1. Johns Hopkins University
IMMPACT-VIII Single-dose and short-term Proof of Concept trials in Neuropathic Pain
Srinivasa N. Raja
Johns Hopkins University

2. Proof of Concept Studies New Drug Development
Early stage clinical drug development of a compound that has shown potential in animal models and early safety testing
Help make an early Go-No Go decision

3. POC studies in Neuropathic Pain Potential uses
Is neuropathic pain resistant to certain drugs?
Opioids in neuropathic pain (PHN and postamputation pains)
Test a new route of therapy/ site of action-Topical lidocaine/capsaicin
Are there predictors of drug effects?
Genetic polymorphisms, Pain mechanisms
Testing novel formulations of an existing drug for better safety
Abuse deterrent opioids
Can neuropathic pain be prevented or the disease modified?
Persistent post-surgical NP pain; Diabetic neuropathy and dietary supplements
Testing and validating objective measures of drug effects
Will use examples from studies we have conducted wherever possible- easier to criticize oneself and less likely to offend anyone.
Early stage clinical drug development of a compound that has shown potential in animal models and early safety testing
Early stage clinical drug development of a promising novel compound (based on animal models and early safety testing)
Determine if a class of drug is effective in neuropathic pain
Opioids in PHN and postamputation pain
Test a new route of therapy/ site of action
Topical lidocaine
Novel formulation of an existing drug
Abuse deterrent opioids
Determine if drug effects can be predicted
Pain mechanisms, genetic polymorphisms
Prevention of neuropathic pain or disease modifying agent
Persistent post-surgical NP pain
Diabetic neuropathy and dietary supplements

4. Pain Score on Numeric Rating Scale (0-10)
Is neuropathic pain resistant to opioids? I.V. morphine and lidocaine infusions in PHN and Phantom Pain: 3-session double-blind cross-over studies
Placebo
Opioid
Lidocaine
Wu et al. Anesthesiology
2002;96:
P<0.001
P=0.08
P=0.88
P=0.001
Diphenhydramine
50 mg
Morphine
0.25mg/kg
5 mg/kg
Pre Post
Pain Score on Numeric Rating Scale (0-10) 2 4 6 8 10
Phantom Pain
Single fixed dose-based on body weight over 60 min
Infusion pain rating correlated with MS blood levels,
but not lidocaine levels
Rowbotham MC et al. Neurol 1991;41:1024
PHN
Infusion pain intensity- Mean pain intensity during 60 min infusion and 60 min postinfusion observation period
19 subjects- 11W, 8M
N=19

5. Single dose infusion cross-over trials: Pros and Cons
Minimizes effects of inter-subject variability
Fewer subjects required
Early signal to help predict efficacy
Short study duration
Slow offset or prolonged duration of effect may lead to carry over effects
May not help predict side effects
No information on oral bio-availability
No dose-response information
May miss effect if inappropriate dose chosen

6. Pain Score on Numeric Rating Scale (0-10)
Postamputation Pain: Oral morphine vs mexiletine on pain intensity ratings (3-period crossover)
Wu et al.
Anesthesiology 2008 (in press)
Placebo Maintenance Pain Score
Opioid Maintenance Pain Score
Mexiletine Maintenance Pain Score 2 4 6 8 10
P<0.001 *
Pain Score on Numeric Rating Scale (0-10) * *
Oral phase pain intensity
Placebo
n=43
Opioid
n=50
Mexiletine
n=42

7. Can topical therapies be effective in neuropathic pain?
Single-dose cross-over design with vehicle control
During the 2 lido patch sessions 5% lidocaine patches (420 cm2) applied for 12 hrs. Vehicle patch similar. Observation only no patches applied
Vehicle and Lidocaine patches for 12 hrs vs Observation alone
Outcome measure: Change in VAS scores of pain
Rowbotham et al., Pain 1996;65:39

8. Single dose cross-over trials with topical agents: Balancing the pros and cons
Helps establish new routes of therapy, mechanistic implications?
Minimizes effects of inter-subject variability
Fewer subjects required
Early signal to help predict efficacy
Short study duration
Short duration of observation may not be predictive of long-term effects
May not help predict side effects with longer term treatment
No dose-response information

9. Predicting responders? Variability in Opioid Response: PHN Trial5 10 15 20 25 30 35 40 45 50 55 60 65
-40
-20 80
100
Decrease in Pain Intensity, %
Subject No.
Side Effects
Lack of response
Tella et al. 2007, Proceedings of 11th World Congress on Pain

10. Predictors of Opioid Response in PHN: Phenotype: Heat pain threshold at unaffected site48 †
P=0.09
P=0.04 47
Quantitative Sensory Testing
Heat pain sensitivity at unaffected site prior to opioid exposure (baseline) * 46 45
Baseline Heat Pain Threshold, °C 44 43 42 41
<30%
Pain
Reduction
≥30%
Pain
Reduction
<30%
Pain
Relief
≥30%
Pain
Relief
Post-hoc analysis prospective study
Is this a phenotype for a genetic polymorphism? e.g., MOR
Edwards R et al. Anesthesiology 2006;104:1243

11. Irritable Nociceptor PHN subgroup of patients more responsive to opioids
Placebo
Opioid
TCA
Pain Intensity
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Deafferentation Subtype
Irritable Nociceptor PHN
(Intact C-fibers)
Placebo
Opioid
TCA
0.0
0.5
1.0
1.5
2.0
2.5
3.0
P<0.001
P=0.04
Pain Intensity
P=0.86
P<0.04
TCA=Tricyclic antidepressant.
Tella et al. IASP 11th World Congress on Pain; 2005.

12. Testing novel formulations of existing drug Abuse-deterrent opioid (ALO-01/Embeda)
3-phase study: Screening/Qualifying, Treatment, Followup
Healthy men and women, non-dependent recreational opioid users aged 18 to 55 years
Ability to tolerate single dose of 120 mg of morphine sulfate and to distinguish morphine from placebo (2-day crossover design)
Screening
MSS
Placebo
Randomization
CSR: p21, 25
Screening: subjects could tolerate 120 mg MS, were able to distinguish MS from placebo
Jones et al. 2008, APS and AAPM
Adapted from Stauffer J. 2008
MSS = morphine sulfate solution.

13. Treatment Phase Study Design Randomized, double-blind, triple-dummy, 4-way crossover
Aim: If study drug taken intact less desirable than crushed capsule or MS sol. for recreation users
ALO-01W
ALO-01C
MSS
Placebo
Session 1
Session 2
Session 3
Session 4
Post-treatment
Follow-up
Washout
14-21d
Safety
Assessments
Capsules, crushed pellets in apple juice, and apple juice
Outcomes: PK and PD measures (subjective and objective measures)
Negative control
Positive control
Source: CSR p 29, 22
Outcome measures: PD- VAS for drug liking, VAS for +, -, and subjective effects, Cole/Addiction Res Ctr Inventory, and Subjective drug value, Pupillometry
As POC study- good positive and negative controls, fewer subjects needed
Single dose study-may not necessarily predict repeated dose effects
PK Measures of interest
Tmax: time to peak plasma concentration
Emax: peak effect
TEmax: time to reach Emax
Jones et al. 2008, APS and AAPM

14. POC Designs: Two phase study Dronabinol as adjuvant for patients on opioid therapy
Placebo
N=29
20 mg DB
10 mg DB
N=30
Randomized
N=30
Screened
N=160
4-wk Multi-dose
N=28
5-60 mg/ day
Phase II
Open-label, extension
Multi-dose study
All patients offered entry
Chronic non-
cancer pain
on stable opioid
Enter Fig 3 and Fig 5
Phase I
Double-blind, randomized,
Placebo-controlled, 3 period
Single-dose croosover study
Three 8-hour visits
Narang et al. J Pain 9;245:2008

15. Pros and Cons
Single dose cross-over design established a POC of effects as an adjuvant analgesic with a small N
Established that higher dose not associated with better pain relief but more common side effects
Pain relief sustained during the open label phase
Limitations
Effectiveness demonstrated only as an adjuvant
Open label phase II could be non-specific- no placebo control
Design useful only for drugs with rapid onset of action

16. Enriched study: Clonidine in diabetic neuropathy
Two stage design- Selection and Efficacy
Stage 1- 3-period crossover
N=41
Stage 2- Four double blind randomized
1 wk treatment periods
12 responders
Byas-Smith et al., Pain 1995

17. N of 1 or single-patient designs
To test statistically within a single patient whether or not an intervention improves clinical outcome
Within patient response vs group response
Active
treatment
Placebo
randomize N
Active
treatment
Placebo
Assessment Patient
preference
Individual
patient
Conditions: random allocations to experimental or control or two treatments, double-blind, measurements of results of Rx, formal statistical analysis. Conventional stats– 5 paired comparisons
Ideal Design:
randomized allocation, blinding, measurements of outcomes, formal statistical analysis
Ideal Drug: Rapid onset, rapid offset, reversible action
Ideal Disease: Stable pain over long duration
Scuffham PA Value in Health 11;97:2008

18. N of 1 trials: the pros and cons
Minimizes effects of inter-subject variability
Potential to identify subset of patients who are responders
Can influence clinical decision for the patient
Has been used for cost-benefit analysis
Slow onset drug effects may lead to long duration study
Slow offset or prolonged duration of effect leads to carry over effects
Potential for drop out and less enthusiasm to continue with paired comparisons
Australian studies: to improve access to selected high cost medications
Celecoxib vs sustained release paracetamol for osteoarthritis
Gabapentin vs placebo for neuropathic pain
Scuffham PA Value in Health 11;97:2008

19. Is an ounce of prevention better than a pound of cure
Is an ounce of prevention better than a pound of cure? NMDA antagonists for postmastectomy pain
Randomized D-B, PC trial in patients undergoing mastectomy, lumpectomy with axillary node dissection
Amantadine 100mg bid, day before to 14 day after surgery
Rescue drugs OK
Eisenberg E. J Pain 2007;8;223

20. Prevention of disease progression and improved pain Acetyl-L-Carnitine in Diabetic Neuropathy
Double-blind placebo-controlled RCT in 333 subjects, 1 yr followup
1 g im for 10 d, 2 g orally for 355 d
NCV (motor and sensory) and amplitude primary OM, pain secondary
6m and12 m- NCV increased in active group in all nerves; decrease or no change in placebo
199 pts had pain at baseline- 39% decrease at 12 m * **
VAS
Cooments- studies from one site, industry sponsored resarch; large trial with long followup; NCV not temp controlled
De Grandis and Minardi Drugs R&D 2002; 3:223

21. Testing drugs for chronic pain Core outcome measures
Physical functioning
Multidimensional Pain Inventory Interference Scale 
Brief Pain Inventory interference items
Emotional functioning
Participant rating of global improvement and satisfaction
Symptoms and adverse events
Patient disposition
IMMPACT recommendations
Dworkin RH et al Pain 2005;113:9

22. Validating a measure of function: Pain relief with Opioids objective increase in activity
Transdermal fentanyl mcg/h
37.4%
Need to do a better job (work in progress) to cross validate with subjective measures of function
Intent to treat: %
decrease in pain
Agarwal S et al.
Pain Medicine
2007; 8:554-62

23. Early stage drug development Sensitivity vs Specificity
Wrong disease state
Wrong dose
Wrong duration of treatment (exposure-response relationship)
Outcome measure- no biomarkers (surrogate endpoint) for pain
Not considering the natural course of the disease- disease progression or regression
Active comparators with proven efficacy to distinguish negative from failed trials
How to increase the signal to noise ration To Avoid Throwing baby out with the bath water of a promising drug
Failure to get the dose right can lead to ineffective Rx in some and excessive toxicity in others
Create slide from p201 - Stanski

24. Dosage changes in new molecular entities approved between 1980-1999
499 NME, 354 evaluable
Dosage changes occurred in 21% of 354 NME post approval
79% safety-motivated dosage decrease
27% neuropharm. drugs
Median
6.5 yr
2.5 yr
Cumulative hazard function for dosage change over time by epoch. The drugs of the most recent epoch were exposed to 3.15 times greater risk of undergoing dosage change (p=0.003) Median time to change following approval fell from 6.5 yr ( ) to 2 yr ( )- possible increased surveillance, hopefully not to less rigid dose determination as a result of time constraints.
Srr Author carl Peck- georgetown Univ
Dosage changes may reflect quality of premarketing development, regulatory review, and post-marketing surveillance.
Smallest dose that produces near maximal effect rather than maximal tolerated dose
Cross et al., 2002
Pharmacoepidemiol & Drug safety

25. Lessons learnt from failed neuropathic pain RCTs Relation to study characteristics
Aim: to identify factors associated with + vs – outcomes of placebo-controlled neuropathic pain trials
106 clinical trials with 123 Rx-group comparisons
+ studies: medication response rates greater, placebo response lower, larger sample size, cross-over design, published earlier (1995 vs )
Greater placebo response:
greater medication response & trial duration, parallel design
-ve vs + outcome: 27% vs 16% placebo responders (>50%)
Katz JK, Finnerup NB, Dworkin RH Neurology 2008;70:263
Polydefkis M, Raja SN Neurology 2008;70:250

26. Rowbotham M Neurology 2005;65:S67
Study Designs
Parallel vs Crossover
Enriched enrollment design
Excluding high placebo responders?
N-of-1 studies
Adaptive designs
Time-to-exit designs (see Galer et al Pain 1999-Lido patch)
Mechanism-based clinical studies (Wallace MS 2002 J Pain)
Genetic screening: e.g. MOR polymorphism
Split-trial strategy- pooled data from few centers with extensive testing
Rowbotham M Neurology 2005;65:S67

27. Summary
Study design- adaptive, depending on nature of question being asked
Consider the balance of pros and cons of the design relative to the question

28. New strategies to test and develop new drugs efficiently for neuropathic pain: A combined effort
Stakeholders
Patients
Health-care providers
Insurers
Industry
Agencies
Academia
NIH
Regulatory
Agencies
Industry
Assuring proper treatment of patients who may benefit from the use of opioids and curtail diversion and abuse of prescription drugs.
29% of pharmacies surveyed by CASA (natl center for addiction and substance abuse at columbia) experienced theft or robbery of controlled substance
NASPER- national All Schedules Prescr. Electronic Reporting Act- 2005