Trained Dogs Sniff Out Thyroid Cancer With High Accuracy

Miriam E Tucker

March 09, 2015

SAN DIEGO — New research shows a trained dog could have a nose for thyroid cancer.

Results from a small study starring a canine named Frankie who smelled the difference between cancerous and benign thyroid tumors with 88% accuracy were presented March 6 here at the Endocrine Society annual meeting, ENDO 2015, by Andrew M Hinson, MD, from the University of Arkansas for Medical Sciences (UAMS), Little Rock, and colleagues.

"The detection and diagnosis of thyroid cancer is often difficult, because you are looking for a very small number of occurrences in a relatively large background of benign thyroid nodules. It is also difficult to say with any certainty that a patient is cancer-free after surgery. We think that this diagnostic modality could therefore be a tremendous advance in thyroid cancer. This could also be of significant benefit in terms of cost savings and preventing unnecessary surgeries," Dr Hinson told Medscape Medical News.

At a press briefing, principal investigator Donald Bodenner, MD, PhD, also from UAMS, said that trained dogs could potentially be utilized in underserved areas where traditional detection methods of biopsy and ultrasound are less accessible or unavailable. "You don't need a big clinic or an ultrasound. You just need a urine sample."

Asked to comment on this nontraditional approach, briefing moderator Jason A Wexler, MD, from Washington Hospital Center, Washington, DC, said "This is at a very early stage, and not ready for any clinical applicability. But if it pans out and you could convince payers to pay for it and it's really as reliable and reproducible as it seems it could be, it would have tremendous potential."

Dr Wexler noted that even in the United States and other advanced nations, the dogs might be helpful for patients who are fearful of needle biopsies or who require multiple procedures.

This is not the first time dogs have been shown to be able to smell cancers, although it is the first report of a canine detecting thyroid cancers. Last year, in the largest study to date on this, two dogs were shown to be able to detect prostate cancer with nearly 100% accuracy in several hundred men, and other trials have shown that dogs are capable of sniffing out lung and colon cancer, too.

Scent of a Cancer

The dog in the current study, Frankie, is a male German Shepard mix, about 7 years old at the time the research was conducted, who had been rescued as a stray puppy from a busy Little Rock street. He had initially been trained as a search-and-rescue dog before becoming the first dog ever trained to differentiate cancerous from benign thyroid disease in urine samples.

Over 6 months, Frankie was imprinted with urine, blood, and thyroid-tissue samples obtained from patients with papillary thyroid cancer (PTC) and trained to discriminate between those with and without PTC and indicate the difference to his handlers by lying down if the sample was positive and turning away if negative.

Of 34 urine samples — 15 with PTC and 19 benign — Frankie's alert matched the final surgical pathology diagnosis in 30 out of 34 cases (88.2% accuracy, two false negatives and two false positives), yielding a sensitivity of 86.6%, specificity of 89.5%, a positive predictive value of 73.3%, and a negative predictive value of 95.3%.

The Kappa coefficient — which accounts for the amount of agreement that could be attributed to chance — was 0.75, indicating "good" agreement, Dr Bodenner said during the press briefing.

By contrast, he said, in experienced hands the accuracy of the gold standard, fine-needle–aspiration biopsy, is about 95%, "so we're not there [yet], but I think we're very close in terms of a noninvasive and inexpensive adjunct to current diagnostic practices."

"Potentially Disrupting Technology"

Dr Bodenner said that the Arkansas team has several research agendas related to canine detection.

For one, they plan to investigate the approach for monitoring high-risk thyroid-cancer patients who have been treated but have a high probability of recurrence. "We'd like to monitor and tell early on if there's a clinically significant recurrence."

The investigators are also hoping to use proteomics or metabolomics to try to determine exactly what it is that Frankie smells.

And, they hope to collaborate with a group at Auburn University that supplies animals to the Department of Defense to try to repurpose bomb-sniffing dogs that have returned from Afghanistan and Iraq to detect a variety of cancers, including thyroid.

"Based on our early results and other preliminary reports in the literature, the utility of using scent-trained dogs to detect cancer is unlikely to be limited to any one cancer," Dr Hinson told Medscape Medical News.

"Basically, thyroid cancer is our template. And, while we don't want to get too far ahead of ourselves, we are currently in the process of expanding our work to see whether these methods could be used in the diagnoses of several other cancers, [such as] ovarian, breast, kidney, bladder, prostate, etc. In short, we're excited about the future."

Dr Wexler commented to reporters, "As a clinician, I'm not wedded to any particular method. To me, it would be the one that's most cost-effective. If you could tell me that a dog could do it most cost-effectively and as reliably as a molecular test could and someone's going to pay for it, I think I'd be happy to use that.…It's hard to say anything definitively, but it could be a disrupting technology."

To further the research agenda, the Arkansas team has established the nonprofit Frankie Foundation.

The authors and Dr Wexler have reported no relevant financial relationships.

ENDO 2015: The Endocrine Society Annual Meeting. Abstract FRI 020-054, presented March 6, 2015.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: