As a prequel to my interest in the sport of competitive nosework, I first became fascinated by the olfactory abilities of my Bouvier when:
As Clinical Director of the Breast Care Center (Cottage Health System), I was asked by several patients to bring my TDI dogs in to the office while they were receiving chemotherapy. The dogs stayed in my private office and did not venture out into the rest of the Care Center unless specifically requested by a patient, but the office door was usually open & the dogs had “olfactory access” to the reception area and the main hallway which accessed 4 exam rooms, two treatment rooms, and three mammography suites as well as a restroom and two dressing areas.
In 1995, just after the dogs began coming to the office, a new patient, MM, came for evaluation. One of the dogs, Gandolf (Ch Madrone Ledge Othello of Fogbank, VBCh TDI, CGC, HCT, NA, TAN) , became very alert & focused, went into the staff reception area, and regarded her intently over the counter, nostrils flaring, sniffing loudly. When she was brought back to an exam room, the dog followed her down the hall & pawed gently at the door when it was shut. At the request of MM, the dog was allowed into the exam room and immediately went to sit by her & leaned against her legs, looking at her face and sniffing loudly – behavior that the dog had not exhibited before and which was not encouraged by the patient. MM unfortunately was found to have a large breast cancer filling the left breast which had spread to lymph nodes under the arm and in the neck. Over the 2-week course of her evaluation, the dog’s behavior was the same each of the several times she came in the office. Gandolf responded to no other patient in this manner (usual patient load was 50 women/day) and his response was so striking that several of the staff and my radiologist partner commented upon it.
MM underwent a mastectomy and lymph node removal; the fresh tissue was sent to the lab in a closed stainless steel bucket. The pathologist was agreeable to evaluating Gandolf’s response to the tissue and when the bucket was placed on the floor and opened, the dog was disinterested and showed no appreciable change in his behavior.
Following surgery and one course of chemotherapy in the hospital, MM came to the office for follow-up treatment and Gandolf ignored her. He continued to remain unengaged with her on multiple subsequent visits until about 10 months later, when once again he exhibited the same intense vigilance – staring at her face, sniffing heavily, and leaning against her body. At that time, she reported symptoms that suggested spread of the cancer to other organs, and testing confirmed metastatic disease. Once again, she underwent chemotherapy; after four monthly cycles of treatment, the dog once again ignored her. For another year, MM did well and the dog ignored her presence, not even coming out of my office when invited to do so. At each office visit MM would tell me “Where’s Gandolf? I don’t want a CAT scan, I want a DOG scan.” However, once again Gandolf alerted on her and once again, she had progressive disease.
Bone marrow transplant was performed; again Gandolf became unengaged. But after only three months, he again “alerted” and MM again did have progressive and rapidly fatal disease. In the few weeks before her death, when she became bedbound, MM asked that Gandolf accompany me on house calls to see her – he would sniff heavily, looking at her face, and lean on the bed or put his paws on the bed & his head on her chest. He was present when she died – took a few sniffs, moved away from the bed, and left the room.
So where did this lead us? Following the observation that removal of the bulk of MM’s disease by mastectomy resulted in a non-response from the dog, we began a two-part study. The first was to observe his behavior with new patients. Since the Breast Care Center was primarily a screening facility, we saw a relatively small number of patients with either metastatic disease or locally far-advanced disease at the time of their first visit to our facility. However, over the next five years, Gandolf correctly identified, in the same manner, all eight patients who presented with metastatic disease. In addition, he identified three of four patients who presented with locally far-advanced disease; the patient who he did not identify was a woman who had a breast lymphoma, not a true breast cancer.
The second part of the study was conducted with the assistance of several surgeons, radiologists, and our pathologist. Gandolf was presented with fresh needle and surgical biopsy samples of cancerous and non-cancerous breast tissue from over 100 patients and never responded to any of them. Since these were fresh specimens as yet unevaluated by the pathologist, neither he nor I nor three other staff knew the patient’s diagnosis and so were blinded observers without bias. We also gave Gandolf samples of urine and whole blood to sniff on all of the 11 patients he alerted on – again, no response, though in this circumstance, the observers all had knowledge of the patient’s diagnosis.
Early in my work with Gandolf, there was a brief paper presented at the American Society of Clinical Oncology annual meeting (1996) that described a group of ketones that could be identified in the breath of women with metastatic breast cancer. These ketones were specific for breast and not other types of metastatic cancer and represented break-down products from the tumor which were metabolized and excreted through the lungs.
More recently, the Georgia Tech Research Institute has been working on defining and measuring compounds in exhaled breath of breast cancer patients that could be used as a simple screening method and follow-up for known cancer patients to evaluate for recurrent disease. In a small study of mid-to-late stage breast cancer (II-IV), their testing method had an accuracy rate of 78% – I surmise that it was this low because of the inclusion of more mid (stage II)- than late –stage (III-IV) cancer patients.
Please note that Gandolf’s behavior was not a trained response, but a naturally-offered and observed response, valid and highly predictive. It was also a response that was not reinforced by reward. In 2002, he identified a fourth locally-far-advanced cancer, this time in a social setting: in my sister-in-law when she came to our home for Christmas dinner. She was treated with surgery, chemotherapy, and radiation, after which the dog no longer alerted on her – nor did he for the remaining two years of his life. At this writing, my sister-in-law is disease-free.
You might wonder about the other two dogs that were in the office. One showed no interest at all in any of the patients. The other, Gandolf’s daughter Tova, alerted in the same manner that he did, but to a different patient population: the developmentally/mentally/psychiatrically handicapped women from a near-by locked-down residential care facility with which the Breast Care Center had a contract to do all of their screening mammograms and breast care. No evaluation was done of just what she was “alerting” on – medications? poor hygiene? the odor of some institutionally-used cleaning product? Who knows?
From this early experience with the remarkable nose that knows, I became interested in the newly-developed competitive sport of canine Nosework, the subject of another article to come.