NDA 21576

1
NDA 21-576
Methyl aminolevulinate Cream (MAL) with Curettage
and Photodynamic Therapy (PDT) for Basal Cell
Carcinoma (BCC)
2
FDA Clinical and Statistical Reviewers
Brenda Vaughan, M.D., Medical Officer Markham C.
Luke, M.D., Ph.D., Dermatology Team
Leader Shiowjen Lee, Ph.D., Mathematical
Statistician Mohamed Alosh, Ph.D., Mathematical
Statistics Team Leader
3
Treatment of Primary Nodular BCC

• Curette-MAL-PDT consists of
• Lesion preparation (Curettage)
• Methyl Aminolevulinate (MAL) Cream
• Combined with Curelight Lamp (Drug/Device
  Combination)

4
Background

• Provided that efficacy and safety are
  established for nodular BCC one independent,
  multi-center, randomized, active controlled study
  conducted in patients with primary superficial
  BCC might be acceptable if supported by strong
  evidence from a solid database. Regulatory
  Guidance Meeting with PhotoCure - March 7, 2000

5
Items for Discussion

• Efficacy
• Adequacy of studies for estimating cure rate?
• Early histology and small numbers in pivotal
  studies
• Minimal recurrence data for nodular BCC
• Adequacy of instructions for lesion preparation?
• Apparent high VEH-PDT response rate
• Center-to-center variability
• Estimate of cure rate for MAL-PDT vs. Surgery?
• Safety
• Pain and minimal information regarding
  anesthesia/pain control
• Contact hypersensitization

6
Measurements of Efficacy for Treatment of Basal
Cell Carcinoma

• Agency Proposed
• Clinical Observation and Excision Histology
• 5 year Recurrence Data (2 year data acceptable
  for filing)
• What was submitted
• Excision Histology Alone
• Clinical Observation Alone
• Recurrence Rates Based on Clinical Observation

7
Clinical Studies

• Studies interpreted for efficacy for nodular BCC
• 2 vehicle controlled randomized studies (307
  308)
• 1 open-label randomized MAL-PDT vs. surgery for
  recurrence rates (303)
• 1 open-label non-randomized MAL-PDT study for
  recurrence rates (205) also had superficial BCC

8
Pivotal Studies

• Study 307 - U.S. Study
• 33 patients randomized to Curette-MAL-PDT
• 32 patients randomized to Curette-VEH-PDT
• Study 308 - Australian Study
• 33 patients randomized to Curette-MAL-PDT
• 33 patients randomized to Curette-VEH-PDT

9
Pivotal Studies Study Design

• First Treatment Cycle - two Curette-(MAL vs.
  VEH)-PDT treatments 7 days apart followed by
  clinical assessment at 3 months
• Second Treatment Cycle - if only partial response
  to First Treatment Cycle - two additional
  treatments 7 days apart

10
Pivotal Studies Study Design

• Clinical evaluation at 3-months to determine
  further management
• Complete response (disappearance of lesions) gtgt
  excision at 6-month for histology.
• Partial response (lesion decreased by gt 50) gtgt
  2nd PDT cycle gtgt excision at 9-month for
  histology.
• No response (lesion decreased by lt 50) or
  Progression (lesion increased by gt 20) gtgt
  excision at 3-month.
• Complete response is not equal to cure.

11
STUDIES 307 and 308
RANDOMIZATION
1st Tx Cycle
7 DAYS APART
MAL or VEH PDT
12

• Biostatistical Analysis of Pivotal Studies

13
Efficacy Evaluation in Pivotal Studies

• Primary Patient complete response rate
• Secondary Lesion complete response rate
• Criterion for assessing response
• a. Clinical and histology - Agency
  recommendation to PhotoCure on March 7, 2000
• b. Histology only PhotoCures protocol
  (August, 2000)
• Complete response is not equal to cure.

14
Overview of Efficacy Findings in Pivotal Studies

• Curette MAL-PDT is superior to Curette VEH-PDT
• Variability in the success rate estimate for
  MAL-PDT which might be attributed to
• a) Relatively small studies ? wide C.I.
  around these estimates
• b) Uncertainty about lesion preparation
  description

15
Overview of Efficacy Findings in Pivotal Studies

• Clues
• High vehicle response rate
• Center-to-center variability
• Small studies with few patients per center

16
Apparent High Vehicle Response Rate Based on
Histological Evaluation
17
Summary of Response rates along with 95
Confidence Intervals
18
Estimates of Response for 1st Treatment Cycle
PhotoCure excluded from the denominator those
lesions which had partial response and were
treated with a 2nd treatment cycle in post-hoc
analysis. ? rate is inflated because not all
lesions treated with 1st PDT cycle are included
in the denominator, as shown in the following
table
19
Treatment Cycle Response

• Study design precludes estimating response rates
  for only one treatment cycle
• Second treatment cycle is based on clinical
  decision regarding first treatment cycle outcome,
  i.e. partial response
• Outcome measure was done at 3 months
• No randomization before second cycle

20
Study 307 Lesion Response Inter-Center
Variability
21
Study 308 Lesion Response Inter-Center
Variability
22
Pivotal Studies Summary of Statistical Analysis

• Curette-MAL-PDT is statistically superior to
  Curette-VEH-PDT for the treatment method used in
  the protocol of nodular BCC.
• For each pivotal study there is relatively high
  Curette-VEH-PDT response rate. There is also
  center-to-center variability in Study 307. This
  might be attributable to
• Small study size
• Lesion preparation for treatment
• Accuracy of clinical and histological evaluations
• The center-to-center variability in the efficacy
  results reduces the reliability in the overall
  point estimates of Curette-MAL-PDT.

23
Efficacy Conclusions from Pivotal Studies

• Curette-MAL-PDT is statistically superior to
  Curette-VEH-PDT using the protocol guided outcome
  assessment
• High response rates with Curette-VEH-PDT
• Effect of curette
• Short follow-up
• Low ability to detect residual nodular BCC by
  histological methods used

24
PhotoCure Video Lesion Preparation
25
Effect of Curettage

• Response may depend on extent and depth of
  curettage
• Efficacy shows center dependence
• High Curette-VEH-PDT response
• Other factors
• Lamp exposure appears to have been applied in a
  consistent manner for each pivotal study

26
Curettage
Page 124 of PhotoCure AC Briefing, Figure 20
27
Recurrence Data Regulatory

• Agency and PhotoCure discussed in the context of
  nodular BCC recurrence data, that a minimum of
  250 subjects were to be submitted. Pre-NDA
  meeting minutes with PhotoCure - June 20, 2002
• A two-year follow-up was recommended for NDA
  submission. Regulatory Guidance Meeting with
  PhotoCure - March 2000
• It was recommended that patients be followed up
  to five years post treatment. Regulatory Guidance
  Meeting with PhotoCure - March 2000

28
Recurrence Rate Database

• PhotoCure provided 2 year recurrence data for 46
  patients (47 lesions) with nodular BCC treated
  with Curette-MAL-PDT (clinical observation for
  recurrence) Study 303/99 (Phase 3, randomized
  open-label).
• Additional data for 30 patients with nodular BCC
  Study 205/98 - Phase 2, non-randomized
  open-label, included both nodular (38 lesions)
  and superficial (39 lesions) BCC. Focus would be
  primarily on nodular.

29
Assessment of Recurrence

• Recurrence of lesions is based on clinical
  assessment (inspection and palpation) confirmed
  by punch biopsy when positive clinically.
• Treated areas that were apparently clinically
  clear were not biopsied and followed.

30
Study 303

• European randomized, open-label, multicenter
  study in 101 subjects
• Curette-MAL-PDT (n52)
• Surgical Excision (n49)
• Initial post-treatment assessment at 3 months
• Followed by 12 and 24 month clinical assessments
  for recurrence

31
Example of Complete Response?
Page 100 of PhotoCure AC Briefing, Figure 19
32
Lesion Recurrence Data Study 303
33
Failure to Cure Analysis Study 303
34
Phase 2 Recurrence Study

• Study 205 Non-randomized, open-label study with
  both nodular and superficial BCC
• 57 patients (79 lesions) were evaluated for
  recurrence at 24 months
• 30 patients with 38 nodular BCC lesions
• 6, 12, and 24 month data available
• 3 patients with 1 superficial/nodular lesion
  each were in the study

35
Factors Affecting Applicability of Recurrence
Data (Study 205)

• Different patient population (high risk)
• Difference in written instructions for lesion
  preparation (curetting below epidermis)
• Difference in MAL application border
• Different lamps used - Some patients had a lamp
  other than that used in the pivotal studies

36
Recurrence Study 205
Lesion recurrence (at 24 months)
Early failures are not included
37
Recurrence Conclusions

• A database of 46 patients with 2 year recurrence
  data for primary nodular BCC was submitted by
  PhotoCure in Study 303.
• The 2 year recurrence rate for MAL-PDT treated
  patients ranged from 9 to 34 depending on how
  missing data were accounted for.
• Failure to treat adequately rate was 45 at 2
  years.
• A larger database was requested.

38
Cosmetic Outcome

• In the pivotal studies, vehicle treated patients
  had as good as or better cosmetic outcome than
  MAL treatment, but poorer response with regard to
  BCC treatment outcome.
• Cosmetic outcome is considered by Agency to be
  secondary to non-recurrence of the basal cell
  carcinoma skin cancer. Recurrence may ultimately
  result in a worse cosmetic outcome due to need
  for further treatment.
• The assessment of cosmetic outcomes across
  treatments were not pre-agreed upon between
  PhotoCure and FDA.

39
Cosmetic Outcome

• Data from the pivotal studies are presented (next
  slide)
• Limited numbers of patients in each study arm
• Photographic basis for assessment was not
  provided by PhotoCure to confirm the data
• Cosmetic assessments could be made prior to
  surgical excision and supported by blinded
  independent review of photographs. March,
  2000 Regulatory Guidance Meeting Minutes.

40
Pivotal Studies Cosmetic Outcome

• Cosmetic Outcome for Lesions Having Histological
  Complete Response -- PhotoCures results

41
Safety Risks Identified

• Local
• Pain, Burning Stinging
• Phototoxic Reactions
• Contact Sensitization

42
Local Adverse Events
43
Local AEs with MAL-PDT
44
Local AE Phototoxicity

• Patients treated with Curette-MAL-PDT could have
  skin ulceration, and blisters that last 1 to 2
  weeks after treatment and in two cases erythema
  lasted more than a year.
• No separate analysis for rates of such
  occurrence.
• PhotoCure summary of non-U.S. labeling and
  Summary of Safety

45
Summary of Local AEs with MAL-PDT

• Curette-MAL-PDT was associated with higher
  incidence rates of Pain, Burning, or Stinging
  than Curette-VEH-PDT.
• The use of local anesthesia with MAL-PDT
  treatment was not studied in a systematic
  fashion.
• Minimal instructions for use of anesthesia
• Tumor fragments from most lesions may be
  removed without damaging normal skin and without
  use of anaesthetics.

46
Sensitization

• Dermal Safety Sensitization Studies
• Subjects during clinical trials
• 2 patients with urticaria/hypersensitivity
  reaction
• Post-Marketing (Europe)
• 2 patients with 1 positive rechallenge

47
Sensitization Study Design

• Induction Phase
• MAL and MAL-vehicle applied for 3 weeks
• 2 Week Rest Period
• Challenge Phase
• 3 hour MAL/VEH application
• 48 hour Aminolevulinate 0.1 in soft paraffin
  (ALA) vehicle cross-sensitization challenge

48
Dermal Safety Sensitization Study
Enrollment stopped due to skin reactions
52 of the 58 Subjects Who Continued and Did Not
Refuse MAL Were Sensitized to MAL Cream
49
ALA 48 hour Cross Sensitization/Challenge
Results (N98)

• None judged positive to 0.1 ALA
• 2 (2) were judged equivocal to ALA
• 2 (2) were judged positive to soft yellow
  paraffin vehicle for ALA

50
Safety Conclusions

• MAL has a high sensitization potential (at least
  52 sensitization rate in a provocative study)
• Cross sensitization to ALA (an endogenous
  substance) cannot be ruled out
• Sensitization to MAL of healthcare providers and
  patients are of concern

51
NDA 21-576 Risk/Benefit

• Efficacy
• Curette-MAL-PDT was shown to be statistically
  superior to Curette-VEH-PDT for the chosen
  outcome assessment in the pivotal studies.
• Adequacy of studies for estimating treatment
  effect?
• Early histology and small numbers in pivotal
  studies
• Minimal recurrence rate data for nodular BCC
• Adequacy of Instructions for Lesion Preparation?
• Apparent high VEH-PDT response rate
• Center-to-center variability
• Numerically higher recurrence rate with MAL-PDT
  vs. Surgery in one small open-label study - exact
  point estimate is uncertain

52
NDA 21-576 Risk/Benefit (continued)

• Safety
• Pain and minimal information regarding
  anesthesia/pain control
• Contact hypersensitization

53
PhotoCure Video