PowerPoint Presentation


Figure 1: IBM Watson Explorys Universe Demographics

Poster presented at the 2017 AMCP Nexus Meeting in Dallas, TX; October 16-19, 2017.

The psoriasis patient journey: progression from topical to biologic 
treatment for psoriasis patients in the United States

Jashin J Wu1; Minyi Lu2; Karen A Veverka2; Maartje Smulders3; Eros 
Papademetriou3; Yunjua Junhua3; Steven R Feldman4

1Department of Dermatology, Kaiser Permanente Los Angeles Medical Center, Los Angeles, CA, USA; 2LEO Pharma Inc., Madison, NJ, 
USA; 3SmartAnalyst, NYC, New York, USA; 4Department of Dermatology, Wake Forest University School of Medicine, North Carolina, USA

Objective Results

Conclusions

Methods

Acknowledgements

References

This study aims to describe the patient treatment 
journey of PsO-diagnosed patients who initially 
receive a topical prescription and then receive a 
biologic treatment during follow-up.

Retrospective observational cohort study

Database: IBM Watson Explorys (Figure 1)
 Electronic health records (EHR) database made up from 39 integrated delivery 

networks (IDNs) across US and comprised of approximately 55 million patients

Patient inclusion
 Index topical date: initial topical prescription b/w Jan 1, 2011 and June 30, 2015 
 Adult patients: ≥ 18 years at index topical date
 PsO diagnosis: ICD-9 CM diagnosis code 696.1 or ICD-10 CM diagnosis codes 

(L40.0, L40.8, L40.9) between 6 months prior to or 2 months post index topical 
date

 Patients had either ≥ 2 PsO diagnoses made by any physician or 1 PsO diagnosis 
made by a dermatologist. Patients were required to not have a topical prescription 
6 months prior to the index topical date. Patients were also required to have activity 
in the EHR for at least 12 months prior to the index topical date and at least 36 
months after the index topical date. Outcomes were assessed in a 36-month 
follow-up period.

Patient exclusion 
 Patients who received an oral or biologic agent prescription in the 12 months prior 

to the index topical date.
 Patients who had a confounding dermatological or non-dermatological condition 

(where a TNF inhibitor is indicated) that was diagnosed during or before the PsO
diagnosis date.

Prescription drugs classes included for analyses
 Topical steroids: grouped into class 1-2 or 3-7 steroids
 Oral agents: cyclosporine, methotrexate, acitretin, hydroxyurea, leflunomide, 

apremilast, azatioprine, dexamethasone, isoretinoin, methoxsalen, 
methylprednisolone, prednisone, tacrolimus, and sulfasalazine

 Biologic agents: etanercept, adalimumab, ustekinumab, infliximab, and 
secukinumab

Statistical analyses 
 Between-group patient characteristics were compared using analysis of variance 

for continuous variables and Chi-square tests for categorical variables without 
adjustment. Multivariable Cox regression analysis was used to assess whether 
the difference between time to biologic was significantly different between patients 
who directly switched to or added on a biologic versus patients who initially 
received an oral agent prescription followed by a biologic prescription

Duration of Treatment Use Definitions
 For all drug types, no distinction was made between a patient switching 

medication(s) or adding on medication(s).
 Topical drugs: each new and refill prescription for a topical drug was assumed to 

be used for 30 days. A gap of 45 days or longer was assumed to be a stop or 
pause in topical treatment

 Oral agents: the duration of use was calculated by dividing the total quantity of 
tablets prescribed by the tablets administered per day, based on dosing 
instructions. When one of these fields was not available for a given prescription, 
we substituted the calculated median duration from prescriptions where it was 
available for each drug

 Biologics: the duration of use was calculated by multiplying the recommended 
maintenance dosing schedule for each medication by the number of syringes 
dispensed. Since infliximab is given intravenously, we assumed a duration of 56 
days for each treatment as per the recommended maintenance dosing schedule

Table 1. Patient Cohort Demographic
Patient Cohort

 There were 33,397,271 adult patients aged ≥ 18 years in 
the Explorys database between January 1, 2011 and June 
30, 2015, of whom 143,133 (0.4 %) had at least one PsO
diagnosis. 

 Of these patients, 38,390 had either ≥2 PsO diagnoses by 
any physician type or 1 PsO diagnosis made by a 
dermatology specialist and no confounding conditions 
before PsO diagnosis.

 Within this cohort, 24,481 patients received a topical 
prescription and 22,484 were defined as being topical 
treatment-naive. Of these patients, 6729 had a psoriasis 
diagnosis and started topical treatment, met all the study 
inclusion criteria and were included in the analysis.     
(Table 1)

Demographics

 Mean age of patients at time of switch to or add on of a 
biologic (48 years, SD 13.3) was significantly lower than 
mean age (54.5 years, SD 15.4) of patients at first topical 
treatment who never received a biologic in the 36-month 
follow-up period (P<0.001) (Table 1)

Duration of Topical Use

 Within the total patient cohort, 1712 (25.4%) stopped topical 
treatment in the 36-month follow-up period. The remaining 
5017 (74.6%) patients continued, or stopped and then 
restarted, topical treatment in the 36-month follow-up 
period.

 The median cumulative duration of any topical use was 
approximately 365 days (IQR: 233-558 days) (Table 2)

 Approximately 1 in 10 people starting topical treatment for 
psoriasis switch from topical to biologics over a 3-year 
period.

 The time to initiating the biologic after topical treatment is 
longer for patients who initially switched to an oral psoriasis 
treatment followed by a biologic treatment.

 Maintaining patients on an effective topical treatment may 
help minimize the need for a switch to biologics

Limitations:

 The results of this study were derived from a large 
retrospective EHR database which primarily covers provider 
administrative data, including written prescriptions. 
However, no medication dispensing data is included. 
Therefore, these results mainly describe provider 
prescribing behavior, rather than actual medication use, 
which could be used by the patient, and could be slightly 
different.

 It is important to note that yearly psoriasis prevalence 
among eligible patients in the database varied from 0.4-0.5%, 
which is a rather small prevalence.

 It is possible that some patients with a biologic do not 
receive a topical treatment first; these patients would been 
excluded by our methodology.

 As the database grows, more patient behavior related results 
can be presented.

Drug Total patients n (%)
Steroids 1-2 (± another drug) 270 (63.7%)

Steroids 3-7 (± another drug) 176 (41.5%)

Betamethasone and Calcipotriene (± another drug) 65 (15.3%)

Vitamin D (± another drug) 51 (12.0%)

* p<0.001 

Table 3A. Topical agents prescribed prior to biologic switch or add-on (N = 424)

Patients were most frequently (63.74%) prescribed a class 1-2 steroid prior to biologic start (Table 3A)

1. Smith J, Cline A, Feldman SR. Advances in Psoriasis. South Med J. 2017 Jan;110(1):65-75.

2. Menter A, Gottlieb A, Feldman SR, Van Voorhees AS, Leonardi CL, Gordon KB, et al. Guidelinesof
care for the management of psoriasis and psoriatic arthritis: section 1. Overview of psoriasisand
guidelines of care for the treatment of psoriasis with biologics. J Am Acad Dermatol2008;58(5):826-50.

This study was based in part on EHR data from the Explorys, an 
IBM Watson database, conducted by SmartAnalyst, and sponsored 
by LEO Pharma. Editorial support was provided by Dharm Patel, 
PhD at LEO Pharma Inc., US.  

Patient 
population 
(N=6,729)

Patients with biologic during 36-
month follow-up:

No (n = 5988) Yes (n=741)

Female sex, n (%) 53.8% 53.5% 56.4%
Mean age, years (SD)* 53.8 (15.3) 54.5 (15.4) 48 (13.3)
Median age, years (IQR) 55(43,65) 56(44,66) 48 (39,58)
Ethnicity

African-American, n (%) 288 (4.3%) 246 (4.1%) 42 (5.7%)

Asian, n (%) 59 (0.9%) 53 (0.9%) 6 (0.8%)

Caucasian, n (%) 5890 (87.5%) 5257 (87.8%) 633 (85.4%)

Hispanic/Latino, n (%) 48 (0.7%) 37 (0.6%) 11 (1.5%)

Multi-racial, n (%) 36 (0.5%) 32 (0.5%) 4 (0.5%)

Other, n (%) 224 (3.3%) 193 (3.2%) 31 (4.2%)

Missing, n (%) 184 (2.7%) 170 (2.8%) 14 (1.9%)

Region
East, n (%) 925 (13.7%) 848 (14.2%) 77 (10.4%)

Midwest, n (%) 3669 (54.5%) 3304 (55.2%) 365 (49.3%)

South, n (%) 1342 (19.9%) 1099 (18.4%) 243 (32.8%)

West, n (%) 550 (8.2%) 502 (8.4%) 48 (6.5%)

Missing, n (%) 243 (3.6%) 235 (3.9%) 8 (1.1%)

Index Topical Agent for Patients who continued 
topical treatment in the 36-month follow-up period 

(N=5017)

Estimated median cumulative duration 
during 36-month follow-up period days 

(IQR)

Steroid Class 1-2 392.5 (247-572.5)
Steroid Class 3-7 368 (246-571)

Vitamin D 437.5 (301-633.5)
Calcipotriol / Betamethasone 401 (261-591)

NSAID 360 (270-590.5)
Vitamin A 396 (253-586)
Anthralin 310 (240-582)

Steroid 1-2 392.5 (247-572.5)

Table 2. Estimated Median Cumulative Duration of Topical 
Use by Class (n = 5017)

Time to switch to/add-on of Biologic

 The median time from topical start date to 
biologic start date was significantly greater for 
patients (N=317) who initially received an oral 
agent prescription followed by a biologic 
prescription (488 days; IQR: 238 - 797 days), 
compared to patients (N=424) who directly 
switched from/added a topical to a biologic 
(206 days; IQR: 71.5 – 523 days) (p<0.0001). 
(Figure 2)
o Patient groups were controlled for age, 

gender and ethnicity 

Patient cohort treatment patterns

 During the 36-month follow up period, 317 
(42.7%) patients received an oral agent 
prescription initially, followed by a biologic 
prescription; while 424 (57.3%) patients 
directly switched to or added on a biologic 
prescription.

 Of this second group, 160 (37.7%) 
subsequently received an oral agent 
prescription after the biologic. (Figure 3)

 Of the patients who directly switched to or 
added a biologic prescription first (N = 741), 
adalimumab was the most common treatment 
(N = 349 [74.1%]). (Figure 4)

Figure 2:  Time from Index Topical to Initiation of Biologic

Figure 4:  Biologics Prescribed First within Patient Cohort (n = 741)

Figure 3:  Patient cohort treatment patterns

Agent Total patients n (%)
Methotrextate 164 (51.7%)
All_prednisone 117 (36.9%)
Dexamethasone 37 (11.7%)
Acitretin 28 (8.8%)
Sulfasalazine 17 (5.4%)
Cyclosporine 15 (4.7%)
Apremilast 14 (4.4%)
Leflunomide 5 (1.6%)
Hydroxyurea 2 (0.6%)
Methoxsalen 2 (0.6%)
Isotretinoin 1 (0.3%)
Methylprednisolone 1 (0.3%)
Tacrolimus 1 (0.3%)

Oral agents were prescribed directly before the switch to/add on of a biologic for a total of 317 patients and 
the top two were methotrexate (51.7%) and prednisone (36.9%) (Table 3B)
The median cumulative duration of use of specific oral agents prescribed before a biologic prescription was 
123 days for methotrexate (IQR: 56-229), and 28 days for prednisone (IQR: 14-58). 

Table 3B. Topical agents prescribed prior to biologic switch or add-on (N = 424)


	Slide Number 1