My clinician started Neurontin 300 mg for back pain. Should I assume it will fail?
Read the two stamps before you trust the hallway
Current US Neurontin label carries two jobs. Everything else is a different claim.
Adult postherpetic neuralgia. Adjunctive treatment of partial-onset seizures, with or without secondary generalization, in adults and in children 3 years and older. That is the licensed pair. A 300 mg capsule is how many of those courses begin. It is not a third indication.
Alpha-2-delta occupancy reduces excitatory release at overactive terminals. That mechanism has a fighting chance in injured sensory axons and in some cortical seizure circuits. It has little to say to a disc, a facet joint, or a Sunday-night worry. Trial results follow that split more honestly than clinic folklore does.
I keep the history of how the hallway got so long on a separate page: how a GABA mimic became Neurontin 300 mg. This page is the evidence card. Handling and renal math sit in the Neurontin 300 mg medicine page.
PHN: two trials, a 300 mg day-one start, a real stamp
Best dataset on the pain side is postherpetic neuralgia. Two randomized, double-blind, placebo-controlled multicenter studies. Intent-to-treat 563 adults. Entry required pain lasting more than three months after the herpes zoster rash healed. That is not 'a bit of lingering ache.' That is a defined nerve-injury syndrome.
Label titration for adult PHN starts with one 300 mg capsule on day one, 600 mg on day two as 300 mg twice, 900 mg on day three as 300 mg three times, then climb as needed toward 1800 mg per day (600 mg three times daily). Efficacy in the studies covered 1800 to 3600 mg per day. Additional benefit above 1800 mg was not demonstrated. Dizziness and somnolence were the common reasons people left the drug arm.
Honest margin: not a cure. Partial relief is the usual win. Some people get little and pay for fog. In a condition that otherwise grinds on, a labeled option with that dataset is still a real option. It is one tool among several, not a default for every post-viral twinge.
Add-on partial-onset seizures: the 1993 job, still the first job
The first stamp was epilepsy. It remains the cleaner of the two licensed lanes.
Effectiveness in people 12 and older rests on three add-on trials in 705 patients, plus a pediatric add-on trial in 247 children aged 3 to 12. Enrolled patients still had frequent partial seizures despite one or more antiepileptics at therapeutic levels. The main readouts were a 50% responder rate and a response-ratio change from a 12-week baseline (6 weeks in the pediatric study).
Adults and adolescents start at 300 mg three times daily. Maintenance on the label is 300 to 600 mg three times daily. Doses up to 2400 mg per day were tolerated in longer work. 3600 mg per day appears only in a small, shorter set. Do not leave more than 12 hours between doses. This is add-on work, not a first monotherapy pitch.
Newer antiepileptics often displace gabapentin because three-times-daily dosing and saturable absorption are inconvenient. The niche that remains is real: almost no hepatic metabolism, few CYP collisions, a 300 mg unit that is easy to count. That is a reason to keep it in a polytherapy pile. It is not a reason to stretch it into psychiatry.
Painful diabetic neuropathy: a fair try, not a Neurontin stamp
Painful diabetic peripheral neuropathy is the next room people walk into. Support is better than for mechanical backache and thinner than for PHN. Pregabalin and duloxetine carry specific US labels for that job. Gabapentin is widely used anyway. I treat it as a time-limited trial with a pain score, not as a forever bottle.
Broader 'peripheral neuropathic pain' reviews show the usual nerve-drug pattern. A minority get a worthwhile drop. Many get dizziness and a small change. If the story is burning, shooting, or allodynia in a nerve territory, a 300 mg start and a climb can be reasonable. If the story is 'my back is stiff,' you are already in the wrong room.
Reassess. Do not assume an ongoing effect because the capsule is familiar. No change after a fair dose in a person with fair kidneys is a reason to taper, not a reason to add a fourth bottle.
Mechanical back pain and sciatica: the molecule is the wrong tool
This is the loudest gap between drawer volume and trial support. Non-specific low back pain reviews find little or no benefit over placebo, with dizziness and sedation still attached. A large pregabalin sciatica trial was negative. Same target class. Same disappointment. Alpha-2-delta drugs do not have a mechanism for a mechanical joint problem.
Radicular leg pain is a slightly kinder story and still uneven. I do not panic when I see a 300 mg start on a discharge list. I do ask what neuropathic feature the prescriber thought they were treating, and I write a review date. A capsule that only makes a person unsteady is not 'covering the pain.'
File: if you cannot name a nerve-injury mechanism, gabapentin is a sedative experiment. Experiments need an end date. Safety numbers for that experiment live in the eGFR-first 300 mg titration note.
Mood, sleep, migraine: leftover scripts from a louder decade
Most of these lines grew in the marketing years. The datasets did not grow with them.
Bipolar disorder, generalized anxiety, primary insomnia, migraine prevention, alcohol-use disorder, fibromyalgia, and ordinary chronic ache sit on thin or negative ground. Sedation can feel like help at 10 p.m. That is not the same as an anxiolytic or hypnotic indication. I say that to people who are afraid I am calling their relief imaginary. The fog can be real. The indication is still missing.
A 300 mg night dose for 'sleep' is one of the commonest leftover patterns I see. It may knock someone out. It also accumulates when eGFR slides, and it still needs a taper if you later decide the job was never there. Do not confuse a side effect with a diagnosis. If the only win is drowsiness, say that out loud on the chart so the next clinician does not inherit a fake nerve-pain indication.
A card that sorts the uses without a speech
Same 300 mg capsule. Different verdicts, because the mechanism only fits some problems.
| Use | What I write in the margin | What the file actually shows |
|---|---|---|
| Adult PHN | Strong, labeled | Two RCTs, n=563 ITT; 2002 stamp; 300 mg day-one start |
| Partial-onset seizures, add-on | Strong, labeled | 1993 adult stamp, later ages 3+; 300 mg TID start if 12+ |
| Painful diabetic neuropathy | Fair try, not stamped | Used often; other drugs hold the specific US labels |
| Other peripheral nerve pain | Uneven | A minority improve; many pay for fog |
| Low back pain / sciatica | Weak to negative | Little or no benefit vs placebo; pregabalin sciatica RCT failed |
| Mood, primary insomnia, migraine | Unproven leftover | Off-label habit from the promotion years |
I keep this table on the blotter when a chart lists four reasons for one bottle. Approved rows rest on RCTs and a stamp. The hallway rows rest on habit. Read down the middle column before you renew.
Pregabalin is a sibling, not a 300 mg swap
| Feature | Gabapentin (Neurontin 300 mg unit) | Pregabalin |
|---|---|---|
| Target | Alpha-2-delta | Alpha-2-delta |
| Absorption | Saturable; fraction falls as daily mg rise | Near-complete, closer to linear |
| Usual rhythm | Three times daily | Often twice daily |
| US federal schedule | Not scheduled (some states watch it) | Schedule V |
| Swap | Not mg-for-mg | Not mg-for-mg |
Shared target. Overlapping nerve-pain and seizure evidence. Different gut. Gabapentin bioavailability falls as the daily total climbs - about 60% at 900 mg per day, about 47% at 1200, about 34% at 2400. Pregabalin stays nearer to complete and linear. That is why clinic talk about 'just switching to Lyrica at a third of the milligrams' is folklore, not a conversion table.
Gabapentin is usually three times daily. Pregabalin is often twice. Pregabalin is Schedule V federally. Gabapentin is watched in some states and not federally scheduled. Cost still pushes many people toward the older 300 mg capsule. Predictability pushes others toward pregabalin. Neither wins every chart.
How I reassess a 300 mg script that never named a target
Ask what symptom should move, by when, and how you will know. For PHN, that is a pain score after a climb to a labeled range in a person whose eGFR can hold it. For add-on seizures, that is a seizure calendar, not a vibe. For leftover uses, the honest answer is often 'we started it because it is familiar.' Familiar is not a trial endpoint.
If the target was never neuropathic and the person is unsteady, I argue for a taper, not a refill. If the target is labeled and the person is still climbing through 300 mg steps, I argue for patience through the titration - day one is not an efficacy test. Abrupt stop is the other error. Withdrawal symptoms and, in epilepsy, extra seizures are on the label.
Patient wording: MedlinePlus gabapentin. Label text: DailyMed Neurontin.
What I write after the evidence pass
Narrow and labeled, gabapentin earns a place. Broad and hopeful, it mostly donates dizziness. The 300 mg capsule is a starting unit, not a diagnosis. PHN and add-on partial-onset seizures have RCTs and stamps. Diabetic nerve pain is a supervised try. Mechanical back pain, primary insomnia, and mood scripts are where I want a better reason than 'we always use this.'
Set a review. Keep the kidney number in the same sentence as the milligrams. If the bottle is not earning its fog, plan a taper with the prescriber. Do not turn a weak indication into a lifelong habit because the yellow 300 mg capsule looks harmless on a discharge list.
Reader mail
Reader questions on this article
Answered by Dr. Helena Vasquez, PharmD · Clinical pharmacology & drug safety
After the evidence card went up, readers circled the uses that still confuse them.
Assume nothing until you name the pain. Mechanical low-back ache has weak evidence and a real dizziness bill. Radicular shooting pain is a slightly better fit and still uneven. I would ask what neuropathic feature they are targeting and how many weeks they want before a review. A 300 mg start is only a first brick. If nothing moves after a fair climb in a person with fair kidneys, there is no prize for staying on it. Taper with the prescriber. Do not stop overnight.
Is anything about gabapentin actually proven, or is the whole bottle off-label?
A proven pair exists. FDA stamps: adult postherpetic neuralgia (2002) and add-on partial-onset seizures from age 3 (first adult stamp 1993). Both rest on randomized add-on or placebo-controlled work. Painful diabetic neuropathy is a common, reasonable try without a Neurontin-specific stamp. Back pain, primary insomnia, and mood uses do not sit in that pair. Trust the stamped jobs most. Primary wording is on FDA and DailyMed, not on a comment thread.
Why do so many people get a 300 mg capsule for things the label never mentions?
Habit plus a 1990s promotion campaign that outlived the guilty plea. A 300 mg start looks small. Perceived risk is low. Those two facts lower the bar to 'try it.' Prescribing volume is not a systematic review. When a chart lists four reasons for one bottle, I ask which reason has a trial. Often only one does, and sometimes none do. That is the hallway this note is about.
For nerve pain, should I push for pregabalin instead of staying on Neurontin 300 mg steps?
Not automatically. Same target. Similar nerve-pain story. Pregabalin is easier to absorb and often twice daily. Gabapentin is cheaper and starts with a familiar 300 mg capsule, but the fraction absorbed falls as the daily total climbs. Insurance and cost decide more switches than pharmacology does. There is no clean milligram bridge. If you change, let the prescriber write a new plan. Do not third-the-dose at the kitchen table.
I take 300 mg at night for anxiety and I sleep. Are you saying that is imaginary?
The sleepiness can be real. The anxiety indication is not. Sedation is a labeled adverse effect, not a stamp for generalized anxiety or primary insomnia. If you and your prescriber still think the trade is worth it, that is a monitored off-label choice - not a proof the molecule is an anxiolytic. Do not stop cold if you later want off. Withdrawal is on the post-marketing list: anxiety, insomnia, nausea, sweating. Review the job on a calendar, not on a single good night.
If I cannot tell that the 300 mg is doing anything, do I just quit?
You raise it. You do not quit in one step. Unclear benefit after a fair trial is a good reason to plan a wean, especially when the original job was a weak-evidence use. Abrupt stop can bring withdrawal. In someone with seizures it can bring extra seizures. The label wants at least a week, and longer after high dose or long use. That conversation is a taper plan, not a permission slip to empty the bottle into the sink tonight.
I heard a big pregabalin sciatica study failed. Does that sink gabapentin too?
It sinks the class for that problem. Pregabalin versus placebo for sciatica did not show a meaningful benefit. Gabapentin reviews for mechanical and radicular back pain point the same way. Shared alpha-2-delta target, shared miss. Defined neuropathic pain is where I still reach for a 300 mg start. Disc and facet pain is where I want a better tool. Condition pages on MedlinePlus are a cleaner first read than a forum argument.
PHN versus 'nerve pain' on my script - why do you treat those as different?
PHN is a labeled adult indication with two specific RCTs and a 300 mg day-one schedule on the Neurontin insert. 'Nerve pain' is a hallway phrase that may mean diabetic neuropathy, chemotherapy neuropathy, or a vague ache. Some of those are fair trials. Some are not. I want the chart to name the syndrome. If it cannot, the 300 mg capsule is being asked to do social work. Social work is not an endpoint.
Where do I put this next to the other notes on your desk?
History and the GABA miss sit in how a GABA mimic became Neurontin 300 mg. The medicine page is Neurontin 300 mg. Titration, eGFR bands, opioid breathing risk, and the one-week taper sit in the eGFR-first titration note. This card is only the use map. Do not let a labeled PHN win talk you into a leftover insomnia script.
General education from a clinician, not personal medical advice. Bring your own history to your own prescriber.
Last Updated