The most common clinical signs in dogs with

Transcription

1 J Vet Intern Med 2007;21: Survival, Neurologic Response, and Prognostic Factors in Dogs with Pituitary Masses Treated with Radiation Therapy and Untreated Dogs Michael S. Kent, David Bommarito, Edward Feldman, and Alain P. Theon Background: Pituitary masses in dogs are not uncommon tumors that can cause endocrine and neurologic signs and, if left untreated, can decrease life expectancy. Hypothesis: Dogs with pituitary masses that received radiation therapy (RT) have more favorable neurologic outcomes and longer survival times compared with untreated dogs. Animals: Nineteen dogs with a pituitary mass identified on CT or MR imaging were irradiated with 48 Gy given in 3 Gy daily-dose fractions. Twenty-seven untreated control dogs had pituitary masses. Methods: Medical records of dogs with pituitary masses were retrospectively reviewed for clinical signs, mass size, and outcome. Results: Median survival time was not reached in the treated group. Mean survival time in the treated group was 1,405 days (95% confidence interval [CI], 1,053 1,757 days) with 1-, 2-, and 3-year estimated survival of 93, 87, and 55%, respectively. Median survival in the nonirradiated group was 359 days (95% CI, days), with a mean of 551 days (95% CI, days). The 1-, 2-, and 3-year estimated survival was 45, 32, and 25%, respectively. Dogs that received RT for their pituitary tumors had significantly longer survival times than untreated dogs (P ). Treated dogs with smaller tumors (based on maximal pituitary-to-brain height ratio or area of tumor to area of brain) lived longer than those with larger tumors (P,.001). Conclusions and Clinical Importance: When compared with untreated dogs, RT increased survival and controlled neurologic signs in dogs with pituitary masses. Key words: Canine; Macroadenoma; Macrotumor; Pituitary tumor; Radiation therapy. The most common clinical signs in dogs with pituitary tumors are related to the secretory nature of the mass (usually ACTH secretion with resultant hyperadrenocorticism), the space-occupying effects of the mass, or both. Neurologic abnormalities attributed to a space-occupying pituitary mass include change in behavior, inappetence, obtundation, aggressiveness, anisocoria, apparent blindness, and seizures. 1,2 Progression of clinical signs can lead to poor quality of life and, if uncontrolled, can lead to death or euthanasia. Pituitary masses can be categorized by size or histopathologically as nodular hyperplasia, cysts, adenomas, adenocarcinomas, or carcinomas. 3 6 Pituitary masses also may be characterized according to function, with ACTH-producing adenomas of the pars distalis or pars intermedia being the most common pituitary tumors in dogs. 6 Dogs with pituitary-dependent hyperadrenocorticism (PDH) by definition have ACTHsecreting pituitary tumors. Up to 30% of dogs with PDH have pituitary tumors large enough to result in mass-related clinical signs within several months of diagnosis of hyperadrenocorticism. 7 Approximately 10% From the Department of Surgical and Radiological Sciences (Kent, Theon), the Veterinary Medical Teaching Hospital (Bommarito) and the Department of Medicine and Epidemiology (Feldman), University of California, Davis, CA. Reprint requests: Michael S. Kent, University of California, Department of Surgical and Radiological Sciences, 2112 Tupper Hall, 1 Shields Avenue, Davis, CA 95616; mskent@ucdavis. edu. Submitted November 22, 2006; Revised February 9, 2007, March 1, 2007; Accepted April 11, Copyright E 2007 by the American College of Veterinary Internal Medicine /07/ /$3.00/0 of dogs with pituitary masses have clinically nonfunctional tumors. 1,7 MR imaging (MRI) and CT are the only antemortem, noninvasive methods for determining the presence and size of a pituitary mass. 8,9 In humans, pituitary tumors.10 mm in height are classified as macroadenomas. 10 The normal height of the pituitary gland in the dog is mm, 11,12 and the definition of macroadenoma used in humans may not be useful in dogs. 13 Radiation therapy (RT), surgery, or a combination of both are the treatments of choice for most intracranial masses RT is the most commonly employed treatment for pituitary tumors in dogs. 14,17,18,19 Most studies evaluating RT demonstrated the potential for decreasing tumor size and prolonging survival. Treatment protocols with a large dose per fraction, however, were associated with brain toxicity, including formation of second tumors and necrosis. 1,20,21 Although microsurgical transsphenoidal hypophysectomy as the sole treatment has proven to be relatively safe and effective for dogs with pituitary masses, it is technically challenging and not widely available. 13 Medical management of hyperadrenocorticism with drugs such as o,p DDD or trilostane does not affect pituitary tumors directly. Use of these drugs does not inhibit mass growth or onset of neurologic signs. 18,22,23 The limited number of studies of clinical-sign progression and survival in dogs with untreated pituitary tumors makes benefit analysis of RT for the treatment of pituitary masses difficult. The purpose of this study was to evaluate survival and neurologic response of dogs with pituitary tumors to a daily 3 Gy per fraction, RT protocol and compare these results with those of an untreated group of dogs. A secondary aim of this study was to analyze prognostic factors in dogs with pituitary tumors. In order to assess RT efficacy, a group

2 1028 Kent et al of dogs with untreated pituitary tumors was used as control. Materials and Methods The electronic medical database of the Veterinary Medical Teaching Hospital of the University of California, Davis was searched for dogs diagnosed with a pituitary mass between January 1998 and December For inclusion, dogs must have had pituitary CT or MRI, and imaging must have demonstrated the presence of a pituitary mass. Although no histologic diagnosis was available for these dogs, all masses were assumed to be pituitary tumors, and the terms pituitary mass and tumor are used interchangeably in this study. Dogs with pituitary masses were divided into 2 groups. Dogs receiving RT were included in the treatment group, and untreated dogs were included in the control group. Information obtained from the medical records included breed; sex; age at diagnosis of a pituitary mass; clinical signs at diagnosis; results of diagnostic tests; and follow-up information including endocrine testing, clinical signs, and date and cause of death. Data from postmortem examinations were collected when available. If follow-up information was not complete in the medical record, local veterinarians and owners were contacted to obtain the missing information. The CT scans or MR images were reviewed to confirm the presence of a mass in the pituitary region. Measurements of maximal pituitary mass height and width as well as maximal brain height and width were taken from the axial CT scan slice in which the mass in the pituitary region was the largest. A dog was defined as having hyperadrenocorticism if it had at least 3 of the following clinical signs: polyuria, polydipsia, polyphagia, excessive panting, weakness, thin skin, pot-bellied appearance, and bilaterally symmetric alopecia as well as at least 1 confirmatory screening test result (post ACTH-stimulation serum cortisol concentration. 22 mg/dl, a serum cortisol concentration. 1.4 mg/dl 8 hours after administration of a low dose [0.01 mg/ Kg] of IV dexamethasone, or a urine cortisol : creatinine ratio. 13.5). PDH was confirmed in dogs either by a serum cortisol concentration, 50% of the baseline concentration or, 1.0 mg/dl 4 hours after administration of a low dose of dexamethasone (0.01 mg/kg), a plasma endogenous ACTH concentration. 45 pg/ml, or presence of 2 relatively equal-sized adrenal glands on abdominal ultrasound examination. 24,25 Dogs were defined as having a nonsecretory pituitary tumor if they had evidence of a mass in the pituitary area without evidence of hyperadrenocorticism. A dog was defined as having neurologic signs possibly caused by a mass in the pituitary region if the owner observed one or more of the following: behavior change, inappetence, decreased alertness (obtundation), apparent blindness, or seizures. In evaluating neurologic response, complete resolution of signs was defined as resolution of the signs with no recurrence; transient resolution of neurologic signs was defined as resolution of neurologic signs, which then returned at some later date; and no change in neurologic signs was defined as continued neurologic signs. Neurologic sign follow-up was based on owner observation or physical examination findings. Irradiation All dogs were treated on a Monday-through-Friday basis with a 4-MV linear accelerator. a Treatment plans were formulated with a commercially available computerized treatment-planning software package, b into which the dog s CT scan was imported, with a source-to-axis distance technique with parallel opposed fields designed to deliver 3 Gy to the planning target volume for each fraction. The planning target volume was defined as 1 cm beyond the gross tumor volume as defined by the imaging study. Dogs were positioned with sand bags and foam wedges. Portal films were taken on the first day of treatment to confirm and document the dog s positioning before treatment. Films also were repeated at least once weekly, and often daily, to confirm the dog s positioning. Seventeen dogs received the planned dose of 48 Gy; 2 dogs that were scheduled to receive 48 Gy died during the course of treatment after receiving 24 and 36 Gy total, respectively. All but 3 of the dogs in the treated group were placed on prednisone ( mg/ kg per day) during RT and had their dosage tapered over several weeks after completion of RT. Data Analysis Using either the CT scans or MR images, we measured pituitary height and width on the transverse image that displayed the maximal pituitary area. The maximal dimensions of brain height and width were measured on this same image. The height of the pituitary divided by brain height was defined as the pituitary-tobrain height ratio. The approximate area of the pituitary divided by the approximate area of the brain was defined as the pituitary-tobrain area ratio. The pituitary and brain were considered to be ovoid in shape, and area was calculated with the following formula: (0.8)(tumor height)(tumor width) / (0.8)(brain height)(brain width). The data were checked for normality, and the pituitary-to-brain area ratio was log-transformed to normalize the data and minimize variance. Age, body weight, sex, presence of PDH, pre- and 1-hour post- ACTH serum cortisol concentrations, presence of neurologic signs, use of mitotane before tumor diagnosis, pituitary-to-brain height ratio, and pituitary-to-brain area ratio in control and irradiated groups were compared. For continuous variables, a Student s t-test was used. For categoric variables a chi-square test was used. The Kaplan-Meier method was used to estimate survival times. The median survival and progression-free survival times represent the estimated time point at which 50% of the animals have died or had progression of their disease. All dogs were evaluated for survival with the date of imaging diagnosis as day 0. Dogs in the untreated group that were euthanized within 1 week of imaging because of their diagnosis were not included in survival analysis. Dogs lost to follow-up were censored on the last day of contact. Dogs identified as dying of a cause other than their pituitary tumor were censored on the date of death only if there was a specific unrelated cause of death identified. Differences in the survival distributions between the treated and untreated groups were analyzed by use of the log-rank test. The effects that age, body weight, pre or post ACTH-stimulation serum cortisol concentrations, presence of neurologic signs, tumor height, pituitary-to-brain height ratio, and pituitary-to-brain area ratio had on overall survival in the treated group were tested by Cox regression analysis. A P value,.05 was considered significant. Where appropriate, data are presented as mean 6 standard deviation or median and range. Statistical analysis was done with a commercially available software package. c Results Dogs Forty-six dogs met the inclusion criteria. Nineteen dogs diagnosed with a pituitary mass on CT or MRI received RT that was started between January 1998 and December There were 3 Boston Terriers; 1 each of Beagle, Boxer, Bullmastiff, Cocker Spaniel, German Shepherd Dog, Golden Retriever, Lhasa Apso, Shih Tzu, and Whippet; and 7 mixed breed dogs. Seven dogs were considered to be brachycephalic. The median weight of these 19 dogs was 21.0 kg (range, kg),

3 Irradiation of Pituitary Masses 1029 and the median age was 10.4 years (range, years). Fourteen of these 19 dogs had been diagnosed with PDH and, therefore, were assumed to have ACTH-secreting pituitary tumors. Five of the 19 dogs were assumed to have endocrine-inactive pituitary tumors. One dog had diabetes insipidus. Dogs started RT a median of 10 days (range, 1 63 days) after imaging. Twenty-seven dogs diagnosed as having a pituitary mass on CT or MRI received no RT. Seven of these dogs were euthanized on the day of diagnosis and were not included in survival analysis. There were 4 Miniature Schnauzers; 2 Boston Terriers; 2 English Springer Spaniels; 1 each of Australian shepherd, Beagle, Bichon Frise, Briard, Cocker Spaniel, Dalmatian, Siberian Husky, terrier, toy poodle, Welsh Springer Spaniel, Wire Fox Terrier; and 8 mixed breed dogs. Three dogs were considered to be brachycephalic. The median weight of the dogs in the control group was 21.1 kg (range, 5 35 kg). The median age of the dogs was 11.3 years (range, years). Seventeen of these 27 dogs had been diagnosed with PDH and, therefore, were assumed to have ACTH-secreting pituitary tumors. Ten of the 27 dogs were assumed to have endocrine-inactive pituitary tumors. Eleven of the 19 dogs in the treated group were diagnosed as having neurologic signs before the time of imaging. Six dogs had periods of obtundation; 5 had decreased appetite; 2 had seizures; 2 had anisocoria; and 1 each displayed aggression, circling, or a head tilt. Sixteen of the 27 dogs in the untreated group had neurologic signs at the time of imaging. Ten dogs had periods of obtundation, 6 had a decreased appetite, 5 had seizures, 2 had anisocoria, 2 were circling, and 1 displayed aggression. The median height of the pituitary tumor for all 46 dogs was 1.0 cm (range, cm). In the treated group, it was 1.3 cm (range, cm), and in the untreated group, it was 0.9 cm (range, cm). For all 46 dogs, mean pituitary-to-brain height ratio was 0.26 (range, ). In the treated group, it was 0.26 (range, ), and in the untreated group, it was 0.26 (range, ). The median pituitary-to-brain area ratio for all 46 dogs was 0.09 (range, ). In the treated group, it was 0.10 (range, ), and in the untreated group, it was 0.08 (range, ). No interaction between tumor size and whether or not a dog was treated was found (P 5.48). Mean pre- and post-acth serum cortisol concentrations at the time of PDH diagnosis in the 10 RT-treated dogs that had this test performed were mg/dl and mg/dl, respectively. Results for the 17 PDH dogs in the control group were mg/dl and mg/dl. Reference intervals for pre and post ACTH-stimulation serum cortisol concentrations are mg/dl and mg/dl, respectively. 24 No significant differences with regard to age (P 5.23), body weight (P 5.36), tumor height (P 5.08), pre and post ACTH-stimulation serum cortisol concentrations (P 5.85, P 5.88), the presence of neurologic signs at diagnosis (P 5.78), diagnosis of PDH (P 5.53), treatment with mitotane before diagnosis (P 5.08), being brachycephalic (P 5.17), pituitary-to-brain height ratio (P 5.82), or pituitary-to-brain area ratio (P 5.47) between the treated and control groups were found. Clinical Endocrine Response Five of the 14 dogs with PDH in the treated group had resolution of clinical signs related to hyperadrenocorticism, as reported by the owner, and at least 1 normal ACTH-stimulation test after RT. Six dogs had no change in clinical signs and no improvement in endocrine testing, 2 dogs died during therapy, and 1 dog had no follow-up hormonal testing information available. The 1 dog that had diabetes insipidus before starting RT had resolution of this condition by the end of RT. Neurologic Signs Twenty-seven dogs with neurologic signs at the time of pituitary imaging had a median tumor height of 1.4 cm (range, cm), median pituitary-to-brain height ratio of 0.33 (range, ), and median pituitary-to-brain area ratio of 0.11 (range, ). Nineteen dogs had no neurologic signs at the time of pituitary imaging. Median tumor height in these dogs was 0.90 cm (range, cm), median pituitary-tobrain height ratio was 0.21 (range, ), and the median pituitary-to-brain area ratio was 0.05 (range, ). There were statistically significant differences for dogs with and without neurologic signs for tumor height (P 5.04), pituitary-to-brain height ratio (P 5.01), and pituitary-to-brain area ratio (P 5.01). Two of 3 dogs without neurologic signs at the time of RT went on to develop neurologic signs at 637 and 1,078 days after RT had been completed. Of the 14 dogs with neurologic signs that completed RT, 5 had complete resolution of signs, 4 had transient resolution of neurologic signs, and 3 had no change in neurologic signs. The median time to development of neurologic signs or to recurrence of neurologic signs was 994 days (Fig 1). None of the untreated dogs had improvement in their neurologic signs. Radiation Adverse Effects Acute adverse effects were minimal in the majority of dogs. Partial alopecia involving the skin exposed to radiation was the most common adverse effect. Leukotrichia in the radiation field was noted on follow-up physical examinations in all dogs. No acute effects involving brain, such as worsening of neurologic signs during or immediately after RT, were noted. One dog developed a thalamic brain mass 217 days after completing RT. This mass was spatially separate from the treated pituitary mass but in the radiation field. No pathology was available on this mass. Survival Median survival in the treated group was not reached. Mean survival in the treated group was 1,405 days (95%

4 1030 Kent et al Fig 1. Kaplan-Meier curve for progression-free period of neurologic signs in dogs treated with radiation therapy (RT) for a mass in the pituitary region. Nineteen dogs were treated with RT. Fourteen of these dogs had or went on to develop neurologic signs. Censored dogs are indicated by tick marks. CI, 1,053 1,757 days). Median follow-up time in the treated group was 702 days (range, 27 1,927 days). The estimated 1-, 2-, and 3-year survival rates were 93% (95% CI, 63 99%), 87% (95% CI, 56 96%), and 55% (95% CI, 22 79%), respectively. Median survival in the untreated group was 359 days (95% CI, days), with a mean of 551 days (95% CI, days). The 1-, 2-, and 3- year estimated survivals were 45, 32, and 25%, respectively. Dogs receiving RT for their pituitary tumor had significantly longer survival times than dogs that were not irradiated (P ) (Fig 2). Ten dogs in the treated group were alive at the time of analysis. Two dogs had been euthanized during therapy, 1 because of aspiration pneumonia and 1 because of suspected pulmonary thromboembolism. One dog was euthanized after being diagnosed as having a splenic mass 738 days after RT. One dog was euthanized because of a new mass in the brain 217 days after RT. Three dogs were euthanized because of progressive neurologic signs at 467, 982, and 994 days after RT. One dog died of drowning 851 days after RT. One dog was euthanized after developing signs of mitotane toxicity 134 days after RT. Prognostic Factors In the treated group, increasing pituitary-to-brain height ratio (P,.001; hazard ratio [HR], 7.71 ) and pituitary-to-brain area ratio (P,.001; HR, 16.67) were both found to negatively impact overall survival, whereas tumor height did not (P 5.31; HR, 2.8). Using a log-rank test, we determined that pituitary-to-brain height ratio. 25% or pituitary-to-brain area ratio. 5% was associated with decreased survival (P 5.04 and P 5.05, respectively). None of the dogs died of their disease during the study period if their pituitary-to-brain height ratio was, 25% or if their pituitary-to-brain area ratio was, 5%. Body weight (P 5.64; HR, 0.98), the use of mitotane before RT (P 5.81), age (P 5.29; HR, 1.3), pre-acth Fig 2. Survival distribution of dogs with pituitary masses treated either with radiation therapy (RT) (dashed line) or untreated (solid line). Nineteen dogs were treated with RT, and 20 dogs received no therapy. These curves were significantly different (log-rank test, P ). Censored dogs are indicated by tick marks. serum cortisol concentrations (P 5.66; HR, 0.95), post- ACTH serum cortisol concentrations (P 5.15; HR, 1.1), presence of neurologic signs at the time of imaging (P 5.86), presence of PDH at the time of diagnosis (P 5.53), or being brachycephalic (P ) was not a significant prognostic factor for survival. In the untreated group, increasing pituitary-to-brain height ratio (P,.001; HR, 4.78), pituitary-to-brain area ratio (P,.001; HR, 10.35), pituitary height (P,.002; HR, 8.03), having an endocrine-inactive tumor (P 5.001), and having an increased pre-acth serum cortisol concentration (P 5.02; HR, 1.35) all were found to negatively impact overall survival. Using a log-rank test, we found that pituitary-to-brain height ratio. 25% or pituitary-to-brain area ratio of. 5% was associated with decreased survival (P,.0001 for both parameters). Body weight (P 5.39; HR, 0.97), the use of mitotane before RT (P 5.09), age (P 5.71; HR, 0.96), post- ACTH serum cortisol concentrations (P 5.16; HR, 1.06), presence of neurologic signs at the time of imaging (P 5.26), or being brachycephalic (P 5.39) was not a significant prognostic factor for survival. Discussion This study was designed to compare the outcome of dogs with pituitary masses that were treated with RT with dogs with pituitary masses that were not treated with RT. Comparing these 2 groups provided a means to assess the benefit of RT. Because this was a retrospective study, dogs could not be randomized to either of the 2 groups. No significant differences were found between the treated and untreated dogs for any of the variables tested. More dogs in the untreated group received mitotane therapy before imaging, but we did not find an effect on the use of mitotane on survival. Mean pituitary tumor height was larger in the treated group, making it unlikely to overestimate survival in this group. Furthermore, tumor size did not affect the decision to irradiate, avoiding a selection bias.

5 Irradiation of Pituitary Masses 1031 Imaging with CT or MRI remains the most accurate means for identifying and characterizing pituitary tumor size. 8,9 Correlation of tumor size to circulating hormone concentrations (eg, endogenous ACTH, ACTH precursors) has been suggested, but these results are not adequate for determining tumor size. 26,27 Although MRI has been reported to give better anatomic resolution and soft-tissue contrast than CT, it has not been proven to enhance tumor recognition. 28 Contrast-enhanced CT can be used to identify masses in the pituitary fossa that would not have been identified on noncontrast CT studies. 3,7 Furthermore, dynamic contrast CT can be used to identify smaller pituitary tumors. 29,30 CT images also are used in most RT-planning systems, making CT adequate for diagnosis and treatment-planning needs. The daily fractionated dose of 3 Gy on a Mondaythrough-Friday basis was used rather than the previously published dose of 4 Gy per fraction on a 3-timesper-week schedule, in an attempt to minimize neurologic complications. Severe late effects in the brain, including necrosis, occur more frequently with increasing dose per fraction. 20 Although the total dose is important in determining the chance of adverse acute effects of radiation, the fractionation schedule, the duration of treatment, and dose per fraction also play a role. No increases in acute effects in these dogs were seen with the use of a daily fractionation protocol as opposed to 3 times per week protocols. Use of corticosteroids in the majority of treated dogs may have decreased the occurrence of acute effects. Because of the retrospective nature of the study, there was a lack of follow-up testing to evaluate for pituitary insufficiency, but no treated patients had information in the medical records suggesting this complication. Previous reports have indicated that neurologic signs only develop in dogs with tumors. 10 mm in diameter. 7 In the present study, we found that neurologic signs could develop in dogs with tumors as small as 3 mm in height, but these neurologic signs occurred in a single dog that developed seizures. There may or may not have been an association between the visualized mass and the seizure disorder, but seizures developed 2 days before imaging. Furthermore, only 4 dogs with tumors. 1cm had no evidence of neurologic signs, and 4 additional dogs with tumors, 1 cm had evidence of neurologic signs. The median height of tumors associated with neurologic signs was 1.4 cm (range, cm), indicating that most dogs with neurologic signs due to a pituitary tumor have masses with heights. 10 mm. The results of this study indicate a variable response to RT with respect to resolving signs of hyperadrenocorticism. This observation is consistent with what has been reported in other studies. 1,21 Because no consistent follow-up endocrine testing was performed in this group of dogs, it was not possible to measure a progressionfree interval for endocrine response. Approximately 50% of dogs with pituitary macrotumors will develop neurologic signs secondary to the space-occupying effect of the mass, regardless of whether the dog has concurrent PDH. 17 This study demonstrated that dogs with larger tumors were more likely to have neurologic signs. Only 6 of the dogs in the treated group went on to develop new neurologic signs or had recurrence of neurologic signs. The progressionfree period for development of neurologic signs in the treated group of dogs with neurologic signs or in those dogs that initially did not have neurologic signs but went on to develop them was 994 days. There was also no significant difference in survival times between dogs with or without neurologic signs in the treated group. As expected, none of the dogs in the untreated group had improvement or resolution of neurologic signs. This finding indicates that RT can be effective in controlling and delaying the onset of neurologic signs in dogs with macrotumors. One study of a smaller number of dogs suggested that external beam RT reduces pituitary tumor size. 21 Another study that evaluated the effects of RT in 46 dogs with brain tumors and neurologic disease (13 dogs had pituitary tumors) also suggested that RT was helpful in reducing neurologic signs and prolonging survival. 14 This same study reported a median survival time of 1,308 days for dogs with pituitary masses, similar to the results of the present study. 14 Median survival was not reached in the treated group, despite the relatively long median follow-up time and high 1-,2-, and 3-year survival rates of 93% (95% CI, 63 99%), 87% (95% CI, 56 96%), and 55% (95% CI, 22 79%), respectively. This finding suggests that irradiation of pituitary masses in 3 Gy fractions on a Mondaythrough-Friday basis is efficacious, and results were similar to those previously reported. 1,14 It is intuitive, given the variation in size of different dog breeds, that tumor size relative to brain size is more likely to yield useful information than tumor size alone. Several published methods have been used to correct for variations in dog size when looking at pituitary glands; these include the ratio of the height of the pituitary gland to the brain area, the ratio of the pituitary area to the brain area, and the volume of the pituitary gland to the brain volume. 1,12 One of these studies found that the ratio of the pituitary area to the brain area was prognostic for survival in dogs with neurologic signs that were treated with RT. 1 Results of our study suggest that both pituitary-to-brain area and ratio of pituitaryto-brain height are prognostic for survival. Interestingly, none of the dogs with tumor heights, 25% of brain height or with tumors, 5% of the cross-sectional area of the brain died of their disease during the study period. Either of these measurements can be readily made at the time of imaging and appears to provide prognostic information. There was no difference in survival for dogs that had neurologic signs versus those without neurologic signs within the treated group. This finding gives additional evidence that RT effectively controls neurologic signs and indicates that even dogs with neurologic signs can have good initial and long-term response to therapy. In the untreated group, several of the possible prognostic factors tested were significant. As in the treated group, increasing pituitary-to-brain area and ratio of pituitary-to-brain height were negative prog-

6 1032 Kent et al nostic indicators. Also found to negatively and significantly impact survival were tumor height, having an endocrine-inactive tumor, and having a higher pre ACTH-stimulation serum cortisol concentration. Although statistically significant, having pituitary height, 25% of brain height or a tumor that was, 5% of the cross-sectional area of the brain no longer predicted survival for all cases as it did in the treated group. These findings are not surprising because they predict that untreated dogs with larger tumors die sooner from their disease. One limitation of this study was the lack of histologic evaluation of the pituitary masses. The pituitary location makes it particularly difficult to obtain samples for histopathology. This study, however, does provide useful information for clinicians and owners because most dogs will not have a histologic diagnosis for a mass in the area of the pituitary when decisions are made. Another limitation of this study is the lack of postmortem examination data, which would have allowed a better evaluation of late effects secondary to irradiation of the surrounding normal brain tissues. Furthermore, because endocrine testing was not done at regular intervals after RT, limited objective data are available regarding the endocrine response to RT. In order to compare survival duration of dogs in the 2 groups, we used the date of imaging for analysis. In most studies, the first or last day of treatment is used for this calculation. Because the median time to starting RT was 10 days, this factor did not have substantial impact on this study. A randomized, controlled clinical trial in which all dogs are examined postmortem would be required to further assess the benefits and risks of RT in dogs with pituitary masses. However, results of the present study demonstrate that RT for the treatment of pituitary macrotumors is effective. Furthermore, a dose of 48 Gy delivered in 3 Gy daily fractions is effective and safe, but additional studies would be needed to determine if lower dose-per-fraction protocols are safer and more effective. Although the radiation dose used in this study may not consistently relieve clinical signs associated with PDH, it was effective at controlling the neurologic signs that occurred because of the space-occupying effect of pituitary masses and led to increased survival. Footnotes a Clinac 4, Varian corporation, Palo Alto, CA b Prowess RTP 3000, Prowess Inc, Chico, CA c STATA Version 9.2, StataCorp, College Station, TX References 1. Theon AP, Feldman EC. Megavoltage irradiation of pituitary macrotumors in dogs with neurologic signs. J Am Vet Med Assoc 1998;213: Feldman EC, Nelson RW. Canine and Feline Endocrinology and Reproduction, 4th ed. St Louis, MO: WB Saunders; 2004: Voorhout G, Stolp R, Lubberink AA, et al. Computed tomography in the diagnosis of canine hyperadrenocorticism not suppressible by dexamethasone. J Am Vet Med Assoc 1988;192: Sarfaty D, Carrillo JM, Peterson ME. Neurologic, endocrinologic, and pathologic findings associated with large pituitary tumors in dogs: Eight cases ( ). J Am Vet Med Assoc 1988;193: Puente S. Pituitary carcinoma in an Airedale terrier. Can Vet J 2003;44: Meuten DJ. Tumors in Domestic Animals, 4th ed. Ames, IA: Iowa State Press; 2002:xii, Ihle SL. Pituitary corticotroph macrotumors: Diagnosis and treatment. Vet Clin North Am Small Anim Pract 1997;27: Turrel JM, Fike JR, LeCouteur RA, et al. Computed tomographic characteristics of primary brain tumors in 50 dogs. J Am Vet Med Assoc 1986;188: Bertoy EH, Feldman EC, Nelson RW, et al. One-year followup evaluation of magnetic resonance imaging of the brain in dogs with pituitary-dependent hyperadrenocorticism. J Am Vet Med Assoc 1996;208: Katznelson L, Bogan JS, Trob JR, et al. Biochemical assessment of Cushing s disease in patients with corticotroph macroadenomas. J Clin Endocrinol Metab 1998;83: Voorhout G. Cisternography combined with linear tomography for visualization of the pituitary gland in healthy dogs: A comparison with computed tomography. Veterinary Radiology 1990;31: Kooistra HS, Voorhout G, Mol JA, et al. Correlation between impairment of glucocorticoid feedback and the size of the pituitary gland in dogs with pituitary-dependent hyperadrenocorticism. J Endocrinol 1997;152: Meij B, Voorhout G, Rijnberk A. Progress in transsphenoidal hypophysectomy for treatment of pituitary-dependent hyperadrenocorticism in dogs and cats. Mol Cell Endocrinol 2002;197: Bley CR, Sumova A, Roos M, et al. Irradiation of brain tumors in dogs with neurologic disease. J Vet Intern Med 2005;19: Heidner GL, Kornegay JN, Page RL, et al. Analysis of survival in a retrospective study of 86 dogs with brain tumors. J Vet Intern Med 1991;5: Axlund TW, McGlasson ML, Smith AN. Surgery alone or in combination with RT for treatment of intracranial meningiomas in dogs: 31 cases ( ). J Am Vet Med Assoc 2002;221: Hanson JM, van t HM, Voorhout G, et al. Efficacy of transsphenoidal hypophysectomy in treatment of dogs with pituitary-dependent hyperadrenocorticism. J Vet Intern Med 2005;19: Mauldin GN, Burk RL. The use of diagnostic computerized tomography and RT in canine and feline hyperadrenocorticism. Probl Vet Med 1990;2: Dow SW, LeCouteur RA, Rosychuk RAW, et al. Response of dogs with functional pituitary macroadenomas and macrocarcinomas to radiation. J Small Anim Pract 1990;31: Brearley MJ, Jeffery ND, Phillips SM, et al. Hypofractionated RT of brain masses in dogs: a retrospective analysis of survival of 83 cases ( ). J Vet Intern Med 1999;13: Goossens MM, Feldman EC, Theon AP, et al. Efficacy of cobalt 60 radiotherapy in dogs with pituitary-dependent hyperadrenocorticism. J Am Vet Med Assoc 1998;212: Nelson RW, Ihle SL, Feldman EC. Pituitary macroadenomas and macroadenocarcinomas in dogs treated with mitotane for pituitary-dependent hyperadrenocorticism: 13 cases ( ). J Am Vet Med Assoc 1989;194:

7 Irradiation of Pituitary Masses Moore TJ, Dluhy RG, Williams GH, et al. Nelson s syndrome: Frequency, prognosis and effect of prior pituitary irradiation. Ann Intern Med 1976;85: Benitah N, Feldman EC, Kass PH, et al. Evaluation of serum 17-hydroxyprogesterone concentration after administration of ACTH in dogs with hyperadrenocorticism. J Am Vet Med Assoc 2005;227: Feldman EC, Nelson RW. Canine and Feline Endocrinology and Reproduction, 4th ed. St Louis, MO: WB Saunders; 2004: Kipperman BS, Feldman EC, Dybdal NO, et al. Pituitary tumor size, neurologic signs, and relation to endocrine test results in dogs with pituitary-dependent hyperadrenocorticism: 43 cases ( ). J Am Vet Med Assoc 1992;201: Granger N, de Fornel P, Devauchelle P, et al. Plasma proopiomelanocortin, pro-adrenocorticotropin hormone, and pituitary adenoma size in dogs with Cushing s disease. J Vet Intern Med 2005;19: Duesberg CA, Feldman EC, Nelson RW, et al. Magnetic resonance imaging for diagnosis of pituitary macrotumors in dogs. J Am Vet Med Assoc 1995;206: van der Vlugt-Meijer RH, Meij BP, Voorhout G. Dynamic computed tomographic evaluation of the pituitary gland in healthy dogs. Am J Vet Res 2004;65: van der Vlugt-Meijer RH, Meij BP, van den Ingh TS, et al. Dynamic computed tomography of the pituitary gland in dogs with pituitary-dependent hyperadrenocorticism. J Vet Intern Med 2003;17: