Institutional review board/informed consent
This study was conducted at Loma Linda University School of Medicine/Medical Center, Department of Internal Medicine, Division of Dermatology, Loma Linda, California. The study protocol, subject informed consent documents, and subject information documents were submitted to and received approval from the study center's Institution Review Board. This study was conducted according to the Code of Federal Regulations of the United States Food and Drug Administration (21 CFR Part 56, Institutional Review Boards, and Part 50, Protection of Human Subjects) and the International Conference on Harmonization Edition 6, Guideline for Good Clinical Practice.
Study Conduct
This was a phase II, double-blind, placebo-controlled, randomized parallel group study. Randomized subjects had at least 5 clinically visible AK lesions within a 25-cm2 area on the balding scalp. Subjects were randomized to imiquimod or placebo cream in a 3:1 ratio and applied study cream to the treatment area 3 times per week for 4 weeks. Study cream was applied prior to normal sleeping hours and remained on the skin for approximately 8 hours before it was removed. Safety evaluations were made at all treatment and post treatment visits, and included monitoring of adverse events and local skin reactions, as well as photographing the treatment area and reviewing any concomitant medications.
At the screening visit, AK lesions were assessed clinically and by confocal microscopy and a representative lesion confirmed by histology. Because AK lesions are in general small in size, histology and gene expression analysis could not be performed on the same biopsy. Therefore, confocal microscopy was performed to establish a correlation of the confocal images and their respective non-sun exposed non-lesional skin, sun exposed non-lesional skin, actinic keratoses lesions, and squamous cell carcinoma. All subjects with lesions histologically identified as having a degree of dysplesia suggestive of squamous cell carcinoma were disqualified from the study. All sites identified as AK lesions were marked and a plastic template of their locations made for exact identification at a later time. Thereafter, changes in AK lesions due to treatment with imiquimod were assessed clinically and by confocal microscopy. Lesions were scored as cleared if the skin exhibited normal epidermis as assessed clinically and by confocal microscopy. Assessment of lesion regression and the results of the confocal microscopy as they relate to aberrant gene expression in AK are discussed in a manuscript submitted for publication (Torres et al, 2006. Micro Array Analysis of Aberrant Gene Expression in Actinic Keratosis: Effect of Imiquimod 5% Cream).
At the treatment initiation visit, a shave biopsy was taken for gene expression analysis from an untreated AK lesion, from a sun-exposed non-lesional site on the head and from a non-lesional sun-unexposed site from under the arm area. An additional biopsy was taken at each subsequent study visit, (treatment period weeks 1, 2, and 4) and at 4 weeks post treatment of either a remaining AK, or if no AKs were present, of nonlesional skin at a previously identified lesional site. Thus each biopsy was of a different AK lesion. In an effort to standardize the amount of tissue that was removed at each biopsy, the same size punch was used to score the skin surrounding all the lesions to be biopsied for gene expression studies with an attempt to shave-biopsy the lesion at the papillary dermis level. Biopsies were taken 8 hr to 16 hr after the third treatment of the treatment period. The nonlesional, sun-unexposed site biopsy was used to establish a baseline for comparison of gene expression changes in AK before and after treatment with imiquimod. Shave biopsies were immediately immersed in RNALater (Ambion, Austin, Texas), equilibrated at room temperature for 1 hour, kept at 4°C for 24 hours, and then stored at -20°C prior to RNA extraction.
RNA extraction and purification
Total RNA from the biopsy samples was extracted and purified using Qiagen RNeasy Mini Kit Protocol for the Isolation of Total RNA from Heart, Muscle and Skin Tissue (Qiagen, Valencia, California) according to manufacturer's instructions. RNA yield varied from 1.5 to 14 υg. The purity of the RNA was determined by the 260 nm/280 nm absorbance ratio. The median 260 nm/280 nm ratio value for the 119 samples was 2.0 (range 1.9 to 2.2). RNA integrity was determined using the Agilent 2100 Bioanalyzer and the RNA 6000 Nano Assay (Agilant Technologies, Palo Alto, CA). All samples gave 28S/18S ratio between 2 and 3 indicating good quality RNA.
Micro array analysis
Samples for micro array analysis were prepared by 2 rounds of linear target amplification according to the Affymetrix instructions for eukaryotic small sample preparation [28,29]. Briefly, double-stranded cDNA was synthesized from 100 ng of total RNA with oligo(dT)24T7 primer (Affymetrix, Santa Clara, California), followed by 2 cycles of in vitro transcription of cRNA. The first cycle of invitro transcription was performed using a T7 polymerase (MEGAscript T7 Kit, Ambion, Austin, Texas) and the second cycle using Enzo BioArray High Yield RNA Transcript Labeling Kit (Affymetrix). The median value for the biotinylated cRNA yield over the 119 samples was 90 (range 42 to 121). The biotinylated cRNA was hybridized to Affymetrix U133A and U133B GeneChip arrays containing 22,253 probes sets each. Each array was hybridized for 16 hours, washed, then stained with streptavidin-phycoerythrin conjugate and scanned according to manufacturer's instructions. Images were analyzed using Micro Array Suite Version 5 (MAS5). Chips were normalized to a global average intensity of 150 to allow chip-to-chip comparison. The quality of the images was ascertained by monitoring the noise, background, percent transcript present, and the 3'/5' ratio for the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) which ranged from 1.5 to 5 and beta actin which ranged from 7 to 24. These values are similar to those reported for double amplification protocols [29,30].
TaqMan™ real time reverse transcriptase polymerase chain reaction
TaqMan real time RT-PCR was performed for a number of the genes to confirm the micro array results. cDNA was reverse transcribed from total RNA using Invitrogen Superscript First-Strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, California). Real time RT-PCR was performed using the Applied Biosystems 7900HT™ sequence detection instrument (Applied Biosystems, Foster City, California), and TaqMan low density custom array micro fluidic cards (Applied Biosystems, Foster City, California) as described previously[31]. The micro fluidic cards consisted of 8 ports with 24 different TaqMan primer pair/probe sets arrayed in duplicate in a 384-well micro plate footprint. Each well contained a gene-specific forward and reverse primer, as well as a gene-specific probe, which is labeled at the 5' position with 6FAM (areporter dye) and at the 3' position with minor groove binder/non-fluorescent quencher. Samples were mixed with TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, California), applied to each port of the card, and analyzed by PCR on the 7900HT instrument using Applied Biosystems Sequence Detection System 2.0 software according to the manufacturer's instructions. A total of 46 selected genes associated with the Toll-like receptor pathway, apoptosis, cell cycle, and immune cell infiltration were analyzed using the TaqMan arrays. The housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used to normalize each sample. The TLR copy number was calculated for 2 ng of RNA.
Analysis of affymetrix gene chip data
Signals from the gene chip images of biopsy samples for untreated normal skin obtained from sun-unexposed sites were used as a control for the calculation of changes in expression in samples of pretreatment AK lesions and AK lesions during and after treatment (treatment weeks 1, 2, and 4 and 4 weeks post treatment). The statistical algorithm in MAS5 evaluates the image for the expression signal, the absent/present call, and the p-value associated with the signal. It also evaluates the fold change of the sample relative to the designated control sample expressed as the signal log2 ratio, the p-value associated with the fold change, and the direction of change (increased, I; decreased, D; or no call, NC). Using the Affymetrix data mining tool software, the expression data from MAS5 were filtered on the basis of specific criteria to identify differentially expressed genes. A given gene was retained if 1sample from the series passed the following criteria: signal detection p-value ≤ 0.01, signal log2 ratio ≤ -2 or ≥ 2, and a change call designation of 'increased' (I) or 'decreased' (D). A total of 1682 genes passed these criteria.
There was a large subject-to-subject variation in the magnitude of changes in expression, as well as the temporal pattern of expression (subject-specific variation in magnitude of response with treatment time) of the 1682 genes. This variation is exemplified in Figure 1, which shows the variation in the temporal expression of IRF7 for selected subjects. Peak expression varied from week 1-treatment to week 4-treatment depending on the subject. This variation may be due to the inherent differences in the responses of subjects and (or) due to variation in the time from treatment to biopsy (range 8 to 16 hr). In order to identify differentially expressed genes in the treatment group without regard to variation in temporal response, the data were reduced in the following manner. The fold change values for samples in the week 1, week 2 and week 4 treatment groups for the 1682 genes for each subject were compared to identify the maximum change in expression (increased or decreased) due to imiquimod or placebo treatment. This value was designated as the treatment-response fold change for imiquimod or placebo and used in the ANOVA described in the Statistical Analysis section.
Cluster analysis was performed with Spotfire DecisionSite-8.1 for Functional Genomics (Spotfire Inc, Somerville, Massachusetts), using the Unweighted Pair-Group Method with Arithmetic mean (UPGMA) and the Euclidean similarity measure. Functional categorization of genes was based on gene ontology analysis using the Ontology Browser in Spotfire and gene descriptions at the National Center for Biotechnology Information (NCBI) website [32]. The ontology browser calculates a Fisher's Exact Test p-value, which reflects the chance that the gene ontology category is represented by random chance [33]. P-value
Statistical analysis
The natural log of the fold change (with respect to sun-unexposed normal skin) from the Affymetrix gene expression and the real time RT-PCR experiments were used in an ANOVA to determine statistically significant changes in expression between sample groups. Due to the small sample size of the vehicle group, 4 subjects, compared to 13 subjects in the imiquimod-treated group, a one way ANOVA comparing vehicle-treated subjects to imiquimod-treated subjects is not expected to yield a reliable determination of imiquimod-response genes. Therefore, a 2-way ANOVA using a blocking factor to account for repeat observations on the same subject was used to compare: (1) the fold change for pretreatment AK (n = 13) and the fold change during treatment with imiquimod (n = 13), and (2) the fold change for pretreatment AK (n = 13) and the fold change for samples taken 4 weeks after the last imiquimod treatment (n = 13). The 'during treatment' fold change is the maximum response fold change value (decreased or increased) selected from week 1, week 2 and week 4 treatment fold changes. Differences between sample groups were considered significant if the p-value for the ANOVA was
Answers to all your Biology Questions
