A tourist finds breast cancer after a thermal scan at Camera Obscura. That doesn’t mean thermography works.

It occurs to me that I haven’t written about thermography in a while; indeed, I don’t think I’ve devoted a whole post to it since 2010, when Dr. Christiane Northrup published a hype-filled article touting thermography for breast cancer detection. Thermography is an imaging technique that produces an image based on the heat signature of the object being imaged. It’s an imaging technique that has many uses, but one of them is definitely not as a screening for breast cancer. Of course, October is Breast Cancer Awareness Month, and the month is almost over, which makes it strange that news or a study about breast cancer that caught my interest has’t shown up on my blog topic radar and piqued my interest enough to write about it. There’s another story out there that I might also want to write about, but it will take a lot more in depth analysis, which is why I’ll probably save it for next week, which, fortunately, will still be October. In the meantime, I’m sure you’ve probably seen this story. It started out as a local story in Scotland a few days ago and found its way to international news, with USA TODAY publishing an article about it just yesterday by N’dea Yancey-Bragg entitled A trip to a popular tourist attraction led this woman to discover she has breast cancer A trip to a popular tourist attraction led this woman to discover she has breast cancer. The attraction was Camera Obscura & World of Illusions in Edinburgh. Here’s the spin:

A family photo snapped at a Scottish tourist attraction led a woman to discover she has breast cancer. When Bal Gill, 41, snapped a photo in the thermal camera at the Camera Obscura and World of Illusions in Edinburgh, she noticed something unusual: a red heat patch over her left breast. It wasn’t until a few days later, when she was going through vacation pictures, that she decided to Google what the image might mean. Gill found several articles about breast cancer and thermal imaging cameras and made an appointment with her doctor, where she learned she had an early-stage breast cancer. She’s since had two surgeries and has to have another to prevent the cancer from spreading. “I just wanted to say thank you: without that camera, I would never have known,” she said in a statement published on the Camera Obscura’s website. “I know it’s not the intention of the camera but for me, it really was a life-changing visit.”

I can see it right now. Naturopaths and other alternative medicine mavens will likely tout this as evidence that thermography “works.”

But is it? Given my specialty as a breast cancer surgeon, you know I can’t resist unpacking this story and taking a closer look. So let’s start out with the post on the Camera Obscura Facebook page on Tuesday:

And, just yesterday:

This post references a blog post on the Camera Obscura website:

Andrew Johnson, General Manager of Camera Obscura & World of Illusions said:

“We did not realise that our Thermal Camera had the potential to detect life-changing symptoms in this way. We were really moved when Bal contacted us to share her story as breast cancer is very close to home for me and a number of our team. It’s amazing that Bal noticed the difference in the image and crucially acted on it promptly. We wish her all the best with her recovery and hope to meet her and her family in the future.” Thermography, also called thermal imaging, is a tool used by breast cancer specialists and uses a special camera to measure the temperature of the skin on the breast’s surface. It is a non-invasive test that involves no radiation. Thermography is based on two ideas; because cancer cells are growing and multiplying very fast, blood flow and metabolism are higher in a cancer tumour; as blood flow and metabolism increase, skin temperature goes up.

Looking at the image above and the image posted to the Camera Obscura website, I must admit to some confusion. The only “red heat patch” I see anywhere near Ms. Gill’s left breast is obviously the face of what I assume to be one of her children, who can easily be identified as a person because he or she is waving at the camera and I can see the outline of the person’s hand. No, seriously. Look at the image. (If I’m off-base on this, let me know.) There is a bit of a yellow patch over Ms. Gill’s right breast, but it’s singularly unimpressive. I suppose it’s possible that the sides are reversed and that yellow patch is really over Ms. Gill’s left breast, but that doesn’t seem too likely either. I suppose it’s also likely that there are other thermal images from Ms. Bal’s trip to Camera Obscura, but if there are why did she choose this one to share with the media? Presumably it’s the one that tipped her off to go to her doctor.

I did find one story that shows the same image, with a circle over the yellow warm patch over the right breast, suggesting that that was the “hot spot” that alerted Ms. Gill to see her doctor. Whatever the case, that is a singularly unimpressive hot spot. It could be due to any number of things that could make that area of her skin or clothing slightly warmer.

I must admit that I was pleasantly surprised that most of the articles reporting this story that I read correctly noted that thermography is not recommended for breast cancer screening. For instance, USA TODAY:

Thermal imaging cameras like the ones used at Camera Obscura have been used in medicine, but experts don’t recommend it for breast cancer screening. Thermography is a noninvasive tool that uses an infrared camera to capture images of heat and blood flood on the surface of the body. Although thermography has been successful at using heat signatures to assess repetitive motion injuries, it “just doesn’t work very well” for breast cancer screenings, according to Robert Smith, Senior Vice President of cancer screening at the American Cancer Society. “The problem with thermography though is that not all breast cancers, especially very little ones, produce a heat signature and other benign conditions in the breast can also produce heat signatures,” Smith said

This is, of course, basically accurate, as is this assessment in a CNN article:

Caroline Rubin, vice president for clinical radiology at The Royal College of Radiologists, said there is no scientific evidence to suggest they are effective tools in screening for breast cancer or other medical conditions. “The principle behind using thermography is that infrared heat cameras can be used to map patterns of heat and blood flow close to the surface of the body,” said Rubin in a statement, who explained that cancer tumors generate heat as they tend to have a good blood supply but they vary hugely in size and position. “Thermography devices are not sensitive or specific enough to be a trusted method to detect breast cancer — in Ms Gill’s case the discovery was serendipitous.”

“Serendipitous”? That’s putting it mildly. I’d say it was more of a happy freak accident based on what was likely a misinterpretation. Consider how thermography works. The idea is simple. As its name implies, thermography measures differences in temperature. Most thermography systems use infrared imaging to achieve these measurements, and there’s nothing magical about it; the technology has been in use for various applications for several decades. The rationale for applying thermography to the detection of breast cancer is that breast cancers (like other cancers) tend to induce angiogenesis, which is nothing more than the ingrowth of new blood vessels into the tumor to supply its nutrient and oxygen needs. A tumor that can’t induce angiogenesis can’t grow beyond the diffusion limit in aqueous solution, which is less than 1 mm in diameter. These blood vessels result in additional blood flow, which results in additional heat. In addition, the metabolism of breast cancer cells tends to be faster than the surrounding tissue, and cancer is often associated with inflammation, two more reasons why the temperature of breast cancers might be higher than the surrounding normal breast tissue and therefore potentially imageable using infrared thermography.

There are many problems, however. For one thing, thermography tends to measure surface temperature; it’s not clear that a tumor deep in the breast would produce enough additional heat to show up on the surface. Another issue is that there aren’t any widely agreed-upon standards for the performance and interpretation of breast thermography. Not only have these standards not been developed, largely thanks to its adoption as a “breast cancer screening tool” that’s “better and safer than mammography” by some physicians, chiropractors, and naturopaths. Still another issue is that it’s a crap study to find breast cancer in an asymptomatic patient. I hadn’t reviewed thermography in several years; so I did a few Pubmed searches.

One study, in particular, was quite damning. Published in Academic Radiology in 2018 by a group at my medical alma mater, the University of Michigan, this retrospective study asked a very simple question: What percentage of women who presented to U. of M.’s breast center with a “positive thermogram” actually turned out to have breast cancer? True, it was a small study, with only 38 patients, but the results are telling:

Ninety-five percent (36 of 38) of patients did not have breast cancer. The two patients diagnosed with breast cancer had suspicious clinical symptoms including palpable mass and erythema. No asymptomatic woman had breast cancer. Negative predictive value was 100%. Of 38 patients, 79% (30 of 38) had Breast Imaging Reporting and Data System (BI-RADS) 1 or 2 assessments; 5% (2 of 38) had BI-RADS 3; and 16% (6 of 38) had BI-RADS 4 (n = 5) or BI-RADS 5 (n = 1) assessments. Two of six patients with biopsy recommendations were diagnosed with breast cancer (Positive predictive value 2 = 33.3%). All findings recommended for biopsy were ipsilateral to the reported thermography abnormality.

The BI-RADS reporting system is a system used to grade how suspicious for breast cancer that patient’s imaging (mammography, ultrasound, MRI). BI-RADS 1 or 2 means normal or benign findings; BI-RADS 3, probably benign findings (chance of cancer ~1-2%); BIRADS 4 and 5, suspicious findings and very suspicious findings, respectively). So, from the above, you can see that the vast majority of women with “positive” thermography had negative mammograms (BI-RADS 1 or 2) and only a minority had suspicious findings (BI-RADS 4 or 5).

These findings led the researchers to conclude:

No cancer was diagnosed among asymptomatic women. The 5% of patients diagnosed with cancer had co-existing suspicious clinical findings. Mammogram and/or ultrasound were useful in accurately characterizing patients with abnormal thermography.

Of course, the very purpose of a screening test is to detect occult disease in an asymptomatic patient. Clearly, this study, at least, shows thermography failing miserably at that task. The only women with “positive thermography” who actually turned out to have disease weren’t asymptomatic!

A 2019 review on thermography notes that thermograms are quite sensitive to many conditions, including the temperature of the room and that current thermography suffers from high false positive and false negative rates. The author argues that these problems can be “effectively reduced by suitable combinations of segmentation, feature extraction and classification techniques” and that specific “outputs from numerical simulation can also act as input to artificial neural networks, thereby integrating thermography and numerical simulation.” In other words, artificial intelligence will be the answer! I’m not so sure. I’m open to potentially using thermography to complement mammography, potentially reducing its false positive and negative rates when the two imaging modalities are used together, as this review article argues to be a promising strategy. Heck, years ago at my previous job I even participated in a clinical trial whose goal was to add a thermography-like machine to mammography. It’s just that thermography by itself is pretty much useless as a screening tool for breast cancer.

I’m actually happy to have seen this story, because there was another story a while back that, for whatever reason, I never wrote about. Looking at the date, I think I know why; it happened right around the time my pain from my cervical radiculopathy was at its worst and I was on track for my first appointment with major surgery in adulthood. Basically it was an FDA warning on February 25, FDA Warns Thermography Should Not Be Used in Place of Mammography to Detect, Diagnose, or Screen for Breast Cancer: FDA Safety Communication. A tourist at Camera Obscura in Edinburgh saw a hotspot on her breast and was diagnosed with breast cancer. Does this validate thermography? In this safety communication, the FDA reiterates that thermography alone is not reliable for breast cancer detection and that only mammography is the only method validated to decrease breast cancer-specific mortality and mentions that it had sent several warning letters to those selling thermography machines and services:

On February 22, 2019, the FDA issued a Warning Letter to Total Thermal Imaging, Inc., in La Mesa, California, for marketing and promoting thermography devices for uses that have not received marketing clearance or approval. The FDA has also issued Warning Letters to the following facilities:

Looking at the letters above, depressingly I see the same sorts of claims, such as that thermography can produce “early detection of the diagnosis of many disorders including breast cancer, inflammatory breast cancer, pre-stroke, heart disease, deep vein thrombosis, reflex sympathetic dystrophy/complex regional pain syndrome, back, leg or headache, and even unexplained pain, TMJ, and other disease” and thus save your life and that it is “an alternative to mammography that doesn’t involve any patient contact (no pain), will not cause cancer (no radiation), and is far more efficient at detecting cancer.” One claimed that breast thermography “has: 99% Sensitivity, 90% Specificity, 90% Accuracy” and is “far more sensitive than mammography,” recommending that all “women 25 and older should get annual Thermography screenings.”

Also found are frequent claims that thermography is “FDA-approved.” This is a deceptive claim in that, yes, in 1982 the FDA approved Class I telethermographic systems for the use of “adjunctive diagnostic screening for detection of breast cancer,” which means that thermography can be used only in conjunction with approved screening methods, not in place of them, to detect breast cancer. Companies and quacks marketing thermography, unfortunately, frequently get away with forgetting to mention that last part. Canada, meanwhile, also warns that thermography is nothing more than ” specialized thermometers” and shouldn’t be used to screen for breast cancer.

Predictably, in response to the stories about her cancer being detected by Camera Obscura (which it almost certainly wasn’t), the quacks are coming out of the woodwork. For instance, here we have someone telling her not to have surgery, which she’s already had, given that her visit was in May:

I’m only surprised there hasn’t been more of this. I’m also pleasantly surprised that most news articles reporting on this story point out that thermography on its own is a highly unreliable method to screen for breast cancer, even as I worry that publicity over this story during Breast Cancer Awareness Month might lead some women to conclude that thermography would be better for them. I’m also very happy for Bal Gill, although it is almost certain that her detection of her breast cancer was not because the Camera Obscura thermal image detected her cancer but because a happy accident that, for whatever reason, that patch of skin (or even her clothings) over her breast was a little hotter than the surrounding skin and clothing, which led her to wonder and to go to see her doctor. She’s a very lucky woman.