Incidence of gastric dilatation-volvulus following a splenectomy in 238 dogs

Abstract

There is contradicting information in the veterinary literature regarding canine splenectomy and the increased risk for subsequent gastric dilatation-volvulus. The main purpose of this study was to determine the rate of occurrence of gastric dilatation-volvulus following splenectomy in medium to large breed dogs compared with a control group undergoing other abdominal procedures. Follow-up was performed by reviewing the medical records and conducting phone interviews. Weight, gender, and presence of a hemoabdomen at the time of surgery were not significantly associated with occurrence of gastric dilatation-volvulus, while increasing age was. Ten of 238 (4%) dogs in the splenectomy group and 3/209 (1.4%) dogs in the control group subsequently developed gastric dilatation-volvulus, which was not significantly different (P = 0.08). While the findings approach significance and support a need for future investigation, the current recommendation for gastropexy at time of splenic removal should be made on a case by case basis and while considering previously documented risk factors.

Introduction

In veterinary medicine, total splenectomies have been performed for treatment of benign or malignant neoplasia, severe trauma, certain immune-mediated diseases, infiltrative disease, and splenic torsion (1). Reported complications following splenectomy include cardiac arrhythmias, damage to the left pancreatic lobe, hemorrhage, disseminated intravascular coagulation, sepsis/infection, and gastric dilatation-volvulus (GDV) (1–4). Of these complications, GDV would theoretically be one of the most preventable by performing a prophylactic gastropexy at the time of splenectomy (5,6). Prevention of GDV would be advantageous, as this disease can result in significant morbidity and mortality for the patient, as well as increased financial and emotional burden to the owner.

Gastric dilatation and volvulus occurs when the pylorus is malpositioned and there is severe gastric distension with gas, food, and/or fluid, resulting in secondary life-threatening complications (7–9). Genetics, breed, food texture, ingredients or size, feeding protocol, thoracic wall dimensions, emotional state, increased hepatogastric ligament length, age, foreign body, time of year, and previous splenectomy have all being proposed as contributing factors to GDV (10–19). Treatment is immediate medical and surgical intervention to stabilize the patient, return the stomach to its normal position, remove devitalized or necrotic tissue and secure the gastric antrum to the right side of the body wall to prevent recurrence (7,9,20,21).

Despite conflicting evidence in the literature regarding the association of splenectomy and subsequent GDV, it has been recommended to perform a gastropexy at the time of splenectomy regardless of breed, age, and underlying disease (1,2,21). Negative consequences following a gastropexy occur at a relatively low frequency and include pneumothorax, accidental penetration into the gastric lumen resulting in suture contamination, increased surgical and anesthetic time, increased financial costs, and possible changes in gastric motility (9,21,22).

The purpose of this study was to determine if the incidence of GDV following splenectomy in medium to large breed dogs differs significantly in comparison to a control group of medium to large breed dogs undergoing an exploratory laparotomy for procedures unrelated to splenic disease. We also aimed to define the time from surgery to occurrence of GDV in affected dogs, and determine if there was an association of age, weight, gender, or presence of a hemoabdomen at time of surgery with occurrence of GDV. Our hypothesis was that there would not be a significant difference in the incidence of GDV between the 2 groups.

Materials and methods

The medical records database at Ocean State Veterinary Specialists was searched for dogs weighing more than 20 kg which had undergone surgery with a vessel sealing or stapling device from 2008 to 2015. Dogs were excluded if a splenectomy was not performed, the dog had history of a GDV, the spleen was not submitted for histopathology, if the dog did not survive the 10 to 14 d until removal of staples, if the medical record was incomplete following surgery, or if the owner or primary care veterinarian could not be reached for follow-up. Dogs were also excluded if a prophylactic gastropexy was performed at any time before, during, or after the splenectomy. Dogs undergoing other procedures at the time of splenectomy were included.

For the control group, the database was searched for dogs that weighed more than 20 kg at adult weight that had undergone an emergency laparotomy for reasons other than treatment for splenic disease, dystocia, pyometra, or GDV from 2008 to 2015. Dogs were excluded if a splenectomy and/or prophylactic gastropexy were performed at any time before, during, or after the emergency laparotomy. Dogs which had abdominal adhesions noted at the time of the initial surgery, had a gastric feeding tube placed, had an abdominal mass larger than 5 cm, and/or had undergone more than 1 surgery for foreign body removal at any time were also excluded. Lastly, dogs were excluded if they did not survive 10 to 14 d until removal of the staples, their medical record was incomplete following surgery, or if the owner or primary care veterinarian could not be reached for follow-up.

For the splenectomy group, medical records were reviewed and data were collected regarding signalment (breed, age, weight, and neuter status at time of surgery), the reason for surgery, presence of a hemoabdomen, type of surgical procedures performed, size of splenic mass, histopathologic diagnosis, co-morbidities and administration of postoperative chemotherapy. For the control group, medical records were reviewed and data were collected regarding signalment (breed, age, weight, and neuter status at time of surgery), the reason for surgery, type of surgical procedures performed, and histopathology or stone analysis when appropriate. For both groups, if the dog was spayed or neutered at the time of surgery, then the intact status was changed to reflect this.

Case follow-up was performed by reviewing medical records, and conducting either an e-mail or telephone interview with the owner and/or the primary care veterinarian. Owners and veterinarians were asked if the dog had any complications following surgery, if the dog had any additional abdominal surgeries, and more specifically if the dog ever displayed clinical signs and/or had been diagnosed with a GDV. The date of occurrence and outcome were noted when indicated. If the dog had not displayed any signs related to a GDV it was confirmed that the dog had never undergone a prophylactic gastropexy. If the animal was no longer alive the owners and veterinarians were asked the date and cause of death. If the cause of death was suspected, but not confirmed, to be a GDV based on the clinical signs then it was considered a GDV.

Statistical analysis

Dogs were divided into control (n = 209) and splenectomy (n = 238) groups. Within those groups dogs were further subdivided into GDV (yes/no). Data were tested for normality (Shapiro-Wilk test and normal probability plot); data not normally distributed were analyzed by means of Wilcoxon rank-sum test. Comparisons between age, gender/neuter status, histopathology results, presence of hemoabdomen at the time of original surgery, survival/follow-up times, and occurrence of GDV were evaluated using the Chi-square test or Wilcoxon rank-sum test. Size of the splenic mass, administration of chemotherapy agents and co-morbidities were not evaluated due to inconsistency in the medical record. Breed was not evaluated for association because there were inadequate numbers of some breeds for proper analysis. Survival times were reported using Kaplan-Meier curve. Data were reported either as mean +/− standard deviation or median (25th and 75th quartiles), as appropriate. Data were analyzed using computer software (SAS 9.3 software; SAS, Cary, North Carolina, USA) and differences were considered significant when P < 0.05.

Results

Overall 238 dogs met the inclusion criteria and were included in the splenectomy group, while 209 dogs met the inclusion criteria and were included in the control group. The splenectomy group included 32 breeds: mixed breeds (n = 58), Labrador retriever (n = 49), golden retriever (n = 34), German shepherd (n =11), basset hound (n = 11), Staffordshire bull terrier (n = 9), rottweiler (n = 8), Weimaraner (n = 7), English bulldog (n = 7), boxer (n = 6), springer spaniel (n = 5), Siberian husky (n = 3), standard poodle (n = 3); 2 each of Airedale, Alaskan malamute, beagle, Bernese mountain dog, Australian cattle dog, English springer, Portuguese water dog, vizsla; and 1 each of American bulldog, American eskimo, border collie, bullmastiff, bearded collie, Doberman pinscher, flat-coated retriever, greyhound, Irish setter, Rhodesian ridgeback, and St. Bernard. The control group included 35 breeds: Labrador retriever (n = 54), mixed breeds (n = 36), golden retriever (n = 34), English bulldog (n = 14), Staffordshire bull terrier (n = 11), boxer (n = 8), Siberian husky (n = 5), Bernese mountain dog (n = 4), Dalmatian (n = 4), Airedale (n = 3), American bulldog (n = 3), Doberman pinscher (n = 3), Plott hound (n = 2), vizsla (n = 2); and 1 each of Alaskan malamute, basset hound, beagle, Australian cattle dog, border collie, bullmastiff, Cane corso mastiff, English springer, flat-coated retriever, German wirehaired pointer, German shepherd, standard poodle, Portuguese water dog, samoyed, greyhound, Irish setter, mastiff, redbone coonhound, Rhodesian ridgeback, Rottweiler, and springer spaniel.

There were 110 females (2 intact), 118 males (8 intact) in the splenectomy group. The control group consisted of 90 females (5 intact), and 119 males (16 intact). There was no significant difference in gender distribution between groups (P = 0.47).

The mean weight at time of surgery was 32.4 kg (range: 21.2 to 56.3 kg) for the splenectomy group and 30.6 kg (range: 12.4 to 55.5 kg) for the control group (Table 1). The mean age at time of surgery was 10.1 y (range: 4.8 to 15.0 y) for the splenectomy group and 5.6 y (range: 4 mo to 14.7 y) for the control group (Table 1). There was a significant difference for both weight (P = 0.01) and age (P < 0.01) between the 2 groups.

Table 1.

Age, gender distribution, and weight for dogs in the splenectomy and control groups. Age of dogs in the control group was a non-normal value and is reported as median value (25th; 75th quartiles). All other values were normal values and reported as mean +/− standard deviation. P-values represent intergroup difference between dogs with and without GDV using Wilson rank-sum test.

Group	Mean age (y) ± SD	Gender	Weight (kg)
Splenectomy (n = 238)	10.1 +/− 2.0	F = 2; M = 18 S = 108; N = 110	32.4 +/− 7.6
No GDV (n = 228)	10.1 +/− 2.0	F = 2; M = 18 S = 103; N = 105	32.4 +/− 7.6
GDV (n = 10)	11.0 +/− 1.8	F = 0; M = 0 S = 5; N = 5	32.4 +/− 7.0
P-value	0.14	0.81	0.97
Control (n = 209)	6.0 (2.0; 8.4)	F = 4; M = 14 S = 51; N = 71	30.6 +/− 7.1
No GDV (n = 206)	6 (2.0; 8.4)	F = 4; M = 14 S = 51; N = 65	30.6 +/− 7.2
GDV (n = 3)	5.4 (4.3; 11.3)	F = 0; M = 0 S = 0; N = 3	30.9 +/− 5.1
P-value	0.51	0.61	0.86

SD — standard deviation of the mean; F — female; M — male; S — female spayed; N — male neutered.

The reason for splenectomy was a splenic mass or nodule (n = 227, 95.3%), splenomegaly (n = 8, 3.4%) and an abnormal splenic appearance on diagnostic imaging (n = 3, 1.3%). No dogs underwent surgery for splenic torsion. Histopathology results revealed a benign underlying disease process for 130 (54.6%) dogs and a malignant neoplastic process for 108 dogs (45.4%). Of the dogs with a benign process, histopathology identified nodular hyperplasia (n = 94, 72.3%), congestion and extramedullary hematopoiesis (n = 14, 10.7%), hematoma (n = 12, 9.2%), unknown benign process (n = 3, 2.3%), diffuse hyperplasia (n = 2, 1.5%), and splenic necrosis (n = 2, 1.5%). Hemangioma, infarction, and myleolipoma were reported in 1 dog each with a benign process. Of the 108 dogs with a malignant neoplastic process, histopathology results included hemangiosarcoma (n = 87, 80.5%), fibrohystiocytic nodule (n = 6, 5.6%), spindle cell sarcoma (n = 4, 3.7%), lymphoma (n = 3, 2.8%), and sarcoma (n = 3, 2.8%). Leiomyosarcoma, liposarcoma, mast cell tumor, neuroendocrine and tubular adenocarcinoma were reported in 1 dog each with a malignant process. The reason for surgery in the control group was gastrointestinal disease (n = 166, 79.4%), urinary disease (n = 25, 12%), reproductive disease (n = 8, 3.8%), hepatobiliary disease (n = 6, 2.9%), and endocrine disease (n = 4, 1.9%).

Of the 238 dogs undergoing a splenectomy 119 (50%) had a hemoabdomen at the time of surgery and 119 (50%) dogs did not. Of the 119 dogs with a hemoabdomen, 72 (60%) had a diagnosis of malignant neoplasia on histopathology and this association was significant (P < 0.01). Of the 119 dogs that did not have a hemoabdomen, 36 (30%) dogs had malignant neoplasia and 83 (70%) had a benign process.

Overall, 10 dogs (4%) in the splenectomy group and 3 dogs (1.3%) in the control group were considered to have had a GDV following surgery (Tables 2 and 3). Results of a Chi-square test showed there was no significant difference in incidence of GDV between the 2 groups (P = 0.08). Of the splenectomy group, 8 dogs (3.3%) had a confirmed GDV and 2 dogs (0.08%) died with a distended abdomen, for which GDV could not be ruled out (Table 2). Of the control group 2 dogs (1%) had confirmed GDV and 1 dog (0.05%) died with a distended abdomen, for which GDV could not be ruled out (Table 3). Two dogs in the splenectomy group had 1 episode each of gastric dilatation without volvulus confirmed on radiographs, and did not go on to develop GDV.

Table 2.

Detailed information for dogs that underwent splenectomy and had a subsequent GDV.

Dog	Breed	Gender	Weight (kg)	Age (y)	Hemoabdomen	Splenic mass (cm)	Histopathology	Time to GDV (d)	Status	Follow-up/survival time (d)	Cause of death
1	American pit bull	FS	26.5	11.7	Yes	6 × 4	Nodular hyperplasia (B)	130	A	328	NA
2	Doberman pinscher	FS	31.4	9.1	No	3.5 to 4	Nodular hyperplasia (B)	371	A	371	NA
3	Mixed breed	FS	22.4	12.0	Yes	6 × 8	Hemangiosarcoma (N)	74	D	74	Unknowna
4	Mixed breed	FS	29.5	12.1	No	15	Nodular hyperplasia (B)	180	D	180	GDV
5	German shepherd	FS	30.3	13.2	Yes	> 20	Spindle cell sarcoma (N)	15	D	15	Unknowna
6	Mixed breed	MN	31.2	7.8	No	15	Fibrohistiocytic stromal tumor (N)	136	D	136	GDV
7	Mixed breed	MN	31.6	12.5	Yes	30	Hematoma (B)	94	D	656	Unknown
8	Labrador retriever	MN	35.0	10.9	Yes	> 20 to 30	Nodular hyperplasia (B)	54	D	898	Unknown
9	Labrador retriever	MN	39.1	11.6	Yes	10	Hemangiosarcoma (N)	118	D	283	Cancer
10	Weimaraner	MN	46.6	8.9	No	> 15	Unknown (B)	1273	D	1273	GDV

F — female; M — male; S — spayed; N — neutered; (B) — benign; (N) — neoplastic; D — Deceased; A — Alive; GDV — gastric dilation and volvulus.

^aPossible GDV.

Table 3.

Detailed information for dogs in the control group that had a subsequent GDV.

Dog	Breed	Gender	Weight (kg)	Age (y)	Surgery/Diagnosis	Time to GDV (d)	Status	Survival time (d)	Cause of death
11	English springer	MN	25	11.3	Gastrotomy/Linear FB	570	D	1350	Paralysis
12	Boxer	MN	33.7	5.4	Enterotomy/Jejunal FB	1663	D	1663	GDV
13	Dalmatian	MN	33.9	4.3	Cystotomy/CaOx stones	1029	D	1029	Suspect GDV

F — female; M — male; S — spayed; N — neutered; FB — foreign body; D — Deceased; GDV — gastric dilation and volvulus; CaOx — calcium oxalate bladder stones.

Overall median time from surgery to subsequent GDV for all dogs was 439 d (range: 15 to 1663 d). Median time from surgery to subsequent GDV was 124 d (range: 15 to 1273 d) in the splenectomy group and 1029 d (range: 570 to 1663 d) in the control group. This difference was significant (P = 0.05).

When evaluating both groups together the mean age and weight for all the dogs that developed a subsequent GDV was 10 y (range: 4.3 to 13.2 y) and 32 kg (range: 22.4 to 46.6 kg), while the mean age and weight for the dogs which did not develop subsequent GDV was 8.0 y (range: 4 mo to 15 y) and 21.6 kg (range: 12.4 to 56.3 kg). The difference in age was significant, and found to be associated with GDV (P = 0.02), while weight was not (P = 0.75). However, when evaluating GDV occurrence with each group separately, neither age, gender, or weight at time of surgery was found to be significantly associated with subsequent GDV (Table 1). Within the splenectomy group 6/10 dogs with subsequent GDV had a hemoabdomen, and this was not significant (P = 1) (Table 2).

At time of follow-up, 48 dogs (20%) were alive and 190 (80%) dogs were deceased in the splenectomy group, and 144 (69%) dogs were alive and 65 (31%) dogs were deceased in the control group. For the splenectomy group, causes of death included splenic neoplasia (n = 80), nonspecific signs/old age (n = 39), unknown cause of death that did not involve a distended abdomen (n = 23), neurologic disease (n = 14), non-splenic neoplasia (n = 13), heart, liver or kidney disease (n = 6), respiratory disease (n = 5), GDV (n = 3), hemoabdomen (n = 3), and possible GDV (n = 2). One dog arrested during surgery for removal of a foreign body, and 1 dog was hit by a car. For the control group, causes of death included neoplasia (n = 35), unknown cause that did not involve a distended abdomen (n = 9), nonspecific signs/old age (n = 9), neurologic disease (n = 5), kidney disease (n = 2), trauma (n = 2), heart disease (n = 1), possible GDV (n = 1) and 1 dog was euthanized due to a GDV.

Overall median survival time for the splenectomy group was 394 d (range: 73 to 1168 d) and for the control group was 2494 d (range: 23 to 2849 d), which was significantly different (P < 0.01). Median follow-up for those surviving in the splenectomy group was 471 d (range: 19 to 1688 d) and in the control group was 1184 d (range: 214 to 2776 d) which was significantly different (P < 0.01).

Median survival time for splenectomized dogs with neoplasia was 102 d (range: 35 to 377 d) compared to 738 d (range: 303 to 1454 d) for benign splenic disease. Median survival time for dogs with a hemoabdomen at time of surgery was 265 d (range: 51 to 898 d) compared to 518 d (range: 101 to 1330 d) for those without a hemoabdomen. Both splenic neoplasia (P < 0.01) and presence of a hemoabdomen (P = 0.01) were significantly associated with a shorter survival time.

Discussion

It has been theorized that splenic disease may cause stretching of the perigastric ligament from expansion secondary to a growing mass and/or hemoperitoneum, which may permit increased gastric mobility (4,7,23,24). Removal of the spleen could result in increased intra-abdominal dead space, allowing the stomach to twist on its longitudinal axis (4,7,23). In the present study, the rate of GDV occurrence was 4% (10/238) for the splenectomy group and 1.4% (3/209) for the control group, which appeared clinically relevant but was not statistically different. We also did not find an association between presence of a hemoabdomen at time of surgery and GDV, which is consistent with previous findings (25). If stretching of the ligament and creation of dead space are underlying contributing factors, then we would expect GDV to also occur at a higher incidence following treatment for dystocias, pyometras, and large liver mass resections. While this is not commonly noted or discussed in the literature, dogs undergoing abdominal surgeries for these procedures were excluded from both groups to avoid confounding results.

We considered dogs that died with a distended abdomen from an unknown cause to have had a GDV to avoid potentially underreporting the overall incidence of GDV. In total, 2 dogs from the splenectomy group (dogs 3, 5) and 1 dog from the control group (dog 13) fell into this category. It is possible that the distension was from another cause, such as a hemoabdomen. If these cases were not a GDV then overall incidence would have been 8/238 for the splenectomy group and 2/209 for the control group, which was still not significantly different (P = 0.09).

There are 2 previous studies that also did not find an association between splenectomy and GDV (25,26), while 1 study and 2 case reports documented an association (3,4,27). Of the 2 studies that found no association between splenectomy and GDV, 1 study included 37 dogs with a follow-up time of 1 y (26), while the other evaluated 172 dogs, of which 88/168 (52%) with histopathology performed were diagnosed with neoplasia and had a median survival time of 83 d (25). The limitation of both is that there may not have been adequate time to allow for subsequent GDV, making it harder to effectively evaluate the risk of GDV following a splenectomy.

The most recent study of 151 dogs by Sartor et al (27) did find a significant association between GDV and splenectomy; they noted that dogs with a GDV were 5.3 times more likely to have had a previous splenectomy compared with a control group. Unlike the 2 previously mentioned studies, in Sartor et al (27) the control group was age, weight, and breed matched, but included dogs that did not undergo a laparotomy. It is also harder to draw a definitive conclusion concerning the association between GDV and splenectomy because all dogs undergoing splenectomy were not evaluated, as association was found by looking back at the incidence of prior splenectomy in dogs known to have a GDV. Lastly, 2 previous case reports described GDV following splenectomy (3,4). They reported occurrence of GDV between 2 d and 17 mo after surgery in a mixed breed, Great Dane and German shepherd. Given that 2 of the 3 dogs were at-risk breeds, it is difficult know if a GDV would have occurred in these dogs regardless of splenectomy, or if removal of the spleen increased the risk for subsequent GDV.

There was no significant difference in gender distribution between our groups, but there was significant difference in both mean age and weight of dogs at the time of surgery. We suspect the difference in age was due to the inclusion criteria for each group, as most dogs in our splenectomy group were undergoing surgery for suspected splenic neoplasia, while in our control group most dogs were undergoing surgery for a gastrointestinal foreign body removal. The difference in weight was because juvenile dogs < 20 kg at time of surgery were included in the control group if they would be at least 20 kg as adults.

The overall significance of differences in both age and weight, and how they would affect accuracy when comparing GDV rates is unknown. Grange et al (25) found that age and weight were not associated with GDV occurrence. However, our results showed that while weight was not significantly associated with subsequent GDV, age was significantly associated with GDV when evaluating both groups together. Age was not significantly associated when each group was evaluated separately. Since the groups differed significantly in their age distribution, the authors felt the association of age and GDV occurrence would be most accurate when evaluating the total population of dogs. Furthermore, previous studies have also documented an association with both age and weight (10,16–18,28,29) and occurrence of GDV. It is plausible that GDV may be linked to a degenerative or age related process. However, since our splenectomy group was significantly older and may have been placed at an increased risk for GDV with removal of the spleen, it is possible that age is merely a correlation, and not causation. Also, our inclusion criteria for the splenectomy group selected for dogs that had not developed a GDV at a younger age. Including younger dogs in the control group allowed for both a longer follow-up time and for comparison between the splenectomy group and dogs undergoing surgery for other reasons. In regard to weight, deBattisti et al (28) found an increased occurrence of GDV with a lower body weight, while Bredal (29) found an increased occurrence of GDV with a higher body weight. However, just as in our study, body condition score was not noted. Future studies are needed to further define the risk of age, weight, and body condition score in relation to GDV.

Breed has been shown to be associated with GDV occurrence (18,30–32). In our study, we elected to not analyze breed, as many breed categories had too few numbers. The largest breed categories in both study and control groups were mixed-breed dogs, Golden retrievers, and Labrador retrievers. It is interesting to note that our splenectomy group contained a larger number of German shepherds, Weimaraners, and basset hound dogs, which are considered at risk breeds for GDV (30–32) and could influence GDV occurrence compared with the control group. However, the dogs that subsequently developed GDV in the splenectomy group were primarily mixed-breed or Labrador retrievers (Table 2).

The median time from surgery to subsequent GDV for dogs in the splenectomy group was significantly shorter than for the control group. This difference is the result of the control group having a significantly younger population with both longer survival and follow-up times, which would then allow more time for development of GDV. We also observed that although the median time from surgery to occurrence of GDV was longer than the survival time for dogs with splenic neoplasia, a prophylactic gastropexy should still be considered in those dogs which belong to an at-risk breed because GDV was documented to occur as quickly as 15 d after splenectomy (Dog 5). This is consistent with other studies that reported intervals of 2 d and 12 d between splenectomy and subsequent GDV (4,25).

The limitations of this study are mainly related to its retrospective nature. Most importantly we relied on the medical records to be both accurate and complete. If owners were unsure if a gastropexy had been performed prior to surgery, the surgery report was then referenced. It is possible that presence of an intact gastropexy site may have been omitted from a surgery report. This would have falsely decreased our overall occurrence of GDV. We were also limited in the details we could obtain from the medical record, and therefore were unable to control for other GDV risk factors, such as anatomy, environment, and diet. Our control group was not weight and age matched, which are factors that may influence both the rate and timing of occurrence of GDV. Lastly, follow-up was performed by contacting owners, which sometimes required them to recall details from up to 7 years earlier. It is possible that some owners may have incorrectly recalled relevant details regarding history of gastropexy, cause of death, and time of death, all of which could have confounded our findings.

In conclusion, we did not find gender, weight, or the presence of a hemoabdomen to be significantly associated with occurrence of GDV. Age was significantly associated with subsequent GDV when both groups were evaluated together, but was not significant when evaluated separately. Presence of hemoabdomen at time of surgery and splenic neoplasia were associated with a shortened median survival time. The difference in occurrence of GDV between dogs that underwent a splenectomy and dogs that underwent exploratory laparotomy for other reasons appeared clinically relevant, but was not statistically significant. Given the limitations of our study, a larger prospective study would be helpful in clarifying these issues. Current recommendation for gastropexy at time of removal of the spleen should be made on a case-by-case basis while taking into consideration previously documented risk factors.