Clinical Review of Protopic Safety, Potential Risk, and Efficacy Martin M. Okun, M.D., Ph.D. Lisa Mathis, M.D. Division of Dermatologic and Dental Drug Products

Protopic Safety Safety Data Base –Adult and Pediatric Patients –3 weeks, 12 weeks, 6 months, 12 months –0.03%, 0.1%, 0.3% ointment

Protopic Safety Data Base Focus on: –Serious Adverse Events (AEs) and AEs  Discontinuation from Study (D/C) –AEs consistent with systemic immunosuppression (e.g., lymphadenopathy, lymphoma, reactivation of VZV, HSV)

Adverse Events--Adults: 3 weeks treatment

Adverse Events--Adults: 12 weeks treatment

Adverse Events--Adults: 6-12 months treatment

Adverse Events--Pediatrics: 3 weeks treatment

Adverse Events--Pediatrics: 12 weeks treatment

Adverse Events--Pediatrics: 1 year treatment

Lymphadenopathy in Clinical Studies Total cases: 33No. of PatientsStrength Adults on 3 weeks therapy: 7 (0.1%) Adults on 1 year therapy: 3 (0.03%) 9 (0.1%) Children on 3 weeks therapy: 1 (0.1%) Children on 1 year therapy: 3 (0.03%) 10 (0.1%)

Lymphadenopathy Etiology 68 year old male developed a parotid lymphoma. Uncertain if lymphoma was a pre-existing condition 59 year old male developed CTCL. 6 cases with no clear etiology: –4 resolved spontaneously –2 discontinued therapy and were lost to follow up. 11 cases were related to skin infections: –9 resolved on antibiotics –3 discontinued therapy but had resolution of the lymphadenopathy. 12 cases related to URIs, resolved on antibiotics. 2 cases related to tooth infections.

Common Adverse Events: 3 Twelve Week Studies Comparing AE rates across vehicle and active treatment arms complicated by the >50% discontinuation of patients in the vehicle arm Application site AEs (e.g., burning, pruritus, erythema, treatment-site infection): –vehicle: 59%.03%: 75%.1%: 72% No obvious increase in local AEs as concentration increases

Common Adverse Events: 2 One Year Studies with 0.1% ointment Application Site AEs: –Pediatric Study 54% (16% of these coded severe) –Adult Study 78% (28% of these coded severe), –Prevalence of Skin Burning declined to <10% by week 4 of long-term studies

Safety Conclusions Most patients tolerated the 0.03% or 0.1% ointment concentrations. –Ointment use associated with application site reactions in ~75% of patients –Most such reactions were mild and transient Tacrolimus 0.03% and 0.1% ointments had similar adverse event profiles

Safety Conclusions (cont’d) Possibility of an increased incidence of herpes zoster, herpes simplex, and Kaposi’s varicelliform eruption may exist Two patients developed lymphoma 6 patients developed lymphadenopathy with no obvious etiology

Efficacy Database 3 efficacy studies, identically designed: –multi-center, double-blinded, randomized –12 weeks in duration –3 arms: vehicle ointment, tacrolimus.03% ointment, tacrolimus.1% ointment Two studies in adult patients (-035, -036) One study in pediatric patients (-037)

Study Subjects Moderate to severe atopic dermatitis extent of involvement: 2%-100% BSA, with most having greater than 1/3 BSA Adult studies –-035: years; -036: years Pediatric studies –037: 2-15 years

Study Protocol Wash out of non-medicated and medicated topical products Subjects applied ointment twice daily to essentially all areas of active disease Patients who cleared continued treatment for an additional week Prespecified treatment success at week 12: complete clearance or  90% improvement

Complete or  90% Clearing

Efficacy Outcomes In all three studies: –0.1%, 0.03% ointments statistically significantly superior to vehicle –0.1% ointment numerically, but not statistically, superior to 0.03% ointment: in -035,-036: treatment differences of 6, 12 percentage points in -037: treatment difference of 5 percentage points Is lack of statistical superiority due to equal efficacy or to Type II error?

Power Analysis Clinical studies may not have been adequately powered to detect statistically significant, clinically relevant differences in treatment effect between 0.03% and 0.1% ointments: –power of 0.18 to detect 10 percentage point difference –power of 0.8 to detect 18 percentage point difference

-035 and -036 Population Subsets: 0.03% vs. 0.1% 0.1% numerically superior to 0.03% ointment: –in males (32%-28%) and in females (40%-27%) –caucasians (40%-33%), african-americans (29%-16%), asian-americans (42%-22%) –in patients who had baseline moderate disease (40%- 38%) and had baseline severe disease (35%-20%) –in patients who are older than 65 years (25%-22%) and who are younger than 65 years (38%-27%)

-037 Population Subsets: 0.03% vs. 0.1% 0.1% numerically superior to 0.03% ointment: –caucasians (45%-40%), african-americans (27%- 25%), asian-americans (83%-43%) –patients with moderate baseline disease (47%-44%) and with severe baseline disease (37%-30%) –ages 2-6 years (48%-38%) –males (44%-33%)

037 Population Subsets: 0.03% vs. 0.1% 0.03% numerically superior to 0.1 % ointment: –ages 7-15 years (32% vs. 30%) –females (39% vs. 38%)

Secondary Endpoint Outcomes: 0.03% vs. 0.1% In all 3 pivotal studies, 0.1% ointment was numerically superior to 0.03% ointment: –% patients with complete clearing at EOT, –reduction in % BSA involvement at EOT, –% patients with  50% improvement at week 1, Patient’s assessment of pruritus: –0.1% ointment numerically superior in -035, numerically inferior in -036, equal in -037

Efficacy Conclusions: Both tacrolimus ointment concentrations were statistically significantly efficacious over vehicle. Tacrolimus 0.1% ointment was numerically efficacious over 0.03%: –overall primary efficacy variable –most clinically relevant population subsets –most clinically relevant secondary efficacy variables

Efficacy Conclusions (2): Clinical studies may have been underpowered to detect clinically relevant, statistically significant differences in treatment outcomes between tacrolimus 0.1% and 0.03% ointments.

Potential Risk What is the potential risk of systemic immunosuppression induced by percutaneous absorption of tacrolimus? –What is the lowest amount of tacrolimus percutaneous penetration at which clinically relevant systemic immunosuppression occurs?

Systemic Immunosuppression in Organ Transplant Recipients May manifest as lymphoproliferative disease, often associated with aggressive immunosuppression Frequently associated with acute or past infection with Epstein-Barr virus May progress to lymphoma or death

Lymphoproliferative Disorder in Adults Prevalence estimated at 0.8% in transplant patients. Spectrum of disease ranges from polymorphic, polyclonal B cell proliferation to frank lymphoma. Most often associated with aggressive immunosuppression and EBV infection. Treatment includes reduction in immunosuppression, antiviral therapy, debulking, chemotherapy and radiotherapy. Spontaneous regression can occur. Mortality is over 50%.

Lymphoproliferative Disorder in Children Incidence estimated to be from 4-8% in transplant patients. 3 major risk factors: allograft type, EBV infection or reactivation, and immunosuppression. Can affect any organ system, and can be diffuse or focal. Abdominal disease is most frequent. Treatment involves reduction in immunosuppression, antivirals, and conventional antineoplastic therapy. Regression after withdrawing immunosuppression ranges from 23% to 65%. Mortality is lower than in adults, approximately %.

Time from Initiation of Immunosuppression to Development of Lymphoproliferative Disease Tacrolimus: 12.6±5.1 months Cyclosporin A: 49.7±20.7 months Tacrolimus Rescue*: 9.8±3 months *Patients initially treated with CsA, but rescued with high-dose tacrolimus secondary to graft rejection

Target Trough Whole Blood Levels of Tacrolimus for Immunosuppression in Transplant Patients Trough blood levels correlates with systemic immunosuppression First month post-transplant: ng/ml Post-transplant months 1-3: ng/ml Post-transplant months 3-6: 5-10 ng/ml After post-transplant month 6: In selected patients, levels <5 ng/ml may be adequate if graft function is stable, but immunosuppression is highly variable.

Tacrolimus Blood Levels in Pivotal Controlled Studies Blood samples collected at weeks 1, 3, and EOT to measure tacrolimus levels. Blood samples collected at random with respect to time of last tacrolimus application (i.e., NOT trough levels)

Maximum Tacrolimus Blood Concentrations: -037

Maximum Tacrolimus Blood Concentrations: -035,-036

What accounts for the isolated elevated tacrolimus blood levels? Why are the two highest levels detected in the 0.03% arm, even though the trend is for more patients in the 0.1% arm to have levels above blq? Possible explanations: –Isolated patients in 0.03% arm absorb more tacrolimus systemically because of less efficacy –Covariate unrelated to treatment arm –Variability in the tacrolimus assay –Chance

Clinical Outcome and Blood Levels

Percentage of Patients with Moderate Improvement or Better by Week 1 By improving patients more quickly, 0.1% ointment may reduce the percentage of patients who are susceptible to absorbing high levels of tacrolimus

Blood Levels in a Long-term Study Blood levels tend to decline over time. Marked variability is noted throughout the study.

Persistent Blood Tacrolimus Levels in the One Year Study 60% of subjects had persistently elevated tacrolimus blood concentrations –This percentage refers to subjects who did not have a specimen measuring blq within any of the following time periods: week 1, week 2, months 1, 3, 6, and 12, –Not all subjects had measurements within each of these periods

Persistent Blood Tacrolimus Levels in the One Year Study Of the 48 subjects who did have 6 specimens collected, 54% had detectible tacrolimus concentrations in all specimens

Potential Risk: Topical Vs. Systemic Exposure Topical: Intermittent exposure Levels only sporadically above blq in most patients Are tacrolimus levels at or near the application site higher than systemic levels? Systemic: Lifetime exposure At least 6 months of >5 ng/ml serum trough levels

Potential Risk: Unresolved Issues The minimum systemic exposure at which there is clinically relevant Tacrolimus ointment-induced immunosuppression is unknown. The possibility of regional immunosuppression induced by topical application of Tacrolimus cannot be excluded.

Tacrolimus 0.03% vs. 0.1% ointment: 0.03% Pros: May have fewer patients with detectible blood levels Most patients with detectible blood levels seem to have lower levels Cons: May have inferior efficacy May be more associated with transient elevations above 5 ng/ml early in treatment.

Tacrolimus 0.03% vs. 0.1% ointment: 0.1% Pros: May have superior efficacy No evidence of an adverse event signal suggesting greater toxicity Cons: Most of those patients with detectible blood levels seem to have higher levels

Advisory Committee Question #1 Is there sufficient evidence for effectiveness of Protopic, 0.03%, in the treatment of atopic dermatitis?

Advisory Committee Question #2 Is there sufficient evidence for superior effectiveness of Protopic 0.1% compared to 0.03%… a)…in adults? b)…in children?

Advisory Committee Question #3 Has the safety profile of Protopic been adequately determined for long term, intermittent therapy in the treatment of moderate to severe atopic dermatitis not adequately responsive to or intolerant of conventional treatment… a)…in adults for both 0.03% and 0.1%? b)…in children for 0.03%?

Advisory Committee Question #4 The proposed indication for Protopic, which would allow –for both 0.03% and 0.1%, –for unrestricted chronic therapy, – as a first-line treatment of atopic dermatitis, –in adults and children 2 years and older may be decomposed into the following elements which may be reconstructed into indications:

Advisory Committee Question #4

Is approval of Protopic recommended and, if so, under what conditions [concentration(s), first- vs. second-line, chronic vs. (time-limited) acute therapy] in which age groups?

Advisory Committee Question #5 Are there additional studies needed to provide information important for the labeling of Protopic? If so, what studies are recommended? (Consider the issues of lymphoma, local suppression of immunity, photocarcinogenesis, etc.)