The Scottish Medicines Consortium ( SMC ) has advised NHS ( National Health Service ) Boards and Area Drug and Therapeutic Committees ( ADTCs ) that Pregabalin ( Lyrica ) is not recommended for use within NHS Scotland for the treatment of peripheral neuropathic pain in adults.

Comparative clinical and cost effectiveness have not been demonstrated.


Evidence on comparative efficacy

Pregabalin is an antiepileptic that decreases central neuronal excitability by binding to an auxillary subunit of a voltage-gated calcium channel in the central nervous system.
It also reduces the release of several neurotransmitters including glutamate, noradrenaline and substance P, though the significance of the latter effects is unknown.

The original submission to the Scottish Medicines Consortium, included data from ten double-blind studies. Five studies recruited patients with diabetes who had glycated haemoglobin ( HbA1C ) less than or equal to 11 percent and distal symmetrical sensorimotor polyneuropathy for more than or equal to 1 year and < 5 years in three studies. Four studies recruited patients with postherpetic neuralgia of more than or equal to 3 and 6 months duration after healing of the herpes zoster skin rash.
A further double-blind trial recruited patients with either diabetic peripheral neuropathy or postherpetic neuralgia.

In all studies patients had a score of more than or equal to 40mm on the 100mm visual analogue pain scale of the Short-Form McGill Pain Questionnaire ( SF-MPQ ) and during the week before randomisation, had at least four daily pain scores more than or equal to 4 assessed on an 11-point scale ( where 0 = no pain and 10 = worst possible pain ).
Patients were randomised to placebo or Pregabalin with one trial including Amitriptyline as an active treatment arm. The primary endpoint, weekly mean pain score ( assessed via the 11-point Likert scale ) at endpoint, was compared in the intention to treat population between placebo and active treatments via an analysis of covariance which adjusted for weekly mean pain score at baseline.

In the trial with an active comparator, Pregabalin 600 mg daily was not associated with a significant difference from placebo in the primary analysis of pain scores at end-point. The difference between Amitriptyline 75mg daily and placebo was significant. This pattern was repeated in two supplementary analyses of pain scores and in a number of secondary end-points including the proportion of patients achieving a response ( more than or equal to 50 percent reduction in pain scores ) and patients’ and clinicians’ global impression of pain. In no endpoint analysis was the result significant for Pregabalin but not Amitriptyline.

Pregabalin 300mg and 600mg were associated with significantly lower weekly mean pain scores at endpoint compared with placebo in all other studies, except for the 300mg dose in a 12-week diabetic neuropathy study. Weekly mean pain scores with Pregabalin 150mg were generally lower than placebo, but differences between this dose and placebo were not consistently significant.
In these studies, responders were defined as patients who had a 50% reduction in mean weekly pain score compared with baseline.
Similar efficacy patterns were observed in this secondary endpoint.

The resubmission and IRP ( Independent Review Panel ) contains additional data from an interim, 15 month, analysis of an ongoing open-label safety study.
Patients had participated in previous Pregabalin studies and had a score of more than or equal to 40mm on the 100mm visual analogue pain scale of the SF-MPQ and were intolerant of or had experienced lack of efficacy after more than or equal to two weeks of at least minimum doses of tricyclic antidepressants, Gabapentin 1800mg or other third line agents. Patients were given Pregabalin 150-600mg daily and could continue to receive other analgesics, including those that may not have provided adequate pain control previously with titration at the investigator’s discretion to optimise pain control.
Pregabalin was discontinued quarterly for 3-28 days until patient’s pain worsened before recommencing. If pain did not worsen or only slightly deteriorated, the patient was discontinued from the study.

Interim analysis data from 45 patients with painful diabetic peripheral neuropathy and 36 patients with postherpetic neuralgia indicate that mean SF-MPQ scores decreased from respective baseline values of 73mm and 75mm to 47mm and 51mm at endpoint.

A published systematic review, new to the IRP, calculates numbers needed to treat ( NNT ) to achieve a 50% reduction in pain scores and numbers needed to harm ( NNH ) and uses these as a basis for an evidence-based treatment algorithm.
Data are included from five of the Pregabalin studies from above: three in painful diabetic peripheral neuropathy and two in postherpetic neuralgia.
The authors report a combined NNT for Pregabalin at doses of 150-600mg in both pain models of 4.2.
This is reported to be similar to the NNT for Gabapentin at all doses and pain models (5.1 ).
The NNH for Pregabalin was considered to represent a relatively high rate of withdrawal and was 11.7 compared with 26.1 for Gabapentin. For a treatment algorithm in peripheral neuropathic pain based on pain relief, the NNT are lowest for tricyclic antidepressants, then opioids, then Tramadol, then Gabapentin/Pregabalin.


Evidence on comparative safety

Pregabalin is commonly associated with central nervous system adverse events similar to most antiepileptic drugs. It is not metabolised in vivo and is eliminated from the systemic circulation primarily by renal excretion of the unchanged drug. It does not bind to plasma proteins and does not induce or inhibit hepatic enzymes. It would thus not be expected to induce or be affected by hepatic pharmacokinetic interactions and did not interact with other antiepileptic drugs or combined oral contraceptives in studies.
Similarly Gabapentin is eliminated renally and does not induce hepatic enzymes or interact with other antiepileptics or oral contraceptives.
However, Carbamazepine is metabolised by and induces the production of hepatic enzymes, and, thus, may be involved in hepatic pharmacokinetic interactions.


Clinical effectiveness issues

There are no direct trials of Pregabalin with Gabapentin in the treatment of neuropathic pain and so relative efficacy is uncertain. In addition, there are limited data in patients unresponsive to Gabapentin since many of the double-blind studies excluded patients who had failed to respond to previous treatment with Gabapentin more than or equal to 1200mg/day.
Since the drugs are considered to act in the same way, this exclusion could potentially favour the results of the Pregabalin studies.
Data represented in the resubmission and IRP include interim results of an open-label trial in patients intolerant of or refractory to other treatments. Although this indicates a reduction in pain with Pregabalin, these results are difficult to interpret due to a number of limitations. The study was descriptive with no statistical testing to assess efficacy; patients could remain on other analgesic medication with doses titrated to optimise control. Results from a new systematic review would suggest comparable efficacy to Gabapentin.

Source: Scottish Medicines Consortium, 2006


XagenaMedicine2006