Quantitative PCR on 5 genes reliably identifies CTCL patients with 5% to 99% circulating tumor cells with 90% accuracy

Abstract

We previously identified a small number of genes using cDNA arrays that accurately diagnosed patients with Sézary Syndrome (SS), the erythrodermic and leukemic form of cutaneous T-cell lymphoma (CTCL). We now report the development of a quantitative real-time polymerase chain reaction (qRT-PCR) assay that uses expression values for just 5 of those genes: STAT4, GATA-3, PLS3, CD1D, and TRAIL. qRT-PCR data from peripheral blood mononuclear cells (PBMCs) accurately classified 88% of 17 patients with high blood tumor burden and 100% of 12 healthy controls in the training set using Fisher linear discriminant analysis (FLDA). The same 5 genes were then assayed on 56 new samples from 49 SS patients with blood tumor burdens of 5% to 99% and 69 samples from 65 new healthy controls. The average accuracy over 1000 resamplings was 90% using FLDA and 88% using support vector machine (SVM). We also tested the classifier on 14 samples from patients with CTCL with no detectable peripheral involvement and 3 patients with atopic dermatitis with severe erythroderma. The accuracy was 100% in identifying these samples as non-SS patients. These results are the first to demonstrate that gene expression profiling by quantitative PCR on a selected number of critical genes can be employed to molecularly diagnosis SS.

Figure 1.

Dendrograms showing gene expression measured by qRT-PCR on RNA from 18 Sézary patients with high blood tumor burden and 12 healthy controls, including 3 untreated and 9 skewed to Th2 phenotype. Hierarchical clustering was applied to both the genes and the samples and was performed using Pearson correlation–based distance metric and Ward linkage. Samples from SS patients, untreated controls, and Th2-skewed controls are colored in red, green, and blue, respectively. Genes up- and down-regulated in the patients as compared with Th2-skewed controls are colored red and green, respectively. For visual enhancement, expression levels for each gene are converted to z scores. (A) Dendrogram of qRT-PCR on 26 of the genes with P < .01 measured by qRT-PCR on amplified RNA. (B) Dendrogram of 5 genes selected for our discriminant model on the same set of 30 samples assayed by qRT-PCR on total RNA derived from the same samples.

Figure 2.

Classification of Sézary patients using qPCR data. Data shown are for analysis using FLDA. Scores for patient samples are shown in panel A, and controls are shown in panel B. The dataset consists of 125 samples: 56 from 49 patients with erythrodermic CTCL and 69 from 65 controls. The controls are untreated (UT), Th1 skewed (Th1), Th2 skewed (Th2), and PHA treated (PHA). The patient samples are sorted according to the blood tumor burden (5% to 99%) indicated by the last 2 digits of the label. Patients in whom percent Sézary was not available are indicated as “NA.” A positive score indicates the sample is classified as a patient; a negative score indicates the sample is classified as a control. The height of each bar represents the average score that a given sample received, when tested approximately 100 times, during 1000 random resamplings. The error bars indicate the standard deviation in the generated set of scores for the sample. Sézary samples (S) are followed by a 3-digit patient donor code. Serial samples taken at different times from the same patient are indicated by an additional number (ie, S151.1, S151.4) and followed by the percent Sézary cells present. None of the controls was sampled more than once, but samples from donors C017, C018, and C019 were also skewed to the Th1 (C017, C018, C019) and Th2 (C017) phenotype. The classifier has a sensitivity of 86% and a specificity of 95% on this dataset.

Figure 3.

Dendrogram of 125 samples on 5 genes selected for our discriminant model. Hierarchical clustering was carried out using Pearson correlation–based distance metric with Ward linkage on both genes and the samples. For visual enhancement, the gene expression values for each gene shown are converted to z scores. Patient samples are shown in red, untreated controls are shown in green, and Th1- and Th2-skewed controls are shown in light and dark blue, respectively. Yellow highlights indicate the samples that were misclassified by FLDA.

Figure 4.

Classification of samples from MF/CTCL, atopic dermatitis, and SS/CTCL in remission. A positive score indicates the sample is classified as an SS; a negative score indicates the sample is classified as a non-SS. The length of each bar represents the average score over 1000 random resamplings, and the error bars indicate the standard deviation in the scores generated by the resamplings. MF indicates mycosis fungoides; RS, SS patient in remission; AD, atopic dermatitis.