E!ect of Ovariohysterectomy at the Time of Tumor Removal inDogs with Benign Mammary Tumors and Hyperplastic Lesions:

A Randomized Controlled Clinical Trial

V.M. Kristiansen, A. Nødtvedt, A.M. Breen, M. Langeland, J. Teige, M. Goldschmidt,T.J. Jonasdottir, T. Grotmol, and K. Sørenmo

Background: Nonmalignant mammary tumors (NMT) are common in intact female dogs. Little is known about theclinical significance of these tumors, and the e!ect of ovariohysterectomy (OHE) on their development.

Hypothesis: Ovarian hormone ablation through OHE decreases the risk of new tumors and thereby improves long-termprognosis for dogs with NMT.

Animals: Eighty-four sexually intact bitches with NMT.Methods: Dogs were allocated to undergo OHE (n = 42) or not (n = 42) at the time of NMT removal in a randomized

clinical trial. Tumor diagnosis was confirmed histologically in all subjects. Information about new tumor development wascollected via follow-up phone calls and recheck examinations. Separate survival analyses were performed with theendpoints new tumor development and death. Cause of death was classified as related or unrelated to mammary tumor. Inaddition to OHE status, the influence of age, body weight, breed, tumor size, tumor number, tumor duration, type ofsurgery, and tumor histology was investigated.

Results: New mammary tumor(s) developed in 27 of 42 (64%) intact dogs and 15 of 42 (36%) ovariohysterectomizeddogs (hazard ratio 0.47, P = .022). Nine of the 42 dogs (21%) which developed new tumors were euthanized because ofmammary tumor. Survival was not significantly di!erent between the 2 treatment groups. In the intact group, nine dogssubsequently developed ovarian–uterine diseases.

Conclusion: Ovariohysterectomy performed at the time of mammary tumor excision reduced the risk of new tumors byabout 50% among dogs with NMT. Survival was not significantly a!ected. Adjuvant OHE should be considered in adultdogs with mammary tumors.

Key words: Canine; Nonmalignant mammary tumors; Cancer; Spaying.

Mammary tumor is the most common tumor insexually intact female dogs.1–5 Close to 60% of

these tumors are benign.6 Only 1% of mammarytumors smaller than 1 cm are malignant and 50% oftumors larger than 3 cm are benign.7 The clinicalimplications of benign mammary tumors (BMT) arestill largely unknown and are commonly considered adisease of limited clinical importance. This is in con-trast to malignant mammary tumors with potential formetastasis and fatal outcome.8–10 However, BMT is arecognized risk factor for developing additionallymammary tumors later.11,12 Mammary tumors in dogsdevelop as a histologic continuum from benign to

malignant, with the malignant tumors representing theend stage of this continuum.7 It therefore seems plausi-ble that a subset of dogs with benign tumors could beat risk of developing malignant, and thus potentiallyfatal, tumors. In comparison, women with a benignbreast tumor or atypical mammary hyperplasia haveincreased risk of subsequently developing breast can-cer, and this risk increases with increasing atypia.13

Factors leading to the subsequent development ofmalignant neoplasms have not been studied in dogswith BMT.

One well-recognized risk factor for mammary tumordevelopment is exposure to ovarian hormones duringthe 1st 2 years of life, as demonstrated by the signifi-cantly reduced risk of mammary tumors in dogs thatundergo ovariohysterectomy (OHE) during this timeperiod.14 OHE performed after the age of four hasbeen found to have limited to no protective e!ectagainst mammary tumors.14–16

New tumors in other mammary glands are commonin dogs with a previous diagnosis of mammarytumors.7,17 One of 4 dogs develops new tumors aftersurgical removal of BMTs.7,18 And importantly, these

From the Department of Companion Animal Clinical Sciences(Kristiansen, Nødtvedt, Grotmol, Jonasdottir, Langeland, Breen)and the Department of Basic Sciences and Aquatic Medicine(Teige), Norwegian School of Veterinary Science, Oslo, Norway;the Department of Pathobiology, University of PennsylvaniaSchool of Veterinary Medicine (Goldschmidt) and the Departmentof Clinical studies, Veterinary hospital of the University ofPennsylvania (Sørenmo), Philadelphia, PA. Selected preliminaryresults were presented as a poster entitled “Ovariohysterectomy atthe time of tumor removal decreases the risk of new tumors in dogswith benign mammary tumors” at the 2nd World VeterinaryCancer Congress, Paris, March 2012.

Corresponding author: V.M. Kristiansen, DVM, NorwegianSchool of Veterinary Science, Department of Companion AnimalClinical Sciences, P.O. Box 8146 Dep., Oslo NO-0033, Norway;e-mail: veronica.kristiansen@nvh.no

Submitted August 15, 2012; Revised March 19, 2013;Accepted April 11, 2013.

Copyright © 2013 by the American College of Veterinary InternalMedicine

10.1111/jvim.12110

Abbreviations:

BMT benign mammary tumor

HR hazard ratio

NMT nonmalignant mammary tumor

NT new tumor

OHE ovariohysterectomy

TVE time-varying e!ect

J Vet Intern Med 2013;27:935–942

later tumors can be malignant even if the initial tumorwas benign,17,18 and thus might pose a risk of metasta-sis and premature death for the a!ected dog. Severalstudies have shown that a significant portion of thebenign epithelial tumors and hyperplastic mammarylesions express estrogen and progesterone receptors,inferring continued hormonal dependence.19,20 Wetherefore hypothesized that ovarian hormone ablationthrough OHE would decrease the risk of new tumorsand thereby improve long-term prognosis for dogswith nonmalignant mammary tumor(s) (NMT). Arecent meta-analysis on the protective e!ect of OHEon the risk for mammary tumor development usingCochrane principles21 concludes that the level of scien-tific evidence is too weak to firmly conclude on thisissue. Therefore, a well-controlled randomized study toanswer the questions regarding the e!ect of OHE indogs with mammary tumors is clearly needed.

The purpose of this study was to determine thee!ect of OHE on new tumor development and survivalin dogs with NMT. A secondary objective was todetermine the association between the initial andsubsequent tumors regarding histology and tumorlocalization.

Materials and Methods

Study Population and Randomization

The study was designed as a randomized controlled clinicaltrial. Intact female dogs with mammary tumors and no previoushistory of mammary malignancy were eligible for inclusion. Acomplete reproductive and health history was collected as part ofthe screening and initial enrollment. Randomization to eithertumor removal or tumor removal and concomitant OHE wasperformed before surgery. In addition, the dogs were stratifiedbased on tumor size (<3 cm, ! 3 cm) and age (<9 years,! 9 years) to ensure equal distribution of these 2 prognostic fac-tors between the treatment groups. A block randomizationscheme was used within each stratum. The allocation sequencewas computer generated, and the treatment allocation was notknown to owners or the investigator until the enrollment hadbeen completed.

All owners of participating dogs signed a written consent formwhere they were given relevant information about the project andagreed to the randomization procedure. There was 1 fixed pricefor all participants irrespective of type of mammary tumorsurgery or if adjunctive OHE was performed or not. The studywas approved by the institutional animal care and use board atthe Norwegian School of Veterinary Science.

Complete staging, including blood work (complete bloodcount and serum biochemistry profile), cytological investigationof enlarged draining lymph nodes, and 3-view thoracic radio-graphs, was performed before surgery. Dogs with distant metas-tasis or other serious diseases were excluded. All tumors wererecorded (number, localization) and size was determined as thelargest diameter measured by caliper. Surgery was performedaccording to standard surgical practices22,23 and involved excisionof at least 1 cm of gross normal appearing tissue surrounding thetumor dependent on size and number of the mammary tumorstreated. No prophylactic mastectomy of normal glands was per-formed. Age, weight, and breed of the dogs were recorded at thetime of surgery. OHE was performed concomitantly with tumorsurgery.

Based on results from the histology review, 2 separate groupswere created: a carcinoma group consisting of dogs with at leastone malignant epithelial tumor, and a nonmalignant group whichincluded dogs with hyperplastic nodular lesions, benign tumorsregardless of tissue of origin, and carcinoma in situ. Here, wereport on this latter group of dogs. Both tumors and hyperplasticlesions in the nonmalignant group are termed nonmalignanttumors (NMT) because they are clinically indistinguishable fromeach other when appearing as discrete well defined “tumorlike”lesions. Cases were enrolled continuously until an adequate num-ber of dogs with malignant epithelial tumors were enrolled in themalignant group. The results from this group will be reportedseparately.

After surgery, the owners of the dogs were instructed to moni-tor for any signs of new mammary tumors and notify the princi-pal investigator (PI) if any signs of recurrence or new tumorswere noted. In addition, they were contacted by the PI (VK)every 6 months through phone to ensure this information regard-ing recurrence or new tumor development (localization and time).Other health issues were recorded as well. Dogs with reported/suspected new tumors were requested to return for clinical exami-nation and confirmation. Surgical excision was recommended forall dogs with new tumors; however, no financial study supportwas provided for any of these subsequent treatments. Completenecropsies were requested regardless of cause on dogs that diedor were euthanized.

Tumor Histopathology

All tumors were fixed in 10% neutral bu!ered formalin andsubmitted for histopathological examination. The slides werestained with hematoxylin and eosin. To minimize bias because ofmisdiagnosis, the slides were evaluated by 2 independent patholo-gists (MHG and JT) who performed the evaluations blinded toeach other’s diagnoses. Dogs diagnosed with only NMT by atleast one of the pathologists, and subsequent agreements, wereincluded. The tumors were further classified according to type oftissue present (epithelial, myoepithelial and/or connective tissue).A complete histopathological description was provided for eachtumor and included information regarding the degree of tumordi!erentiation, presence of cellular atypia, carcinoma in situ, andtumor margins.22

Outcome Variables

The 2 main outcome variables in the analysis were time to 1stnew mammary tumor and time to death/euthanasia. Cause ofdeath/euthanasia was also recorded.

New tumors were defined as any tumor arising in the mammarytissue after excision of the original tumor(s). Tumors occurring inthe same gland or close to the original tumor were also classifiedas new tumors, rather than recurrences. All subsequent tumorsdetected by owner, during rechecks and/or at necropsies wererecorded. If the dogs were not available for recheck examination,information about new tumors was based on owner’s description.Localization of new tumors was recorded referring to closest asso-ciated mammary gland. Their relation to the original removedtumors were also recorded in terms of adjacency and whether thenew tumors were located ipsi- or contralateral to these. The newtumor was, if available for histology, diagnosed by the samepathologists and categorized as benign or malignant.

Cause of death (including euthanasia) was classified as mam-mary tumor specific or not-mammary tumor specific. To be classi-fied as mammary tumor specific, the owners’ decision to euthanizehad to be directly related to the tumor, or its metastases, and hadto be confirmed by necropsy or by diagnostic imaging. Cause of

936 Kristiansen et al

death in dogs with mammary tumors in which these tumors didnot cause clinical problems was classified as not-mammary tumorspecific.

Explanatory Variables

In addition to the intervention variable (spayed or not at the timeof tumor removal: OHE/non-OHE), the following variables wereexamined for potential influence on new tumor development andoverall survival: age at the time of surgery as continuous and cate-gorized (<9 years, ! 9 years) variable, body weight (<22 kg,! 22 kg), breed (pure, mixed), tumor duration (<6 months,! 6 months), number of tumors (multiple, single), tumor size(<3 cm, ! 3 cm), extent of surgery (lumpectomy/simple mastec-tomy, regional mastectomy/radical mastectomy), and tumor histol-ogy. Tumor histology was categorized into (1) hyperplasia andcysts; (2) benign tumors (adenomas, complex adenomas, benignmixed) without atypia; and (3) benign tumors with atypia, necrosisor with carcinoma in situ. In dogs with multiple tumors, the tumorwith the most atypical histological changes determined the category.

Statistical Analysis

Descriptive Statistics. The distribution of potential risk factorsby each OHE group was calculated and compared for the vari-ables age, body weight, breed, tumor size, tumor histology, num-ber of tumors, type of surgery, and tumor duration to ensuresimilar group characteristics at baseline, using chi-squared test.When categorized, the cut-o! values were determined by themedian derived from all dogs. For age and tumor size, however,the predetermined cut-o! values for the stratification procedurewere used. Dogs lost to follow-up or still alive without any of theevents, were censored at the date of last known status. Dogs thatdeveloped ovarian or uterine disease requiring OHE were cen-sored at the date of such surgery. The Kaplan–Meier methodwas used to compute survival curves and estimate remission–andsurvival time of dogs by OHE group. Di!erences in survivalbetween di!erent groups were tested using the log-rank test.Uni- and Multivariable Analyses. To evaluate the e!ect of

OHE on the two events (1) new tumor development; and (2)

death/euthanasia, whereas adjusting for other possible risk fac-tors, separate Cox proportional hazards models were built foreach of the events. Two outcomes were modeled for death/eutha-nasia: death of any cause (overall survival) and death attributableto mammary tumor (tumor-specific survival). Time at risk wasdefined as months from surgery date to the event or censoring.All the clinical variables were initially screened for e!ect ontumor development or survival by applying univariable Cox pro-portional-hazards models with adjustment for age and using acut-o! of P < .10 for o!ering variables to the multivariablemodel. Predictors were retained in the final model if P < .05 or ifassumed to have great biological interest a priori. Finally, an esti-mate of the baseline hazard was derived, conditional upon the setof coe"cients in the model.Validation of the Model. The assumption of proportional

hazards was evaluated based on Schoenfeld residuals for the vari-able OHE in each of the 3 models. If this assumption was vio-lated and graphical assessment indicated a time-varying e!ect(TVE) of this variable, an interaction term between this variableand time (on either a linear or logarithmic scale) was included inthe model. The importance of the assumption of independentcensoring was evaluated by sensitivity analyses based on bothcomplete positive and complete negative correlation between cen-soring and outcome. The amount of explained variation wasevaluated by an overall r2 statistic for proportional hazard mod-els. Plots of the deviance residuals, score residuals, and scaledscore residuals against time at risk were used to identify outlyingobservations with influence on the model, and the model was fitwith and without any outlying observations.24 All analyses wereperformed using the software package Stata.a

Results

Study Population, Randomization, and Follow-Up

Of 330 dogs initially screened for eligibility, 84 dogshad nonmalignant tumors and were included in thestudy (Fig 1). Histopathologically, 108 (51%) oftotally 210 tumors were classified as complex or mixedtumors. Forty-four (21%) tumors were adenomas or

Assessed for eligibility (n=330)

Excluded (n=246)Did not meet inclusion criteria (n=180)Malignant histology on excisional biopsy (n=66) (included in a different project)

Included and randomized (n= 84)

Non-OHE group (n=42) OHE group (n=42)

Allocated totumor removal without OHE

(n= 42)

Allocated totumor removal with OHE

(n=42)

Fig 1. Flow diagram of the enrollment and randomization procedure in a randomized controlled clinical trial investigating the e!ect ofadjuvant ovariohysterectomy (OHE) in dogs with nonmalignant mammary tumors.

Ovariohysterectomy and Canine Nonmalignant mammary Tumors 937

variants thereof, and 4 (2%) were diagnosed as carci-noma in situ. The remaining 54 (26%) tumors werediagnosed as di!erent types of hyperplastic or dysplas-tic lesions. All dogs were evaluated, staged, surgicallytreated, and whenever feasible rechecked at theDepartment of Companion Animal Clinical Sciences atthe Norwegian School of Veterinary Sciences in theperiod from February 2005 to May 2012. The medianfollow-up time for all dogs, including the censoredones was 31.5 months (range 3.5–87.5). For the dogsthat were still alive at the end of the study (n = 24),the median follow-up was 34 months (range 19.0–77.9).

Forty-two dogs underwent OHE concomitantly withtumor surgery and 42 remained sexually intact. Forthe non-OHE group, the median follow-up time was31.2 months (range 3.5–87.5) and for the OHE group,31.8 months (range 4.2–78.2). There was no di!erencebetween the treatment groups in terms of signalment,tumor characteristics, and type of surgery, see Table 1.Postsurgically, all owners were able to provide follow-up information on their dog by the regular phone-interviews. In addition, 1 or more clinical rechecks,necropsy, or both were performed in 36 dogs in thenonOHE group, and in 34 dogs in the OHE groupduring the follow-up period.

Uni- and Multivariable Analysis

E!ect of OHE on Time to New Tumor. Dogs fromthe non-OHE group developed significantly more newtumors (P = .022) than the dogs from the OHE group.Of the 42 dogs that developed new tumors, 27 (64%)and 15 (36%) were from non-OHE and OHE group,respectively.

Twenty-six of the 27 dogs in the non-OHE groupwith new tumors, and all the 15 dogs in the OHE grouphad their tumors confirmed clinically by a veterinarian.Only in 1 dog reported to have new tumor, this infor-mation was based solely on owners’ description. Med-ian time to new tumor development was 20.8 months(range 3.8–80) and 19.6 months (range 2.3–47.2) for thenon-OHE and the OHE group, respectively. The Kap-lan–Meier survival curves according to OHE status withnew tumor as endpoint (Fig 2) illustrate a protectivee!ect of OHE. The OHE variable was found to be sta-tistically significant (P = .022) in the univariable screen-ing with a protective e!ect of OHE (HR 0.47, 95% CI0.25–0.89). None of the other variables analyzed werestatistically significant in the univariable screening. Inthe final Cox proportional-hazards model, OHE-statuswas the only variable that had a statistically significante!ect on the hazard of new tumor with a hazard ratioof 0.47 (95% CI 0.25–0.90, P = .022) for the OHEgroup. However, age at time of surgery was added tothe model because of its possible biological importance.The hazard ratio for age was 1.12 per 1-year increase(95% CI 0.94–1.33, P = .27).

The protective e!ect of OHE against new tumordevelopment was found to vary with time. It increasedon a logarithmic time scale (log-time). Thus, an inter-

action between OHE status and log-time was includedin the model. The protective e!ect of OHE appearedapproximately 5 months after surgery and increasedup to 48 months after surgery when it reached a morestable level (Table 2).

E!ect of OHE on Survival

By the end of the observation period (May 31,2012), 60 (30 non-OHE and 30 OHE) of the 84 dogshad died; of these 33 underwent necropsy. Figure 3

Table 1. Distribution of patient characteristics bytreatment group in 84 dogs with nonmalignant mam-mary tumors randomized to ovariohysterectomy (OHEgroup) or to remain intact (non-OHE group) at thetime of tumor removal.

Non-OHEGroup(n = 42)

OHEGroup(n = 42) P value

Age .28! 9.0 years 19 24<9.0 years 23 18

Body weight .38! 22 kg 18 22<22 kg 24 20

Small breed versus other 1.00Small breed (" 10 kg) 7 7Other breed (>10 kg) 35 35

Pure versus mix breed .29Pure breeda 35 31Mix breedb 7 11

Tumor durationc .38! 6 months 17 21<6 months 25 21

Multiple tumors .23Yes 27 31No 15 11

Tumor size .83! 3 cm 20 21<3 cm 22 21

Type of surgery .12Local surgeryd 14 21Extensive surgerye 28 21

Tumor typef

Hyperplasia 4 6 .78Benign 31 30Benign with atypiag 7 6

aComprised of 30 di!erent breeds with English Setter (n = 8),German Shepherd (n = 6), German Pointer (n = 5), and Dachs-hund (n = 5) as the most common.

bDominated by middle-sized dogs (median weight = 22 kg,range: 3–53).

cTime from the tumor was first discovered by owner to thetime of project inclusion.

dLumpectomy or simple mastectomy.eRegional or radical mastectomy.fCases with multiple tumors and more than 1 histopathological

diagnose are grouped according to the most neoplastic character-ization of the tumors.

gIncludes benign tumors with necrosis, sclerosis, and carcinomain situ.

938 Kristiansen et al

illustrates the Kaplan–Meier survival curves byOHE status with death/euthanasia of any cause asthe endpoint. There was no statistically significantdi!erence in overall survival between the 2 groups.Median time to death was 31.2 months (range3.5–87.5) and 31.0 months (range 4.2–78.2) for thenon-OHE and the OHE group, respectively. How-ever, the hazard of dying regardless of cause wasinfluenced by the dogs’ age at the time of surgerywith a hazard ratio of 1.4 per 1 year increase in age(95% CI 1.19–1.62, P < .001). It did not di!er signif-icantly between the other investigated variables,including the OHE variable when applying a Coxproportional-hazard model with adjustment for age.The same result was seen when the endpoint wasdeath from mammary tumor.

Nine of the 42 dogs (21%) that developed newmammary tumors were euthanized because of clinicalproblems related to it (metastatic disease [n = 6], largeor ulcerated tumor compromising normal function[n = 2], development of inflammatory carcinoma[n = 1]). Six dogs belonged to the non-OHE group,and three dogs to the OHE group (95% CI: 0.098–1.59, P = .19). In Figure 4, clinical outcome by OHE-status is summarized for new tumor development,death because of mammary tumors, and death relatedto other causes. Of deaths related to other causes, 19

were because of other tumor types and 32 were relatedto nonneoplastic causes (multifactorial causes [“old”;n = 10], chronic degenerative joint disease [n = 8],heart disease [n = 4], acute abdomen [n = 3], trauma[n = 3], pyometra [n = 2], kidney failure [n = 1], andidiopathic epilepsy [n = 1]). The assumption of propor-tional hazards was valid for the OHE variable in themodel when death/euthanasia was the endpoint.

Diagnosis and Localization of Subsequent Tumors

In total, 42 dogs (50%) developed new tumors dur-ing the follow-up-period. Only 4 dogs had the newtumors surgically excised, but histopathological analy-sis was available in a total of 27 dogs because an addi-tional 23 dogs were returned for euthanasia and

0.00

0.25

0.50

0.75

1.00

Prop

ortio

n of

tum

or-fr

ee d

ogs

0 20 40 60 80Time after OHE (months)

Fig 2. The Kaplan–Meier curve shows time to new tumor devel-opment in 84 dogs with nonmalignant mammary tumors random-ized to ovariohysterectomy (OHE; dashed line, n = 42) or toremain intact (non-OHE; solid line, n = 42) at the time of surgery(P = .022). Median time to new tumor development was 19.6 and20.8 months for the OHE and the non-OHE groups, respectively.

Table 2. The time-varying e!ect of ovariohysterecto-my (OHE) on the hazard of new tumor developmentat di!erent time points estimated from a survivalmodel in a randomized trial among dogs with nonma-lignant mammary tumors.

Months after OHE 5a 12 24 36 48 60Hazard ratio 1.00 0.61 0.41 0.32 0.27 0.24

aOHE starts to be protective at this time (ratio >1.00).

0.00

0.25

0.50

0.75

1.00

Prop

ortio

n of

dog

s al

ive

0 20 40 60 80Time after OHE (months)

Fig 3. Kaplan–Meier survival curve on time to death from anycause after removal of nonmalignant mammary tumors in 84dogs randomized to ovariohysterectomy (OHE, dashed line) orto remain intact (solid line) at the time of surgery (P = .19).Median survival time was 31.0 and 31.2 months for the OHEand the non-OHE groups, respectively.

27

6

24

15

3

27

0

5

10

15

20

25

30

35

new tumor death due to mammary tumor

death due to other causes

No.

of d

ogs intact

spayed

* P = 0.009

Fig 4. Distribution of postsurgical clinical outcome (newtumors, death because of mammary tumors, and death becauseof other causes) in 84 dogs with nonmalignant mammary tumorsrandomized to ovariohysterectomy or to remain intact at thetime of tumor removal (42 in each treatment group). Significantresults are indicated by an asterisk (*).

Ovariohysterectomy and Canine Nonmalignant mammary Tumors 939

necropsy. Sixteen dogs had at least 1 malignant newtumor and 10 dogs were found to have only benignnew tumors. The distribution of these dogs by OHEstatus is shown in Figure 5.

Of the 12 dogs which initially had 1 single tumor, 7dogs (58%) developed new tumors ipsilaterally, ofwhich 5 appeared in the adjacent gland. The other 5dogs (42%) developed new tumors in the contralateralchain, of which 4 were found in the most adjacent ofthe contralateral glands.

Thirty dogs with new tumor development had multi-ple tumors initially, thereof 22 cases with bilateral dis-tribution. In the remaining dogs where the initialtumors were unilateral (n = 8), 5 developed newtumors in the contralateral mammary chain.

Impact of OHE on Other Health Issues

Six dogs in the intact group developed pyometra forwhich 3 were treated with OHE (censored at that time)and 3 were euthanized. Moreover, clinically significantuterine/ovarian tumors were found in 2 other dogs; 1dog developed uterine leiomyosarcoma and was ovar-iohysterectomized because of this (censored at thattime). A 2nd dog was euthanized as a result of acuteabdomen because of torsion of a cystic ovary. Also,on necropsy, an ovarian granulosa cell tumor and avaginal leiomyofibroma were revealed in one of thedogs.

Discussion

This study shows a significant risk reduction for newmammary tumor development in dogs treated withOHE adjuvant to NMT removal. Based on previouspublications,14–16 it has been assumed that OHE laterin life has little e!ect on mammary tumor risk. How-ever, none of these former studies were prospective,randomized, or stratified according to malignancy norwas the development of new tumors used as endpointin the analysis. These findings strengthen the similari-ties between women and dogs in term of the role ofhormonal exposure and breast cancer risk; in womenthe breast cancer risk is associated with the cumulative

life time exposure of breast tissue to hormones, ourresults suggest a similar situation in the dog, the e!ectof hormones are continuous and contributes toincreased risk also after the age of 2.

The hazard ratio of 0.47 (OHE/nonOHE) for newtumor development reflects that bitches spayed at thetime of tumor removal, on average run approximatelyhalf the risk of developing new tumors at any giventime compared with the bitches remaining intact. Ahazard can therefore be interpreted as an instanta-neous risk, and the term risk will be used in the fol-lowing discussion because it is probably a morefamiliar concept to most readers.

The advantage of OHE on time to new tumor devel-opment did not become apparent until 5 months afterthe OHE intervention. This somewhat delayed e!ectand then increasing benefit over time makes sense froma biological point of view and probably reflect the timeneeded for ovarian hormones in mammary tissue to bewashed out in the dogs undergoing OHE. Also, thefact that most of the dogs in the non-OHE group hadtime to experience estrus during the 5 months periodmight explain the increasing benefit of OHE over time.The temporarily increased endogenous hormonal expo-sure might have accelerated new mammary tumordevelopment in the non-OHE group compared withthe OHE group. This may also explain why this bene-fit becomes apparent after 5 months. Normal mam-mary epithelial tissues and most benign tumors expresshormone receptors. With time, continued exposure tohormones is likely to contribute to early subclinicalstages of mammary tumorigenesis, resulting in addi-tional mammary tumors in the subset of dogs thatremained sexually intact.

Relatively few (n = 9) dogs died because of newmammary tumors, and all of them occurred late rela-tive to the initial tumor surgery. Consequently, theincreased rate of new tumors in the intact group didnot translate into a significant di!erence in overall sur-vival between the 2 treatment groups. In addition tothe increased risk of new mammary tumors, nine ofthe intact dogs also experienced significant uterine andovarian diseases necessitating emergency surgery oreven euthanasia in some cases.

In this study, cause of death was only attributed tomammary tumors if the presence of mammary tumorsand metastasis was the direct cause of death/euthana-sia. The majority of the dogs in this study were eutha-nized because of other causes; however, many (n = 32)of these dogs did have documented new mammarytumors at the time of euthanasia. Because of the com-plexities of issues a!ecting decisions regarding eutha-nasia, survival may represent a “soft” and biasedendpoint in this study. In contrast, the development ofnew tumors represents an objective and clearly definedendpoint and our results show that OHE significantlydecreases the risk for new tumors in dogs with NMT.

Alternatively or in addition to OHE, prophylacticmastectomies of normal mammary glands could beconsidered to prevent new mammary tumors. To makegood decisions regarding which glands to prophylacti-

84 NMT

42 non-OHE

15 NT+

27 NT-

27 NT+

15 NT-

1 benign and malignant

3 benign

3 malignant

6 benign and malignant

11 unknown

5 unknown

7 benign

6 malignant

42 OHE

Fig 5. Distribution of new tumor development (NT) in 84 dogsafter surgical removal of nonmalignant mammary tumors with orwithout concomitant ovariohysterectomy (OHE).

940 Kristiansen et al

cally remove, the likely localization of such new tumorsmust be known. According to a previous study, dogswith 1 mammary tumor were likely to develop newtumors ipsilaterally.17 The majority of dogs with mam-mary tumors, however, have more than 1 tumor. In thisstudy, those dogs were also included when determiningthe localization of new tumors. Our results showed thatmost new tumors developed adjacent to a previouslya!ected gland; however, almost half of the dogs (n = 5/12) with a single mammary tumor developed a newtumor in the contralateral chain. These findings are inconflict with Stratmann et al’s findings,17 and reflectthat the exact localization of new tumors may be some-what unpredictable. Clearly, a bilateral radical mastec-tomy would e!ectively prevent any new mammarytumors in dogs. This is, however, an aggressive proce-dure and may be di"cult to justify as 36% of the intactdogs and 64% of the spayed dogs did not develop addi-tional mammary tumors. In contrast, most dogsrecover quickly from OHE. Furthermore, the benefitsare 2-fold: OHE significantly reduces the risk for newtumors and prevents uterine/ovarian diseases later on.As many as nine of the 42 dogs (21%) in the nonOHEgroup developed uterine–ovarian diseases, of whichfour were euthanized. The previously published dataregarding the incidence of pyometra in dogs with mam-mary tumors are unclear. Some studies suggest that asignificant percentage of bitches with mammary tumorshave concomitant uterine–ovarian disease, and thatthey are likely to experience clinical signs (pyometra,mucometra) in the future if OHE is not performed.25

Others, however, did not find any increased risk of py-ometra in dogs with mammary tumors.26 In this study,6 of 42 dogs (14%) developed pyometra.

Nonmalignant mammary tumors are generally asso-ciated with a long postsurgical survival, thus allowingnew tumors and other diseases to develop. As age isadvancing, the risk of uterine diseases increases.26,27

Some owners choose not to do more surgery in an olddog if it already has been put through 1 or more sur-geries earlier. Otherwise treatable uterine and ovariandiseases, may for this reason lead to the decision toeuthanize the dog rather than “putting an older dogthrough another surgery”. OHE at the time of tumorremoval may decrease the risk of new tumors, but alsoprotect the dog from later uterine/ovarian diseaseswhich again might save some dogs from being eutha-nized because of the negative influence such concurrentdiseases might have on the owner’s decision to pursueadditional surgery in an old dog. Our study shows thatOHE confers a double benefit in these dogs.

Methodological Considerations

This study’s main strength lies in the design. Pro-spective randomized clinical trials are considered themost reliable form of acquiring scientific evidencebecause they reduce spurious causality and bias byensuring comparable groups. Thus, any significantdi!erences between groups in the outcome event canbe attributed to the intervention (OHE) and not to

some other unidentified factor. Unfortunately, thisstudy was not published in time to be included in themeta-analysis on the e!ect of OHE on mammarytumor risk performed by Beauvais et al (2012).21 Infact, this study is the first investigating this questionusing this design. Additional strength of this study isthe long follow-up (median 31.5 months, range3.5–87.5). Few studies include more than 2 years offollow-up.

The main limitation of this study is that the postsurgi-cal information for some dogs was retrieved solelythrough phone (17%). However, most of these ownershad detected the initial tumors by themselves, thus con-firming their ability to recognize such tumors. More-over, we were only able to obtain the histopathologicaltumor diagnosis in 26 of the 42 dogs with new tumors.The fact that only 33 dogs were necropsied is also alimitation, but owner contact through phone ensuredreliable information regarding why the dog was eutha-nized. Furthermore, these limitations are not likely tointroduce bias between the groups and thus a!ect theoutcome as these sources of bias were evenly distributedbetween the 2 intervention groups. Finally, the smallsample size is a potential weakness when consideringmultiple variables because the power of detecting e!ectthat really are present decreases with increasing numberof variables. However, the significant e!ect of the OHE-status variable was maintained in the multivariableanalysis, indicating that the power of detecting an e!ectof this variable was indeed acceptable.

In conclusion, OHE at the time of tumor removalreduced the risk for subsequent mammary tumors by47% in bitches with NMT in this study. This protec-tive e!ect was evident from 5 months after the OHEintervention and increased until a steady state wasreached after approximately 4 years. No significante!ect of OHE on overall or tumor-specific survivalwas detected for this sample of dogs. Adjunctive OHEshould be considered in the treatment of adult dogswith mammary tumors.

Footnotea Stata 11; Stata Corporation, College Station, TX

Acknowledgments

The authors thank all the dog owners who partici-pated with their dog in the study and gratitude is givento Randi Krontveit for assistance with the survivalanalysis.

This work was supported by Morris Animal Foun-dation, the Norwegian “Research foundation forcanine cancer”, the Scientific Foundation from theNorwegian Small Animal Veterinary Association, andthe Norwegian School of Veterinary Science.

Conflict of Interest Declaration: Authors disclose noconflict of interest.

Ovariohysterectomy and Canine Nonmalignant mammary Tumors 941

References

1. Dobson JM, Samuel S, Milstein H, et al. Canine neoplasiain the UK: Estimates of incidence rates from a population ofinsured dogs. J Small Anim Pract 2002;43:240–246.

2. Dorn CR, Taylor DO, Schneider R, et al. Survey of animalneoplasms in Alameda and Contra Costa Counties, California.II. Cancer morbidity in dogs and cats from Alameda County.J Natl Cancer Inst 1968;40:307–318.

3. Egenvall A, Bonnett BN, Ohagen P, et al. Incidence of andsurvival after mammary tumors in a population of over 80,000insured female dogs in Sweden from 1995 to 2002. Prev Vet Med2005;69:109–127.

4. Merlo DF, Rossi L, Pellegrino C, et al. Cancer incidence inpet dogs: Findings of the Animal Tumor Registry of Genoa,Italy. J Vet Intern Med 2008;22:976–984.

5. Moe L. Population-based incidence of mammary tumoursin some dog breeds. J Reprod Fertil Suppl 2001;57:439–443.

6. Goldschmidt M, Pena L, Rasotto R, Zappulli V. Classifica-tion and grading of canine mammary tumors. Vet Pathol2011;48:117–131.

7. Sorenmo KU, Kristiansen VM, Cofone MA, et al. Caninemammary gland tumours; a histological continuum from benignto malignant; clinical and histopathological evidence. Vet CompOncol 2009;7:162–172.

8. Chang SC, Chang CC, Chang TJ, Wong ML. Prognosticfactors associated with survival two years after surgery in dogswith malignant mammary tumors: 79 cases (1998–2002). J AmVet Med Assoc 2005;227:1625–1629.

9. Clemente M, Perez-Alenza MD, Pena L. Metastasis ofcanine inflammatory versus non-inflammatory mammarytumours. J Comp Pathol 2010;143:157–163.

10. Sorenmo KU, Shofer FS, Goldschmidt MH. E!ect ofspaying and timing of spaying on survival of dogs with mam-mary carcinoma. J Vet Intern Med 2000;14:266–270.

11. Ferrara A. Benign breast disease. Radiol Technol2011;82:4472.

12. Goldacre MJ, Abisgold JD, Yeates DG, Vessey MP.Benign breast disease and subsequent breast cancer: Englishrecord linkage studies. J Public Health (Oxf) 2010;32:565–571.

13. Ashbeck EL, Rosenberg RD, Stauber PM, Key CR.Benign breast biopsy diagnosis and subsequent risk of breastcancer. Cancer Epidemiol Biomarkers Prev 2007;16:467–472.

14. Schneider R, Dorn CR, Taylor DO. Factors influencingcanine mammary cancer development and postsurgical survival.J Natl Cancer Inst 1969;43:1249–1261.

15. Misdorp W. Canine mammary tumours: Protective e!ectof late ovariectomy and stimulating e!ect of progestins. Vet Q1988;10:26–33.

16. Taylor GN, Shabestari L, Williams J, et al. Mammary neo-plasia in a closed beagle colony. Cancer Res 1976;36:2740–2743.

17. Stratmann N, Failing K, Richter A, Wehrend A. Mam-mary tumor recurrence in bitches after regional mastectomy. VetSurg 2008;37:82–86.

18. Morris JS, Dobson JM, Bostock DE, O’Farrell E. E!ectof ovariohysterectomy in bitches with mammary neoplasms. VetRec 1998;142:656–658.

19. Queiroga FL, Perez-Alenza MD, Silvan G, et al. Role ofsteroid hormones and prolactin in canine mammary cancer.J Steroid Biochem Mol Biol 2005;94:181–187.

20. Rutteman GR, Misdorp W, Blankenstein MA, van denBrom WE. Oestrogen (ER) and progestin receptors (PR) inmammary tissue of the female dog: Di!erent receptor profile innon-malignant and malignant states. Br J Cancer 1988;58:594–599.

21. Beauvais W, Cardwell JM, Brodbelt DC. The e!ect ofneutering on the risk of mammary tumours in dogs–a systematicreview. J Small Anim Pract 2012;53:314–322.

22. Misdorp W, Else RW, Hellmen E, Lipscomb TP. Histolo-gic classification of mammary tumors in the dog and cat. WorldHealth Organization International Histological Classification ofTumors of Domestic Animals. 2nd Series, Vol. 7. WashingtonDC: Armed Forces Institute of Pathology; 1999: 1–59.

23. Withrow SJ, Vail DM, Stringer S, Winkel A. Withrowand MacEwen’s Small Animal Clinical Oncology, 4th ed.St. Louis, MO: Elsevier; 2007.

24. Dohoo I, Martin W, Stryhn H. Modelling survival data,chapt. 19. In: McPike SM, ed. Veterinary Epidemiologic Research,2nd ed. Charlottetown, PEI, Canada: Ver Inc; 2009:467–527.

25. Ferguson HR. Canine mammary gland tumors. Vet ClinNorth Am Small Anim Pract 1985;15:501–511.

26. Hagman R, Lagerstedt AS, Hedhammar A, Egenvall A. Abreed-matched case-control study of potential risk-factors forcanine pyometra. Theriogenology 2011;75:1251–1257.

27. Fukuda S. Incidence of pyometra in colony-raised Beagledogs. Exp Anim 2001;50:325–329.

942 Kristiansen et al