Literature search results
Our search retrieved 1359 records after removing duplications, 61 of them were eligible for full-text screening according to our inclusion criteria. We included 28 trials in our final analysis.16,20−47 Kragballe et al. 2006 reported data from the same population in two papers; one reported efficacy,29 and the other reported safety.28 The flow of data collection and screening process are shown in figure.1.
Summary of the included studies
Most of the trials included adult patients with plaque psoriasis for more than 6 months duration and more than 2% of the body or 10% of the scalp. They excluded patients with other forms of psoriasis rather than plaque psoriasis and excluded patients on current or recent treatment regimens with phototherapy, biological therapy, or any other medications that may affect psoriasis. Gel, ointment, aerosol foam, scalp formulation, and topical suspension are the topical therapy forms in the included studies. Summary of the included trials and baseline characters of enrolled subjects are shown in tables.1 and 2, respectively.
Quality assessment
All included single-arm trials were of good to fair quality according to the NIH assessment tool (Table S.1). Most of the included RCTs had low or unclear risk of bias regarding selection and detection biases, low risk of bias regarding performance, attrition, and reporting biases, and high risk of bias in the "other bias" domain. Figures S.1 and S.2 show the risk of bias graph and summary for the included RCTs, and Table S.2 shows the authors' judgment of each risk of bias domain with justification.
Outcomes
- Efficacy outcomes
PASI score change (body psoriasis)
The pooled estimate showed that Cal/BD significantly reduces PASI score at all measured time-points, but the results were heterogeneous (P < 0.0001, I2 > 75%). At one week; MD=-3.444, 95% CI [-2.554, -4.334], P < 0.0001. At two weeks; MD=-4.301, 95% CI [-4.018, -4.584], P < 0.0001. At four weeks; MD=-5.841, 95% CI [-5.120, -6.561], P < 0.0001. At eight weeks; MD=-5.215, 95% CI [-3.169, -7.260], P < 0.0001 (Fig. 2).
PASI 75 (body psoriasis)
This analysis showed a low or insignificant rate of PASI 75 (patients achieving 95% PASI score improvement) with Cal/BD at different time points, and the results were heterogeneous (P < 0.02, I2 > 80%) except at eight weeks (P = 0.776, I2 = 0%). At one week; ER = 0.037, 95% CI [0.008, 0.146], P < 0.0001. At four weeks; ER = 0.415, 95% CI [0.296, 0.545], P = 0.199; heterogeneity was solved by excluding Langley et al. 201132 (P = 0.233, I2 = 29.85%) and the results remained insignificant; ER = 0.482, 95% CI [0.433, 0.532], P = 0.483. At eight weeks; ER = 0.342, 95% CI [0.303, 0.383], P < 0.0001. At 12 weeks; ER = 0.529, 95% CI [0.380, 0.674], P = 0.706 (figure.S.3).
Treatment success according to PGA/IGA
For body psoriasis, PGA/IGA –based treatment success rate was low or insignificant at measured timepoints, and the pooled trials were heterogeneous (P < 0.0001, I2 > 40%) except for one week (P = 0.113, I2 = 43.91%). At one week; ER = 0.042, 95% CI [0.027, 0.065], P < 0.0001. At two weeks; ER = 0.173, 95% CI [0.098, 0.287], P < 0.0001; heterogeneity was solved by excluding Koo et al. 201627 (P = 0.865, I2 = 0%) and the results remained significant; ER = 0.101, 95% CI [0.066, 0.151], P < 0.0001. At four weeks; ER = 0.522, 95% CI [0.432, 0.611], P = 0.628. At eight weeks; ER = 0.411, 95% CI [0.298, 0.534], P = 0.153. At 12 weeks; ER = 0.505, 95% CI [0.291, 0.718], P = 0.964 (figure S.4).
For scalp psoriasis, the pooled estimate showed that PGA/IGA-based treatment success rates with Cal/BD was low at one week, insignificant at two weeks, and high at four and eight weeks. The drug effect appeared to be duration-dependent, and the pooled trials were heterogeneous (P < 0.03, I2 > 80%). At one weeks ER = 0.333, 95% CI [0.205, 0.491], P = 0.039. At two weeks; ER = 0.564, 95% CI [0.473, 0.650], P = 0.168. At four weeks; ER = 0.664, 95% CI [0.566, 0.749], P = 0.001. At eight weeks; ER = 0.710, 95% CI [0.644, 0.769], P < 0.0001 (figure.S.5).
Treatment success according to PaGA
For body psoriasis, the pooled estimate showed high PaGA treatment success rate with Cal/BD at four weeks but it was insignificant at eight weeks of treatment. Pooled studies were heterogeneous (P < 0.0001, I2 > 90%). At four weeks; ER = 0.658, 95% CI [0.564, 0.741], P = 0.001. At eight weeks; ER = 0.497, 95% CI [0.26, 0.73], P = 0.981 (figure S.6).
For scalp psoriasis, the pooled estimate showed high rates of PaGA-based treatment success with Cal/BD at all measured timepoints, but the results were heterogenous (P < 0.0001, I2 > 85%). At two weeks; ER = 0.660, 95% CI [0.515, 0.780], P = 0.031. At four weeks; ER = 0.751, 95% CI [0.600, 0.859], P = 0.002. At eight weeks; ER = 0.766, 95% CI [0.698, 0.822], P < 0.0001 (figure S.7).
TSS change (scalp psoriasis)
The pooled estimate showed that Cal/BD significantly reduces TSS score at eight weeks; MD=-5.035, 95% CI [-5.443, -4.628], P < 0.0001. Pooled trials were heterogenous (P < 0.0001, I2 = 98.35%) (Fig. 3).
Treatment success according to TSS (scalp psoriasis)
This analysis showed insignificant TSS-based treatment success rate with Cal/BD at eight weeks of treatment; ER = 0.469, 95% CI [0.307, 0.638], P = 0.725. Pooled trials were heterogenous (P = 0.001, I2 = 85.07%). Heterogeneity was solved by excluding Gooderham et al. 201423 (P = 0.995, I2 = 0%) and the results became significantly low; ER = 0.387, 95% CI [0.327, 0.450], P = 0.001 (figure S.8).
- Safety outcomes
The analysis results showed that Cal/BD has low rates of all studied adverse events for both body psoriasis and scalp psoriasis populations. Table 3 shows the detailed results of each studied adverse event.
Table 2
Baseline table. Data are presented as mean ± SD or number (percentage). Abbreviations: Cal/BD, calcipotriol/betamethasone dipropionate; PASI, psoriasis area and severity index. PGA/IGA, physician/investigator's global assessment.
Study ID
|
Subgroups
|
Subjects
|
Age (years)
|
White/Caucasian
|
Males
|
PASI score
|
Duration of psoriasis (years)
|
Total body surface psoriatic area (percentage)
|
PGA/IGA
|
Mild
|
Moderate
|
Severe/very severe
|
Lebwohl 201633
|
-
|
100
|
47.4 ± 14.8
|
93 (93)
|
53 (53)
|
8.8 ± 4.6
|
14.6 ± 13.8
|
6.7 ± 4.9
|
9 (9)
|
77 (77)
|
14 (14)
|
Paul 201740
|
-
|
185
|
54 ± 14.5
|
122 (86.5)
|
126 (68.1)
|
7.1 ± 4.5
|
19.3 ± 14.1
|
7.1 ± 5.7
|
54 (29.2)
|
109 (58.9)
|
22 (11.9)
|
Koo 201627
|
Foam
|
141
|
51 ± 10.5
|
122 (86.5)
|
87 (61.7)
|
7 ± 4.2
|
16.1 ± 8.3
|
7.7 ± 4.7
|
22 (15.6)
|
112 (79.4)
|
7 (5)
|
Ointment
|
135
|
52 ± 11.7
|
118 (87.4)
|
87 (64.4)
|
6.7 ± 3.3
|
16.3 ± 8.5
|
7.4 ± 4.7
|
22 (16.3)
|
106 (78.5)
|
7 (5.2)
|
Ma 201636
|
-
|
120
|
39.9 ± 13.1
|
-
|
73 (60.8)
|
-
|
7.4 ± 7.3
|
-
|
-
|
83 (69.2)
|
37 (30.9)
|
Queille-Roussel 201541
|
-
|
24
|
50.25 ± 13.5
|
-
|
16 (66.7)
|
-
|
22.75 ± 11.5
|
-
|
-
|
-
|
-
|
Reich 201542
|
-
|
1795
|
50.9 ± 15.22
|
173 (94.5)
|
1034 (57.6)
|
4.5 ± 2.19
|
17.7 ± 14.7
|
|
600 (33.4)
|
1117 (62.2)
|
78 (4.3)
|
Langley 201132
|
-
|
183
|
50.9 ± 14.3
|
75 (55.6)
|
117 (63.9)
|
8.93 ± 5.75
|
21.2 ± 13.2
|
9 ± 7.7
|
-
|
130 (71)
|
53 (28.9)
|
Tyring 201045
|
-
|
135
|
44.4 ± 9.5
|
158 (97.5)
|
84 (62.2)
|
-
|
10.5 ± 7.3
|
-
|
-
|
110 (81.5)
|
25 (18.5)
|
Fleming 201022
|
-
|
162
|
50.1 ± 14.9
|
205 (99)
|
93 (57.4)
|
7.7 ± 4.6
|
18.5 ± 13.8
|
-
|
31 (19.1)
|
95 (58.6)
|
36 (22.2)
|
Kragballe 200930
|
-
|
207
|
50.8 ± 15.3
|
559 (98.4)
|
90 (43.5)
|
-
|
18.4 ± 13.8
|
-
|
|
113 (54.6)
|
94 (45.4)
|
Vandekerkhof 200946
|
-
|
568
|
48.5 ± 16.4
|
404 (96.4)
|
238 (41.9)
|
-
|
15.7 ± 13.3
|
-
|
56 (9.9)
|
311 (54.8)
|
201 (35.4)
|
Luger 200835
|
-
|
419
|
48.5 ± 15.3
|
518 (95.7)
|
185 (44.2)
|
-
|
17.7 ± 13.6
|
-
|
|
233 (55.6)
|
186 (44.4)
|
Jemec 200825
|
-
|
541
|
47.9 ± 15.4
|
105 (97.2)
|
259 (47.9)
|
-
|
15.4 ± 13.5
|
-
|
47 (8.7)
|
304 (56.2)
|
190 (35.1)
|
Buckley 200820
|
-
|
108
|
48.4 ± 16.5
|
-
|
47 (43.5)
|
-
|
16 ± 15.5
|
-
|
-
|
-
|
-
|
Saraceno 200743
|
-
|
75
|
49.1 ± 15.9
|
-
|
45 (60)
|
9.44 ± 5.37
|
11.9 ± 12.5
|
19 ± 8
|
-
|
-
|
-
|
White 200647
|
Calcipotriol
|
383
|
49.7 ± 11.13
|
369 (97.9)
|
223 (58.2)
|
8.5 ± 3.5
|
15 ± 12
|
11.9 ± 4.5
|
-
|
299 (78.1)
|
84 (21.9)
|
Alternating
|
377
|
51.4 ± 11.16
|
358 (95.2)
|
232 (61.5)
|
9.1 ± 3.9
|
15 ± 12.3
|
12.3 ± 4.7
|
-
|
288 (76.4)
|
89 (23.6)
|
Vehicle
|
376
|
51.1 ± 11.83
|
204 (96.2)
|
235 (62.5)
|
9 ± 3.9
|
17.5 ± 11.17
|
12.2 ± 4.7
|
-
|
276 (73.4)
|
100 (26.6)
|
Kragballe 200628,29
|
Cal/BD
|
212
|
48
|
210 (98.6)
|
128 (60.4)
|
-
|
19.6
|
-
|
-
|
149 (70.3)
|
56 (26.4)
|
Alternating
|
213
|
49.8
|
203 (97.1)
|
125 (58.7)
|
-
|
19.3
|
-
|
-
|
151 (70.9)
|
58 (27.2)
|
Cal/BD then Cal
|
209
|
48.4
|
236 (94.8)
|
134 (64.1)
|
-
|
20.1
|
-
|
-
|
138 (66)
|
63 (30.1)
|
Ortonne 200438
|
-
|
249
|
50.8 ± 14.9
|
311 (96.6)
|
134 (53.8)
|
9.7 ± 6.2
|
20.2 ± 15.2
|
-
|
-
|
-
|
-
|
Kragballe 200431
|
Group 1
|
322
|
47.8 ± 13.2
|
309 (95.7)
|
201 (62.4)
|
10.3 ± 7.83
|
18.3 ± 11.7
|
-
|
-
|
-
|
-
|
Group 2
|
323
|
47.8 ± 11.7
|
|
206(63.8)
|
10.4 ± 6.67
|
19 ± 10
|
-
|
-
|
-
|
-
|
Papp 200339
|
-
|
304
|
47.5
|
473 (96.5)
|
173 (56.9)
|
|
19.2
|
-
|
-
|
-
|
-
|
Kaufmann 200226
|
-
|
490
|
47.6 ± 10.7
|
-
|
308(62.9)
|
9.9 ± 6.93
|
18.3 ± 11
|
-
|
-
|
-
|
-
|
Guenther 200224
|
Once/d
|
152
|
47.9
|
369 (99.2)
|
90 (59.2)
|
9.9
|
18.3
|
-
|
-
|
-
|
-
|
Twice/d
|
237
|
49.3
|
165 (69.6)
|
10.6
|
18.3
|
-
|
-
|
-
|
-
|
Douglas 200216
|
-
|
372
|
47.6 ± 11.8
|
28 (90)
|
216 (58.1)
|
10.8 ± 5.98
|
19 ± 10.3
|
-
|
-
|
-
|
-
|
Eichenfield 201521
|
-
|
31
|
15 ± 1.25
|
707 (90)
|
12 (39)
|
-
|
4.7 ± 2.5
|
5.2 ± 3
|
-
|
21 (68)
|
10 (33)
|
Gooderham 201423
|
-
|
78
|
14.6 ± 0.83
|
-
|
35 (45)
|
-
|
5.9 ± 2.83
|
5 ± 8.17
|
-
|
58 (74)
|
20 (26)
|
Lee 201634
|
-
|
193
|
48.74 ± 13.61
|
-
|
104 (53.89)
|
-
|
-
|
4.32 ± 2.6
|
93 (48.19)
|
91 (47.15)
|
9 (4.66)
|
Ortonne 201437
|
-
|
159
|
46.7 ± 13.6
|
30 (81)
|
99 (62.3)
|
-
|
-
|
72.5 ± 45
|
-
|
-
|
-
|
Taraska 201644
|
-
|
37
|
48 ± 14
|
-
|
20 (54)
|
-
|
16 ± 12
|
21 ± 5
|
-
|
28 (76)
|
9 (24)
|
Table 3
Safety outcomes. Abbreviations: AEs, adverse events; ER, event rate; CI, confidence interval.
Adverse event
|
Psoriasis type
|
Timepoint
|
ER [95% CI]
|
P value
|
ER significance
|
Heterogeneity
|
Chi2 P value
|
Figure
|
Any AEs
|
Body psoriasis
|
4 weeks
|
0.194 [0.147, 0.252]
|
P < 0.0001
|
Low rate
|
Heterogenous
|
P < 0.0001
|
Figure 4
|
8 weeks
|
0.346 [0.238, 0.473]
|
P = 0.018
|
Low rate
|
Heterogenous
|
P < 0.0001
|
After excluding Lee et al. 2016
|
0.409 [0.358, 0.462]
|
P = 0.001
|
Low rate
|
Homogenous
|
P = 0.581
|
12 weeks
|
0.306 [0.169, 0.489]
|
P = 0.039
|
Low rate
|
Heterogenous
|
P < 0.0001
|
After excluding Saraceno et al. 2007
|
0.434 [0.385, 0.485]
|
P = 0.012
|
Low rate
|
Homogenous
|
P = 0.484
|
Scalp psoriasis
|
8 weeks
|
0.365 [0.341, 0.390]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.35
|
Figure 5
|
AEs leading to discontinuation
|
Body psoriasis
|
4 weeks
|
0.008 [0.004, 0.018]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.369
|
Figure S.12
|
Diarrhea
|
Body psoriasis
|
4 weeks
|
0.013 [0.004, 0.043]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.281
|
|
12 weeks
|
0.017 [0.008, 0.037]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.409
|
|
Dry skin
|
Body psoriasis
|
8 weeks
|
0.014 [0.005, 0.036]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.285
|
|
Scalp psoriasis
|
8 weeks
|
0.004 [0.001, 0.013]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.56
|
|
Erythema
|
Body psoriasis
|
8 weeks
|
0.005 [0.001, 0.023]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.632
|
|
Scalp psoriasis
|
8 weeks
|
0.001 [0.000, 0.006]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.98
|
|
Headache
|
Body psoriasis
|
4 weeks
|
0.092 [0.011, 0.473]
|
P = 0.040
|
Low rate
|
Heterogenous
|
P = 0.047
|
|
Scalp psoriasis
|
8 weeks
|
0.024 [0.004, 0.137]
|
P < 0.0001
|
Low rate
|
Heterogenous
|
P = 0.088
|
|
Lesional or perilesional AEs
|
Body psoriasis
|
4 weeks
|
0.078 [0.058, 0.103]
|
P < 0.0001
|
Low rate
|
Heterogenous
|
P = 0.013
|
|
After excluding Ortonne et al. 2004
|
0.086 [0.070, 0.106]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.191
|
|
Scalp psoriasis
|
8 weeks
|
0.031 [0.016, 0.057]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.829
|
|
Pharyngitis
|
Body psoriasis
|
4 weeks
|
0.018 [0.009, 0.035]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.576
|
|
12 weeks
|
0.031 [0.017, 0.055]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.351
|
|
Scalp psoriasis
|
8 weeks
|
0.055 [0.025, 0.118]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.785
|
|
Pruritis
|
Body psoriasis
|
4 weeks
|
0.041 [0.030, 0.058]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.128
|
|
8 weeks
|
0.021 [0.011, 0.040]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 153
|
|
12 weeks
|
0.021 [0.010, 0.043]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.26
|
|
Scalp psoriasis
|
8 weeks
|
0.023 [0.017, 0.033]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.455
|
|
Psoriasis flaring
|
Body psoriasis
|
4 weeks
|
0.004 [0.001, 0.015]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.764
|
|
8 weeks
|
0.006 [0.001, 0.023]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.927
|
|
12 weeks
|
0.031 [0.017, 0.054]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.379
|
|
Scalp psoriasis
|
8 weeks
|
0.002 [0.001, 0.008]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.599
|
|
Serious AEs
|
Body psoriasis
|
8 weeks
|
0.019 [0.009, 0.039]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.767
|
Figure S.13
|
12 weeks
|
0.004 [0.001, 0.021]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.688
|
Scalp psoriasis
|
8 weeks
|
0.005 [0.001, 0.037]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.875
|
|
Alopecia
|
Scalp psoriasis
|
8 weeks
|
0.004 [0.001, 0.012]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.297
|
|
Skin burning sensation
|
Scalp psoriasis
|
8 weeks
|
0.002 [0.001, 0.007]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.984
|
|
Skin inflammation
|
Scalp psoriasis
|
8 weeks
|
0.004 [0.001, 0.019]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.208
|
|
Skin irritation
|
Scalp psoriasis
|
8 weeks
|
0.004 [0.002, 0.011]
|
P < 0.0001
|
Low rate
|
Homogenous
|
P = 0.612
|
|
- Publication bias
Publication bias assessment was applicable for PASI score change, PGA/IGA treatment success, and any adverse events; each in body psoriasis patients. No significant publication bias were detected for PASI score change (P = 0.17986) (figure S.9) and any adverse events (P = 0.11314) (figure S.10) as shown by the symmetry of the funnel plot and Egger's test. While publication bias was significant for PGA/IGA treatment success (P = 0.04876) (figure S.11) as shown by the asymmetry of the funnel plot and Egger's test.