Patient-friendly summary
If you read nothing else
Bottom line
Quality evidence to support pregabalin or gabapentin for chronic low back pain without radiculopathy or neuropathy is lacking; pregabalin was not superior to other analgesics, while gabapentin showed mixed results that may suggest it as a viable option pending more research.
Preliminary evidencePublished
Evidence hierarchy
Study participants
Adults aged 18 and older with chronic low back pain (at least 2 months) without radiculopathy or neuropathy
Study Summary
This systematic review examined whether the antiepileptic-derived drugs gabapentin and pregabalin are effective and safe for chronic low back pain (CLBP) in patients who do NOT have radiculopathy or neuropathy. From 2,230 articles screened, only 5 randomized controlled trials (242 included participants) met criteria, and just one trial studied a "pure" non-radicular sample. Pregabalin was slightly less effective than comparator drugs (amitriptyline, tramadol/acetaminophen, celecoxib), and adding pregabalin to celecoxib gave no benefit over celecoxib alone (all very low certainty evidence). Gabapentin findings were mixed: one trial showed a modest reduction in pain and improved mobility (moderate certainty), while another in a broader population showed no advantage over placebo; no serious adverse events occurred in any study, though gabapentin caused more adverse events than placebo.
Key Findings
| Finding | Detail | Impact |
|---|---|---|
| Very limited eligible evidence exists | Of 2,230 articles identified, only 5 randomized controlled trials qualified, contributing 242 included participants. Only one trial (McCleane, 2000) had a sample composed entirely of patients with non-radicular, non-neuropathic CLBP; for the others, only data subgroups could be used. | High |
| Pregabalin was not superior to comparator drugs | In no comparison was pregabalin monotherapy superior to its comparator for pain relief, and it was sometimes inferior to amitriptyline (50% pain reduction in 32.6% vs 53.2%; reduction in Oswestry Disability Index >20% in 39.1% vs 65.96%, p=0.01) and slower to act than tramadol/acetaminophen. All such results were graded very low certainty. | High |
| Adding pregabalin to celecoxib gave no extra benefit | Romanò et al. found the pregabalin/celecoxib combination was not superior to celecoxib alone for pain reduction (VAS mean 45.1 to 32.9 vs 43.8 to 32.5; p=0.9, very low certainty). | Medium |
| Gabapentin results were mixed but one trial was favorable | McCleane (2000), the only pure-sample trial, found a subtle pain reduction on a 0-10 verbal numeric scale during gabapentin use (7.10 to 6.39; p<0.05) and improved mobility (4.65 to 5.46; p<0.01), both moderate certainty (GRADE). Atkinson et al. found pain fell in both gabapentin and placebo groups with no significant between-group difference. | High |
| No serious adverse events, but gabapentin caused more adverse events than placebo | No serious adverse events were reported in any study. A meta-analysis of two gabapentin-vs-placebo trials showed more adverse events with gabapentin (risk ratio 1.52; 95% CI 1.20-1.91; p=0.0004). | Medium |
Of 2,230 articles identified, only 5 randomized controlled trials qualified, contributing 242 included participants. Only one trial (McCleane, 2000) had a sample composed entirely of patients with non-radicular, non-neuropathic CLBP; for the others, only data subgroups could be used.
In no comparison was pregabalin monotherapy superior to its comparator for pain relief, and it was sometimes inferior to amitriptyline (50% pain reduction in 32.6% vs 53.2%; reduction in Oswestry Disability Index >20% in 39.1% vs 65.96%, p=0.01) and slower to act than tramadol/acetaminophen. All such results were graded very low certainty.
Romanò et al. found the pregabalin/celecoxib combination was not superior to celecoxib alone for pain reduction (VAS mean 45.1 to 32.9 vs 43.8 to 32.5; p=0.9, very low certainty).
McCleane (2000), the only pure-sample trial, found a subtle pain reduction on a 0-10 verbal numeric scale during gabapentin use (7.10 to 6.39; p<0.05) and improved mobility (4.65 to 5.46; p<0.01), both moderate certainty (GRADE). Atkinson et al. found pain fell in both gabapentin and placebo groups with no significant between-group difference.
No serious adverse events were reported in any study. A meta-analysis of two gabapentin-vs-placebo trials showed more adverse events with gabapentin (risk ratio 1.52; 95% CI 1.20-1.91; p=0.0004).
Strengths
- Comprehensive, multi-database search (CENTRAL, MEDLINE, EMBASE, LILACS, Web of Science) with no language restriction, plus manual reference and trial-protocol searches
- Pre-registered protocol and methodology aligned with the Cochrane Handbook
- Standardized appraisal using Cochrane RoB 2 and GRADE certainty grading
- Careful focus on the non-radicular/non-neuropathic subpopulation, extracting only relevant subgroups where possible
Limitations
- Only 5 small trials (242 participants) met criteria, and just one had a pure non-radicular sample; for the others only population subgroups could be used, compromising applicability
- Most results were graded very low certainty on GRADE; only two outcomes in one trial reached moderate certainty
- Trials were clinically heterogeneous (different drugs, comparators, populations, and 4-14 week follow-ups), preventing pooled efficacy meta-analysis
- Grey literature was not searched, raising risk of non-reporting/publication bias
- Initial screening and data extraction/risk-of-bias assessment were each done by a single reviewer, increasing risk of human error
Key Takeaways for Patients
What This Means for You
- 01For ordinary (non-nerve-related) chronic low back pain, there is little high-quality evidence that gabapentin or pregabalin reliably reduces pain.
- 02Pregabalin did not outperform other commonly used pain medicines (such as amitriptyline, tramadol/acetaminophen, or celecoxib) in these studies, and adding it to celecoxib gave no extra benefit.
- 03Gabapentin showed mixed results: one study suggested a modest improvement in pain and mobility, but another found it no better than a placebo.
- 04No serious side effects occurred in any of the studies, but gabapentin caused more minor-to-moderate side effects than placebo, so the risks and benefits should be weighed with your clinician.
- 05The evidence base is very small and uncertain, so these drugs should not be assumed to work for low back pain that does not involve nerve (radicular/neuropathic) pain.
Read the Full Paper
Access the complete peer-reviewed study from Arquivos de Neuro-Psiquiatria
View Full Study