclic antidepressants (nortriptyline) or SNRIs (duloxetine) — an alternative or adjunct for neuropathic pain, particularly useful when comorbid depression or chemotherapy-induced peripheral neuropathy is present
Radiotherapy remains one of the most effective single interventions for localized bone metastasis pain, with response rates of 60–80% and complete relief in a substantial minority, typically within 1–4 weeks. Nerve blocks (e.g. celiac plexus block for pancreatic cancer pain) and vertebroplasty/kyphoplasty for vertebral compression fractures offer targeted, opioid-sparing relief. Psychological support, physical therapy, and patient/family education about pain as a treatable symptom (not an inevitable feature of cancer) meaningfully improve outcomes independent of drug therapy.
Escalation to Step 2 or directly to low-dose Step 3 is indicated when: pain remains NRS ≥4 despite an adequately dosed and adequately trialed (24–72h) Step 1 regimen with appropriate adjuvant; pain is rapidly worsening; or breakthrough episodes are frequent and severe despite optimized background therapy. The ladder is not meant to be climbed slowly through under-treatment — severe pain at presentation justifies starting higher.
Moderate pain (NRS 4–6) unresponsive to optimized Step 1 traditionally adds a weak (“mild-to-moderate”) opioid — codeine or tramadol — on top of the non-opioid base. In practice this step has become the most debated rung of the ladder: pharmacogenomic variability in codeine metabolism and a low analgesic ceiling have led many contemporary guidelines to favor low-dose strong opioids instead, especially where strong opioids are readily available and affordable.
Codeine is itself a weak mu-opioid agonist but functions largely as a prodrug: hepatic CYP2D6 O-demethylates roughly 10% of a codeine dose into morphine, which accounts for most of its analgesic effect. This creates striking inter-patient variability: CYP2D6 poor metabolizers (~7–10% of people of European descent, with wide variation by ethnicity) derive minimal analgesia even at maximal doses, while ultra-rapid metabolizers can experience unexpectedly pronounced morphine-like effects and toxicity — a risk that led regulators to contraindicate codeine in breastfeeding mothers and children after fatal neonatal morphine toxicity cases.
Tramadol has a dual mechanism: a weak mu-opioid agonist effect (again partly dependent on CYP2D6 activation of its active metabolite, O-desmethyltramadol) plus serotonin and norepinephrine reuptake inhibition, giving it a genuinely useful role in mixed nociceptive-neuropathic pain. Its serotonergic activity mandates caution when combined with SSRIs, SNRIs, or MAO inhibitors due to serotonin syndrome risk, and it lowers seizure threshold more than other opioids.
Weak opioids are started at the lower end of the dosing range (e.g. codeine 30 mg every 4–6 hours, tramadol 50 mg every 6 hours) and titrated upward against ongoing NRS scores and breakthrough consumption, continuing the Step 1 non-opioid and any indicated adjuvant concurrently — the ladder is additive, not a drug-substitution sequence. Because weak opioids have a relatively low analgesic ceiling (further dose increases beyond the maximum yield disproportionately more side effects than pain relief), persistent pain at or near the maximum recommended dose is the trigger to move to Step 3 rather than to keep escalating a weak opioid indefinitely.
Common side effects mirror strong opioids in kind if not degree — constipation, nausea, sedation — and should be anticipated and managed proactively (see Stage 5) rather than treated only after they emerge.
The WHO's 2018 update to its cancer pain guidance formally endorsed an alternative pathway: for moderate-to-severe cancer pain, clinicians may move directly from Step 1 to a low starting dose of a strong opioid (e.g. oral morphine 2.5–5 mg every 4 hours) rather than trialing a weak opioid first. The evidence review underlying this change found no consistent analgesic advantage for weak opioids over low-dose strong opioids, while low-dose strong opioid regimens avoid the CYP2D6 variability problem, offer a smoother and more predictable titration curve toward effective doses, and eliminate the need for an intermediate drug switch.
In practice, the choice between a traditional three-step climb and a two-step (skip Step 2) approach often comes down to local prescribing culture, regulatory access to strong opioids, and patient/family comfort with the word "morphine" — which in many settings still carries a stigma of imminent death that requires explicit counter-education.
Severe pain (NRS ≥7), or pain that has failed an adequately dosed Step 2 trial, is managed with a strong opioid — most commonly oral morphine, oxycodone, or transdermal fentanyl. Step 3 is where the greatest technical skill in palliative pharmacotherapy lives: choosing a starting dose, titrating safely against breakthrough consumption, converting between routes and drugs using equianalgesic ratios, and sizing breakthrough (rescue) doses correctly.
In an opioid-naive patient, immediate-release oral morphine is typically started at 5–10 mg every 4 hours, with an equivalent breakthrough (rescue) dose available every 1 hour as needed — patients starting an opioid should never be left without a rescue option. Elderly, frail, renally impaired, or opioid-sensitive patients are started at the lower end (e.g. 2.5–5 mg q4h).
After 24–48 hours, the regimen is reviewed: total opioid consumed (scheduled doses plus rescue doses) is summed, and if pain control remains inadequate, the scheduled dose is increased by roughly 25–50% — a proportional increase, not a fixed increment, since the same 5 mg step means little at a 200 mg/day baseline. Once the 24-hour requirement stabilizes, the patient is converted from an immediate-release q4h schedule to an equivalent sustained-release (long-acting) formulation dosed every 8–24 hours depending on the product, which improves adherence and overnight symptom control, while immediate-release breakthrough doses continue to be available on top.
Switching opioid or route requires converting the current total daily dose into an equianalgesic dose of the new agent — approximate ratios (individual variation is substantial, and conservative rounding down is standard practice):
Oral morphine is the reference standard (ratio 1:1). Oral oxycodone is roughly 1.5–2× more potent than oral morphine by weight (e.g. 30 mg oral morphine ≈ 15–20 mg oral oxycodone). Transdermal fentanyl is dosed in mcg/hr and converted from oral morphine at approximately 100:1 (mg/day oral morphine : mcg/hr fentanyl), so 60 mg/day oral morphine corresponds to roughly a 25 mcg/hr patch — conversion tables vary and cross-checking two independent sources before a switch is standard safe practice. Parenteral (IV/SC) morphine is roughly 2–3× more potent than oral morphine because it bypasses first-pass hepatic metabolism.
Transdermal fentanyl patches carry an FDA and international black-box-equivalent warning against use in opioid-naive patients: because the patch's slow onset (12–24h to steady state) and long elimination half-life prevent rapid dose adjustment, initiating fentanyl in a patient without established opioid tolerance has caused fatal respiratory depression. Fentanyl patches are for patients already stabilized on a known equivalent daily opioid dose — never a first-line strong opioid.
The standard rule of thumb sets each breakthrough (rescue) dose at 10–15% of the total 24-hour scheduled opioid dose, given as an immediate-release formulation of the same or a pharmacokinetically similar opioid. For a patient on 120 mg/day of oral morphine equivalent, a reasonable breakthrough dose is roughly 12–18 mg, available as needed.
Redosing intervals depend on route and time-to-peak effect: oral immediate-release morphine or oxycodone can be redosed roughly hourly if needed (time to peak effect ~60 minutes); subcutaneous or intravenous doses act within 15–30 minutes and can be redosed accordingly; transmucosal/sublingual fentanyl products act within minutes and have their own separate titration schedule, not derived by the 10–15% rule, because of fentanyl's distinct absorption kinetics.
Consistent use of more than 3–4 breakthrough doses per day is a signal that the background (scheduled) dose is undertreated and should be increased at the next review — rescue-dose frequency is itself one of the most useful titration signals available.
Long-term opioid therapy for cancer pain is rarely a "start and forget" prescription. Tolerance develops to some effects (notably respiratory depression and euphoria) but not others (notably constipation), toxicity can emerge as doses climb, and a substantial minority of patients need to be switched — "rotated" — to a different opioid to regain analgesic control or resolve intolerable side effects. Managing this phase safely requires distinguishing expected pharmacological adaptation from true addiction, and treating side effects prophylactically rather than reactively.
Constipation is near-universal and does not improve with tolerance the way sedation or nausea do — it requires prophylactic treatment from the first opioid prescription, not reactive treatment once it occurs. A stimulant laxative (senna, bisacodyl) with or without an osmotic agent is first-line; bulk-forming fiber agents are specifically discouraged in opioid-induced constipation because they can worsen obstruction symptoms in a gut with slowed motility. Peripherally acting mu-opioid receptor antagonists (methylnaltrexone, naloxegol) are reserved for refractory cases, since they reverse opioid effects on the gut without crossing into the CNS to affect analgesia.
Nausea occurs in roughly 30% of patients starting an opioid, typically resolving within 3–7 days as tolerance develops; a short course of an antiemetic (metoclopramide, haloperidol) covers this window without needing to abandon the opioid. Sedation is usually most pronounced at initiation or after a dose increase and improves with tolerance over days; persistent sedation warrants review for accumulation (especially in renal impairment, where morphine's active metabolites accumulate) before assuming it is simply expected.
Escalating doses without proportionate analgesic benefit can produce a neurotoxicity syndrome rather than better pain control: myoclonus (involuntary muscle jerks), hyperalgesia (pain that paradoxically worsens with more opioid), vivid dreams or hallucinations, and delirium. This is more common with morphine at high doses in renal impairment, where the active metabolite morphine-6-glucuronide accumulates, but can occur with any opioid at sufficiently high or rapidly escalated doses.
Recognizing this syndrome is critical because the intuitive response — increasing the dose further to chase uncontrolled pain — worsens it. The correct response is opioid rotation to a different agent at a reduced equianalgesic dose, aggressive hydration where appropriate, and treatment of myoclonus if severe (benzodiazepine) while the rotation takes effect.
Rotation is indicated for: inadequate analgesia despite appropriate titration; intolerable or dose-limiting side effects; neurotoxicity as above; or a need to change route (e.g. patient can no longer swallow, requiring conversion to transdermal or subcutaneous delivery).
Protocol: (1) calculate the current total 24-hour dose of the existing opioid; (2) convert to the equianalgesic dose of the new opioid using a published conversion table; (3) reduce this calculated dose by 25–50% to account for incomplete cross-tolerance — tolerance built up to one opioid does not fully transfer to a different mu-receptor agonist, so a "full" equianalgesic dose of the new drug can cause overdose; (4) titrate the new opioid upward from this reduced starting point using the same 25–50% titration principle as initial Step 3 dosing; (5) provide an appropriately sized breakthrough dose of the new agent from the outset.
Three distinct phenomena are frequently and harmfully conflated in opioid prescribing:
• Tolerance — a pharmacological adaptation where a given dose produces a diminishing effect over time, requiring dose increases to maintain analgesia; expected and not pathological • Physical dependence — the body's adaptation such that abrupt discontinuation produces a withdrawal syndrome; also expected with regular use and managed by tapering rather than abrupt cessation, not evidence of addiction • Addiction (opioid use disorder) — a behavioral disorder characterized by compulsive use despite harm, loss of control over use, and craving; large studies of appropriately monitored cancer-pain opioid therapy consistently find true addiction rates below 1%
"Pseudo-addiction" describes drug-seeking behavior that resembles addiction but is actually driven by undertreated pain — the behavior resolves once analgesia is adequately titrated, distinguishing it from true addiction, where escalating access does not resolve the underlying craving. This distinction directly affects clinical decision-making: assuming addiction in an undertreated patient leads to further undertreatment, worsening the very behavior that raised concern.