Sensitivity and area under the curve (AUC) analyses of thermography that is combined with diagnostic software demonstrate “the efficacy of the tool for breast cancer screening,” concludes an observational, comparative study from India published online Oct. 1 in JCO Global Oncology, a publication of the American Society of Clinical Oncology.
Siva Teja Kakileti of Niramai Health Analytix, Koramangala, Bangalore, India, and colleagues said that the product, Thermalytix, is potentially a good fit for low- and middle-income countries because it is portable and provides automated quantitative analysis of thermal images – and thus can be conducted by technicians with “minimal training.”
Conventional thermography involves manual interpretation of complex thermal images, which “often results in erroneous results owing to subjectivity,” said the study authors.
That manual interpretation of thermal images might involve looking at 200 color shades, which is “high cognitive overload for the thermographer,” explained Mr. Kakileti in an interview.
However, an American mammography expert who was approached for comment dismissed thermography – even with the new twist of software-aided diagnostic scoring by Thermalytix – as wholly inappropriate for the detection of early breast cancer, owing to inherent limitations.
“Thermal imaging of any type has no value in finding early breast cancer,” Daniel Kopans, MD, of Harvard University and Massachusetts General Hospital, both in Boston, said in an interview. He said that thermal imaging only detects heat on the skin and perhaps a few millimeters beneath the skin and thus misses deeper cancers, the heat from which is carried away by the vascular system.
The new study included 470 women who presented for breast screening at two centers in Bangalore, India. A total of 238 women had symptoms such as breast lump, nipple discharge, skin changes, or breast pain; the remaining 232 women were asymptomatic.
All participants underwent a Thermalytix test and one or more standard-of-care tests for breast cancer screening (such as mammography, ultrasonography, biopsy, fine-needle aspiration, or elastography). A total of 78 women, or 16.6% of the group overall, were diagnosed with a malignancy. For the overall group of 470 women, Thermalytix had a sensitivity of 91.02% (symptomatic, 89.85%; asymptomatic,100%) and a specificity of 82.39% (symptomatic, 69.04%; asymptomatic, 92.41%) in detection of breast malignancy. Thermalytix showed an overall AUC of 0.90, with an AUC of 0.82 for symptomatic and 0.98 for asymptomatic women.
The study authors characterized both the sensitivity and AUC as “high.”
The results from the study, which the authors characterized as preliminary, encouraged the study sponsor, Niramai, to start planning a large-scale, multicountry trial.
But Dr. Kopans, who serves as a consultant to DART, which produces digital breast tomosynthesis units in China, suggested that this research will be fruitless. “Thermal imaging seems to raise its head every few years since it is passive, but it does not work and is a waste of money,” Dr. Kopans reiterated.
“Its use can be dangerous by dissuading women from being screened with mammography, which has been proven to save lives,” he stressed.
Thermalytix compared with mammography
Investigators also compared screening results in the subset of 242 women who underwent both Thermalytix and mammography. Results showed that Thermalytix had a higher sensitivity than did mammography (91.23% vs. 85.96%), but mammography had a higher specificity than Thermalytix did (94.05% vs. 68.65%).
In the asymptomatic group who underwent both tests (n = 95), four cancers were detected, and Thermalytix demonstrated superior sensitivity than mammography (100% vs. 50%), Mr. Kakileti and colleagues state.
Thermalytix evaluates vascularity variations too
In the subset of 228 women who did not undergo mammography (owing to dense breasts, younger age, or other reasons), Thermalytix detected tumors in all but 3 of 21 patients who went on to be diagnosed with breast cancer. The authors state that, because their artificial intelligence–based analysis uses vascularity, as well as temperature variations on the skin, to complement hot-spot detection, it is able to detect small lesions.
In the current study, 24 malignant tumors were less than 2 cm in diameter, and Thermalytix was able to identify 17 of the tumors as positive, for a 71% sensitivity rate for T1 tumors. This compared with a 68% sensitivity rate for mammography for detecting the same T1 tumors. Thermalytix also showed promising results in women younger than 40 years, for whom screening mammography is not usually recommended. The automated test picked up all 11 tumors eventually diagnosed in this younger cohort.
“Thermalytix is a portable, noninvasive, radiation-free test that has shown promising results in this preliminary study,” the investigators wrote, “[and] it can be an affordable and scalable method of screening in remote areas,” they added.
“We believe that Thermalytix ... is poised to be a promising modality for breast cancer screening,” Mr. Kakileti and colleagues summarized.
The FDA warns about thermography in place of mammography
The US Food and Drug Administration fairly recently warned against the use of thermography as an alternative to mammography for breast cancer screening or diagnosis, noting that it has received reports that facilities where thermography is offered often provide false information about the technology that can mislead patients into believing that it is either an alternative to or a better option than mammography.
Dr. Kopans says that other groups have invested in thermography research. “The Israelis spent millions working on a similar approach that didn’t work,” he commented.
The new software from Thermalytix, which is derived from artificial intelligence, is a “gimmick,” says the Boston radiologist. “If the basic information is not there, a computer cannot find it,” he stated, referring to what he believes are deeper-tissue tumors that are inaccessible to heat-detecting technology.
Mr. Kakileti is an employee of Nirami Health Analytix and owns stock and has filed patents with the company. Other investigators are also employed by the same company or receive research and other funding or have patents filed by the company as well. Dr. Kopans serves as a consultant to DART, which produces digital breast tomosynthesis units in China.
A version of this article originally appeared on Medscape.com.