Clindamycin phosphate with benzoyl peroxide, gel, 10 mg (base) - 50 mg per gm (1% – 5%), 25 gm, Duac® Once Daily Gel, March 2009

Public summary document for Clindamycin phosphate with benzoyl peroxide, gel, 10 mg (base) - 50 mg per gm (1% – 5%), 25 gm, Duac® Once Daily Gel, March 2009

Page last updated: 03 July 2009

PDF printable version of Clindamycin phosphate with benzoyl peroxide, gel, 10 mg (base) - 50 mg per gm (1% – 5%), 25 gm, Duac® Once Daily Gel (PDF 30 KB)

Public Summary Document

Product: Clindamycin phosphate with benzoyl peroxide, gel, 10 mg (base) - 50 mg per gm (1% – 5%), 25 gm, Duac® Once Daily Gel
Sponsor: Stiefel Laboratories Pty Ltd
Date of PBAC Consideration: March 2009

1. Purpose of Application

The submission sought a Restricted benefit listing for clindamycin phosphate 1% with benzoyl peroxide 5% gel for the treatment of moderate acne vulgaris with inflammatory lesions in patients with an inadequate response to a minimum of two over-the-counter products.

2. Background

The PBAC has not previously considered this product.

3. Registration Status

Duac Once Daily Gel® was TGA registered on 30 January 2006 for the topical treatment of comedo, papules and pustular acne vulgaris.

4. Listing Requested and PBAC’s View


Restricted Benefit
Treatment of moderate acne vulgaris with inflammatory lesions in patients with an inadequate response to a minimum of two over-the-counter medications.

In the Pre-PBAC response, noting comment made during the evaluation, the sponsor proposed the following restriction wording:

Moderate acne vulgaris with a minimum of 20 inflammatory lesions (papules or pustules) in a patient with an inadequate response to at least two other topical therapies.

NOTE: Duration should not exceed 11 weeks.

For PBAC’s view, see Recommendation and Reasons.

5. Clinical Place for the Proposed Therapy

The availability of clindamycin 1% and benzoyl peroxide 5% gel would provide a PBS-subsidised treatment for patients with moderate acne vulgaris.

6. Comparator

The submission nominated no treatment (placebo) as the main comparator as there are currently no treatments for moderate acne listed on the PBS. The submission also nominated topical retinoids, topical antibiotics and benzoyl peroxide as secondary comparators.

Whether the product would remain as a third-line treatment if PBS listed was a concern, with a potential for use earlier if available as a subsidised treatment, and also in mild acne, given there are no products specifically listed for this indication.

7. Clinical Trials

The submission presented:
- five direct randomised trials with four treatment arms: four trials of clindamycin + benzoyl peroxide gel (i.e. Duac), clindamycin alone, benzoyl peroxide alone, and placebo gel, and one trial of clindamycin + benzoyl peroxide gel, clindamycin alone and benzoyl peroxide alone, without a placebo arm;
- one direct randomised trial with two treatment arms, comparing clindamycin plus benzoyl peroxide gel with the topical retinoid adapalene (Langner 2008); and
- one supportive study of quality of life measures with the use of clindamycin plus benzoyl peroxide gel in Korean patients in Korea.

The trials published at the time of submission are detailed below:

Trial ID/First Author Protocol title/ Publication title Citation
Direct randomised trial
Langner 2008 A randomized, single-blind comparison of topical clindamycin + benzoyl peroxide and adapalene in the treatment of mild to moderate facial acne vulgaris. Brit J Derm 2008 158(1): 122-9
Study 150 Lookingbill D.P. et al Treatment of acne with a combination clindamycin/benzoyl peroxide gel compared with clindamycin gel, benzoyl peroxide gel and vehicle gel: Combined results of two double-blind investigations Journal of the American Academy of Dermatology 1997; 37(4): 590-595.

8. Results of Trials

The pooled results of the percentage reduction in counts of inflammatory and non-inflammatory lesions in the five direct randomised trials with four treatment arms demonstrated that treatment with Duac results in a statistically significant reduction in both inflammatory and non-inflammatory lesion counts at 11 weeks compared with once per day treatment with benzoyl peroxide, clindamycin or placebo, though with wide confidence intervals.

The pooled results of the physician assessed global improvement, proportion of “good” or “excellent” improvement in the same studies was increased with Duac treatment compared with benzoyl peroxide (risk ratio 1.32 [95%CI 1.13, 1.56]), clindamycin (risk ratio 1.66 [95%CI 1.25, 2.20]) and placebo (risk ratio 3.85 [95%CI 1.56, 9.45]).

The most common adverse event for Duac was local skin irritation comparable with benzoyl peroxide alone.

The PBAC noted the TGA Resistance Risk Assessment report concerning Duac was prepared for review by the Expert Advisory Group on Antimicrobial Resistance. The report concluded that there was no evidence that the combination product would make the emergence of resistance in targeted or other bacteria more likely than topical clindamycin alone, and that the combination of clindamycin and benzoyl peroxide may make the emergence of clindamycin resistant bacteria less likely. However, the report also recommended that a maximum course of treatment with Duac should be 12 weeks, without further continuation of clindamycin containing topical therapies.

As with topical clindamycin, Duac has rarely been associated with pseudomembranous colitis.

9. Clinical Claim

The submission claimed that Duac was superior to benzoyl peroxide, clindamycin, placebo and adapalene in terms of reduction in inflammatory and non-inflammatory lesions and global improvement in acne vulgaris of the face, and equivalent to these treatments in terms of safety and tolerance.

The PBAC agreed with the clinical claims that clindamycin with benzoyl peroxide was more effective than benzoyl peroxide, clindamycin, placebo and adapalene, and was equivalent to these treatments in terms of safety and tolerance. It was noted that four of the key trials included patients with mild acne.

10.Economic Analysis

A stepped economic evaluation was presented comparing the costs and consequences of Duac compared with no treatment in patients with moderate acne. The submission estimated the base case incremental cost per QALY gained was less than $15,000.

The submission presented the results of a number of univariate sensitivity analyses. The model was most sensitive to alternative utilities for the responder and non-responder health states, with the incremental cost per QALY gained less than $15,000 when utilities from Chen, et al (2008) were used (assuming responders achieve 50% acne clearance). The model was also sensitive to including the proportion of responders in the ‘no treatment’ arm from the meta-analysis of key trials (16.8%; base case: 0%), and included the costs of GP visits to obtain scripts for Duac. Combining these three analyses resulted in an incremental cost per QALY gained in the range $15,000-$45,000.

The PBAC noted the results of multivariate sensitivity analyses conducted during the evaluation. These analyses demonstrated that the model is sensitive to the utility assumptions, the assumption of no responders in the placebo arm (rather than 16.8% as in the meta-analysis) and the inclusion of GP costs associated with clindamycin.

11. Estimated PBS Usage and Financial Implications

The financial cost per year to the PBS was estimated to be less than $10 million in Year 5 of listing.

12. Recommendation and Reasons

The PBAC noted the amended restricted benefit listing sought, as proposed in the Pre-PBAC Response, however, it considered a definition of ‘moderate acne’ remained problematic, and the requirement for the use of two prior topical preparations while reasonable, would be impractical to implement. Further, that a Note stating use per treatment was limited to 11 weeks would be of limited value in a restricted benefit listing, and there was a high potential for long-term use of the product. While the TGA approved Product Information recommended that a course of treatment be limited to a maximum duration of 11 weeks, the PBAC agreed that patients would want to continue treatment after this time, if it is effective.

Whether the product would remain as a third-line treatment if PBS listed was a concern, with a potential for use earlier in the treatment algorithm if available as a subsidised treatment, and also in mild acne, given no products are specifically listed for this indication.

The PBAC agreed with the clinical claims that clindamycin with benzoyl peroxide is more effective than benzoyl peroxide, clindamycin, placebo and adapalene, and is equivalent to these treatments in terms of safety and tolerance. It was noted that four of the key trials included patients with mild acne.

The PBAC noted that given survival benefit is not an issue in this condition all the benefits of treatment are derived from an improvement in quality of life (QOL). The Committee noted there were several areas of uncertainty with the economic modelling.

A key concern to the PBAC was is in relation to the translation of the QOL outcomes observed in the trials to QALY weights and QALYs. The QOL results were drawn from only one study (which included patients with mild and moderate acne), as this was the only study presented that included a QOL scale, the Dermatology Life Quality Index (DLQI). The trial was carried out in a Korean population and there were concerns about the generalisability of these QOL results to the PBS population. The PBAC noted the Pre-Sub-Committee response considered that the applicability of the DLQI results to the Australian population was supported by the available evidence, citing the results of two studies (Fernandez-Obregon et al. 2003 and Chen et al. 2008) which the sponsor regards as showing that there are similar improvements in QOL and utility scores for Asian compared to Caucasian patients. However, the PBAC noted that the Chen et al. study compares groups of Asian-American and Caucasian teenagers who both live in America where there was likely to be greater cultural similarity than across Korean and Australian populations. Thus the finding of similarity is not necessarily generalisable. The PBAC considered that the applicability of the results from the Korean study to the Australian population remains uncertain, noting that there are well documented differences across countries in QOL assessment and utility assessment with the same instrument.

Another key issue was that the DLQI scores in acne were then translated to EQ-5D utility scores with the use of a mapping algorithm for psoriasis patients from Woolacott et al. (2006) [EQ-5D utility score = 0.956 – [0.0248 x (DLQI total score)]] to translate the DLQI scores to a utility score for acne patients. The PBAC considered there are likely to be other factors that affect the QOL in psoriasis patients compared to acne patients that would not be captured by this mapping algorithm. Further, the precision of the mapping algorithm was uncertain as there is no information concerning the confidence interval around the reported coefficients from the regression based mapping.

The PBAC noted the Pre-Sub-Committee response argued that in the model the attributing of all the utility gain observed to responders did not favour clindamycin with benzoyl peroxide, however, the PBAC was not convinced as there were more responders in the clindamycin with benzoyl peroxide arm. The PBAC agreed there was uncertainty associated with the modelling assumption that the utility gain for responders occurs for half of the first period of treatment (6 weeks of 12) but for all subsequent treatment periods and that the total utility gain is implicitly assumed to occur for the whole of the “on treatment” period. This not only assumes that utility gains from initial treatment are sustained in subsequent treatments over a 12 month period but also implies a greater utility gain in the later treatment periods than in the first period.

Overall, the PBAC considered the approach to measuring utilities to be highly uncertain and appeared to overestimate the utility gain associated with treatment because of (1) use of a psoriasis-based mapping algorithm (2) the assumptions in relation to non-responders (3) the assumption that the Korean QOL results apply to Australian population (4) the assumption that these results apply to moderate to severe acne.

Other factors contributing to the economic uncertainty associated with the requested listing were in relation to costs and the assumptions relating to the proportion of responders in the placebo arm and the extent of acne clearance in responders.

The PBAC considered the base case ICER of less than $ 15,000 per QALY gained was sensitive to a number of assumptions, and the multivariate sensitivity analyses conducted during the evaluation produced values in the range $ 15,000 to $ 45,000. However, the PBAC was advised that these may have been applied inappropriately. The base case non-responder utility in this sensitivity analysis is 0.932, on the basis that this was the “own” health state utility for Chen participants who had seen a GP for their acne. The responder utility is 0.978 (assuming 100% acne clearance in responders) – which corresponds to the “clearance” utility for the whole Chen population. If the value for the responder state is based on the whole Chen population then it is also appropriate for the base case/non-responder utility to be based on the same sample, that is the “own” utility for the whole Chen population (=0.961). The approach taken in the submission is to selectively pick utilities from different populations at different time points, which is not appropriate. Recalculating the ICER with this value gives an ICER within the range of $45,000-$75,000 per QALY but higher than when using the “responder” utility value. When a similar adjustment is made to the sensitivity analyses that use the Chen utilities assuming 50% acne clearance in responders, the ICER falls in the range $75,000-$105,000 per QALY as opposed to less than $15,000 per QALY. When the additional assumptions queried by the PBAC are tested in sensitivity analyses these ICERs will become higher.

Therefore, the PBAC agreed there is considerable uncertainty surrounding the ICERs as a result of the approach to modelling QALY gains, leading to an ICER that is highly uncertain and likely to be much higher than the base case in the submission.

The Committee noted that no other drug had been recommended for PBS listing for the treatment of moderate (or mild) acne vulgaris, and in the absence of a clear definition of mild to moderate questioned whether PBS subsidisation was appropriate at this time. If mild is interpreted without qualification, the guidelines for submissions to the PBAC state in Box 1.2 that low priority is given to a listing for a drug for the treatment of clinically minor or trivial conditions.

The PBAC also considered there was not a high clinical need for the product, and was concerned that the wider availability and potential for long term use may contribute to the increase in community acquired MRSA resistance given the significant market for the gel.

The PBAC rejected the application on the basis of unacceptably high and uncertain cost effectiveness, uncertainty with its clinical place and the potential for the development of antibiotic resistance.

The PBAC noted that the submission meets the criteria for an Independent Review.

13. Context for Decision

The PBAC helps decide whether and, if so, how medicines should be subsidised in Australia. It considers submissions in this context. A PBAC decision not to recommend listing or not to recommend changing a listing does not represent a final PBAC view about the merits of the medicine. A company can resubmit to the PBAC or seek independent review of the PBAC decision.

14. Sponsor’s Comment

The Sponsor has no comments.