strategies for analgesic development and the fda guidance for analgesic indications
TRANSCRIPT
Strategies for Analgesic Development and the FDA
Guidance for Analgesic Indications
Brett Gordon• Project Director
Ben Vaughn• Senior Statistical
Scientist
Topics• Context• New guidance• Patient Reported Outcomes: Paper vs. Device• Case Study
Note: Case studies were developed from publicly available information and do not include information about Rho’s clients or studies.
Context• Prior analgesic guidance was from 1992• For context, OxyContin was approved in 1995• 1998 Effectiveness Guidance clarified agency’s
expectations for establishing efficacy and extrapolating to a new population, dose, regimen or form
• Broad indications have lead to historic inconsistencies in labeling
ContextThe new draft guidance:• Addresses specific requirements for both narrow
and broad indications, and acute and chronic pain indications
• Falls in line with agency’s recent approvals• Reflects understanding gained over the past
decade in where extrapolation of narrow indication trials to broad claims may not be appropriate
New GuidanceGuidelines for establishing indications and Claims are organized into:• New Molecular Entities• Reformulations of approved products• Add-on or Adjunctive Indications• Additional claims
New Molecular Entities (NMEs)• Requirements build in a logical manner• Must meet requirements for narrower
indications to achieve broad claims (Think Monopoly: you need houses before hotels)
• All requirements are a minimum; additional trials may be needed
• Typical product development will probably focus on getting approval a specific indication and slowly expand broader claims
New Molecular Entities (NMEs)Specific/ Narrow indications: • Typically two trials are sufficient• Relatively narrow indications may be granted as a
result of efficacy being shown in limited settings or substantial safety concerns exist
Breakthrough pain:• Typically two trials• Population should be on around the clock opioid
therapy of at least 60mg ME with persistent pain
New Molecular Entities (NMEs)General acute pain: • Typically two trials, one visceral, one non-visceral; at
least one in the intended setting (in-patient/outpatient)
• Visceral pain– Acute pancreatitis, renal colic, postoperative visceral
surgery• Non-visceral– Postoperative orthopedic surgery, fractures, and other
acute musculoskeletal pain
New Molecular Entities (NMEs)Peripheral Neuropathic Pain (diabetic; post-herpetic neuralgia; HIV-associated; post-traumatic /postoperative; chemotherapy-associated):• Two trials in specific condition for claim on that
condition• May add an additional condition with one additional
trial in that condition• At least one trial each in three conditions required for
general claim of treatment of peripheral neuropathic pain
New Molecular Entities (NMEs)Central Neuropathic Pain:• Two trials in different conditions (pain of
spinal cord injury, post-stroke neuropathic pain, or the pain of multiple sclerosis)
General Neuropathic Pain:• Must meet requirements for general
Peripheral Neuropathic Pain and complete one central Neuropathic Pain trial (four trials total)
New Molecular Entities (NMEs)Chronic Musculoskeletal Pain:• Two trials in one indication for labeling in the
condition; one additional trial in a separate condition for claim of Chronic Musculoskeletal Pain
New Molecular Entities (NMEs)Chronic Pain:• Meet General Neuropathic Pain requirements (four
trials)• Non-neuropathic Pain:– Two trials in one condition– Two trials in TWO other conditions– Suitable conditions include osteoarthritis, chronic low back
pain, chronic visceral pain, cancer pain, and fibromyalgia• Eight trials in 7 conditions, total
Other ScenariosReformulations of approved products• One trial may be sufficient for reformulation– Includes modifications to release from IR products
• Two trials needed for new routes, new populations, new indications– Exceptions are made with adequate justification
Add-On or Adjunctive Indications• Indication match the setting under which the
product was studied
Additional Claims• Review the Patient Reported Outcome
Measures Guidance• Assess whether reliable instruments exist and
whether it is feasible to create and validation measures where they do not
• Discuss with agency intended claims• Pre-specify and include statistical plans for
multiplicity
Efficacy Considerations• Again, try to use established and reliable instruments;
Discuss with the agency EARLY in development if a novel instrument is to be used
• Minimize recall:– Avoid “pain relief” since it requires recall of prior state– No more than 24h recall for worst pain intensity is
recommended• Observer reported outcomes should be limited to
events, behaviors and signs that can be detected; an observer can report wincing or crying, but not intensity
Efficacy Considerations• Avoid composite scales that span multiple
domains for primary outcome• Multiple items within a domain are acceptable• Responder definitions that incorporate pain
intensity, rescue and study completion may be acceptable for primary
• Pain intensity outcomes must take rescue medication use into account
Safety• No major surprises in the guidance• Analgesics are likely to be sedating and cause
respiratory depression; special monitoring required while risk is assessed
• Naltrexone blocking should be used in early phase testing for opioids, especially at higher doses
• Follow-up periods for safety are likely to extend past the time of the efficacy endpoint
Safety• Abuse liability should be assessed as part of
development• REMS requirements will expand to more analgesic
products where the risks are similar to ER/LA opioids• Post-marketing studies will be required to assess risk of
misuse, abuse, addiction, hyperalgesia, overdose, tolerance and death
• NSAIDs should discuss with agency assessment of CV risk including myocardial infarction, sudden cardiac death, and cerebrovascular accident
Trial ConsiderationsDrop outs are a major challenge• Allow use of rescue meds; take assessment prior to issuing
RM• Use Add-on design (but might only support adjunctive
therapy indication)• Allow for sufficient titration periods; opioids especially can
be challenge starting and stopping• Titrate to effect to tailor to individuals needs and tolerance• Enrichment design: Removes subjects that are not
candidates for the product prior to randomization
Trial Considerations• Time to onset of pain relief and time to rescue or
re-medication should be characterized; this can be in a single- or multi-dose setting
• Acute pain, multi-dose trials– Parenteral drugs should be followed a minimum of 24h;
48h in second trial if required– Oral drugs should have longer follow up that parallels
the planned usage; assessments can be pain right now and primary is time-weighted average over follow-up (SPID24/SPID48)
Trial ConsiderationsChronic pain• Typically 12 week minimum duration (an
exception might be terminal cancer)• Missing data strategies must be carefully
considered• Primary efficacy should be change from baseline
to end of DB period• Products should enter phase III with adequate
information to determine best trial design
Trial ConsiderationsPopulation• Should mirror market population• Narrow entry criteria=Narrow label; second trial
with broader criteria may support generalizability• Try to keep any underlying disease and other
treatments stable• If including subjects with prior substance abuse
history, specific monitoring should be included
Trial ConsiderationsConcomitant medications• PK/PD studies may be needed to characterize
interactions between IP and likely conmeds• Narrow entry criteria=Narrow label; second trial with
broader criteria may support generalizability• Continuing usage of previous analgesic meds =
adjunctive indication• Consider stratification when subjects continue existing
varying regimens for underlying disease• Be certain to consider all therapies including non-drug
Trial ConsiderationsPediatrics• Unmet need, sponsors are encouraged to
perform trials to support pediatric indication• See guidance• Discuss with review division
Trial ConsiderationsShould be superiority trials, thus comparators include:• Placebo• Lower dose of investigational drug• Approved drug where the investigational drug
is expected to be superiorEven when primary is placebo, consider including active comparator.
OutcomesPain intensity• Must be a PRO• NRS, VAS or categorical; each has advantages
and disadvantages that must be considered• Disease-specific instruments may be
preferable (WOMAC for OA pain, for example)
OutcomesFunction• Pain may impact physical and emotion
function; also, drug related AEs may impact function
• Important to collect for informing risk/benefit• Again, use established instruments where
available
OutcomesHealth Related Quality of Life• Multi-domain• Not appropriate as primary endpoint• Unlikely to gain label claims as a secondary• “welcome as an exploratory endpoint”
OutcomesRescue Medication• If allowed, MUST be recorded and analyzed• Specify drugs allowed• Specify scenarios under which their usage is
triggered• Specify timing of pain measures with respect
to RM usage
OutcomesGlobal Single-Item Assessment• Highly problematic: attempt to measure overall
experience, but contributing factors from subject to subject vary wildly
• Not to be used for claims of treatment benefitOpioid Sparing• May provide evidence of efficacy• If product is intended only for concomitant use
with opioids, must show additional benefit
OutcomesSleep• Any claim must be clinically and statistically
meaningful beyond the analgesic effect• Single-item assessments will not support label
claims• Use established instruments that measure
specific domains: difficulty falling asleep, staying asleep, and waking up refreshed
Statistical Considerations• Should always have a single, primary analysis on
which to claim trial success• Analyses should fully describe the drug’s efficacy:
Effect size, variability, onset, duration• Responder analyses are easy to interpret, but
may be lower-powered• Including a cumulative distribution function plot
in the package insert is suggested to better inform prescribers of the trial outcome (!)
Statistical Considerations• Include descriptive statistics for all clinically
relevant outcomes• Always present some means for assessing
variability (CDFs, Boxplots, SDs)• If there is not an established clinical
significance for your primary outcome, use conservative estimates wherever possible to maximize power [personal recommendation]
Statistical ConsiderationsMissing data• Single Imputation (LOCF, BOCF, etc) is no
longer acceptable by default but may be used as sensitivity analysis
• See “The Prevention and Treatment of Missing Data in Clinical Trials” for recommendations; final method will be specific to the protocol
• Method must be pre-specified
Tablets (and other e-devices) vs. Paper An assortment of new technology is available for collecting patient reported outcomesExamples:• IVRS• IWRS• Tablets• Smartphones
Key advantages of Electronic Collection
• Data are available much faster (sometimes instantly)• Devices can trigger alarms to prompt patients (or an
IVRS may call at a specific time), giving consistency in time of collection and minimizing missing data
• Instructions may be programmed into the device to guide subjects for more consistent answers
• Site staff does not have to key data or manually measure VAS scores
• Elimination of digit bias
Disadvantages of Electronic Collection• Staff training, staff re-training• Validation of the device• Population level of comfort with device• Cost• Accuracy (?) Finger size vs. pen mark• If your pen breaks, you don’t have to call tech
support
Digit Bias• Can generally be ignored but interesting to
observe in action• On a 100 point VAS, there should be roughly
equal numbers of multiples of 5 and multiples of 5 +/- 1 (i.e. 44, 45 and 46 should appear roughly equally: 33% for each)
• Very consistently, hand-measured VAS outcomes return 40% for multiples of 5
Digit Bias• On an 11 point NRS, it’s not uncommon for
subjects to use just 2 or three numbers repeatedly
• Hand held devices can eliminate the issue by presenting a slider with no numbers associated
Key Considerations for Electronic vs. Paper PRO collection• Is the site comfortable and experienced with
the device?• Is the subject population comfortable with the
device?• Will you be collecting daily measures or in-
clinic?• How many outcomes will be collected?
Case Study- Cymbalta (duloxetine)• 1994- Originally approved for Major
Depressive Disorder• 2006- Approved for Diabetic Painful
Neuropathy– Peripheral Neuropathic Pain – Two trials in specific condition for claim on that
condition• 2007- Approved for Generalized Anxiety
Disorder
Case Study- Cymbalta (duloxetine)• 2009- Approved for Fibromyalgia – Specific/ Narrow indications– Two trials in specific condition for claim on that
condition• 2010- Approved for Chronic Musculoskeletal
Pain– Chronic Low Back Pain- two trials– Osteoarthritis • Needed minimum of one trial• Conducted two trials
Case Study- Cymbalta (duloxetine)• What about a Treatment of Chronic Pain
indication?– Meets Non-Neuropathic criteria (At least four
trials in three conditions)- Fibromyalgia, LBP, OA– Need General Neuropathic Pain • Need one trial in one central neuropathic condition • Need Peripheral Neuropathic Pain
» Meets two trials in any one condition- DPN» Need one additional trial in each of two conditions
– Would still need to conduct minimum of 3 trials
Pain
Acute Chronic
Narrow2 trials in 1 condition
Breakthrough2 trials in 1 condition
General2 trials in nociceptive- 1 visceral- 1 non-visceral
Musculoskeletal Neuropathic
General8 trials- General Neuropathic Indication (4 trials)- 2 trials in 1 non-neuropathic condition- 1 trial in each of 2 other non-
neuropathic conditions
Narrow2 trials in 1 condition
General3 trials - 2 condition A- 1 condition B
Peripheral Central
General4 trials- Peripheral Indication (3 trials)- 1 central trial
Narrow1st Condition- 2 trials
2nd Condition- 1 trial
General3 trials- 1 condition A- 1 condition B- 1 condition C
Narrow2 trials in 1 condition
General2 trials- 1 condition A- 1 condition B
Case Study- CDRH Comparison • MONOVISC- Approved February 25, 2014– Hyaluronic acid(HA) visco-supplementation– Indication- treatment of pain in osteoarthritis (OA)
of the knee– Trials completed for initial PMA submission: 1 – Trials demonstrating superiority for the primary
effectiveness endpoint: 0– New trials completed: 0
Case Study- CDRH Comparison • MONOVISC- How? – Post-hoc non-inferiority analysis • Historical comparison to data used for ORTHOVISC
approval in 2004 • ORTHOVISC (3 syringes) vs MONOVISC (1 syringe)• Closer look at ORTHOVISC approval
– Trials completed: 3– Trials demonstrating superiority for the primary effectiveness
endpoint: 0– Approval granted on combined effectiveness subgroup
population from 2 of 3 trials
– What would CDER do?
References
• Guidance for Industry Analgesic Indications: Developing Drug and Biological Products(Feb 2014).
• Guidance for Industry Providing Clinical Evidence of Effectiveness for Human Drug and Biological Products (May 1998)
• Guideline for the Clinical Evaluation of Analgesic Drugs (Dec 1992)
• Analgesic Indications A Historical and Regulatory Perspective (Cymbalta Ad Com, Aug 2010)
• MONOVISC Summary of Safety and Effectiveness Data (Feb 2014)
• ORTHOVISC Summary of Safety and Effectiveness Data (Feb 2004)