SEARCH

SEARCH BY CITATION

Contents

  1. Top of page
  2. Contents
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Author Contributions
  9. References

Benign prostatic hyperplasia (BPH) is a spontaneous and age-related condition in humans and intact male dogs. A symptom index for BPH in men was created by the American Urological Association. In this study, it has been developed and statistically validated as a model to assign an objective score to canine BPH severity based on clinical signs observed and/or subjectively reported to the veterinarian by dog owners. The medical records of the Animal Reproduction Unit of University of Bologna (Italy) were used to select dogs with a clinical diagnosis of BPH. A data set was built up, and the animals were included in the statistical analysis as dependent variables. A score of 1–3 was assigned to the disease severity of each case based on signs annotated, graded using a scale ranging from 1 to 4. Signs of BHP were entered as predictors while disease severity as dependent variable to generate the predictive model. The model was finally used to re-classify each case of the data set, and the percentage of corrected predictions calculated. Overall, 373 subjects were entered in the model. Between them, 243, 107 and 23 animals have been represented based on medical records with a BPH severity score of 1, 2 and 3, respectively. The model correctly predicted the response variable in 97.3% of the cases. In this study, a BPH symptom index was created for the first time in dogs, which may be useful to standardize BPH severity with an objective score and to evaluate the necessity, the kind and the effectiveness of treatment.


Introduction

  1. Top of page
  2. Contents
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Author Contributions
  9. References

Benign prostatic hyperplasia (BPH) is a spontaneous and age-related condition in humans and intact male dogs (Krawiec and Heflin 1992). With time, almost all intact male dogs will develop BPH, with >95% affected by 9 years of age (Gobello and Corrada 2002). Even if symptoms related to BPH are relatively common, particularly in dogs >6 years, there are typically no clinical signs displayed by dogs with BPH until the condition has progressed to the point that the enlarging prostate causes them (Krawiec and Heflin 1992; Gobello and Corrada 2002). However, in men, there is an absence of any relationship between symptoms and objective measures of prostate with benign hyperplasia. Thus, the prostate volume alone is not useful for the estimation of disease severity (Vesely et al. 2003). Symptomatology in BPH-affected dogs could include sanguinous preputial/urethral discharge, haematuria or hemospermia, tenesmus and dysuria. A presumptive diagnosis is based on detection of the above-mentioned symptoms and on detecting uniform prostatic enlargement by palpation, radiography and ultrasonography (Smith 2008). Depending on the findings of physical examination and diagnostic imaging, additional procedures such as cytology and microbiology of prostatic fluid, and prostatic aspiration or biopsy can be performed (Gobello and Corrada 2002; Smith 2008). It is recommended to treat dogs with BPH only if clinical symptoms are present (Johnston et al. 2001).

A symptom index for BPH in men was created by the American Urological Association (AUA). AUA has developed a questionnaire including seven questions about BPH-related clinical signs in humans. The aim was to objectify BPH symptoms severity with a score, to help men determine how serious their symptoms are and to check the necessity or the effectiveness of treatment. This questionnaire is worldwide known as the International Prostate Symptom Score (IPSS). The AUA symptom index is clinically sensible, reliable, valid and responsive. It is very useful in practice and for inclusion in research protocols (Barry et al. 1992). In this study, we developed and statistically validated a model symptom index to assign an objective score to canine BPH severity based on clinical signs observed and subjectively reported to the veterinarian by dog owners.

Materials and Methods

  1. Top of page
  2. Contents
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Author Contributions
  9. References

The medical records of the Animal Reproduction Unit of the Department of Veterinary Medical Sciences-University of Bologna (Italy) were used to retrieve all cases of dogs diagnosed with prostatic disease, including hyperplastic, neoplastic and inflammatory diseases between January 2001 and January 2007. Only dogs with a clinical diagnosis of BPH were selected to build up a data set and included in the statistical analysis as the dependent variable. The data set included 477 cases. Based upon the medical records, a score of 1 (low), 2 (medium) or 3 (high) was given to the disease severity of each case. Signs annotated in the medical records of each case were anorexia, weight loss, dyschezia/tenesmus (severity), dysuria (severity), urinary incontinence, urethral leakages (amount), urethral leakages (frequency), urethral leakages (duration), haematuria (amount), haematuria (frequency) and haematuria (duration) and were graded using a scale ranging from 1 to 4, where 1 = absent, 2 = light and/or short duration, 3 = moderate severity and/or middle duration and 4 = severe and/or long lasting. In particular, scores ranging from 1 to 4 were assigned as described in Table 1.

Table 1. Criteria applied for BPH signs grading
GradeBPH signs
AnorexiaWeight lossDyschezia/Tenesmus (severity)Dysuria (severity)Urinary incontinenceUrethral leakageHaematuria
AmountFrequencyaDurationAmountFrequencyaDuration
  1. a

    number of episodes from event onset; event = urethral leakage or hematuria; each episode is composed by several events; if events are <1 every 15 days, they are considered belonging to different episodes.

1AbsentAbsentAbsentAbsentAbsentAbsentAbsentAbsentAbsentAbsentAbsent
2Since 1 dayLightRegular, but difficult and painful defecationUrination with weaker or interrupted flowPresentFew drops1 episode per yearSince 1–7 daysSometimes pinkish urine1 episode per yearSince 1–7 days
3Since 1 day to 1 weekModerateIrregular, difficult and painful defecationUrination with very weak or interrupted flow/Many drops2–3 episodes per yearSince 7–15 daysPinkish urine2–3 episodes per yearSince 7–15 days
4Since >1 weekSevereDefecation impossible or absent during some daysUrinary retention with urinary drip or paradoxal urinary incontinence/Stains on the floor>3 episodes per yearSince >15 daysRed urine>3 episodes per yearSince >15 days

Statistical analysis

To build up a BPH data set, signs and disease severity were consistently assessed and annotated by the same clinician or under his supervision. Signs of BPH were entered as predictors while disease severity as dependent variable to generate the predictive model. Breed, weight and age were also recorded in the database and evaluated as predictors as well. The ordinal regression procedure (PLUM) of the spss 12.0 (IBM, SPSS Italia, Bologna, Italy) for Windows software package was used to analyse the data set with the ordinal response. In particular, all predictors of the data set were initially entered in the model; some were removed based on the empirical evaluation of the contribution to increase the Nagelkerke pseudo r2 coefficient (Nagelkerke 1991). To select the link function included in the model, besides the Nagelkerke pseudo r2 value, the evaluation of the Chi-square value of the difference of the log-likelihood between the intercept only model and the final model and the assessment of the percentage of correct prediction were used.

The model was finally used to re-classify each case of the data set, and the percentage of corrected predictions calculated.

Results

  1. Top of page
  2. Contents
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Author Contributions
  9. References

Overall, 373 subjects were entered in the model. Between them, 243, 107 and 23 animals have been represented based on medical records with a BPH severity score of 1, 2 and 3, respectively (Table 2). The model was built using the Cauchit link function and the following predictors: age, dyschezia/tenesmus (severity), dysuria (severity), urinary incontinence, urethral leakage (amount), urethral leakage (frequency), urethral leakage (duration), haematuria (amount), haematuria (frequency), haematuria (duration). No variables were entered as covariates. The Nagelkerke pseudo r2 coefficient was 0.946. The model correctly predicted the response variable in 97.3% of the cases in the data set. Erroneous predictions were represented by 10 cases (2.7%). Errors were represented by type 2 predicted as type 1 (five cases), type 1 predicted as type 2 (three cases) and type 3 predicted as type 2 (two cases).

Table 2. BPH severity score assigned based on medical records and relative signs entered as predictors in the model
 CategoryNumber in category (animals)Percentage in category (animals)
Diagnosis (BPH severity)124365.1%
210728.7%
3236.2%
Age (years)3143.8%
4123.2%
5328.6%
64712.6%
75013.4%
85013.4%
9308.0%
104411.8%
114010.7%
12195.1%
13184.8%
1482.1%
1571.9%
1620.5%
Dyschezia/tenesmus, severity128275.6%
2379.9%
34512.1%
492.4%
Dysuria, severity128877.2%
2287.5%
34612.3%
4112.9%
Urinary incontinence137199.5%
220.5%
Urethral leakage, amount117647.2%
210327.6%
36818.2%
4267.0%
Urethral leakage frequency117747.5%
27921.2%
38923.9%
4287.5%
Urethral leakage duration117847.7%
212934.6%
35715.3%
492.4%
Haematuria, amount129779.6%
23910.5%
3287.5%
492.4%
Haematuria, frequency129779.6%
23810.2%
3267.0%
4123.2%
Haematuria, duration129679.4%
24211.3%
3338.8%
420.5%

Discussion

  1. Top of page
  2. Contents
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Author Contributions
  9. References

In practice, diagnosis of BPH is usually based on anamnesis, symptoms and prostatic ultrasonographic features. Symptomatic BPH severity has always been established in a subjective manner, making a discussion or a comparison of clinical cases between clinicians often very difficult. A predictive model could permit assignment of a severity score to this pathology based on signs objectively detected and assessed. The model applied in this study showed a high accuracy (97.3% – 363 cases), and only 2.7% of cases (10/373 subjects) were represented by a type not correctly predicted by the model. However, besides the low percentage, the prediction error was never between type 1 and 3, but between close types (1 and 2 or 2 and 3). In conclusion, in this study, a BPH symptom index was created for the first time in dogs. As already proposed by AUA for symptomatic BPH in humans (Barry et al. 1992), results obtained in this study allow to include the predictors used in a questionnaire to submit to dogs owners or to be used by practitioners. In addition, the purpose of the authors is to use the formula which is at the basis of the predictive model developed in this study to create a WEB application utilizable by all practitioners. Both questionnaire and/or WEB application will be useful to standardize BPH severity with an objective score and to evaluate the necessity, the kind (medical or surgical) and the effectiveness of treatment. In the author's opinion, it is important to consider that above-mentioned questionnaire and/or WEB application will represent a simple tool to share with colleagues data about our patients affected by BPH, using a simple, objective and standardized method.

Conflicts of interest

  1. Top of page
  2. Contents
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Author Contributions
  9. References

None of the authors have any conflicts of interest to declare. None of the authors have received grants, speakers fees, etc., from any commercial body within the past 2 years.

Author Contributions

  1. Top of page
  2. Contents
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Author Contributions
  9. References

Each of the authors contributed to the design and conduct of the research reported herein as original, and each contributed to the writing of the manuscript.

References

  1. Top of page
  2. Contents
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conflicts of interest
  8. Author Contributions
  9. References
  • Barry MJ, Fowler FJ Jr, O'Leary MP, Bruskewitz RC, Holtgrewe HL, Mebust WK, Cockett AT, 1992: The American Urological Association symptom index for benign prostatic hyperplasia. The Measurement Committee of the American Urological Association. J Urol, 148, 15491564.
  • Gobello C, Corrada Y, 2002: Noninfectious prostatic diseases in dogs. Compend Contin Educ Vet 24, 99107.
  • Johnston SD, Root Kustritz MV, Olson PNS, 2001: Disorders of the canine testes and epididymes. In: WB Saunders Company (ed.), Canine and Feline Theriogenology, Philadelphia, PA, pp. 312332.
  • Krawiec DR, Heflin D, 1992: Study of prostatic disease in dogs: 177 cases (1981–1986). J Am Vet Med Assoc 200, 11191122.
  • Nagelkerke NJD, 1991: A note on the general definition of the coefficient of determination. Biometrika, 788, 691692.
  • Smith J, 2008: Canine prostatic disease: a review of anatomy, pathology, diagnosis, and treatment. Theriogenology 70, 375383.
  • Vesely S, Knutson T, Damber JE, Dicuio M, Dahlstrand C, 2003: Relationship between age, prostate volume, prostate-specific antigen, symptom score and uroflowmetry in men with lower urinary tract symptoms. Scand J Urol Nephrol 37, 322328.