PET Imaging of the Placebo Response: The Role of the Endogenous Opioid System in Pain

J. James Frost, M.D., Ph.D.

Johns Hopkins University School of Medicine, Baltimore, MD

Grant Program:

David Mahoney Neuroimaging Program

Funded in:

May 2001, for 2 years

Funding Amount:


Investigator Biographies

J. James Frost, M.D., Ph.D.

Professor of Radiology and Neuroscience, Johns Hopkins University School of Medicine



Mu opioid receptor availability will be lower during a capsaicin-induced pain plus placebo condition compared to capsaicin-induced pain alone.

To use a randomized crossover design to measure brain mu-opioid receptor binding using PET in response to capsaicin-evoked pain in healthy human subjects under the influence of placebo, compared to a session without placebo.

PET imaging of mu opioid receptors in 10 healthy subjects will be performed using C-11 carfentanil, a mu-selective ligand. Subjects will first undergo a training session in which they will experience topical capsaicin (10%) on the back of the right hand. A thermode placed over the capsaicin site regulates the actual pain. They will be told that the placebo will significantly lower the pain. At the time the subjects receive an IV solution of saline (the placebo), the temperature of the thermode will be lowered from 37°C to 43°C. The pain from the topical capsaicin typically decreases in response to this temperature manipulation.

The next two sessions are held on separate days and each is conducted with PET scanning. The topical capsaicin will be placed on the dorsum of the left hand. The site will be different for the two test sessions. In one the same placebo is given (temperature maintained at 37°C), while in the other no placebo manipulation is done. The order of the placebo versus non-placebo condition will be randomized. Subjects will continuously rate pain intensity using a mouse-driven computer VAS display on a monitor visible during PET scanning.

The imaging results of the two conditions will be compared using Statistical Parametric Mapping (SPM) analysis software in order to identify brain regions where the mu receptor availability is lower in the placebo condition compared to the non-placebo condition; the converse analysis will also be performed. Since the magnitude of the placebo response will likely vary among the subjects, we will also perform SPM analysis using the placebo response magnitude as a confounding covariate in identifying regions of significant change between the baseline and placebo conditions.

The study indicated that several areas of the brain are activated under conditions of placebo analgesia, which lends support to the remarkable hypothesis that the opioid pathways mediate placebo analgesia. Moreover, we learn about where this effect is mediated. Of the regions identified, a particularly interesting area is the periaqueductal area. Activation in this region supports our hypothesis that this region plays a key role in control of processing of noxious information. This area is rich in opioid receptors and microinjection of opioids induces analgesia. This area also gives rise to a descending modulatory input to the rostral-medial medulla, a center linked to processing of noxious inputs at the level of the dorsal horn. The haplotype data provide support for the hypothesis that polymorphisms in the enzyme, GCH1, account for a significant degree of the variation in pain ratings. This supports the view that genetic variation plays a key role in whether patients are likely to suffer with pain.