References
1. Migden MR et al. Cancer Treat Rev. 2018;64:1–10.

2. Lear JT et al. Br J Cancer. 2014;111:1476–1481.

3. Kudchadkar R et al. Semin Oncol. 2012;39:139–144.

4. American Cancer Society. Treating basal cell carcinoma; 2019. Available at: https://www.cancer.org/cancer/
basal-and-squamous-cell-skin-cancer/treating/basal-cell-carcinoma.html. [Accessed July 24, 2020].

5. Trakatelli M et al. Eur J Dermatol. 2014;24:312–329.

6. National Comprehensive Cancer Network. NCCN Clinical practice guidelines in oncology: basal cell skin cancer 
version 1.2020; 2019. Available at: https://www.nccn.org/professionals/physician_gls/pdf/nmsc.pdf. [Accessed 
June 2, 2020].

7. Cancer Council Australia/Australian Cancer Network. Basal cell carcinoma, squamous cell carcinoma (and 
related lesions) – a guide to clinical management in Australia; 2008; 2008. Available at: https://www.cancer.org.
au/content/pdf/HealthProfessionals/ClinicalGuidelines/Basal_cell_carcinoma_Squamous_cell_carcinoma_Guide_
Nov_2008-Final_with_Corrigendums.pdf. [Accessed July 24, 2020].

8. Goldenberg G et al. J Am Acad Dermatol. 2016;75:957–966.

9. Sekulic A et al. J Am Acad Dermatol. 2015;72:1021–1026 e1028.

10. Basset-Seguin N et al. Eur J Cancer. 2017;86:334–348.

Acknowledgments
The study was funded by Regeneron Pharmaceuticals, Inc. and Sanofi. Editorial writing support was provided  
by Atif Riaz, PhD, of Prime, Knutsford, UK, funded by Regeneron Pharmaceuticals, Inc. and Sanofi.

Disclosures
Jessica J Jalbert, Chieh-I Chen, Ning Wu, Matthew G Fury, and Wenzhen Ge are employees and 
shareholders of Regeneron Pharmaceuticals, Inc. Emily Ruiz has received consulting fees from PellePharm, 
Sanofi, Regeneron Pharmaceuticals, Inc., and Leo Pharmaceuticals, and is on the consulting and advisory board 
for Checkpoint Therapeutics.

Summary and Conclusion
• The majority (>99%) of patients initiated vismodegib. 

• Most patients (~70%) used HHIs alone versus using HHIs as potential 
neoadjuvant or adjuvant treatment.  

• Although variability was observed based on the selected grace period, median 
treatment duration was <180 days even when using a 120-day grace period.

 - Median duration of HHI treatment was considerably shorter than the median 
treatment duration (8.6–13.3 months) reported in the pivotal clinical trials  
of vismodegib.8,9  

 - Median duration of use was shorter (by >20 days) for patients using HHIs  
as potential neoadjuvant or adjuvant treatment compared to patients using 
HHI monotherapy.

• Few patients reinitiated HHIs after discontinuation. 

 - The probability of HHI reinitiation within 12 months of discontinuation was 
20% using a grace period of 60 days. 

 - The probability of reinitiation at 12 months was similar among patients using 
HHIs only and those taking HHIs potentially neoadjuvantly. 

• Characterizing real-world persistence on HHIs is challenging due to the likely use of 
drug holidays, as treatment discontinuations, treatment duration, and reinitiations 
were sensitive to the duration of the grace period selected. Moreover, results were 
sensitive to the type of HHI use (i.e. potential adjuvant or neoadjuvant treatment).

• Real-world studies of medications with tolerability issues should consider use  
of drug holidays, and in oncology, the intent of treatment (i.e. adjuvant/
neoadjuvant) use should be considered when characterizing drug exposure.

Synopsis
• In the US, it is estimated that >2 million patients are diagnosed with basal cell 

carcinoma (BCC) annually.1 

• Most cases of BCC are treatable by surgery; however, in approximately 1% of 
patients who develop advanced BCC (including metastatic and locally advanced 
BCC) additional therapies are required.1-8 

• Oral targeted therapies such as the hedgehog inhibitors (HHIs) vismodegib and 
sonidegib have shown clinical benefit in advanced BCC.8,9 However, patients 
treated with HHIs commonly experience side effects, which may require permanent 
or temporary treatment discontinuation.9,10 

• We hypothesized that HHI treatment patterns may vary according to the selected 
grace period.

Objective
• To assess the impact of grace period selection on patterns of HHI treatment 

discontinuation, treatment duration, and reinitiations among patients with BCC.

Methods
• This retrospective, observational cohort study used data from the IBM MarketScan® 

Commercial and Medicare claims database.

• We identified new users of HHIs (vismodegib or sonidegib) by selecting the first  
HHI dispensation (date of first dispensation = index date) during the study period 
(January 1, 2013 and March 30, 2019) and requiring ≥6 months of continuous 
health plan enrollment with medical and pharmacy benefits (i.e. baseline period).  
To be included in the study, patients were also required to be ≥18 years on the 
index date and to have ≥1 diagnosis of BCC during the baseline period (including 
the index date).

• HHI use was identified using National Drug Codes, and BCC diagnoses were 
identified using International Classification of Diseases (ICD)-9 and ICD-10 codes.

• HHI use was categorized according to whether it was HHI only, potentially 
adjuvant, neoadjuvant, or both neoadjuvant and adjuvant use:

 - HHI only: no surgery or radiotherapy within 60 days before or after initiation of HHIs
 - Potential adjuvant: surgery or radiotherapy followed by initiation of HHIs within  
60 days

 - Potential neoadjuvant: initiation of HHIs followed by surgery and radiotherapy 
within 60 days

 - Both: surgery or radiotherapy both within 60 days before and after HHI initiation.
• Surgery and radiotherapy were identified using ICD-9 and ICD-10 procedure codes, 

Current Procedural Terminology, 4th Edition codes, or Healthcare Common 
Procedure Coding System codes.  

• HHI treatment discontinuations were defined as the lack of an HHI dispensation 
before the exhaustion of the days’ supply and allotted grace period. A grace period 
is the gap between the exhaustion of the days’ supply and a subsequent refill that 
is permitted for treatment to be considered continuous. 

 - HHIs are generally dispensed in 30-days’ supply, benefit of treatment is continuously 
assessed, and treatment interruptions are commonly employed during HHI therapy.10 
We, therefore, used grace periods of 14, 30, 60, 90, and 120 days.

• HHI treatment duration was estimated as days between the index date to the end 
of treatment (last dispensation + days’ supply).

• Reinitiation was defined as ≥1 HHI dispensation after treatment interruption.

• The risk of HHI treatment discontinuation at 12 months, and median time (95% 
confidence intervals [CIs]) to treatment discontinuation was estimated via Kaplan–
Meier survival analysis using different grace periods and stratified by type of HHI 
use (i.e. potential adjuvant, neoadjuvant, or both neoadjuvant and adjuvant use) 
using a 60-day grace period. 

Patterns of hedgehog inhibitor (HHI) treatment interruptions and  
reinitiations among patients with basal cell carcinoma (BCC) 

in real-world clinical practice  
Jessica J Jalbert,1 Chieh-I Chen,1 Ning Wu,1 Matthew G Fury,1 Emily Ruiz,2 Wenzhen Ge1

1Regeneron Pharmaceuticals, Inc., Tarrytown, NY, USA; 2Brigham and Women’s Hospital, Boston, MA, USA

• Among patients with HHI treatment discontinuation, the risk of 12-month HHI 
reinitiation, median time (95% CIs) to reinitiation, and the risk of treatment 
discontinuation following second HHI use at 12 months was estimated via  
Kaplan–Meier analysis and stratified by type of HHI use (i.e. potential adjuvant, 
neoadjuvant, or both neoadjuvant and adjuvant use) using a 60-day grace period.

Results
Patient demographics
• A total of 526 patients who initiated HHIs were included in this study (Table 1).

• The mean (standard deviation [SD]) age was 67.0 (15.8) years; 65.4% were men.

• The proportion of patients covered by commercial insurance was 53.4%.

• The majority (75.9%) of patients had BCC of the head and neck.

Presented at the 2020 Fall Clinical Dermatology Conference, October 29-November 1, Virtual Scientific Meeting (encore of ICPE 2020 poster presentation).

Table 1. Baseline characteristics of patients with BCC initiating HHI treatment

N=526

Age, years

Mean (SD) 67.0 (15.8)

Median (Q1–Q3) 64.0 (56.0–81.0)

Male, n (%) 344 (65.4)

Payer, n (%)

Commercial 281 (53.4)

Medicare 245 (46.6)

Geographic US region, n (%)†

Midwest 107 (21.7)

Northeast 72 (14.6)

South 233 (47.3)

West 81 (16.4)

HHI index year, n (%)

2013–2014 166 (31.6)

2015–2016 176 (33.5)

2017–2018 145 (27.6)

2019 39 (7.4)

Vismodegib, n (%) 522 (99.2)

Potential adjuvant or neoadjuvant HHI 
use, n (%)

HHI only 366 (69.6)

Potential neoadjuvant 20 (3.8)

Potential adjuvant 117 (22.2)

Both‡ 23 (4.4)

BCC location, n (%)#

Head and neck 399 (75.9)

Limb 140 (26.6)

Trunk 166 (31.6)

CCI score, n (%)

0 518 (98.5)

≥1 8 (1.5)
†Geographic data is missing for 33 (6.3%) patients. Percentages are calculated per number of patients with available data. 
‡Surgery or radiotherapy both within 60 days before and after HHI initiation. 
#Tumor locations are not mutually exclusive. Patients can have tumors at multiple sites.
CCI, Charlson comorbidity index (excluding cancer); Q, quintile.

Table 3. Patterns of HHI discontinuation and reinitiation using a 60-day grace period, stratified by type of HHI use

First episode of HHI use
HHI reinitiation following  

first discontinuation
Second episode  

of HHI use

Starting 
sample  
size, n

Patients  
discontinuing 

HHIs, n

Median  
treatment 

duration, days,  
(95% CI)

Risk of  
discontinuation  
at 6 months, % 

(95% CI)

Patients 
reinitiating 

HHIs, n

Probability of  
reinitiation at 
12 months, % 

(95% CI)

Patients  
discontinuing 

HHIs, n

Median 
treatment 

duration days,  
(95% CI)

Overall 526 360 144 (131–162) 60.1 (55.0–64.6) 75 19.7 (15.0–24.2) 53 118 (89–172)

Potential neoadjuvant use 20 16 107 (86–NA) 52.0 (23.4–69.9) 2 16.9 (0–35.9) 2 183 (114–NA)

Potential adjuvant use 117 82 128 (118–171) 63.3 (52.1–71.9) 21 22.0 (11.7–31.1) 16 118 (63–NA)

Both 23 15 90 (40–NA) 71.7 (39.3–86.9) 3 13.3 (0–28.9) 2 89 (51–NA)

HHIs only 366 247 148 (134–172) 59.0 (52.7–64.4) 49 19.4 (13.7–24.9) 33 122 (87–174)
†Surgery or radiation both within 60 days before and after HHI initiation. NA, not applicable due to small patient numbers. 

Table 2. Patterns of HHI reinitiation among patients discontinuing HHIs 

Grace 
period

Probability of reinitiation  
following first 

HHI discontinuation 

Time to second HHI 
discontinuation among 

those who reinitated HHIs

Patients 
discontinuing 

HHIs, n

Patients 
reinitiating 

HHIs, n

Probability  
of reinitiation  
at 12 months,  

% (95% CI)

Patients 
discontinuing 

HHIs, n

Median 
treatment 
duration, 

days, 
(95% CI)

14-day 425 170 40.9 (35.6–45.7) 145 64 (59–81)

30-day 387 105 27.0 (21.9–31.7) 80 97 (81–136)

60-day 360 75 19.7 (15.0–24.2) 53 118 (89–172)

90-day 341 61 15.8 (11.4–20.1) 43 137 (112–189)

120-day 322 53 13.6 (9.3–17.6) 34 171 (114–318)

1.00

0.75

0.50

0.25

0

0 90 180 270 360 450

14-day
30-day
60-day
90-day
120-day

94 (90–109)
123 (115–135)
144 (131–162)
156 (142–178)
172 (155–190)

78.8 (74.5–82.4)
68.3 (63.4–72.5)
60.1 (55.0–64.6)
56.1 (51.0–60.7)
52.5 (47.4–57.2)

Grace 
period

Median treatment
duration, days (95% CIs)

Risk of discontinuation at
6 months, % (95% CIs)

540 630 720 810 900 990 1080 1170

526V30::All 302 112 50 21 9 8 7 3 1 1 1 1 1
526V14::All 255 79 30 16 7 6 5 3 1 1 1 1 1
526V60::All 324 137 67 17 11 10 7 5 3 3 2 1 1
526V90::All 334 149 82 34 15 13 10 6 4 3 2 1 1
526V120::All 343 160 87 38 19 16 11 7 4 3 2 1 1

C
u

m
u

la
ti

ve
 r

is
k

 o
f 

H
H

I 
tr

e
a

tm
e

n
t 

d
is

c
o

n
ti

n
u

a
ti

o
n

Time in days
Number at risk

S
tr

a
ta

0 90 180 270 360 450 540 630 720 810 900 990 1080 1170
Time in days

V30::All V14::All V60::All V90::All V120::AllStrata

Figure 1. HHI treatment discontinuation using different grace periods

V14, 14-day grace period; V30, 30-day grace period; V60, 60-day grace period; V90, 90-day grace period; V120,  
120-day grace period.

Limitations
• Unable to distinguish between permanent versus temporary discontinuation of  

HHI treatment.

• Information pertaining to reasons for discontinuing HHI treatment as well as 
disease severity were not available. 

• There was no information regarding indication of HHI treatment, i.e. whether  
it was used as neoadjuvant or adjuvant therapy. 

Type of HHI treatment 
• The majority of patients (99.2%) initiated HHI therapy with vismodegib. 

• Most patients (69.6%) were treated solely with HHIs. HHIs were used as adjuvant 
therapy in 22.2% of patients and neoadjuvant therapy in 3.8% of patients.

Patterns of HHI discontinuations using different grace periods 
• Risk of treatment discontinuation at 6 months ranged from 78.8% when requiring  

a 14-day grace period to 52.5% when requiring a 120-day grace period (Figure 1).

• Median treatment durations were 94 days (95% CI: 90–109) using a grace period  
of 14 days, 144 days (95% CI: 131–162) using a grace period of 60 days, and  
172 days (95% CI: 155–190) using a grace period of 120 days (Figure 1). 

Patterns of HHI reinitiation using different grace periods
• The risk of HHI reinitiation at 12 months was 40.9% (95% CI: 35.6–45.7) using  

a grace period of 14 days, 19.7% (95% CI: 15.0–24.2) using a grace period  
of 60 days, and 13.6% (95% CI: 9.3–17.6) using a grace period of 120 days 
(Table 2).

• Among patients reinitiating HHIs, median time to subsequent discontinuation 
was 64 days (95% CI: 59–81) using a grace period of 14 days, 118 days (95% 
CI: 89–172) using a grace period of 60 days, and 171 days (95% CI: 114–318) 
using a grace period of 120 days (Table 2). 

Patterns of HHI discontinuations using a 60-day grace period, stratified 
by type of HHI use
• Risk of treatment discontinuation at 6 months was 52.0% (95% CI: 23.4–69.9) with 

potential neoadjuvant use and 63.3% (95% CI: 52.1–71.9) with potential adjuvant 
use (Table 3). 

• Median treatment duration from HHI initiation to first treatment discontinuation  
was 107 days (95% CI: 86–not applicable [NA]) with potential neoadjuvant use,  
and 128 days (95% CI: 118–171) with potential adjuvant use (Table 3). 

Patterns of HHI reinitiation using a 60-day grace period, stratified by 
type of HHI use 
• The risk of subsequent HHI reinitiation was 16.9% (95% CI: 0–35.9) with potential 

neoadjuvant HHI use, and 22.0% (95% CI: 11.7–31.1) with potential adjuvant HHI 
use (Table 3).

• Median treatment duration following HHI reinitiation was 183 days (95% CI: 114–NA) 
with potential neoadjuvant HHI use and 118 days (95% CI: 63–NA) with potential 
adjuvant HHI use (Table 3).