Reference
Kontos M, Wilson R, Fentiman I. Digital infrared thermal imaging (DITI) of breast lesions: sensitivity and specificity of detection of primary breast cancers. Clin Radiol. 2011;66(6):536-539.
Design
Prospective cohort
Participants
63 female patients with symptomatic breast mass(es) who presented to a diagnostic breast clinic
Intervention
The breast masses were evaluated using digital infrared thermal imaging (DITI) before core biopsy or surgical excision.
Outcome Measures
Accuracy of DITI as compared to pathology from core biopsy or surgical excision of breast mass
Key Findings
Digital thermography had 90 true-negative, 16 false-positive, 15 false-negative and 5 true-positive results. The sensitivity was 25%, specificity 85%, positive predictive value 24%, and negative predictive value 86%.
Practice Implications
It’s time to reconsider the use of breast thermography in breast cancer screening and breast mass assessment. Thermography remains a controversial topic in integrative oncology. Although digital thermography is FDA-approved for adjunctive breast imaging to clarify results of mammography or breast MRI, it is often employed for broader applications. In particular, some practitioners and patients use thermography for cancer screening, to evaluate palpated masses, and as an alternative to mammography.
The key question is: Can we ethically use thermograms as a screening tool to detect breast cancer? The answer depends on how accurate digital infrared thermograms are when used to screen asymptomatic women or to assess breast masses.
Early studies on thermography did not produce favorable results. Hitchcock et al reported in 1968 that thermography was ineffective as a screening test for breast cancer. They used thermograms to screen 2,523 female volunteers for breast cancer. The technique identified only 1 of the 4 confirmed breast cancers that were detected by mammography.1 A study from 1977 reported that analog thermography detected only 39% of biopsy-proven tumors, while clinical examination detected 55% and analog mammography found 78%.2
Thermography advocates have discounted these early results, arguing that in recent years technological improvements in digital imaging, infrared technology, and use of sympathetic stimulation have contributed to improving the results obtained via DITI. Some practitioners and patients, despite these earlier studies, still consider DITI as an alternative to mammography.
This current publication is timely as it provides a measure of the accuracy of modern digital thermography. Kontos et al found that thermography’s primary purported application, screening for breast masses, is the application for which it is least accurate, with sensitivity of 25% and positive predictive value of 24%. This means that DITI detected only ¼ of malignancies and that, given a positive thermogram, the probability of a true malignancy was 24%. Thus even modern thermography is a poor substitute for a mammogram or breast MRI in the detection of malignancy.
Thermography is FDA-approved for adjunctive breast imaging and may have applications for clarifying the significance of breast masses detected on mammogram or MRI. For example, a prospective clinical trial that followed an abnormal mammogram with DITI found that thermography had an 82% negative predictive value.3 Given an abnormal mammogram followed by a normal thermogram, there was a 82% probability that the lesion was truly benign.
Thermograms may be useful for clarifying the significance of an abnormal mammogram or for adjunctive imaging of a highly suspicious palpated mass.
Similarly, Kontos et al found a specificity of 85% and negative predictive value of 86%.4 The increased accuracy of thermography in this application is due in part to an increase of prior probability of malignancy in a population of patients with abnormal mammograms, compared to a screening population. Based on Bayes theorem, the prior probability of malignancy significantly contributes to the accuracy of a screening test.
How should we utilize this information? Based on the data from Kontos and others, I do not recommend digital thermograms to patients for breast cancer screening or for primary assessment of a breast mass. Thermograms may be useful for clarifying the significance of an abnormal mammogram or for adjunctive imaging of a highly suspicious palpated mass. However, the clinician needs to be prepared to translate the accuracy statistics for the patient and to carefully document informed consent in the event that the patient chooses to forego biopsy of a breast mass.
There are other applications that might be more appropriate for thermograms that this article did not measure. These uses include the detection of precancerous dysplasia, the accurate assessment of long-term risk of developing breast cancer, and the evaluation of possible dysplasia developing in response to hormone replacement therapy.
