Olfactory detection of human bladder cancer by dogs: proof of principle study
http://www.100md.com
《英国医生杂志》
1 Department of Dermatology, Amersham Hospital, Amersham HP7 0JD, 2 Hearing Dogs for Deaf People, Saunderton, Princes Risborough HP27 9NS, 3 Centre for Statistics in Medicine, Institute of Health Sciences, Oxford OX3 7LF
Correspondence to: C M Willis carolyn.willis@sbucks.nhs.uk
Abstract
The hypothesis that dogs may be able to detect malignant tumours on the basis of odour was first put forward by Williams and Pembroke in a letter to the Lancet in 1989.1 Their thinking arose from a consultation with a woman who claimed to have sought medical help as a direct result of her dog's inordinate interest in a skin lesion, which subsequently proved to be a malignant melanoma. Since then similar anecdotal claims of detection of skin cancer, and of malignancies of internal organs such as breast and lung, have appeared in the press and in a further letter to the Lancet.2-4
Although these anecdotal events remain unsupported by experimental evidence, the concept that dogs can "smell" cancer is not unreasonable. Tumours produce volatile organic compounds, which are released into the atmosphere through, for example, breath and sweat.5-9 Some of these volatile organic compounds are likely to have distinctive odours; even when present in minute quantities, they could be detectable by dogs, with their exceptional olfactory acuity.10-13
Interest in the exploitation of volatile organic compounds for diagnostic purposes is growing,5 6 suggesting that dogs have the potential to make a contribution in the field of oncology, providing that a scientific basis to the anecdotal reports can be established. With this in mind, we designed a study to determine whether dogs can detect cancer by olfactory means. We chose human bladder cancer as the experimental model,14 on the basis that tumour related volatile organic compounds are released into urine,8 which can be readily collected and presented to dogs for training and testing purposes. Our aim was to train dogs to recognise an odour, or combination of odours (an "odour signature"), characteristic of bladder cancer but distinct from those associated with the secondary effects of the tumour, such as bleeding, inflammation, infection, and necrosis. These factors are present in a multitude of non-malignant conditions of the urinary tract and elsewhere in the body and must be ignored by the dogs if discrimination is to be attained. We assessed the dogs' abilities to detect bladder cancer, once trained, by comparison of their success rate with that expected by chance alone, in choosing one cancer urine placed randomly among six controls in blinded experiments.
We should emphasise that our objective at this stage was to conduct a simple, yet stringent, "proof of principle" study to answer the question, "Can dogs be trained to detect bladder cancer more successfully than would be expected by chance alone?" This was not an attempt to assess or predict the clinical usefulness of this hypothesised capability of dogs.
Methods
Table 2 gives the results for the formal test runs. Taken as a single group, the dogs correctly selected the positive bladder cancer urine on 22 of 54 occasions. This gave a mean success rate of 41% (95% confidence intervals 23% to 58% under assumptions of normality and 26% to 52% using bootstrap methods), compared with 14% expected by chance.
Table 2 Urine samples selected by the six dogs during evaluation
The association between presence of cancer and selection by the dogs was slightly stronger in the multivariate model, which also included presence of blood and ketones, than in the univariate model. This indicated that the association was not due to confounding with factors measured on urinalysis.
The four dogs trained on wet urine specimens (50% correct) seemed to perform better than the two dogs trained on dried samples (22% correct; P=0.03 by t test, P=0.06 by rank sum test). However, the small numbers involved limit confidence in the certainty of this observation.
Discussion
Williams H, Pembroke A. Sniffer dogs in the melanoma clinic? Lancet 1989;1: 734.
Fraser L. Scientists put sniffer dogs on the scent of men with cancer. Sunday Telegraph 2002 June 2.
Dobson R. Dogs can sniff out first signs of men's cancer. Sunday Times 2003 Apr 27: 5.
Church J, Williams H. Another sniffer dog for the clinic? Lancet 2001;358: 930.
Phillips M, Gleeson K, Hughes JM, Greenberg J, Cataneo RN, Baker L, et al. Volatile organic compounds in breath as markers of lung cancer: a cross-sectional study. Lancet 1999;353: 2897-8.
Di Natale C, Macagnano A, Martinelli E, Paolesse R, D'Arcangelo G, Roscioni C, et al. Lung cancer identification by the analysis of breath by means of an array of non-selective gas sensors. Biosens Bioelectron 2003;18: 1209-18.
Phillips M, Cataneo RN, Ditkoff BA, Fisher P, Greenberg J, Gunawardena R, et al. Volatile markers of breast cancer in the breath. Breast J 2003;9: 184-91.
Spanel P, Smith D, Holland TA, Al Singary W, Elder JB. Analysis of formaldehyde in the headspace of urine from bladder and prostate cancer patients using selected ion flow tube mass spectrometry. Rapid Commun Mass Spectrom 1999;13: 1354-9.
Yamada K, Walsh N, Hara H, Jimbow K, Chen H, Ito S. Measurement of eumelanin precursors metabolites in the urine as a new marker for melanoma metastases. Arch Dermatol 1992;128: 491-4.
Schoon GAA, De Bruin JC. The ability of dogs to recognise and cross-match human odours. Forensic Sci Int 1994;69: 111-8.
Schoon GAA. A first assessment of the reliability of an improved scent identification line-up. J Forensic Sci 1998;43: 70-5.
Komar D. The use of cadaver dogs in locating scattered, scavenged human remains: preliminary field test results. J Forensic Sci 1999;44: 405-8.
Lorenzo N, Wan T-L, Harper RJ, Hsu Y-L, Chow M, Rose S, et al. Laboratory and field experiments used to identify Canis lupus var. familiaris active odor signature chemicals from drugs, explosives, and humans. Anal Bioanal Chem 2003;376: 1212-24.
Pashos CL, Botteman MF, Laskin BL, Redaelle A. Bladder cancer: epidemiology, diagnosis and management. Cancer Pract 2002;10: 311-22.
Mills DS. Learning, training and behavioural modification techniques. In: Horwitz DF, Mills DS, Heath S, eds. BSAVA manual of canine and feline behavioural medicine. Gloucester: British Small Animal Veterinary Association, 2002.
Carpenter JR, Bithell JF. Bootstrap confidence intervals: when? which? what?—A practical guide for medical statisticians. Stat Med 2000;19: 1141-64.(Carolyn M Willis, senior )
Correspondence to: C M Willis carolyn.willis@sbucks.nhs.uk
Abstract
The hypothesis that dogs may be able to detect malignant tumours on the basis of odour was first put forward by Williams and Pembroke in a letter to the Lancet in 1989.1 Their thinking arose from a consultation with a woman who claimed to have sought medical help as a direct result of her dog's inordinate interest in a skin lesion, which subsequently proved to be a malignant melanoma. Since then similar anecdotal claims of detection of skin cancer, and of malignancies of internal organs such as breast and lung, have appeared in the press and in a further letter to the Lancet.2-4
Although these anecdotal events remain unsupported by experimental evidence, the concept that dogs can "smell" cancer is not unreasonable. Tumours produce volatile organic compounds, which are released into the atmosphere through, for example, breath and sweat.5-9 Some of these volatile organic compounds are likely to have distinctive odours; even when present in minute quantities, they could be detectable by dogs, with their exceptional olfactory acuity.10-13
Interest in the exploitation of volatile organic compounds for diagnostic purposes is growing,5 6 suggesting that dogs have the potential to make a contribution in the field of oncology, providing that a scientific basis to the anecdotal reports can be established. With this in mind, we designed a study to determine whether dogs can detect cancer by olfactory means. We chose human bladder cancer as the experimental model,14 on the basis that tumour related volatile organic compounds are released into urine,8 which can be readily collected and presented to dogs for training and testing purposes. Our aim was to train dogs to recognise an odour, or combination of odours (an "odour signature"), characteristic of bladder cancer but distinct from those associated with the secondary effects of the tumour, such as bleeding, inflammation, infection, and necrosis. These factors are present in a multitude of non-malignant conditions of the urinary tract and elsewhere in the body and must be ignored by the dogs if discrimination is to be attained. We assessed the dogs' abilities to detect bladder cancer, once trained, by comparison of their success rate with that expected by chance alone, in choosing one cancer urine placed randomly among six controls in blinded experiments.
We should emphasise that our objective at this stage was to conduct a simple, yet stringent, "proof of principle" study to answer the question, "Can dogs be trained to detect bladder cancer more successfully than would be expected by chance alone?" This was not an attempt to assess or predict the clinical usefulness of this hypothesised capability of dogs.
Methods
Table 2 gives the results for the formal test runs. Taken as a single group, the dogs correctly selected the positive bladder cancer urine on 22 of 54 occasions. This gave a mean success rate of 41% (95% confidence intervals 23% to 58% under assumptions of normality and 26% to 52% using bootstrap methods), compared with 14% expected by chance.
Table 2 Urine samples selected by the six dogs during evaluation
The association between presence of cancer and selection by the dogs was slightly stronger in the multivariate model, which also included presence of blood and ketones, than in the univariate model. This indicated that the association was not due to confounding with factors measured on urinalysis.
The four dogs trained on wet urine specimens (50% correct) seemed to perform better than the two dogs trained on dried samples (22% correct; P=0.03 by t test, P=0.06 by rank sum test). However, the small numbers involved limit confidence in the certainty of this observation.
Discussion
Williams H, Pembroke A. Sniffer dogs in the melanoma clinic? Lancet 1989;1: 734.
Fraser L. Scientists put sniffer dogs on the scent of men with cancer. Sunday Telegraph 2002 June 2.
Dobson R. Dogs can sniff out first signs of men's cancer. Sunday Times 2003 Apr 27: 5.
Church J, Williams H. Another sniffer dog for the clinic? Lancet 2001;358: 930.
Phillips M, Gleeson K, Hughes JM, Greenberg J, Cataneo RN, Baker L, et al. Volatile organic compounds in breath as markers of lung cancer: a cross-sectional study. Lancet 1999;353: 2897-8.
Di Natale C, Macagnano A, Martinelli E, Paolesse R, D'Arcangelo G, Roscioni C, et al. Lung cancer identification by the analysis of breath by means of an array of non-selective gas sensors. Biosens Bioelectron 2003;18: 1209-18.
Phillips M, Cataneo RN, Ditkoff BA, Fisher P, Greenberg J, Gunawardena R, et al. Volatile markers of breast cancer in the breath. Breast J 2003;9: 184-91.
Spanel P, Smith D, Holland TA, Al Singary W, Elder JB. Analysis of formaldehyde in the headspace of urine from bladder and prostate cancer patients using selected ion flow tube mass spectrometry. Rapid Commun Mass Spectrom 1999;13: 1354-9.
Yamada K, Walsh N, Hara H, Jimbow K, Chen H, Ito S. Measurement of eumelanin precursors metabolites in the urine as a new marker for melanoma metastases. Arch Dermatol 1992;128: 491-4.
Schoon GAA, De Bruin JC. The ability of dogs to recognise and cross-match human odours. Forensic Sci Int 1994;69: 111-8.
Schoon GAA. A first assessment of the reliability of an improved scent identification line-up. J Forensic Sci 1998;43: 70-5.
Komar D. The use of cadaver dogs in locating scattered, scavenged human remains: preliminary field test results. J Forensic Sci 1999;44: 405-8.
Lorenzo N, Wan T-L, Harper RJ, Hsu Y-L, Chow M, Rose S, et al. Laboratory and field experiments used to identify Canis lupus var. familiaris active odor signature chemicals from drugs, explosives, and humans. Anal Bioanal Chem 2003;376: 1212-24.
Pashos CL, Botteman MF, Laskin BL, Redaelle A. Bladder cancer: epidemiology, diagnosis and management. Cancer Pract 2002;10: 311-22.
Mills DS. Learning, training and behavioural modification techniques. In: Horwitz DF, Mills DS, Heath S, eds. BSAVA manual of canine and feline behavioural medicine. Gloucester: British Small Animal Veterinary Association, 2002.
Carpenter JR, Bithell JF. Bootstrap confidence intervals: when? which? what?—A practical guide for medical statisticians. Stat Med 2000;19: 1141-64.(Carolyn M Willis, senior )