8Topical therapy

8.1.2.1. Evidence profile

Table

8.1.2.2. Evidence statements

In people with psoriasis, topical vitamin D or vitamin D analogue treatment was statistically significantly better than placebo for:

• Investigator’s assessment (clear/nearly clear on PGA) at 4–10 weeks for calcipotriol once daily, calcipotriol twice daily, calcitriol once daily, calcitriol twice daily or tacalcitol once daily [11 studies (7 between- and 4 within-patient studies); 2387 participants (2594 randomised units); low to moderate quality evidence]^{25,81,102,132,148,179,208,210,211,302,311}
• Patient assessment (clear/nearly clear on PGA) at 4–8 weeks for calcipotriol once daily or calcipotriol twice daily [3 between-patient studies; 1432 participants; moderate quality evidence]^132,141,179
• Percentage change in PASI at 4 weeks for calcipotriol twice daily [1 within-patient study; 60 participants (120 randomised units); moderate quality evidence]⁸¹
• Withdrawal due to adverse events at 4–8 weeks [11 studies (6 between- and 5 within-patient); 2367 participants (2791 randomised units); low quality evidence] ^{25,132,141,148,179,208,210,211,311,354}
• Withdrawal due to lack of efficacy at 4–8 weeks [7 studies (4 between- and 3 within-patient); 1207 participants (1477 randomised units); moderate quality evidence]^{25,132,141,210,211,311,354}
• Relapse at 8 weeks post treatment with tacalcitol once daily [1 between-patient study; 36 participants; very low quality evidence]²⁰⁸.

In people with psoriasis, there was no statistically significant difference between topical vitamin D or vitamin D analogue treatment and placebo for:

• Patient assessment at 8 weeks (clear/nearly clear) with tacalcitol once daily [1 between-patient study; 227 participants; very low quality evidence]²⁰⁸
• Skin atrophy at 4 weeks for calcipotriol twice daily [2 between-patient studies; 851 participants; very low quality evidence]^132,302

8.1.2.3. Heterogeneity

• There was significant heterogeneity between data regarding the investigator’s assessment of efficacy. This heterogeneity was removed by creating subgroups based on the specific agent and treatment frequency of the vitamin D or vitamin D analogue. Nevertheless, all agents and frequencies demonstrated a clinically significant benefit compared with placebo.
• There was significant heterogeneity between data regarding the patient’s assessment of efficacy. This heterogeneity was removed by creating subgroups based on the specific agent within the vitamin D or vitamin D analogue class, while treatment frequency did not explain the differences. It appeared that tacalcitol was not more effective than placebo based on patient’s assessment, whereas calcipotriol was more effective. However, the heterogeneity may also have been caused by the tacalcitol study having a higher risk of bias as it was only investigator blinded (although this may be more likely to increase the effect estimate in favour of the active intervention) and had a 30% drop-out rate in the placebo group.
• There was no significant heterogeneity for the remaining outcomes

8.1.3.1. Evidence profile

Table

8.1.3.2. Evidence statements

In children with psoriasis, there was no statistically significant difference between calcipotriol twice daily and placebo for:

• Investigator’s assessment (clear/nearly clear) at 8 weeks [1 between-patient study; 77 participants; low quality evidence]²⁹⁵
• Patients assessment (clear/nearly clear) at 8 weeks [1 between-patient study; 77 participants; very low quality evidence²⁹⁵
• % change in PASI at 8 weeks [1 between-patient study; 77 participants; low quality evidence²⁹⁵

8.1.3.3. Heterogeneity

• Not applicable as only one study assessed vitamin D or vitamin D analogues compared with placebo in children

8.1.4.1. Evidence profile

Table

8.1.4.2. Evidence statements

In people with psoriasis, topical potent corticosteroid treatment was statistically significantly better than placebo for:

• Investigator’s assessment (clear/nearly clear) at 3–8 weeks for mometasone furoate once daily, hydrocortisone butyrate twice daily and betamethasone dipropionate once or twice daily [6 between-patient studies; 1507 participants; moderate quality evidence]^{102,179,246,302,360,428}
• Patient’s assessment (clear/nearly clear) at 3–4 weeks for hydrocortisone butyrate twice daily or betamethasone dipropionate once daily [2 between-patient studies; 794 participants; moderate quality evidence]^179,360
• Withdrawal due to adverse events at 3–4 weeks for potent corticosteroid (mometasone furoate or betamethasone dipropionate) once daily [2 between-patient studies; 693 participants; moderate quality evidence]^179,246

In people with psoriasis, there were no events with either topical potent corticosteroid treatment or placebo for:

• Withdrawal due to lack of efficacy at 3 weeks for betamethasone dipropionate twice daily [1 between-patient study; 76 participants; high quality evidence]⁴²⁸

In people with psoriasis, there was no statistically significant difference between topical potent corticosteroid treatment and placebo for:

• Withdrawal due to adverse events at 3–12 weeks for potent corticosteroid (hydrocortisone butyrate, betamethasone valerate or betamethasone dipropionate) twice daily [3 studies (2 between- and 1 within-patient); 285 participants (325 randomised units); very low quality evidence]^360,381,428
• Skin atrophy [2 between-patient studies; 516 participants; very low quality evidence]^246,302

8.1.4.3. Heterogeneity

• There was significant heterogeneity between data regarding withdrawals due to adverse effects. This heterogeneity was removed by creating subgroups based on treatment frequency. It was considered clinically more likely that the treatment frequency was causing the heterogeneity rather than the specific agent within the potent corticosteroid class.
• There was no significant heterogeneity for the remaining outcomes

8.1.5.1. Evidence profile

Table

8.1.5.2. Evidence statements

In people with psoriasis, topical very potent corticosteroid treatment was statistically significantly better than placebo for:

• Investigator’s assessment (clear/nearly clear)at 2–4 weeks for clobetasol propionate once or twice daily [5 between-patient studies; 859 participants; very low quality evidence]^{73,125,167,216,227}
• Patient’s assessment (clear/nearly clear) at 2 weeks for clobetasol propionate twice daily [2 between-patient studies; 124 participants; low quality evidence]^125,216

In people with psoriasis, there was no statistically significant difference between topical very potent corticosteroid treatment and placebo for:

• Withdrawal due to adverse events at 2–4 weeks for clobetasol propionate once or twice daily [7 between-patient studies; 984 participants; very low quality evidence]^{32,73,125,167,174,216,227}
• Withdrawal due to lack of efficacy at 4 weeks for clobetasol propionate once or twice daily [3 studies (2 between- and 1 within-patient); 360 participants (385 randomised units); very low quality evidence]^32,73,167
• Skin atrophy at 4 weeks for clobetasol propionate once or twice daily [4 studies (3 between- and 1 within-patient); 439 participants (464 randomised units); very low quality evidence]^{32,73,167,174}

8.1.5.3. Heterogeneity

• For the outcome of investigator’s assessment of achieving clear/nearly clear status high heterogeneity was present between the results for the five studies. The heterogeneity could not be explained by any of the pre-specified subgroups for investigation or by excluding studies at high/very high risk of bias. It is likely to be caused by the small size of three of the studies^167,216,227. The two sufficiently powered studies demonstrated a clear clinical benefit of very potent steroids compared with placebo.
• There was no significant heterogeneity for the remaining outcomes.

8.1.6.1. Evidence profile

Table

8.1.6.2. Evidence statements

In people with psoriasis, placebo was statistically significantly better than tazarotene applied once daily for:

• Withdrawal due to adverse events at 12 weeks [3 between-patient studies; 1573 participants; low quality evidence] ^418–420

In people with psoriasis, there were no events with either tazarotene or placebo for:

• Skin atrophy at 12 weeks [1 between-patient study; 324 participants; moderate quality evidence]^418,420

In people with psoriasis, there was no statistically significant difference between tazarotene and placebo applied once daily for:

• Investigator’s assessment (clear/nearly clear) at 12 weeks [2 between-patient studies; 1303 participants; very low quality evidence]⁴¹⁹
• Withdrawal due to lack of efficacy at 12 weeks [1 between-patient study; 324 participants; very low quality evidence]^418,420

8.1.6.3. Subgroups and heterogeneity

• For the outcome of investigator’s assessment of achieving clear/nearly clear status heterogeneity was present between the results. The heterogeneity could not be explained by any of the pre-specified subgroups for investigation or excluding studies at high risk of bias.
• There was no significant heterogeneity for the remaining outcomes.

8.1.7.1. Evidence profile

Table

8.1.7.2. Evidence statements

In people with psoriasis, intermittent twice daily topical potent corticosteroid (betamethasone dipropionate) was statistically significantly better than placebo for the maintenance of remission for:

• Investigator’s assessment (clear/slight) at 24 weeks [1 between-patient study; 90 participants; low quality evidence]¹⁷⁸
• Time-to-relapse after a maximum follow-up of at 24 weeks [1 between-patient study; 90 participants; low quality evidence]¹⁷⁸

In people with psoriasis, there were no events with either intermittent twice daily topical potent corticosteroid (betamethasone dipropionate) or placebo for the maintenance of remission for:

• Withdrawal due to adverse events at 24 weeks [1 between-patient study; 86 participants; moderate quality evidence]¹⁷⁸
• Skin atrophy at 24 weeks [1 between-patient study; 90 participants; moderate quality evidence]¹⁷⁸

8.1.7.3. Heterogeneity

Not applicable as only one study assessed potent corticosteroid compared with placebo for the maintenance of remission.

8.1.8.1. Evidence profile

Table

8.1.8.2. Evidence statements

In people with psoriasis, potent corticosteroid was statistically significantly better than vitamin D or vitamin D analogue for:

• Investigator’s assessment (clear/nearly clear) at 4–8 weeks for calcipotriol once or twice daily or calcitriol twice daily compared to betamethasone dipropionate once or twice daily or betamethasone valerate twice daily [6 between-patient studies; 3136 participants; very low quality evidence]^{49,79,102,179,255,302}
• Patient’s assessment (clear/nearly clear) at 4 weeks for calcipotriol once or twice daily compared to betamethasone dipropionate once or twice daily [2 between-patient studies; 1684 participants; low to moderate quality evidence]^79,179
• Withdrawals due to adverse events at 4–8 weeks for calcipotriol once or twice daily or calcitriol twice daily compared to betamethasone dipropionate once or twice daily, betamethasone valerate twice daily or fluocinonide twice daily [7 studies (6 between- and 1 within-patient); 3082 participants (3427 randomised units); low quality evidence]^{47,49,66,79,179,198,255}

In people with psoriasis, vitamin D or vitamin D analogue was statistically significantly better than potent corticosteroid for:

• Patient’s assessment (clear/nearly clear) at 6 weeks for calcipotriol twice daily compared to betamethasone valerate twice daily [2 studies (1 between- and 1 within-patient); 743 participants (1085 randomised units); low quality evidence]^66,198
• % change in PASI at 6–8 weeks for calcipotriol twice daily compared to betamethasone valerate twice daily [2 studies (1 between- and 1 within-patient); 754 participants (1096 randomised units); moderate quality evidence]^198,255
• Relapse rate (requiring re-treatment [not maintaining clear/nearly clear] within 8-weeks post treatment) for calcitriol twice daily compared with betamethasone dipropionate twice daily [1 between-patient study; 131 participants; very low quality evidence]⁴⁹

In people with psoriasis, there was no statistically significant difference between potent corticosteroid and vitamin D or vitamin D analogue for:

• Withdrawals due to lack of efficacy at 6 weeks for calcipotriol or calcitriol twice daily compared with betamethasone dipropionate or valerate twice daily [3 studies (1 between- and 2 within-patient); 976 participants (1321 randomised units); very low quality evidence] ^49,66,198
• Skin atrophy at 4–8 weeks for calcipotriol twice daily vs betamethasone dipropionate or valerate twice daily [2 between-patient studies; 1038 participants; very low quality evidence]^255,302

8.1.8.3. Heterogeneity

• For the outcome of investigator’s assessment of achieving clear/nearly clear status heterogeneity was present. The heterogeneity could not be explained by any of the pre-specified subgroups for investigation or by excluding studies at higher risk of bias.
• For the outcome of patient’s assessment of achieving clear/nearly clear status heterogeneity was present. The heterogeneity was explained by creating subgroups based on treatment frequency and the specific agent, suggesting that betamethasone valerate may be less effective than betamethasone dipropionate.
• There was no significant heterogeneity for the remaining outcomes.

8.1.9.1. Evidence profile

Table

8.1.9.2. Evidence statements

In people with psoriasis, concurrent vitamin D or vitamin D analogue and potent corticosteroid treatment (one applied in the morning and one in the evening) was statistically significantly better than vitamin D or vitamin D analogue alone for:

• Investigator’s assessment (clear/nearly clear)at 6–8 weeks for calcipotriol and betamethasone valerate compared with calcipotriol once or twice daily [2 between-patient studies; 682 participants; low to moderate quality evidence]^197,344
• Investigator’s assessment (clear/nearly clear among those who did not respond to calcipotriol after 2 weeks) at 6 weeks for calcipotriol and betamethasone valerate compared with calcipotriol twice daily [1 between-patient study; 88 participants; low quality evidence]³⁴⁴
• Patient’s assessment (clear/nearly clear) at 8 weeks for calcipotriol and betamethasone valerate compared with calcipotriol once or twice daily [1 between-patient study; 518 participants; low to moderate quality evidence]¹⁹⁷

In people with psoriasis, there was no statistically significant difference between concurrent vitamin D or vitamin D analogue and potent corticosteroid treatment (one applied in the morning and one in the evening) and vitamin D or vitamin D analogue alone for:

• Withdrawals due to adverse events at 4–8 weeks for calcipotriol and betamethasone valerate or diflucortolone valerate compared with calcipotriol once or twice daily [3 studies (2 between- and 1 within-patient); 711 participants (774 randomised units); very low quality evidence]^197,344,346
• Withdrawals due to lack of efficacy calcipotriol and betamethasone valerate or diflucortolone valerate compared with calcipotriol once or twice daily [2 studies (1 between- and 1 within-patient); 563 participants (626 randomised units); very low quality evidence]^197,346

8.1.9.3. Heterogeneity

• For the outcomes of investigator’s and patient’s assessment of achieving clear/nearly clear status heterogeneity was present. The heterogeneity was removed by separating into subgroups based on frequency of administration of vitamin D or vitamin D analogue, suggesting that concurrent use of vitamin D or vitamin D analogue and potent steroid (one applied in the morning and one in the evening) is clinically more effective than once daily vitamin D or vitamin D analogue alone, but the effect in favour of the concurrent use is smaller compared with twice daily vitamin D or vitamin D analogue application.
• There was no significant heterogeneity for the remaining outcomes but OD and BD subgroups were kept separate where necessary to avoid double counting data from the Kragballe1998 study.

8.1.10.1. Evidence profile

Table

8.1.10.2. Evidence statements

In people with psoriasis, a combined product containing calcipotriol monohydrate and betamethasone dipropionate once daily was statistically significantly better than calcipotriol once or twice daily or tacalcitol once daily for:

• Investigator’s assessment (clear/nearly clear)at 4–8 weeks [6 between-patient studies; 1249 participants; moderate quality evidence]^{102,132,179,201,208,298}
• Patient’s assessment (clear/nearly clear) at 4–8 weeks [4 between-patient studies; 2182 participants; very low quality evidence]^{132,179,208,298}
• Percentage change in PASI at 4–8 weeks [5 between-patient studies; 2334 participants; moderate quality evidence]^{102,132,179,201,208}
• Withdrawals due to adverse events at 4–8 weeks [3 between-patient studies; 1636 participants; moderate quality evidence]^132,179,208

In people with psoriasis, there was no statistically significant difference between a combined product containing calcipotriol monohydrate and betamethasone dipropionate once daily and vitamin D or vitamin D analogue once or twice daily for:

• Relapse rate at 8 weeks post-treatment for the combination product compared with tacalcitol once daily [1 between-patient study; 98 participants; very low quality evidence]²⁰⁸
• Withdrawals due to lack of efficacy at 4 weeks for the combination product compared with calcipotriol twice daily [1 between-patient study; 378 participants; very low quality evidence]¹³²
• Skin atrophy at 4–12 weeks for the combination product compared with calcipotriol twice daily [2 between-patient studies; 1027 participants; very low quality evidence]^132,201

8.1.10.3. Heterogeneity

• For the outcome of investigator’s assessment of achieving clear/nearly clear status heterogeneity was present. The heterogeneity was removed by separating into subgroups based on frequency of administration of vitamin D or vitamin D analogue, suggesting that use of combined vitamin D or vitamin D analogue and potent steroid is clinically more effective than once daily vitamin D or vitamin D analogue alone, but the effect in favour of the combined use was smaller compared with twice daily vitamin D or vitamin D analogue application.
• For the outcome of patient’s assessment of achieving clear/nearly clear status high heterogeneity was present. The heterogeneity was not fully explained by any of the pre-specified subgroups although for the comparison with once daily vitamin D or vitamin D analogue the combination was clearly clinically more effective in all studies, but again the effect in favour of the combined use was smaller compared with twice daily vitamin D or vitamin D analogue application.
• There was no significant heterogeneity for the remaining outcomes.

8.1.11.1. Evidence profile

Table

8.1.11.2. Evidence statements

In people with psoriasis, a combined product containing calcipotriol monohydrate and betamethasone dipropionate was statistically significantly better than potent corticosteroid (betamethasone dipropionate once daily) for:

• Investigator’s assessment (clear/nearly clear) at 4–8 weeks [2 between-patient studies; 1211 participants; moderate quality evidence]^102,179
• Patient’s assessment (clear/nearly clear) at 4 weeks [1 between-patient study; 966 participants; moderate quality evidence]¹⁷⁹
• Percentage change in PASI at 4–8 weeks [2 between-patient studies; 1211 participants; moderate quality evidence] ^102,179

In people with psoriasis, there was no statistically significant difference between a combined product containing calcipotriol monohydrate and betamethasone dipropionate and potent corticosteroid (betamethasone dipropionate once daily) for:

• Withdrawals due to adverse events at 4 weeks [1 between-patient study; 932 participants; very low quality evidence]¹⁷⁹

8.1.11.3. Heterogeneity

• There was no significant heterogeneity for the any of the outcomes.

8.1.12.1. Evidence profile

Table

8.1.12.2. Evidence statements

In people with psoriasis, a combined product containing calcipotriol monohydrate and betamethasone dipropionate then vitamin D or vitamin D analogue was statistically significantly better than topical vitamin D or vitamin D analogue for:

• Investigator’s assessment (clear/nearly clear) for a combined product once daily for 4 weeks then calcipotriol once daily for 4 weeks compared to tacalcitol once daily for 8 weeks; a combined product once daily for 8 weeks then calcipotriol once daily for 4 weeks vs. calcipotriol BD for 12 weeks [2 between-patient studies; 1150 participants; moderate quality evidence]^201,298
• Patient’s assessment (clear/nearly clear) for a combined product once daily for 4 weeks then calcipotriol once daily for 4 weeks compared to tacalcitol once daily for 8 weeks [1 between-patient study; 501 participants; moderate quality evidence]²⁹⁸
• Percentage change in PASI for a combined product once daily for 4 weeks then calcipotriol once daily weekdays/a combined product once daily at weekends for 8 weeks compared to calcipotriol twice daily for 12 weeks; a combined product once daily for 8 weeks then calcipotriol once daily for 4 weeks vs. calcipotriol twice daily for 12 weeks [1 between-patient study; 972 participants; moderate quality evidence]²⁰¹
• Percentage change in PASI for a combined product once daily for 4 weeks then calcipotriol once daily for 4 weeks compared to tacalcitol once daily for 8 weeks [1 between-patient study; 501 participants; moderate quality evidence]²⁹⁸

In people with psoriasis, there were no events with either a combined product containing calcipotriol monohydrate and betamethasone dipropionate then vitamin D or vitamin D analogue or topical vitamin D or vitamin D analogue for:

• Skin atrophy for a combined product once daily for 4 weeks then calcipotriol once daily weekdays/a combined product once daily at weekends for 8 weeks compared to calcipotriol twice daily for 12 weeks [1 between-patient study; 649 participants; moderate quality evidence]²⁰¹

In people with psoriasis, there was no statistically significant difference between a combined product containing calcipotriol monohydrate and betamethasone dipropionate then vitamin D or vitamin D analogue and topical vitamin D or vitamin D analogue for:

• Investigator’s assessment (clear/nearly clear) for a combined product once daily for 4 weeks then calcipotriol once daily weekdays/a combined product once daily at weekends for 8 weeks compared to calcipotriol twice daily for 12 weeks [1 between-patient study; 650 participants; low quality evidence] ²⁰¹
• Withdrawal due to adverse events for a combined product once daily for 4 weeks then calcipotriol twice daily for 8 weeks compared to calcipotriol twice daily for 12 weeks [1 between-patient study; 101 participants; very low quality evidence]³⁵¹
• Withdrawal due to adverse events for a combined product once daily for 4 weeks then calcipotriol once daily for 4 weeks compared to tacalcitol once daily for 8 weeks [1 between-patient study; 451 participants; very low quality evidence]²⁹⁸
• Withdrawal due to lack of efficacy for a combined product once daily for 4 weeks then calcipotriol twice daily for 8 weeks compared to calcipotriol twice daily for 12 weeks [1 between-patient study; 100 participants; very low quality evidence]³⁵¹
• Withdrawal due to lack of efficacy for a combined product once daily for 4 weeks then calcipotriol once daily for 4 weeks compared to tacalcitol once daily for 8 weeks [1 between-patient study; 445 participants; very low quality evidence]²⁹⁸
• Skin atrophy for a combined product once daily for 8 weeks then calcipotriol once daily for 4 weeks compared to calcipotriol twice daily for 12 weeks [1 between-patient study; 649 participants; very low quality evidence]²⁰¹

8.1.12.3. Heterogeneity

• Not applicable as the studies assessed slightly different comparisons and so were not a combined

8.1.13.1. Evidence profile

Table

8.1.13.2. Evidence statements

In people with psoriasis, there was no statistically significant difference between the maintenance regimens for 52 weeks maintenance for:

• Investigator’s assessment of treatment success (absent, very mild or mild disease) at 52 weeks [1 between-patient study; 297 participants; low to very low quality evidence]¹⁹⁶
• Skin atrophy at 52 weeks [1 between-patient study; 634 participants; very low quality evidence]¹⁹⁵
• Withdrawal due to adverse events at 52 weeks [1 between-patient study; 485 participants; very low quality evidence]^195,196
• Withdrawal due to lack of efficacy at 52 weeks [1 between-patient study; 552 participants; low to very low quality evidence]^195,196

8.1.13.3. Heterogeneity

• Not applicable as this study assessed multiple comparisons and combining all results would lead to double counting of data.

8.1.14.1. Evidence profile

Table

8.1.14.2. Evidence statements

In people with psoriasis, vitamin D or vitamin D analogue was statistically significantly better than dithranol for:

• Investigator’s assessment (clear/nearly clear) at 8–12 weeks for calcipotriol twice daily compared to dithranol once daily [3 between-patient studies; 908 participants; very low quality evidence]^29,58,417
• Patient’s assessment (clear/nearly clear) at 8–12 weeks for calcipotriol twice daily compared to dithranol once daily [2 between-patient studies; 742 participants; moderate quality evidence]^29,417
• Withdrawals due to adverse events at 8–12 weeks for calcipotriol or calcitriol twice daily compared to dithranol once daily [5 between-patient studies; 1085 participants; moderate quality evidence] ^{29,58,156,410,417}

In people with psoriasis, dithranol was statistically significantly better than vitamin D or vitamin D analogue for:

• Relapse rate at 8 weeks post treatment for calcipotriol twice daily compared to dithranol once daily [1 between-patient study; 95 participants; very low quality evidence]⁵⁸

In people with psoriasis, there was no statistically significant difference between dithranol and vitamin D or vitamin D analogue for:

• Investigator’s assessment (clear/nearly clear) at 8 weeks for calcitriol twice daily compared to dithranol once daily [1 between-patient study; 114 participants; very low quality evidence]¹⁵⁶
• Percentage change in PASI at 8 weeks for calcipotriol twice daily compared to dithranol once daily [1 between-patient study; 86 participants; low quality evidence]⁴¹⁰
• Withdrawals due to lack of efficacy at 8 weeks for calcipotriol twice daily compared to dithranol once daily [1 between-patient study; 96 participants; low quality evidence] ⁴¹⁰

8.1.14.3. Heterogeneity

• For the outcome of investigator’s assessment of achieving clear/nearly clear status heterogeneity was present. The heterogeneity was greatly reduced by separating into subgroups based on the specific vitamin D or vitamin D analogue used; suggesting that calcitriol may be less effective than dithranol but calcipotriol may be more effective. However, there was still some heterogeneity among the studies using calcipotriol, although all showed the same direction of effect.
• There was no significant heterogeneity for the remaining outcomes.

8.1.15.1. Evidence profile

Table

8.1.15.2. Evidence statements

In people with psoriasis, vitamin D or vitamin D analogue treatment was statistically significantly better than coal tar for:

• Investigator’s assessment (clear/nearly clear) at 6–8 weeks for calcipotriol twice daily compared to 15% coal tar solution in aqueous cream once daily; calcipotriol twice daily compared to coal tar polytherapy (coal tar 5%/allantoin 2%/hydrocortisone cream 0.5%) twice daily [2 studies (1 within- and 1 between-patient); 149 participants (176 randomised units); low to moderate quality evidence]^313,400
• Percentage change in PASI at 6 weeks for calcipotriol twice daily compared to 15% coal tar solution in aqueous cream once daily [1 within-patient study; 27 participants (54 randomised units); moderate quality evidence]⁴⁰⁰

In people with psoriasis, coal tar was statistically significantly better than vitamin D or vitamin D analogue for:

• Investigator’s assessment (clear/nearly clear) at 12 weeks for calcipotriol twice daily compared to coal tar solution (liquor carbonis distillate (LCD 15%, equivalent to 2.3% coal tar) twice daily [1 between-patient study; 55 participants; very low quality evidence]¹⁵
• Percentage change in PASI at 12 weeks for calcipotriol twice daily compared to coal tar solution (liquor carbonis distillate (LCD 15%, equivalent to 2.3% coal tar) twice daily [1 between-patient study; 55 participants; low quality evidence]¹⁵
• Relapse rate at 6 weeks post-treatment for calcipotriol twice daily compared to coal tar solution (liquor carbonis distillate (LCD 15%, equivalent to 2.3% coal tar) twice daily [1 between-patient study; 25 participants; very low quality evidence]¹⁵

In people with psoriasis, there were no events with either vitamin D or vitamin D analogue or coal tar for:

• Withdrawals due to adverse events at 12 weeks for calcipotriol twice daily compared to coal tar solution (liquor carbonis distillate (LCD 15%, equivalent to 2.3% coal tar) twice daily [1 between-patient study; 55 participants; low quality evidence]¹⁵

In people with psoriasis, there was no statistically significant difference between vitamin D or vitamin D analogue and coal tar for:

• Withdrawals due to adverse events at 6 weeks for calcipotriol twice daily compared to 15% coal tar solution in aqueous cream once daily [1 within patient study; 25 participants (50 randomised units); very low quality evidence]⁴⁰⁰
• Withdrawals due to adverse events at 8 weeks for calcipotriol twice daily compared to coal tar polytherapy (coal tar 5%/allantoin 2%/hydrocortisone cream 0.5%) twice daily [1 between-patient study; 116 participants; very low quality evidence]³¹³

8.1.15.3. Heterogeneity

• Heterogeneity was present for all outcomes. The heterogeneity was removed by separating into subgroups based on treatment duration. However, it is also possible that the coal tar formulation caused the heterogeneity, although this was thought to be clinically less likely to be the source of the inconsistency.

8.1.16.1. Evidence profile

Table

8.1.16.2. Evidence statements

In people with psoriasis, calcipotriol twice daily was statistically significantly better than calcipotriol once daily for:

• Investigator’s assessment (clear/nearly clear) at 8 weeks [1 within-patient study; 344 participants; low quality evidence]¹⁹⁷
• Patient’s assessment (clear/nearly clear) at 8 weeks [1 within-patient study; 344 participants; low quality evidence]¹⁹⁷

In people with psoriasis, there was no statistically significant difference between calcipotriol once daily and calcipotriol twice daily for:

• Withdrawal due to adverse events at 8 weeks [1 within-patient study; 348 participants; very low quality evidence]¹⁹⁷
• Withdrawal due to lack of efficacy at 8 weeks [1 within-patient study; 348 participants; very low quality evidence]¹⁹⁷

8.1.16.3. Heterogeneity

• Not applicable as only one study was available for this comparison

8.1.17.1. Vitamin D or vitamin D analogues

Evidence profile

Table

8.1.17.2. Potent corticosteroids

Evidence profiles

Table

8.1.17.3. Very potent corticosteroids

Evidence profile

Table

8.1.17.4. Combined product containing vitamin D or vitamin D analogue and potent corticosteroid (calcipotriol plus betamethasone dipropionate)

Evidence profile

Table

8.1.17.5. Concurrent potent corticosteroid and vitamin D or vitamin D analogue (one applied in the morning and one in the evening)

Evidence profile

Table

8.1.17.6. Coal tar

Evidence profile

Table

8.1.17.7. Dithranol

Evidence profile

Table

8.1.17.8. Tazarotene

Evidence profile

Table

8.1.18.1. Results of NMA for investigator assessed outcome: clear/nearly clear (IAGI/PGA)

Thirty-four studies^{15,25,29,49,58,73,79,81,102,125,132,148,156,167,178,179,197,208,210,211,216,227,246,255,298,302,311,313,344,360,400,417,419,428} met the inclusion criteria for the base case network meta-analysis of the investigator assessed outcome of clear/nearly clear. Three further studies^251,295,401 were included in a sensitivity analysis, the details and results of which can be found in Appendix K. Based on the GRADE quality ratings from the review of direct comparisons (section 8.1), the evidence included in the network meta-analysis ranges in quality from very low to moderate.

Figure 3 presents all the interventions included in the NMA as well as shows where there is direct evidence for a particular comparison and the number of studies that have included that comparison. For example, there are 7 studies reporting the outcome ‘clear’ or ‘nearly clear’ as measured by IAGI or PGA for the comparison of twice daily vehicle/placebo and twice daily vitamin D or vitamin D analogue. The diagram also highlights where there are gaps in the direct evidence. For example, there are no studies comparing a combined product containing vitamin D or vitamin D analogue and potent corticosteroid to concurrent vitamin D or vitamin D analogue and potent corticosteroid (one applied in the morning and one in the evening).

Figure 3

Clear or nearly clear – IAGI and PGA. Note: Solid lines indicate direct head-to-head comparisons and the colour indicates the number of trials per comparison included in the base case. Dashed lines indicate all head-to-head comparisons included (more...)

The results of the network meta-analysis in terms of the relative risk of each intervention compared to twice daily vehicle/placebo are presented in Table 62. It also gives a probability that the intervention is the most effective overall.

Table 62

Relative risks of clear/nearly clear on IAGI/PGA for all interventions compared to twice daily vehicle/placebo.

8.1.18.2. Results of NMA for patient assessed outcome: clear/nearly clear (PAGI)

Fourteen studies^{29,66,79,125,132,141,179,197,198,208,216,298,360,417} met the inclusion criteria for the base case network meta-analysis of the patient assessed outcome of clear/nearly clear. Two further studies^295,302 were included in a sensitivity analysis, the details and results of which can be found in Appendix L. Based on the GRADE quality ratings from the review of direct comparisons (section 8.1), the evidence included in the network meta-analysis ranges in quality from very low to moderate.

Figure 4 presents all the interventions included in the NMA as well as shows where there is direct evidence for a particular comparison and the number of studies that have included that comparison. From the diagram, one can see that fewer studies have reported PAGI. There are 4 studies reporting the outcome of ‘clear’ or ‘nearly clear’ as measured by PAGI (in contrast to 7 studies reporting for IAGI or PGA) for the comparison of twice daily vehicle/placebo and twice daily vitamin D or vitamin D analogue.

Figure 4

Clear or nearly clear - PAGI. Note: Solid lines indicate direct head-to-head comparisons and the colour indicates the number of trials per comparison included in the base case. Dashed lines indicate all head-to-head comparisons included in the sensitivity (more...)

The results of the network meta-analysis in terms of the relative risk of each intervention compared to twice daily vehicle/placebo are presented in Table 63. It also gives a probability that the intervention is the most effective overall.

Table 63

Relative risks of clear/nearly clear with PAGI for all interventions compared to twice daily vehicle/placebo.

8.1.19.1. Economic evidence – literature review

An economic evaluation should ideally compare all relevant alternatives. No applicable studies of good enough methodological quality were identified comparing all interventions of interest –vitamin D or vitamin D analogues, potent or very potent corticosteroids, coal tar, dithranol and retinoids – in the treatment of patients with mild to moderate chronic plaque psoriasis.

Three studies^18,37,289 were identified that included two or more of the relevant comparators. These are summarised in the economic evidence profile below (Table 64 and Table 65). See also the full study evidence tables in Appendix I.

Table 64

Calcipotriol versus short contact dithranol – Economic study characteristics.

Table 65

Calcipotriol versus short contact dithranol – Economic summary of findings.

Six studies were selectively excluded, four due to very serious methodological limitations^{110,140,234,359} and two due to the availability of more applicable economic evidence^21,310. Reasons for their exclusion are provided in Appendix G.

Ashcroft and colleagues present a simple decision tree analytic model to explore the relative cost effectiveness of topical calcipotriol and short contact dithranol. Caution should be exercised when interpreting the results of this study as it is unclear if the best possible sources were used to inform the parameters, and the short time horizon means that the costs of treatment failure may have not been fully accounted for.

Ashcroft et al. did not perform a quality of life assessment which limits its usefulness in determining cost effectiveness of the interventions studied. The below table shows the results of Ashcroft et al., with estimates of the possible incremental cost effectiveness ratio over a 1-year time horizon had quality of life measurements been incorporated. The ICERs presented below show that if utility gains of 0.03 or 0.09 are assumed (based on estimates used by other authors^37,289 in the economic review) the additional cost of calcipotriol is very unlikely to be offset by the additional benefits associated with this treatment.

Table 66

Economic summary of Ashcroft et al. findings with quality of life incorporated.

Both studies identified had potentially serious limitations with their chosen methodology. Bottomley and colleagues used an NHS provider perspective and was directly applicable, but is limited by the method used to generate estimates of treatment effect. The authors used performed an unadjusted indirect comparison which may introduce bias. The sensitivity analyses conducted by Bottomley et al. provide some indication that once daily product containing calcipotriol monohydrate and betamethasone dipropionate may be a cost effective strategy provided that the difference in utility between baseline and that experienced on the waiting list is small (i.e. 0.075). Interestingly, Bottomley and colleagues found concurrent but separate treatment with vitamin D or vitamin D analogue and potent corticosteroids to be the most expensive strategy and provided the least QALYs.

Table 67

Vitamin D or vitamin D analogues vs potent corticosteroids vs combined and concurrent vitamin D or vitamin D analogues and potent corticosteroids (one applied in the morning and one in the evening) - Economic summary of findings.

Oh and colleagues compared calcipotriol and different durations of clobetasol after first line treatment of a potent corticosteroid (betamethasone valerate) failed. Their evidence suggests that where the needed dosage and length of treatment of calcipotriol is similar or less than the ultra high potency corticosteroid clobetasol propionate, then calcipotriol might be the more cost effective second line treatment, however its incremental cost effectiveness compared to 2 weeks of very potent steroid was over the NICE £20,000 per QALY threshold. In a second analysis, they found that calcipotriol performed better as a second line treatment for psoriasis which had returned following prior treatment with betamethasone valerate or other agents, with increased utility due to lower side effects compared to fluocinonide.

8.1.19.2. Economic evidence – original economic analysis

The review of clinical evidence for topical therapies used in the treatment of individuals with mild to moderate plaque psoriasis showed that there were a wide variety of options – emollients, tars, dithranol, retinoids, corticosteroids (potent and very potent), vitamin D or vitamin D analogues and combination products – each associated with certain advantages and disadvantages. The results of the network meta-analysis suggested that some interventions, such as combined or concurrent vitamin D analogue and potent corticosteroid, were more likely to induce clearance or near clearance than others. Given that these combined and concurrent application strategies carry additional cost compared to both their individual constituent parts and compared to other topical alternatives, it was important to consider whether these additional costs are justified by additional health benefits in terms of improved quality of life.

The choice of which topical therapy to offer patients with mild to moderate psoriasis in primary care was identified as among the highest economic priorities by the GDG because the greatest proportion of psoriasis patients are managed at this point in the care pathway. Even if the unit costs of the interventions are quite modest, the population affected is relatively large; therefore the health economic impact of any recommendation is likely to be substantial.

Three cost-effectiveness analyses were identified in the published literature, but each had methodological limitations that called its conclusions into question. The analysis by Ashcroft and colleagues¹⁸ was based on only one trial and included only two of the interventions of interest (dithranol and calcipotriol). The analysis by Oh and colleagues²⁸⁹ was quite old and had analysed economic outcomes for different lines of treatment within separate models each having different comparators, thus making it difficult to identify a clearly cost-effective sequence of topical therapies. The analysis by Bottomley and colleagues,³⁷ although the most applicable of the included studies, used an unadjusted indirect comparison to inform the treatment effect estimates, which likely overestimated the effectiveness of some interventions and underestimated the effectiveness of others. Bottomley and colleagues also did not include all the possible comparators of interest. Due to the methodological limitations of the published economic analyses, there was still substantial uncertainty as to which topical therapy or therapies represented the best value for NHS resources. In order to reduce this uncertainty, an original cost-effectiveness analysis was undertaken by the guideline health economist in collaboration with the GDG. Below is a summary of the analysis that was undertaken. For full details please see Appendix M: Cost-effectiveness analysis.

8.1.19.3. Methods

An analysis was undertaken to evaluate the relative cost-effectiveness of different topical therapy sequences used in the treatment of individuals with mild to moderate chronic plaque psoriasis. A Markov model was used to estimate 12-month costs and quality-adjusted life years (QALYs) from a current UK NHS and personal social services perspective. A 12-month time horizon was considered clinically relevant and sufficiently long enough to capture important costs and consequences of first-line treatment in primary care. Uncertainty was explored through probabilistic analysis and sensitivity analysis. The performance of alternative treatment sequences was estimated using incremental cost-effectiveness ratios (ICERs), defined as the added cost of a given strategy divided by its added benefit compared with the next most expensive strategy. A threshold of £20,000 per QALY gained was used to assess cost-effectiveness.

The aim of the analysis was to identify the most cost-effective sequence of first, second and third line topical therapies. It was important to model sequences given that most patients will commence treatment with one topical and then try others before moving on to more intensive treatments such as phototherapy and/or systemic therapy. In all, 118 sequences were compared in the base case analysis. Table 68 presents the list of possible first, second and third line treatments which may be combined in a sequence.

Table 68

All possible sequences of first, second and third line interventions.

The following conditions were placed on the sequences, ensuring that they represented logical clinical practice:

• Concurrent treatment with vitamin D or vitamin D analogue and potent corticosteroid (one applied in the morning and one in the evening) would not come after a failure of once daily combined product containing calcipotriol monohydrate and betamethasone dipropionate;
• Once daily treatment with a given topical would not come after a failure of twice daily treatment with the same topical;
• Once daily treatment with potent steroid or vitamin D or vitamin D analogue would not come after concurrent treatment with vitamin D or vitamin D analogue and potent corticosteroid (one applied in the morning and one in the evening) or once daily combined product containing calcipotriol monohydrate and betamethasone dipropionate;
• No strategy could include potent corticosteroids among all three lines of treatment (including as part of concurrent vitamin D or vitamin D analogues and potent corticosteroid (one applied in the morning and one in the evening) or combined product containing calcipotriol monohydrate and betamethasone dipropionate).

Most comparators focus on evaluating a trial of three different treatments before referral for specialist review, but the GDG was also interested in whether earlier escalation of care might be more cost-effective. To test this, strategies have also been combined into two-treatment sequences with referral following a failure of second line treatment.

Due to the unacceptability of dithranol and coal tar as routine treatments (difficult application, risk of staining, strong and unpleasant odours, etc), these treatments were reserved for third line treatment only. This reflects their current placement in primary care given the availability of more acceptable and effective topicals such as those being compared as first and second line topicals. In a series of sensitivity analyses, other restrictions were placed on the potential sequences, namely due to concerns about the safety of continued use of potent corticosteroids.

The structure of the model developed by the NCGC was adapted from the model developed by Bottomley and colleagues³⁷ and was validated by the GDG as a reasonable reflection of current clinical practice. The Markov model and how patients move through the pathway is illustrated in Figure 5. Key model assumptions (these are discussed in more detail in the full write-up in Appendix M):

Figure 5

Markov model of treatment with topical therapy.

• All hypothetical patients commence treatment with a given topical and experience one of two outcomes after 4 or 8 weeks:
  • response (defined as clearance/near clearance of their psoriasis)
  • no response (defined as something less than clearance/near clearance of their psoriasis).
• Patients who respond stop treatment and they either maintain response in the absence of treatment or they relapse.
  • Patients who relapse resume treatment with the same topical and again face a probability of responding or not responding.
• Patients who do not respond to a given topical after 8 weeks of treatment are assumed to return to their GP and receive a prescription for an alternative topical therapy.
• Patients can receive up to three different topical therapies before being referred by the GP to a specialist review in an outpatient dermatology clinic where second-line treatment options could be considered.
  • Some proportion of these referred patients will be kept on topical therapies, receive support and advice at the review consultation and be discharged back to their GP for long-term management.
  • The remaining proportion undergo a course of phototherapy:

    –

    If they respond to phototherapy they are then discharged to their GP for long-term management.

    –

    If they do not respond to phototherapy they continue to be managed by a specialist.

Movement between various health states is governed by transition probabilities, derived from the systematic review of clinical effectiveness data. Thirteen 4-week cycles were modelled, resulting in a 1-year time horizon for the analysis, with a half-cycle correction applied.

Model inputs were based on the clinical effectiveness review undertaken for the guideline, other published data and expert opinion where required. These are described in full in the technical report in Appendix M. All model inputs and assumptions were validated by the GDG.

8.1.19.4. Results

This analysis found that, given a NICE willingness-to-pay threshold of £20,000 per QALY gained, the most cost-effective strategy is likely to be one of starting with twice daily potent corticosteroid and moving to concurrent potent corticosteroid and vitamin D or vitamin D analogue (one applied in the morning and one in the evening) and then twice daily coal tar. This strategy was also the least costly strategy among the 118 modelled. Base case results for non-dominated and non-extendedly dominated strategies are presented Table 69.

Table 69

Incremental analysis of base case results – psoriasis of trunk and limbs.

Results showed that starting with concurrent potent corticosteroid and vitamin D or vitamin D analogue (one applied in the morning and one in the evening) and switching to twice daily potent corticosteroid and then twice daily coal tar is £9 more costly over 1 year and only produces 0.00041 more QALYs than the least costly strategy mentioned above. This gives it an incremental cost-effectiveness ratio (ICER) of £22,658 which is just above the NICE £20,000 per QALY threshold.

The most effective strategy (once daily combined product containing calcipotriol monohydrate and betamethasone dipropionate then twice daily potent corticosteroid then twice daily coal tar) costs an additional £192 per year compared to the next most costly non-dominated strategy (concurrent steroid and vitamin D or vitamin D then twice daily potent steroid then twice daily coal tar), yet produces just 0.00107 additional QALYs for an ICER of over £179,000. Based on the results of this model, it appears that starting with once daily combined product containing calcipotriol monohydrate and betamethasone dipropionate, although most effective, is very unlikely to be cost-effective.

Results of the analysis showed that a strategy of using vehicle or emollient with no active agent only was the most costly and least effective, largely driven by the cost of referrals and specialist management for non-responders. Strategies that included once or twice daily vitamin D or vitamin D analogue were not cost-effective regardless of where they were included in the sequence. This is largely due to their relatively low rank in terms of effectiveness and their relatively high acquisition cost. Strategies that included dithranol were also all dominated, that is more costly and less effective than alternatives. Finally, strategies in which patients were referred after non-response to only 2 topicals were all dominated, thus not cost effective.

The probabilistic analysis indicates that there is a great deal of uncertainty as to which sequence is optimal (i.e. most cost effective). There appears to be very little difference between initial potent corticosteroid followed by concurrent potent corticosteroid and vitamin D or vitamin D analogue (one applied in the morning and one in the evening) and vice versa, with the difference in their net monetary benefits (NMB) being only £1 (£16,748 and £16,747 respectively) and both having an equal probability of being optimal at a £20,000 willingness to pay threshold. Generally, it looks as though a strategy of starting with either potent corticosteroids or concurrent treatment with potent corticosteroid and vitamin D or vitamin D analogue (one applied in the morning and one in the evening) is most likely to be cost-effective, whereas starting with once daily combined product containing calcipotriol monohydrate and betamethasone dipropionate is very unlikely to be cost-effective.

A series of sensitivity analyses suggested that the conclusions from the base case are sensitive to changes in some parameters and/or assumptions.

8.1.19.5. Interpretation and limitations

In assessing the relative cost-effectiveness of alternative topical therapies in patients with mild to moderate psoriasis limited evidence was available from the published economic literature. The evidence that was identified and included in the health economic review had potentially serious limitations and therefore the GDG considered it a priority to undertake original evaluation for the guideline in order to inform recommendations. This analysis showed that there were relatively small differences in terms of benefit between different topical sequences, but the differences in terms of cost were quite substantial. Based on the mean costs and benefits, the analysis suggests that initial treatment with potent corticosteroids followed by concurrent treatment with potent corticosteroid and vitamin D or vitamin D analogue (one applied in the morning and one in the evening) and followed then by twice daily coal tar therapy is likely to represent the most cost-effective sequence for implementation in primary care. Uncertainties in the analysis were explored through sensitivity analysis which showed that in some scenarios

• Once daily potent corticosteroid or concurrent treatment should come first in the sequence
• Twice daily vitamin D or vitamin D analogue should come second or third in the sequence, after concurrent treatment
• Combined product containing calcipotriol monohydrate and betamethasone dipropionate should be offered third in the sequence, after potent corticosteroids and concurrent treatment

Sequences starting with once daily combined product containing calcipotriol monohydrate and betamethasone dipropionate were slightly more effective than the same sequence starting with concurrent potent corticosteroid and vitamin D or vitamin D analogue (one applied in the morning and one in the evening); however, the very modest additional benefit (0.0011) would only be considered potentially cost-effective if willingness to pay thresholds were between £ 100,000 and £ 500,000 per QALY gained.

The analysis has several limitations which were considered carefully by the GDG. Firstly, the analysis evaluates treatment sequences even though the available trial data compares single topicals head to head without sequencing. In order to apply the treatment effects within the sequencing model, we assumed that treatment effects were independent. That is, we assumed the effectiveness of combined product containing calcipotriol monohydrate and betamethasone dipropionate as a second or third line topical was equal to its effectiveness as a first line agent and that this was true regardless of other topicals it may follow. The GDG did not believe this to be a significant limitation given that the patients included in the overwhelming majority of RCTs were reported to have psoriasis for longer than 5 years, during which the can be assumed to have previously tried, succeeded and/or failed various topical treatments.

The analysis only captured the efficacy of topicals and did not capture the costs or consequences of adverse events. Although the RCT evidence on adverse events was sparse, the GDG is conscious of the risks associated with the long-term use of potent and very potent corticosteroids. They carefully considered whether the added effect in terms of clearance was worth the potential risks of adverse effects.

The model was also focused on the induction of disease clearance as opposed to the maintenance of clearance. Trials focusing on maintenance were limited in number and inadequately reported for use in the economic model. In particular, there was uncertainty as to how maintenance treatments were applied in the trials and therefore incorporating such evidence and assumptions into the model was considered too difficult and unlikely to be valid.

The model also takes a relatively short time horizon considering that psoriasis is a chronic, long term condition for which patients may undergo treatment for many years of their lives. Frequency and severity of relapse, selection for and speed of onward referral, methods of self-management and long-term safety are all issues inadequately addressed in the evidence base and therefore translate into limitations of the economic analysis.

The model estimated the health gain for each treatment by mapping the change in PASI score to the EQ-5D based on observational evidence. However, it has been noted that several important areas of health-related quality of life for people with psoriasis are not directly assessed by the EQ-5D questionnaire²²⁶. Therefore it is possible that the EQ-5D may lack content validity for these patients. Research is ongoing in this area. But we note that even using a £ 30,000 per QALY threshold rather than £ 20,000 would not change the conclusions of our analyses. Therefore only if the EQ-5D is under-estimating health gain of one treatment compared to another by a considerable extent, could this pose a serious limitation.

8.1.19.6. Comparison with published studies

The findings from the NCGC original economic analysis are quite different from the results of the most similar published study by Bottomley and colleagues³⁷. Bottomley and colleagues found 8 weeks of once daily combined product containing calcipotriol monohydrate and betamethasone dipropionate to dominate other modelled strategies including once and twice daily vitamin D or vitamin D analogue followed by potent corticosteroid, potent corticosteroid followed by vitamin D or vitamin D analogue and 8 weeks of concurrent treatment with vitamin D or vitamin D analogue and potent corticosteroid (one applied in the morning and one in the evening). Although the analysis appears to have been executed well, the estimates of effect and resource use had limitations which called the conclusions of the analysis into question.

The biggest differences in the results of the NCGC analysis presented here and the analysis undertaken by Bottomley has to do with the treatment effect sizes used. In their analysis, concurrent treatment was found to be very ineffective, with just 14.9% of patients responding with a PASI75 compared to the combined product containing calcipotriol monohydrate and betamethasone dipropionate to which 50.3% of patients responded (RR=3.38). The NCGC analysis showed a much small difference between these treatments, with 65.1% of patients responding to concurrent treatment and 70.7% responding to The combined product containing calcipotriol monohydrate and betamethasone dipropionate (RR=1.09).

In addition, the estimate they used for quantity of topical used per 4-week treatment period was 92.6 g, compared to the estimate used in the NCGC analysis 134.0 g. Based on these estimates of resource use, the NCGC analysis assumes 4 weeks of the combined product containing calcipotriol monohydrate and betamethasone dipropionate costs £ 29.26 more than Bottomley and colleagues did. Furthermore, the difference between the combined product containing calcipotriol monohydrate and betamethasone dipropionate and concurrent treatment is different between the analyses. The additional cost of the combined product containing calcipotriol monohydrate and betamethasone dipropionate was £ 36.91 in Bottomley and more than twice that, £ 76.34, in the NCGC analysis. We performed a sensitivity analysis in which we assumed the same quantity of the combined product containing calcipotriol monohydrate and betamethasone dipropionate used by Bottomley and colleagues (i.e. 92.6 g, £ 61.27). The ICER for the combined product containing calcipotriol monohydrate and betamethasone dipropionate improved compared to the base case (£ 124,400 vs £ 174,545), but was still well above the NICE cost-effectiveness threshold of £ 20,000 per additional QALY.

The one thing that Bottomley and colleagues were able to capture that the NCGC analysis was not had to do with the potential disutilities associated with adverse events; however these inputs were not reported, were not included in their base case and, their impact on the results were not reported in full. The authors simply state that the influence of AEs ‘had no impact on the results.’

8.1.19.7. Evidence statements

• One partially applicable study with potentially serious limitations found that short-contact dithranol may be more cost-effective than calcipotriol.
• One directly applicable study with potentially serious limitations found that a combined product containing calcipotriol monohydrate and betamethasone dipropionate administered once daily may be more cost effective than concurrent but separate treatment with vitamin D or vitamin D analogue and potent corticosteroids (one applied in the morning and one in the evening) and both vitamin D or vitamin D analogue alone (once daily and twice daily) and potent corticosteroids alone (once daily).
• One partially applicable study with potentially serious limitations found that six weeks of vitamin D or vitamin D analogue offered after a trial of potent corticosteroids is likely to be cost effective compared to four or six weeks of very potent corticosteroids offered after a trial of potent corticosteroids; however, it is less likely to be cost effective compared to two weeks of very potent corticosteroids.
• One partially applicable study with potentially serious limitations found that vitamin D or vitamin D analogue offered after failure of potent corticosteroid is likely to be cost effective compared to continued treatment with alternative potent corticosteroids.
• New economic analysis from a current UK NHS and PSS perspective comparing 118 different sequences of topical therapies found twice daily potent corticosteroids or concurrent treatment (one in the morning and one in the evening) with potent corticosteroid and vitamin D or vitamin D analogue to be the most cost-effective options for the first and second line treatment of patients with mild to moderate chronic plaque psoriasis. This conclusion was robust to the majority of sensitivity analyses undertaken.
  • The base case and sensitivity analyses showed that the choice of third line treatment in a given sequence was highly uncertain. Depending upon the data used and assumptions made, third line treatment with twice daily coal tar, twice daily vitamin D or vitamin D analogue or once daily combined product containing calcipotriol monohydrate and betamethasone dipropionate was likely to be most cost-effective.

8.2.2.1. Vitamin D or vitamin D analogue vs. placebo

Evidence profile

Table

8.2.2.2. Potent corticosteroid vs. placebo

Evidence profile

Table

8.2.2.3. Very potent corticosteroid vs. placebo

Evidence profile

Table

8.2.2.4. Combined product containing potent corticosteroid and vitamin D analogue (betamethasone dipropionate and calcipotriol) vs. placebo

Evidence profile

Table

8.2.2.5. Very potent corticosteroid vs. placebo for maintenance of remission

One study assessed the efficacy and safety of clobetasol propionate compared with placebo as a maintenance treatment for up to 6 months among those who had achieved clear, very mild or mild disease during a 4-week induction phase with once-daily clobetasol propionate. During the maintenance phase clobetasol propionate was used twice-weekly (3 days apart), but once daily dosing was permitted for up to 4 weeks if relapse occurred.

Evidence profile

Table

8.2.2.6. Vitamin D or vitamin D analogue vs. potent corticosteroid

Evidence profile

Table

8.2.2.7. Vitamin D or vitamin D analogue vs. very potent corticosteroid

Evidence profile

Table

8.2.2.8. Combined product containing vitamin D analogue and potent corticosteroid (betamethasone dipropionate and calcipotriol) vs. potent corticosteroid

Evidence profile

Table

8.2.2.9. Combined product containing vitamin D analogue and potent corticosteroid (betamethasone dipropionate and calcipotriol) vs. vitamin D or vitamin D analogue

One study²²⁸ assessed long-term (52 weeks) treatment for this comparison. This study used a once daily administration schedule as required by the participants and the mean treatment duration was 44 weeks and 37 weeks for the combination and vitamin D or vitamin D groups, respectively (mean weekly weight used: 10.6g in two compound group and 12.8g in calcipotriol group; mean weight used over whole study period 470.8g and 440.0g, respectively).

Evidence profile

Table

8.2.2.10. Very potent corticosteroid vs. coal tar polytherapy

Evidence profile

Table

8.2.2.11. Vitamin D analogue vs. coal tar polytherapy

Evidence profile

Table

8.2.3.1. Vitamin D or vitamin D analogues

Evidence profile

Table

8.2.3.2. Potent corticosteroids

Evidence profile

Table

8.2.3.3. Very potent corticosteroids

Evidence profile

Table

8.2.3.4. Combined product containing potent corticosteroid and vitamin D analogue (betamethasone dipropionate and calcipotriol monohydrate)

Evidence profile

Table

8.2.3.5. Coal tar

Evidence profile

Table

8.2.4.1. Results of NMA for investigator assessed outcome: clear/nearly clear (IAGI/PGA)

A total of 13 studies^{108,109,128,166,168,189,199,245,292,330,377,405,408} from the original evidence review met the inclusion criteria for the network. Based on the GRADE quality ratings from the review of direct comparisons (section 8.2.2), the evidence included in the network meta-analysis ranges in quality from very low to moderate.

Figure 6 presents all the interventions included in the NMA as well as shows where there is direct evidence for a particular comparison and the number of studies that have included that comparison. For example, there are 3 studies reporting the outcome ‘clear’ or ‘nearly clear’ as measured by IAGI or PGA for the comparison of twice daily vehicle/placebo and twice daily very potent corticosteroid. The diagram also highlights where there are gaps in the direct evidence. For example, there are no studies comparing a combined product containing vitamin D or vitamin D analogue and potent corticosteroid to very potent corticosteroid.

Figure 6

Clear or nearly clear – IAGI and PGA. Note: Solid lines indicate direct head-to-head comparisons and the colour indicates the number of trials per comparison included in the analysis.

The results of the network meta-analysis in terms of the relative risk of each intervention compared to twice daily vehicle/placebo are presented in Table 71. It also gives a probability that the intervention is the most effective overall.

Table 71

Relative risks of clear/nearly clear on IAGI/PGA for all interventions compared to twice daily vehicle/placebo.

8.2.5.1. Economic evidence – literature review (scalp psoriasis)

One study⁶ was included that included relevant comparisons. It is summarised in the economic evidence profile below (Table 72 and Table 73). See also the full study evidence tables in Appendix I. No studies were excluded.

Table 72

Economic study characteristics.

Table 73

Economic summary of findings.

Although not presented in the above profile because they were dominated, it is worth noting themes from the overall analysis of all 12 treatment comparators. Overall, strategies that did not include combined or concurrent vitamin D or vitamin D analogue and potent corticosteroids (one applied in the morning and one in the evening) generated fewer QALYs and higher costs than those that did. In fact, the analysis showed that a strategy of starting with vitamin D or vitamin D analogue once daily and escalating to twice daily and then moving finally to Capasal (salicylic acid and coal tar shampoo) once daily was the most costly and the least effective of all 12 strategies.

There was little difference between the overall effectiveness (QALYs gained) of strategies depending upon when in the sequence the combined product containing calcipotriol monohydrate and betamethasone dipropionate came (first-, second- or third-line). Costs also did not seem to follow a pattern based on where combination product came in the sequence, but seemed to be driven more by what other treatments were in the sequence (e.g. once or twice daily vitamin D or vitamin D analogue and/or potent corticosteroid).

8.2.5.2. Economic evidence – original economic analysis (scalp psoriasis)

The review of clinical evidence for topical therapies used in the treatment of individuals with moderate to severe scalp psoriasis showed that there were several treatment options – tars, corticosteroids (potent and very potent), vitamin D or vitamin D analogues and combination products – each associated with certain advantages and disadvantages. The results of the network meta-analysis indicated that some interventions, such as very potent corticosteroid as well as combined product containing calcipotriol monohydrate and betamethasone dipropionate, were more likely to induce clearance or near clearance than others. Given that these combined and concurrent application strategies carry additional cost compared to both their individual constituent parts and compared to other topical alternatives, it was important to consider whether these additional costs are justified by additional health benefits in terms of improved quality of life.

The choice of which topical therapy to offer patients with moderate to severe scalp psoriasis in primary care was identified as among the highest economic priorities by the GDG because scalp psoriasis affects a large proportion of patients and is typically managed in primary care. As with topicals used to treat other body sites, even if the unit costs of the interventions are quite modest, the population affected is relatively large; therefore the health economic impact of any recommendation is likely to be substantial.

One cost-effectiveness analysis was identified in the published literature, but it had methodological limitations that called its conclusions into question. The analysis by Affleck⁶ did not include all of the relevant comparators under consideration for the guideline, namely very potent corticosteroids. Furthermore, the treatment effects used in their analysis differed from those found in the NCGC clinical review and network meta-analysis, and this difference was considered likely to affect the conclusion of the analysis. Due to these methodological limitations, there was still substantial uncertainty as to which topical therapy or therapies represented the best value for NHS resources in the treatment of scalp psoriasis. In order to reduce this uncertainty, an original cost-effectiveness analysis was undertaken by the guideline health economist in collaboration with the GDG. Below is a summary of the analysis that was undertaken. For full details please see Appendix N.

8.2.5.3. Methods

An analysis was undertaken to evaluate the relative cost-effectiveness of different topical therapy sequences used in the treatment of individuals with moderate to severe scalp psoriasis. A Markov model was used to estimate 12-month costs and quality-adjusted life years (QALYs) from a current UK NHS and personal social services perspective. A 12-month time horizon was considered clinically relevant and sufficiently long enough to capture important costs and consequences of first-line treatment in primary care. Uncertainty was explored through probabilistic analysis and sensitivity analysis. The performance of alternative treatment sequences was estimated using incremental cost-effectiveness ratios (ICERs), defined as the added cost of a given strategy divided by its added benefit compared with the next most expensive strategy. A threshold of £ 20,000 per QALY gained was used to assess cost-effectiveness.

The aim of the analysis was to identify the most cost-effective sequence of first, second and third line topical therapies for scalp psoriasis. It was important to model sequences given that most patients will commence treatment with one topical and then try others before moving on to more intensive treatments such as specialist applied topicals and/or systemic therapy. Table 74 presents the list of possible first, second and third line scalp treatments which may be combined in a sequence.

Table 74

Possible sequences of first, second and third line treatment.

The following conditions were placed on the sequences, ensuring that they represented logical clinical practice:

• Once daily treatment with a given topical would not come after a failure of twice daily treatment with the same topical;
• Once daily treatment with potent steroid or vitamin D or vitamin D analogue would not come after once daily combined product containing calcipotriol monohydrate and betamethasone dipropionate;
• Once or twice daily treatment with potent corticosteroid would not come after once or twice daily with very potent corticosteroid.

Most comparators focus on evaluating a trial of three different treatments before referral for specialist review, but the GDG was also interested in whether earlier escalation of care might be more cost-effective. To test this, strategies have also been combined into two-treatment sequences with referral following a failure of second line treatment.

Due to the unacceptability of coal tar as a routine treatment (strong and unpleasant odours), this treatment was reserved for third line treatment only. This reflects their current placement in primary care given the availability of more acceptable and effective topicals such as those being compared as first and second line topicals.

The structure of the model developed by the NCGC was adapted from the model developed by Affleck and colleagues⁶ and was validated by the GDG as a reasonable reflection of current clinical practice. The Markov model and how patients move through the pathway is illustrated in Figure 7. Key model assumptions (these are discussed in more detail in the full write-up in Appendix N):

Figure 7

Patient flow diagram for the Markov model of topical treatments for scalp psoriasis.

• All hypothetical patients commence treatment with a given topical and experience one of two outcomes after 4 or 8 weeks:
  • response (defined as clearance/near clearance of their scalp psoriasis) or
  • no response (defined as something less than clearance/near clearance of their scalp psoriasis).
• Patients who respond stop treatment and they either maintain response in the absence of treatment or they relapse.
  • Patients who relapse resume treatment with the same topical and again face a probability of responding or not responding.
• Patients who do not respond to a given topical after 8 weeks of treatment are assumed to return to their GP and receive a prescription for an alternative topical therapy.
• Patients can receive up to three different topical therapies before being referred by the GP to a specialist review in an outpatient dermatology clinic where second-line treatment options could be considered.
  • Some proportion of these referred patients will be kept on topical therapies, receive support and advice at the review consultation and be discharged back to their GP for long-term management.
  • Some will be treated by a specialist over 3 appointments in outpatient dermatology
  • The remaining proportion undergo a supervised scalp treatment with intensive topical therapy over the course of 3 dermatology day centre appointments:

    –

    If they respond to intensive topical therapy they are then discharged to their GP for long-term management.

    –

    If they do not respond to intensive topical therapy they continue to be managed by a specialist.

Movement between various health states is governed by transition probabilities, derived from the systematic review of clinical effectiveness data and network meta-analysis. Thirteen 4-week cycles were modelled, resulting in a 1-year time horizon for the analysis, with a half-cycle correction applied.

Model inputs were based on the clinical effectiveness review undertaken for the guideline, other published data and expert opinion where required. These are described in full in the technical report in Appendix N. All model inputs and assumptions were validated by the GDG.

8.2.5.4. Results

This analysis found that, given a NICE willingness-to-pay threshold of £ 20,000 per QALY gained, the most effective and cost-effective strategy is likely to be one of starting with once daily very potent corticosteroid and then escalating to twice daily very potent corticosteroid and then trying once daily TCF product if very potent steroids alone are insufficient to induce clearance or near clearance. This conclusion was based on the comparison of mean costs and mean QALYs across 169 modelled sequences. Base case results for non-dominated and non-extendedly dominated strategies are presented in Table 75. By starting with very potent corticosteroids once and then twice daily followed by TCF product was expected to generate 0.0014 more QALYs for an additional cost of £ 26.80 compared to the least costly sequence (once daily potent corticosteroid followed by once and then twice daily very potent corticosteroids). This gives and ICER of £ 19,143 per QALY gained, which is just under the NICE cost-effectiveness threshold. Based on total net monetary benefits and probabilities of being most cost-effective, there is little difference between the two strategies.

Table 75

Incremental analysis of base case results – scalp psoriasis.

Complete results for all 169 comparators can be found in Appendix N. Overall, results of the analysis showed that the most effective (and cost-effective) strategies involved use of potent and very potent corticosteroids in at least two lines of treatment.

Results also showed that a strategy of using vehicle gel or emollient with no active agent only was the most costly and least effective strategy, largely driven by the cost of referrals and specialist management for non-responders. Similarly, a strategy of prescribing coal tar polytherapy for ongoing management was only slightly more effective than continued use of vehicle gel and cost the third most of any treatment sequence. Compared to strategies relying heavily on corticosteroids, strategies that included once or twice daily vitamin D analogue were unlikely to be cost-effective regardless of where they came in a treatment sequence. This finding is driven by their relatively low rank in terms of effectiveness and their relatively high acquisition cost relative to potent and very potent corticosteroids. Two compound formulation product, although third most effective in the network meta-analysis, was found to be cost-effective only as a third line intervention following very potent corticosteroids. Like vitamin D analogues, its high unit cost compared to other cheaper and effective topicals makes it unlikely to represent reasonable value for NHS resources.

The probabilistic analysis indicates that there is a great deal of uncertainty as to which sequence is optimal (i.e. most cost-effective). No single sequence was most cost-effective at a £20,000 per QALY willingness to pay threshold in more than 30% of simulations; however, looking across strategies indicates that those starting with once daily potent corticosteroid were optimal in 43% of simulations. In 33% of all simulations, following once daily potent with once or twice daily very potent corticosteroid was optimal. In another 44% of simulations, a sequence starting with either once or twice daily very potent corticosteroid was likely to be most cost-effective. The remaining 13% of simulations indicated that twice daily potent corticosteroids was an optimal first line strategy. These trends can also be seen by looking at the rank order of strategies in Table 36 of Appendix N, which shows that those starting with potent and very potent corticosteroids have the highest mean net benefits. These statistics indicate that we can be reasonably confident that starting with once daily potent or very potent corticosteroid is going to bring the greatest benefit for resources used, and that escalating to a twice daily very potent corticosteroid is likely to provide further benefit at reasonable extra cost.

A series of scenario analysis suggested that the conclusions from the base case are somewhat sensitive to changes in assumptions made.

8.2.5.5. Interpretation and limitations

In assessing the relative cost-effectiveness of alternative topical therapies in patients with moderate to severe scalp psoriasis limited evidence was available from the published economic literature. The evidence that was identified and included in the health economic review had potentially serious limitations and therefore the GDG considered it a priority to undertake original evaluation for the guideline in order to inform recommendations.

Original decision modelling undertaken for the guideline showed that there were relatively small differences in terms of benefit between 169 different topical sequences, but the differences in terms of cost were quite substantial. Based on the mean costs and benefits, the analysis suggests that initial treatment with once daily very potent corticosteroid followed by twice daily very potent corticosteroid and then once daily TCF product if very potent corticosteroids alone are insufficient to induce clearance or near clearance is likely to represent the most cost-effective sequence for moderate to severe scalp psoriasis. Uncertainties in the analysis were explored through sensitivity analysis which showed that in some scenarios in which restrictions were placed on the comparators

• Once daily potent corticosteroid is likely to be the optimal first line treatment if very potent corticosteroids are considered too aggressive.
• Once or twice daily vitamin D or analogues are likely to be cost-effective second in the sequence, after trials of potent or very potent corticosteroids, particularly where continuous corticosteroids are to be avoided
• Once or twice daily very potent corticosteroids is likely to be the most cost-effective third line treatment if potent corticosteroid and vitamin D have not worked
• TCF product may be cost-effective, but only after potent and/or very potent corticosteroids have failed and when only once daily applications of topicals is being considered

In general, sequences including once daily TCF product were slightly more effective than the same sequence including alternatives such as vitamin D analogue or potent corticosteroid; however, the very modest additional benefits (<0.001 and dependent on comparator) would only be considered potentially cost-effective if willingness to pay thresholds were substantially greater than £20,000 per QALY gained. If, however, the amount of TCF product used by patients is less than reported in the clinical trial evidence, such that a single 60 g pack is needed for 4 weeks, then TCF product may be cost-effective as a second or third line treatment following potent corticosteroids. Under no conditions was first line use of TCF product likely to represent better value for NHS resources than potent or very potent corticosteroids.

The analysis has several limitations which were considered carefully by the GDG. Firstly, the analysis evaluates treatment sequences even though the available trial data compares single topicals head to head without sequencing. In order to apply the treatment effects within the sequencing model, we assumed that treatment effects were independent. That is, we assumed the effectiveness of the combined product containing calcipotriol monohydrate and betamethasone dipropionate as a second or third line topical was equal to its effectiveness as a first line agent and that this was true regardless of other topicals it may follow. The GDG did not believe this to be a significant limitation given that the patients included in the overwhelming majority of RCTs were reported to have psoriasis for longer than 5 years, during which they can be assumed to have previously tried, succeeded and/or failed various topical treatments.

The analysis only captured the efficacy of topicals and did not capture the costs or consequences of adverse events. Although the RCT evidence on adverse events was sparse, the GDG is conscious of the risks associated with the long-term use of potent and very potent corticosteroids. They carefully considered whether the added effect in terms of clearance was worth the potential risks of adverse effects.

The model was also focused on the induction of disease clearance as opposed to the maintenance of clearance. No trials focusing on maintenance were identified in the clinical evidence review and therefore no evidence was available for use in the economic model.

The model also takes a relatively short time horizon considering that psoriasis of the scalp is a chronic, long term condition for which patients may take up treatment intermittently for many years of their lives. Frequency and severity of relapse, selection for and speed of onward referral, methods of self-management and long-term safety are all issues inadequately addressed in the evidence base and therefore translate into limitations of the economic analysis. Longer time horizons of up to 5 years were explored in sensitivity analyses and conclusions were insensitive to these extensions.

The model estimated the health gain for each treatment by mapping the change in PASI score to the EQ-5D based on observational evidence. However, it has been noted that several important areas of health-related quality of life for people with psoriasis are not directly assessed by the EQ-5D questionnaire²²⁶. Therefore it is possible that the EQ-5D may lack content validity for these patients. Research is ongoing in this area. But we note that even using a £30,000 per QALY threshold rather than £20,000 would not change the conclusions of our analyses. Therefore only if the EQ-5D is under-estimating health gain of one treatment compared to another by a considerable extent, could this pose a serious limitation.

This analysis of the treatment of psoriasis of the scalp is distinct from the analysis of the treatment of scalp of the trunk and/or limbs largely because it is based on a different evidence base and as such has given rise to site-specific recommendations. In clinical practice, healthcare professionals are likely to see patients who are dealing with psoriasis at a variety of sites, including their face and flexures. It is quite possible that healthcare professionals will need to prescribe different topicals for different sites, meaning that patients may have several different agents at a time. Indeed, even if they are using the same product (i.e. potent corticosteroid) on different sites, they may be prescribed different formulations for each site (i.e. creams or ointments for the trunk and limbs; gels or foams for the scalp). It would be simpler to prescribe one single treatment for all sites, but as the clinical and cost-effectiveness has shown, such an approach may not represent the most effective or efficient use of NHS resources.

8.2.5.6. Comparison with published studies

The findings from the NCGC original economic analysis are quite different from the results of the most similar published study by Affleck and colleagues⁶. Affleck and colleagues found a sequence starting with twice daily potent corticosteroids followed by concurrent treatment with vitamin D or vitamin D analogue and potent corticosteroid corticosteroids (one applied in the morning and one in the evening) and then once daily combined product containing calcipotriol monohydrate and betamethasone dipropionate to be most cost-effective. Although the analysis appears to have been executed well, the included comparators and the estimates of effect and resource use had limitations which called the conclusions of the analysis into question.

The biggest differences in the results of the NCGC analysis presented here and the analysis undertaken by Affleck has to do with the comparators included, namely the inclusion/exclusion of very potent corticosteroids. The NCGC analysis included very potent corticosteroids as the network meta-analysis demonstrated them to be highly efficacious in the short term treatment of psoriasis of the scalp. The GDG confirmed that although very potent corticosteroids are not normal management for the treatment of the trunks and limbs, they constitute a reasonable, short-term option for treating the scalp.

The second key difference between the analyses relates to the relative treatment effects used. Affleck and colleagues derived their treatment effects from an adjusted indirect comparison³⁸, which, when compared to the NCGC network meta-analysis, appears to have overestimated the effectiveness of TCF product compared to other topicals. For example, in their analysis TCF product was found to be 2.45 times more likely to induce response than once daily calcipotriol (RR=2.45, 95% CI: 1.84 to 3.27). The NCGC network meta-analysis found the risk ratio to be lower, around 1.857. This translates into an absolute risk difference between the two comparators of 35.54% using Affleck’s estimates and 29.65% using the NCGC estimates. Differences such as these add up when synthesised in economic models and could lead to biased conclusions.

In addition, the estimate they used for quantity of TCF product used per 4-week treatment period was 60 g, compared to the estimate used in the NCGC analysis 71.4 g. Based on these estimates of resource use, the NCGC analysis assumes 4 weeks of TCF product costs £31.29 more than Affleck and colleagues did. We performed a sensitivity analysis in which we assumed the same quantity of TCF product used by Affleck and colleagues (i.e. 60 g, £36.50). The ICER for TCF product as a third line treatment improved significantly compared to the base case, making it potentially cost-effective given the NICE willingness to pay threshold. However, there remains a great deal of uncertainty in this conclusion.

One thing that Affleck and colleagues were able to capture that the NCGC analysis was not had to do with the potential disutilities associated with adverse events. They included these in their base case, and unfortunately did not report a sensitivity analysis wherein they were removed altogether with which to compare. However, the authors did state that variation in the incidence of adverse events, upwards and downwards, did not change the conclusions of their analysis.

8.2.5.7. Evidence statements

• One directly applicable study with potentially serious limitations found that a sequence of potent corticosteroid followed by concurrent vitamin D or vitamin D analogue and potent corticosteroid corticosteroids (one applied in the morning and one in the evening) and followed by the combined product containing calcipotriol monohydrate and betamethasone dipropionate to be the most cost-effective strategy to treat chronic scalp psoriasis.
• One directly applicable study with potentially serious limitations found that treatment sequences that do not include combined or concurrent vitamin D or vitamin D analogue and potent corticosteroids (one applied in the morning and one in the evening) are among the least effective and most costly in the treatment of chronic scalp psoriasis.
• New economic analysis from a current UK NHS and PSS perspective comparing 169 different sequences of topical therapies found sequences beginning with once daily very potent corticosteroids to offer the best value for NHS resource in the treatment of patients with moderate to severe scalp psoriasis; however, this conclusion was sensitive to many sensitivity and scenario analyses undertaken.
  • The most consistently cost-effective first line treatment when very potent corticosteroids were excluded was once daily potent corticosteroid. This conclusion was robust to the majority of sensitivity and scenario analyses undertaken.
  • Choice of second and third line treatments was more uncertain, but very potent corticosteroids, once or twice daily, were generally shown to be most cost effective, followed in rank order by once or twice daily vitamin D or analogue and then once daily two-compound formulation product. This conclusion was somewhat sensitive to alternative assumptions regarding suitability and acceptability of certain comparators.

    –

    Sensitivity analyses in which continuous or consecutive use of topicals containing steroids was restricted found that once and twice daily vitamin D analogues are cost-effective as second line treatments in sequences with potent and very potent corticosteroids.

    –

    Sensitivity analyses in which only once daily applications were considered found that initial treatment with potent steroids was optimal, followed by either very potent corticosteroid and then two-compound formulation product if steroids could be used continuously or followed by vitamin D analogue and very potent corticosteroid if continued use of steroids was to be avoided.

8.2.6.1. Tacrolimus vs. placebo

Evidence profile

Table

8.2.6.2. Pimecrolimus vs. placebo

Evidence profile

Table

8.2.6.3. Tacrolimus vs. vitamin D or vitamin D analogue

Evidence profile

Table

In people with chronic plaque psoriasis affecting the face and/or genitofemoral areas, there were no events with either tacrolimus twice daily or vitamin D (calcitriol twice daily) for:

• Withdrawal due to adverse events at 6 weeks [1 study; 46 participants; moderate quality evidence]²²⁰
• Withdrawal due lack of efficacy at 6 weeks [1 study; 46 participants; moderate quality evidence]²²⁰

In people with chronic plaque psoriasis affecting the face and/or genitofemoral areas, there was no statistically significant difference between tacrolimus twice daily and vitamin D (calcitriol twice daily) for:

• Investigator’s assessment (clear/nearly clear) at 6 weeks [1 study; 49 participants; low quality evidence]²²⁰

8.2.7.1. Tacrolimus

Evidence profile

Table

8.2.7.2. Pimecrolimus

Evidence profile

Table

8.2.7.3. Summary

The evidence suggests that maximum response may be achieved by 8 weeks, with the continued response rate increasing only slightly between weeks 6 and 8¹²⁹. However, the majority of those who will respond within 8 weeks had done so by week 4.