Brain tumors cause mechanical destruction of brain tissue and secondary alterations in intracranial physiology, reducing relative brain volume and resulting in physical invasion and destruction of neurons, metabolic alterations in neurons or glial cells, damage to the vascular supply, impairment of autoregulation, irritation, edema formation, and increased ICP.1 Currently, there is a paucity of objective information about optimal treatment for dogs with intracranial tumors, and data that are available are limited to results of retrospective studies and studies that combined intra- and extra-axial tumors or involved small numbers of cases.2–4 Dogs undergoing palliative treatment (ie, treatment with corticosteroids or AEDs alone or in combination) for intracranial tumors have a poor prognosis, with a median survival time of 69 days (range, 18 to 201 days) from the time of diagnosis.5 Furthermore, median survival time was not significantly improved with chemotherapy (93 days), compared with survival time following palliative treatment (60 days).6 Radiation treatment for intracranial tumors has been associated with a median survival time of 351 days (range, 139 to > 900 days), and surgical treatment has been associated with a median survival time of 312 days (range, 27 to 2,104 days).2 However, reported median survival times vary substantially,2,7–11 likely because of differences in types and grades of tumors being treated as well as surgical techniques and equipment used.3
In human medicine, the standard of care for treatment of primary intracranial tumors is surgical cytoreduction alone or in combination with adjunctive treatments, depending on the type of tumor.12–15 Complete surgical removal is attempted, but within the constraints of preserving the patient's neurologic function and overall health. Also, in humans, the gross total resection (as assessed with postoperative MRI) of primary brain tumors (eg, glioma and meningioma) has been associated with longer survival times than has incomplete resection.16,17
Intra- and postoperative complications can greatly impact patient recovery and quality of life. In human medicine, a widely adopted definition of postoperative neurosurgical complications is any deviation from the normal postoperative course occurring within 30 days after surgery.18 Complications following surgical cytoreduction of brain tumors in people have been investigated and reported in detail, with the incidence of complications during the 30 days after surgery ranging from 14% to 37% and the overall death rate ranging from 1.2% to 4.3%.18–23 Risk factors identified for complications in human patients have included tumor location, previous radiotherapy, reoperation, older age, and severe concomitant disease.19,24
There is currently little information concerning specific postoperative complications and associated risk factors for complications in dogs undergoing cytoreductive surgery of intracranial masses.2,25 Complications that have been reported are limited to findings in case reports and small case series and include aspiration pneumonia,26 tension pneumocephalus,27,28 and cerebral and cerebellar artery thrombosis.9 The prevalence of aspiration pneumonia as a complication in dogs following craniotomy has been reported to be 24%, with presurgical vomiting, regurgitation, and megaesophagus identified as significant risk factors for this complication.26 Further, a 100% death rate within hours after surgery has been reported for dogs that developed cyclic hyperthermia and hypernatremia syndrome after resection of diencephalic tumors.29 Unfortunately, most reports on surgical management of brain tumors in veterinary medicine focus on long-term outcome, with cases censored if death was unrelated to tumor recurrence, adding to the paucity of information regarding incidence and types of surgical complications after craniotomy in animals. Thus, the purposes of the study reported here were to determine the incidence of and risk factors for major complications occurring in dogs within 30 days after cytoreductive surgery performed by a single pair of surgeons for treatment of suspected primary intracranial masses.
Materials and Methods
Case selection
Medical records of the University of Minnesota Veterinary Medical Center for January 2009 through December 2015 were searched to identify all dogs that underwent cytoreductive surgery performed by the same pair of surgeons (GEP, MAH) for treatment of a solitary intracranial mass identified by advanced diagnostic imaging and suspected to be a primary tumor of the brain. Dogs were eligible for inclusion in the study if a complete medical record and follow-up information for at least 30 days after surgery (or until the time of death for dogs that died within 30 days after surgery) were available.
Medical records review
Data collected from the medical records included patient signalment (age at the time of surgery, breed, and sex); original reason for examination as reported by the owner; results of neurologic examination (normal or abnormal), diagnostic imaging, preanesthetic clinicopathologic analyses, and histopathologic evaluation; date of surgery; duration from onset of clinical signs to surgery (categorized as < 4 weeks, 4 to 12 weeks, and > 12 weeks); and any treatments administered prior to surgery (categorized as no treatment, prednisolone alone, AEDs alone, and prednisolone combined with AEDs). Surgical, anesthetic, and intensive care unit records were reviewed and evaluated for complications, including death.
Diagnostic imaging
Characteristics of solitary brain masses evident by diagnostic imaging (MRI or CT) were recorded and included intra- vs extra-axial location, right vs left side, associated lobe or region, degree and quality of gadolinium enhancement, presence of associated cysts, presence and severity of associated edema, and signs of herniation. Masses were categorized on the basis of associated lobe or region (ie, the olfactory and frontal lobes, temporal and parietal lobes, occipital lobe, cerebellar region, brainstem region, or interthalamic adhesion). For dogs that underwent postoperative MRI to assess extent of resection of the mass, the extent of resection was categorized as gross total resection if no residual contrast enhancement was evident for tumors that were gadolinium enhancing or if no T2 fluid attenuating inversion recovery signal abnormality was evident for non-contrast-enhancing tumors, subtotal resection if ≤ 20% of the preoperative lesion remained evident, or partial resection if > 20% of the preoperative lesion remained evident.
Anesthesia and surgery
The anesthetic protocol was determined by the individual anesthesiologist on the basis of each dog's preoperative status. All dogs received cefazolin sodium (22 mg/kg [10 mg/lb], IV) 30 minutes prior to surgery and every 90 minutes during surgery. After placement of a urinary catheter, mannitol (1 g/kg [0.45 g/lb], IV), followed by furosemide (1 mg/kg, IV), then methylprednisolone sodium succinate (30 mg/kg [13.6 mg/lb], IV), were given for diuresis and to reduce cerebral edema and ICP prior to surgery.30 Dogs were maintained in sternal recumbency with their head elevated to prevent occlusion of the jugular veins throughout the procedure and during recovery from anesthesia.
Dogs with a brainstem or cerebellar mass underwent craniectomy, whereas dogs with a supertentorial intracranial mass underwent craniotomy. The surgical approaches used were bilateral transfrontal, right or left lateral rostrotentorial, and suboccipital. Surgical approaches were determined on the basis of lesion location. For the transfrontal and rostrotentorial approaches, a bone flap was created with a saline (0.9% NaCl) solution-cooled high-speed burr to mark the corners of the flap followed by use of a saline solution-cooled oscillating saw to create the flap, which was replaced in all craniotomies. A saline solution-cooled high-speed burr and rongeurs were used to remove bone from the floor of the frontal sinuses for the transfrontal approach and from the caudal aspect of the skull for the suboccipital approach and to extend the ventral margin of the craniotomy for the rostrotentorial approach. Abnormal tissue was identified then excised by blunt dissection with Penfield elevators, pituitary rongeurs, and suction until normal parenchyma was seen at the margins of the resection cavity. Hemostasis was achieved with bipolar electrocautery, bone wax, and sterile compressed absorbable gelatin sponges.a The dura was apposed whenever possible.
Hospitalization
All dogs received maintenance IV fluid therapy with lactated Ringer solution supplemented with KCl (16 mEq/L) and pain control with hydromorphone hydrochloride (0.05 mg/kg [0.023 mg/lb], IV) on an as-needed basis during hospitalization. Dogs for which treatment with an AED had been initiated prior to surgery continued to receive the same medication for a minimum of 30 days after surgery, whereas the corticosteroid dosages were tapered and discontinued over 10 to 14 days in all dogs. The morning after surgery, PCV and serum total protein concentration were assessed. The duration of hospitalization was recorded. Prior to hospital discharge, owners were specifically asked to contact the surgeon if complications were observed at home or if the pet died or was euthanized for any reason.
Follow-up information
Medical records were reviewed for follow-up information recorded from the time of hospital discharge to 30 days after surgery to determine whether any complications occurred in that timeframe. Follow-up information consisted of data from recheck examinations at the University of Minnesota Veterinary Medical Center, medical record information provided by referring veterinarians, and telephone or electronic communications with owners. In addition, if an owner indicated that a pet was examined for a complication or died during the 30 days after surgery, the review extended to medical records of the referring veterinarian as well as logged telephone conversations and email communications between the surgeon and the referring veterinarian or owner.
Postoperative complications
Major complications, including death, were recorded. A major postoperative complication was defined as an unexpected event requiring medical attention or directly affecting the animal in a negative manner. Timing of the complication was recorded as occurring during surgery, during hospitalization, or after discharge but ≤ 30 days after surgery. Deterioration of neurologic status after surgery, compared with neurologic status before surgery, was considered a complication. Seizure activity after surgery was recorded, but not considered a major complication unless seizures had not been noted prior to surgery, medical attention with hospitalization was required to resolve the seizure activity, or seizures led to death or euthanasia. Any complication that did not require medical attention was considered minor and was not included in the analyses. Further, superficial surgical site infections were not considered a major complication and thus were not included in the analyses.
Additional treatment
All dogs included in the study were also enrolled in various prospective clinical trials that assessed outcome, including survival time, after surgical debulking of brain tumors in combination with other novel treatments (eg, tumor lysate vaccines, gene therapy, and administration of temozolomide [an alkylating chemotherapy agent]). Any complication that occurred as a direct result of an additional treatment related to these clinical trials was recorded. The University of Minnesota Institutional Animal Care and Use Committee approved all experimental portions of the protocols for all studies.
Statistical analysis
Descriptive data were presented as percentages and median or mean (range). Data were analyzed with available statistical software.b For dogs in which > 1 surgery was performed, only the information related to the first surgery was included in the analyses. Associations between predictor variables and the response variables of death, aspiration pneumonia, pneumocephalus, seizures, neurologic worsening, cyclic hyperthermia and hypernatremia syndrome, and lethargy were analyzed with logistic regression to determine ORs. The logistic regression models used to determine statistical significance were fit with a single response variable and a single predictor variable. Most risk factors were evaluated as categorical variables and coded as yes or no; age was evaluated as a quantitative variable. For categorical variables, significance was assessed on the basis of estimated ORs and their 95% confidence intervals. Values of P ≤ 0.05 were considered significant.
Results
Animals
The medical records review identified 160 dogs that had undergone cytoreductive surgery by a single pair of surgeons for treatment of a suspected primary intracranial mass during the study period, and all 160 dogs were included in the study. Eight of the 160 dogs had tumor recurrence and underwent a second cytoreduction procedure; however, only data pertaining to the first surgical procedure were included in analyses.
Dogs (75 brachycephalic and 85 nonbrachycephalic) were reported as Boxer (n = 31 [19.4%]); Golden Retriever (18 [11.3%]); Labrador Retriever (14 [8.8%]); Boston Terrier (12 [7.5%]); French Bulldog (9 [5.6%]); American Staffordshire Terrier (8 [5.0%]); Poodle (6 [3.8%]); German Shepherd Dog (4 [2.5%]); Chihuahua, Bulldog, Jack Russell Terrier, English Springer Spaniel, and Weimaraner (3 [1.9%] each); American Bulldog, Beagle, Havanese, Pug, Shetland Sheepdog, Shih Tzu, and Yorkshire Terrier (2 [1.3%] each); and 29 other breeds (1 [0.6%] each]. Mean age at the time of diagnosis of an intracranial mass was 8.7 years (range, 10 months to 14 years). There were 84 neutered males, 62 spayed females, 9 sexually intact males, and 5 sexually intact females.
Patient history and physical examination findings
Mean time from initial observation of clinical signs to surgery was 10.2 weeks (range, 1 to 52 weeks). Seventy dogs were examined < 4 weeks after the onset of clinical signs, 59 dogs between 4 and 12 weeks, and 31 dogs > 12 weeks. All dogs were referred by veterinary neurologists who had initially examined the dogs and diagnosed primary intracranial lesions. Clinical signs that prompted use of diagnostic imaging included generalized seizures (n = 131 [81.9%]), abnormal mentation (20 [12.5%]), ataxia (15 [9.4%]), paresis (12 [7.5%]), circling (10 [6.3%]), head tilt (9 [5.6%]), blindness (8 [5.0%]), neck pain (3 [1.9%]), and aggression (2 [1.3%]), alone or in combination. Medical treatment initiated by the referring neurologists to stabilize the dogs prior to surgery included a combination of corticosteroids and AEDs (n = 108 [67.5%]), corticosteroids alone (25 [15.6%]), AEDs alone (23 [14.4%]), or no treatment (4 [2.5%]).
Neurologic examination findings
Results of neurologic examinations prior to surgery were reported to have been normal for 103 of the 160 (64.4%) dogs. The remaining 57 dogs had abnormalities, including postural reaction deficits (n = 20/160 [12.5%]), cranial nerve deficits (20 [12.5%]), circling (18 [11.3%]), paresis (15 [9.4%]), abnormal mentation (10 [6.3%]), head tilt (9 [5.6%]), ataxia (7 [4.4%]), and nonambulatory paraparesis (4 [2.5%]), alone or in combination.
Diagnostic imaging findings
All 160 dogs in the study underwent advanced diagnostic imaging (158 [98.8%] by MRI and 2 [1.3%] by CT) of the brain prior to surgery, and all available diagnostic images were assessed prior to surgical intervention because of the innate differences in imaging characteristics of each type of tumor. On the basis of imaging findings, solitary intracranial masses were diagnosed in the olfactory or frontal lobes (n = 86 [53.8%]), temporal or parietal lobes (54 [33.8%]), cerebellum (10 [6.3%]), occipital lobe or brainstem (4 [2.5%] each), or interthalamic region (2 [1.3%]). An extra-axial mass (meningioma tentatively suspected) was evident in 58 (36.3%) dogs, whereas an intra-axial mass (glioma tentatively suspected) was evident in 102 (63.8%). There were equal numbers (n = 80) of masses that involved the right versus left sides of the brain.
Immediately after surgery, 155 of the 160 (96.9%) dogs underwent MRI to assess the extent of resection, and 126 dogs were considered to have gross total resection of the tumor, 24 dogs had subtotal resection with 5% to 20% of the initial tumor volume remaining after surgery, and 5 dogs had partial resection with 40% to 70% of the initial tumor volume remaining. Five dogs did not undergo MRI immediately after surgery.
Anesthesia and surgery
Anesthesia records were available for 159 of the 160 (99.4%) dogs. The anesthetic protocol used for dogs varied and was determined by the individual anesthesiologist on the basis of each dog's preoperative status. Complications reported during anesthesia included hypothermia (40/160 [25.0%]), hypotension (19 [11.9%]), bradycardia (13 [8.1%]), hypertension (9 [5.6%]), hyperthermia (5 [3.1%]), cardiac arrhythmias (5 [3.1%]), prolonged recovery (4 [2.5%]), a combination of bradycardia and hypertension (3 [1.9%]), and seizure activity during recovery (1 [0.6%]). None of the dogs died or were euthanized during the anesthetic procedure.
Seventy-eight of the 160 (48.8%) dogs underwent a bilateral transfrontal surgical approach, 66 (41.3%) underwent a lateral rostrotentorial approach, 14 (8.8 %) underwent a suboccipital approach, and 2 (1.3%) underwent an interhemispheric approach to the third ventricle or interthalamic adhesion as described by Bagley et al.31 As part of an additional clinical trial, 27 dogs received multiple intraparenchymal injections of adenoviral-mediated gene therapy around the resection cavity. Mean duration of surgery was 109 minutes (range, 60 to 195 minutes).
Hospitalization
Intensive care unit treatment sheets were available for 143 of the 160 (89.4%) dogs. Mean duration of hospitalization was 1.4 days (range, 1 to 5 days) for dogs surviving to discharge. Of the 149 dogs that were discharged from the hospital, 115 (77.2%) were discharged < 24 hours after surgery, 24 (16.1%) were discharged between 24 and 48 hours after surgery, 8 (5.4%) were discharged between 48 and 72 hours after surgery, 1 (0.7%) was discharged between 72 and 96 hours after surgery, and 1 was discharged > 96 hours after surgery. Eighty-three of the 102 (81.4%) dogs treated during the last 4 years of the study period were discharged the day after surgery.
Eleven dogs (6.9%) died or were euthanized shortly after surgery during hospitalization in the intensive care unit. Dogs that did not survive to discharge died a mean of 1.9 days (range, 0 to 5 days) after surgery. Six dogs died ≤ 24 hours of surgery, 2 dogs died the second day after surgery, and 1 each died on days 3, 4, and 5 after surgery.
Serum total protein concentration and PCV were measured for 122 of the 160 (76.3%) dogs the morning after surgery. Mean PCV was 38.9% (range, 25% to 58%; reference range, 37.5% to 60.3%). Five dogs had a PCV between 25% and 30%, but none required a blood transfusion. Mean serum total protein concentration was 6.5 g/dL (range, 4.0 to 9.4 g/dL; reference range, 5.0 to 6.9 g/dL).
Histologic evaluation
A tissue sample of the lesion from each dog was submitted for histologic evaluation by board-certified veterinary pathologists. Definitive histologic diagnoses included 79 (49.4%) gliomas, of which 72 were high-grade; 58 (36.3%) meningiomas, of which 18 were grade II or III; 13 (8.1%) nontumorous lesions (8 granulomatous meningoencephalitides, 4 unspecified nontumorous lesions, and 1 vascular infarct); 4 (2.5%) primitive neuroectodermal tumors; 2 (1.3%) each of nerve sheath tumors and histiocytic tumors; and 1 (0.6%) each of a cavernoma and a choroid plexus tumor.
Complications
Death—Twenty-one of the 160 (13.1%) dogs died during the first 30 days after surgery (Supplementary Table S1 available at avmajournals.avma.org/doi/suppl/10.2460/javma.253.12.1594). Eleven dogs died before discharge from the hospital, and 10 dogs died after discharge. Five of the 11 dogs that died in the hospital were given a poor prognosis before surgery because of severe neurologic deficits, and the owners were warned of the strong possibility that their dogs might not survive surgery. When these 5 dogs were excluded from analyses, the death rate decreased from 13.1% to 10.3%, with an in-hospital death rate of 3.9% (6 in-hospital deaths/155 surgeries). Four dogs that were discharged from the hospital died within 5 days after surgery, and the remaining 6 dogs died between 6 and 30 days after surgery. A necropsy on 1 dog that died 26 days after surgery revealed brain herniation and extensive tumor infiltration that was likely attributable to residual disease and progression of the aggressive glioma. Complete postmortem examinations were performed on 17 of the 21 dogs that died, and the brains were examined on the remaining dogs.
All 4 dogs that had infratentorial tumors (ie, tumor below the tentorium cerebelli) involving a region other than the cerebellum died within 30 days after surgery. The tumors of these dogs were extra-axial and ventrolateral in location and consisted of a meningioma just rostral to the foramen magnum (n = 2), meningioma at the petro-occipital junction (1), and a single large vestibular schwannoma (1). Of these 4 dogs, 1 was nonambulatory and stuporous prior to surgery. The owners of this dog were discouraged from pursuing surgery, but elected surgery rather than euthanasia. After surgery, this dog progressed from stupor to coma, developed systemic hypertension, and required a ventilator; 18 hours after surgery, it was euthanized. Two other dogs in this small subset had cardiopulmonary arrest after an episode of vomiting within 24 hours after surgery and could not be resuscitated; 1 of the 2 had been determined to have megaesophagus prior to surgery. The last dog in this subset had a papillary meningioma, recovered well, and was healthy 2 weeks after surgery; however, the dog developed pneumonia, and the owners elected euthanasia after the dog's condition deteriorated despite 3 days of treatment.
Two of the 10 dogs with cerebellar lesions died. One developed rapidly progressing necrotizing fasciitis of the pelvic limbs and perineum and was euthanized 10 days after surgery; the other had severe vestibular signs after surgery that progressed to nonambulatory paraparesis. This dog developed aspiration pneumonia, suffered respiratory arrest 5 days after surgery, and was found to have diffuse necrotizing leukoencephalopathy at necropsy.
Two of the 21 dogs that died had systemic signs of hyperadrenocorticism that contributed to the cause of death. Another dog died 2 weeks after surgery of severe hematemesis that was suspected to have been a result of steroid-induced gastric perforation. In addition, 1 dog with inflammatory meningoencephalitis died of disease progression 20 days after surgery.
Results of multivariate analysis identified multiple risk factors associated with death. Dogs having an abnormal result on preoperative neurologic examination were significantly (OR = 3.7; P = 0.008) more likely to die within 30 days after surgery. Among the specific neurologic abnormalities, head tilt and abnormal mentation were significantly (P = 0.003 and P = 0.04, respectively) associated with death. In addition, dogs that underwent a suboccipital surgical approach were significantly (OR = 5.1; 95% confidence interval, 1.87 to 8.85; P = 0.01) more likely to die. The odds of death with a suboccipital approach increased if the dog also had a brainstem tumor (OR = 16; 95% confidence interval, 11.2 to 23.7; P = 0.02). In fact, all 4 dogs with a brainstem tumor died within 30 days after surgery. The only dog that underwent surgical resection of a histologically confirmed choroid plexus tumor developed cyclic hyperthermia and hypernatremia syndrome and subsequent severe CNS signs and died; otherwise, a specific tumor type was not associated with increased odds of death. In addition, the amount of residual tumor after surgery was not associated with the odds of death; 13 of the 21 dogs that died had gross total resection. However, a large volume of residual tumor was seen on postoperative MRI and confirmed on necropsy in a dog with a butterfly glioma that died 26 days after surgery.
Other major complications—One or more major complications other than death were reported within 30 days after surgery in 30 of the 160 (18.8%) dogs. Major complications included seizures (18 [11.3%]); worsening of neurologic signs including nonambulatory hemiparesis, severe vestibular signs, blindness and abnormal mentation status (6 [3.8%]); and aspiration pneumonia (6 [3.8%]). One of the 160 (0.6%) dogs developed status epilepticus and was euthanized; however, no tumor was found in the brain on necropsy. Another major complication that resulted in death or euthanasia immediately after surgery was cyclic hyperthermia and hypernatremia syndrome (4 [2.5%]).
The only variables associated with a significant (P = 0.04 and P = 0.015, respectively) increase in the odds of developing major complications other than death were an abnormal result on the preoperative neurologic examination (abnormal mentation in most dogs) and tumor location in the brainstem. The extent of resection was not associated with odds of developing major complications.
Potential risk factors were analyzed for associations with specific major complications. For dogs that developed the major complication of worsening neurologic status, significant risk factors included a suboccipital surgical approach (P = 0.02) and an abnormal result on preoperative neurologic examination (P = 0.001), specifically postural reaction deficits (P = 0.04) and cranial nerve deficits (P = 0.04). For dogs that developed the major complication of aspiration pneumonia, significant risk factors were postural reaction deficits (P = 0.03) and a tumor located in the brainstem necessitating a suboccipital approach (P = 0.04). Furthermore, vomiting after surgery was significantly (P = 0.03) associated with development of aspiration pneumonia as a major complication; however, it was not significantly (P = 0.06) associated with the likelihood of death.
Minor and transient complications—Minor complications (eg, mild epistaxis or serosanguineous nasal discharge, signs of dysphoria, and mild inflammation of the surgical site) were observed. These complications were self-limiting, including epistaxis (n = 12/160 [7.5%]) after a surgical approach through the frontal sinuses, and were not included in analyses. One dog was suspected to have had a local postoperative infection; however, on necropsy 2.5 years later when the dog was euthanized after developing lymphoma, results of bacterial culture were negative, and no evidence of local infection was noticed. In addition, neurologic deficits after surgery included transient hemiparesis that resolved within hours to days (n = 14/160 [8.8%]), long-term minor cranial nerve deficits with no obvious compromise of quality of life (12 [7.5%]), and long-term hemiparesis necessitating a support harness (1 [0.6%]). Tension pneumocephalus was diagnosed in 1 dog after a bilateral transfrontal approach, and this dog was successfully treated by surgical closure of the small dural defect.
Additional treatments
All dogs in the study were enrolled in clinical trials that involved various treatments in addition to surgical debulking of the tumor. Related to those clinical trials, 27 received gene therapy delivered intraparenchymally around the resection cavity, and 127 received autologous tumor lysate vaccines beginning 10 to 14 days after surgery. Complications attributed to additional treatments of a clinical trial, rather than the surgery, included 48 to 72 hours of lethargy after receiving immunotherapy vaccines (n = 12) and euthanasia after development of severe encephalitis following administration of a high dose of interferon-γ gene therapy (1).
Discussion
Results of the present study suggested that in dogs undergoing cytoreductive surgery for treatment of suspected primary intracranial masses, the incidences of death and major complications other than death during the first 30 days after surgery were relatively low (21/160 [13.1%] and 30/160 [18.8%], respectively). Importantly, dogs with abnormal findings on preoperative neurologic examination were more likely to develop complications or die, and dogs undergoing a suboccipital approach were more likely to die, suggesting that these factors may be useful in case selection.
Previous studies of complications following craniotomy in dogs focused on a single complication,26 involved a relatively small number of cases,7–11 or both.27–29,32 In contrast, the present study included a relatively large number of dogs with solitary intracranial masses that underwent cytoreductive surgery, included dogs with a wide variety of intracranial diseases, and evaluated all major complications that occurred in the first 30 days after surgery.
Overall, 30 of the 160 (18.8%) dogs in the present study developed major complications other than death. This was similar to morbidity rates reported in human medicine, which range from 22.8% to 36.9%.12,23,33 In addition, although the overall incidence of death during the first 30 days following surgery for dogs in the present study was 13.1% (21/160), it decreased from 19% (15/80) for the first 80 consecutive dogs to 8% (6/80) for the second 80 consecutive dogs. This lower incidence of death in the later half may have been caused by the development of greater surgical expertise for these difficult cases as well as the decision to encourage owners of dogs with brainstem tumors to seek options other than surgery. Further, because the present study assessed complications that occurred during the first 30 days after surgery, deaths specifically caused by tumor recurrence were minimized.
Findings from the present study indicated that the most important and clinically relevant risk factors for death in dogs following surgery to treat a primary intracranial mass were abnormal results on a neurologic examination performed prior to surgery and specific location of the tumor, especially in the brainstem. Many factors could contribute to abnormal results of neurologic examination and increase the risk of perioperative death. Cranial nerve deficits may imply a disease or tumor involving structures of the brainstem. This region of the brain is also responsible for numerous vital functions such as wakefulness and respiration.1,34 A head tilt can be associated with disease of the vestibular or cerebellar systems.1,34 There is a dense population of vital structures in these neuroanatomical locations, making surgical access challenging, and small physical changes or minor surgical trauma may induce major vital alterations or complications. Further, abnormal mentation could result from a brain tumor because of increased ICP, ischemia, edema, metabolic alterations, or ventricular outflow obstruction.1 In addition, damage to or compression of the ascending reticular activating system can result in altered consciousness.34 Small increases in cerebral blood flow can cause a dangerous increase in ICP because of the limited compliance of the cranial vault in patients that already have a high ICP as a result of a mass effect.35 Similar to the findings of the present study, abnormal mentation was the only significant risk factor in a study36 evaluating risk factors for perianesthetic complications in dogs with intracranial disease. In the future, it may be useful to try to separate complications believed to be associated with anesthesia to provide information on ways to reduce the risk of aspiration pneumonia and other complications that may be associated with anesthesia.
In addition to abnormal results on neurologic examination, results of the present study indicated that brain tumor location and, somewhat relatedly, a suboccipital surgical approach, were risk factors for postoperative complications. However, a study37 of necropsy results for dogs with intracranial neoplasia shows that 79 of 168 tumors occupied > 1 anatomic division of the brain and that assigning a single neuroanatomical location for tumors may be inappropriate in many cases. In the present study, it was likely that a suboccipital approach itself was not a risk factor for death and other major complications. Rather, a suboccipital approach was necessary to access tumors in the brainstem and cerebellum, and all dogs with tumors involving the brainstem in the present study died. Relatedly, a study38 that evaluated a ventral approach to the brainstem in healthy research dogs found that 1 of 4 dogs died within 12 hours after the procedure. The high complication rate associated with surgery of the brainstem could be attributable to the substantial neurologic dysfunction resulting from small iatrogenic lesions in a region without functional redundancy.38 Therefore, it currently seems reasonable to advise against surgical debulking of brainstem tumors in dogs. Surgical debulking of cerebellar tumors, however, was not a significant risk factor for death in the present study.
Thirteen of the 160 (8.1%) suspected primary intracranial tumors were determined by histologic examination to have been nontumorous lesions. This finding was consistent with results of other studies39–50 showing it is impossible to make a definitive diagnosis of solitary intracranial lesions on the basis of MRI characteristics.
Interestingly, neither tumor type nor tumor grade was significantly associated with death or major complications in the present study. It was possible that intra-axial tumors (eg, high-grade gliomas) could have been associated with more complications than extra-axial tumors (eg, low-grade meningiomas). In people, infratentorial tumors have a significantly higher complication rate (44.0%), compared with supratentorial tumors (22.8%), regardless of histopathologic diagnosis.22 Our findings supported that intra-axial tumors, with the possible exception of choroid plexus tumors, had similar rates of morbidity and death as extra-axial tumors. The single dog that underwent resection of a choroid plexus tumor died after developing severe cyclic hyperthermia and hypernatremia syndrome and, subsequently, severe CNS signs. It was likely, as is common during resection of these tumors, that the organum vasculosum of the lamina terminalis, or supraoptic crest, in the anterioventral aspect of the third ventricle, which controls basal functions (eg, electrolyte balance and thermoregulation), suffered thermal injury during electrocoagulation to control hemorrhage. Three other dogs with lesions near the diencephalon developed the same clinical signs immediately after surgery and died within days afterward. Similarly, this constellation of clinical signs has been reported in 4 dogs and 2 cats that died within 24 hours after surgical resection of diencephalic lesions, and damage to the hypothalamus was reported to have been the cause.29 On the basis of results in the present study, we suspected that dogs with abnormal results on neurologic examination and a tumor of the piriform cortex of the cerebrum or of the temporal lobe were more likely to develop cyclic hyperthermia and hypernatremia syndrome. However, the small sample size prevented statistical analysis of the potential association.
Although the extent of resection has been identified as a significant factor in overall survival times in people undergoing surgical resection of brain tumors, it was not associated with perioperative complication rate or neurologic outcome in patients having a first or second craniotomy for glioma.24 Similarly, the extent of resection was not significantly associated with short-term complications for dogs in the present study.
Further and similar to human studies,19 worsening neurologic signs after surgery were the most common (33/160 [20.6%]) complication in the present study. Transient neurologic deficits related to manipulation of nervous tissue are almost unavoidable sequelae during surgical procedures on brain tissue, and transient neurologic deficits were observed in the present study. The incidence of postoperative seizures as a complication was challenging to determine because 81.9% (131/160) of dogs in the present study had generalized seizures prior to surgery. In addition, all dogs in the present study continued to receive AEDs indefinitely after surgery, whereas 69.3% of people with epilepsy secondary to an intracranial meningioma were seizure-free after tumor removal.51 Because seizure activity originates from the tissue surrounding the tumor and not the tumor itself, continued seizures following surgical resection should be expected and should not be considered a complication.52 However, development of new seizures, which occurred in 5 of the 160 (3.1%) dogs in the present study, should be considered a complication of surgery and is most likely attributed to damage of the surrounding tissue during tumor resection. Other surgically related potential causes for the development of seizures could include increased ICP, scar formation, hematoma formation, obstructive hydrocephalus, infection, and pneumocephalus. Determining whether worsening of seizures is a postoperative complication or a progression of disease is more difficult. A prospective randomized trial53 in humans who had brain tumors but no seizures and underwent brain tumor removal showed no difference in the rate of seizures after surgery (8%) for patients who were given prophylactic AEDs, compared with those who were not.
Although most dogs in the present study had generalized seizures before surgery and had tumors of the cerebral hemispheres, only 7 of the 21 dogs that died had these findings. On the contrary, most dogs that died underwent diagnostic MRI for signs other than seizures, had abnormal results on neurologic examination, and had tumors in locations other than the cerebral cortex. These findings could be useful when discussing potential treatments with owners.
A rapid and predictable recovery without excitement is important to minimize the risk that ICP will increase after intracranial surgery.54 In addition, use of postoperative analgesia has evolved over time, with fewer doses of opioid drugs given when the drugs are ordered to be administered as needed, compared with when they are ordered to be administered on a set schedule. Pain after craniotomy is described but poorly understood in people and can lead to nausea, vomiting, and high blood pressure.55 In addition, such pain in humans has been suggested to be less severe, compared with pain following other surgeries (eg, rotator cuff surgery, thoracotomy), with 90% of patients reporting satisfactory pain control.56 This was probably true for dogs in the present study also, because recently treated dogs appeared comfortable and rarely received any additional doses of analgesic medications after recovery from anesthesia.
We noted in the present study that dogs with shorter durations of hospitalization had lower rates of morbidity and death. Given the conservative use of analgesic medications combined with the short duration of hospitalization for dogs in the present study, it was not surprising that the percentage of dogs that developed aspiration pneumonia (3.8% [6/160]) was lower than that previously reported (24%).26 In addition, a history of preexisting neurologic disease is 1 of 3 factors in dogs that contribute to the development of postanesthetic aspiration pneumonia, and dogs with intracranial disease that also develop aspiration pneumonia have the lowest survival rate.57 Two related risk factors for postoperative aspiration pneumonia in the present study were tumors of the brainstem and a suboccipital approach, for which the resulting procedure may affect the nerves that control the gag reflex and swallowing. Although development of postoperative vomiting in dogs of the present study was rare, results indicated that it was a predisposing factor for aspiration pneumonia, consistent with findings of a previous report.26
Nonpathological pneumocephalus is reported in 100% of human patients after craniotomy, but tension pneumocephalus is a life-threatening condition.58 Tension pneumocephalus has been reported in 2 dogs at 7 and 12 days after transfrontal craniotomy to excise olfactory lobe meningiomas.27,28 Although nonpathological pneumocephalus was a frequent finding in the present study, tension pneumocephalus was diagnosed only in a single dog after a bilateral transfrontal approach. This dog was successfully treated by surgical closure of the small dural defect.
In the present study, no dogs were identified that died or were hospitalized because of complications related to local postoperative infection (eg, encephalitis or meningitis). In addition, although the PCV decreased after surgery in many dogs, none required a transfusion, whereas a national systematic review of the scientific literature regarding humans undergoing a craniotomy identified hemorrhage severe enough to warrant a blood transfusion as the most common complication.33 In the present study, epistaxis or serosanguineous nasal discharge was fairly common in dogs that underwent a bilateral transfrontal approach, but this blood loss was not considered to be a complication because it was self-limiting. The findings of the present study concurred with those in the case series32 that first described the bilateral transfrontal surgical approach, in which 60% of dogs were reported to have mild, self-limiting postoperative epistaxis.
Several limitations as well as directions for future investigations were identified in the present study. Importantly, none of the authors were board-certified veterinary neurologists; however, the senior author had expertise and experience in treating dogs with intracranial tumors. For each dog, the initial evaluation was performed and the diagnosis was made by a board-certified veterinary neurologist, who then referred the dog for clinical trial enrollment at our veterinary teaching hospital. Many dogs had neurologic deficits that led to the diagnosis of an intracranial mass, but had a normal neurologic status by the time of surgery. Another limitation was that all dogs in the present study were enrolled in clinical trials that involved various treatments in addition to surgical debulking of the tumor. It was known that some of the observed postoperative complications were secondary to those treatments, rather than the surgery itself. In addition, limitations were inherent in the retrospective design of the present study. Although owners and referring veterinarians were instructed to inform the investigators of any complications, some data could have been missing.
Another limitation of the present study was that the calculation for the incidence of death included all dogs, which may have resulted in inclusion of some dogs for which the cause of death was not specifically related to the surgery. For example, 1 dog died 2 weeks following surgery after developing severe hematemesis that was suspected to have resulted from steroid-induced gastric perforation. Several dogs had severe mentation changes prior to surgery, and their owners were advised against pursuing surgical debulking. However, these clients elected to pursue surgery despite understanding the higher risks of complications and death. Nonetheless, most (139/160 [86.9%]) dogs survived for at least 30 days after tumor resection.
Case selection in the present study was biased because dogs with severe neurologic deficits were excluded on the basis of criteria of the concomitant clinical trials. However, some dogs were treated despite having systemic disease or severe neurologic deficits, which may have affected the incidence of death within 30 days of surgery in the present study. For example, pathologists identified the probable cause of death of 2 dogs in the present study as signs related to hyperadrenocorticism. Therefore, prior to surgical debulking of an intracranial mass, it would seem prudent to perform further diagnostic testing on any dog suspected of having hyperadrenocorticism. Another confounding factor was that we did not deny surgery to the few dogs that were stuporous or semicomatose; however, we tried to dissuade their owners from pursuing surgery. Although it was important to show compassion to owners of dogs in this condition, it may not have been the best option, given the results of the present study.
As the field of optimizing treatments for veterinary patients with brain tumors advances, surgery will likely remain an integral component, either alone or in combination with other treatment modalities. Results of the present study provided valuable information on predisposing factors, odds of major complications or death, and incidences of major complications or death in dogs during the first 30 days after undergoing cytoreductive surgery for treatment of suspected primary intracranial masses. Careful case selection and attention to the neurologic examination and location of the tumor prior to surgery may improve outcomes and minimize complications.
Acknowledgments
The authors declare that there were no conflicts of interest. The authors thank Dr. Aaron Rendahl for statistical analysis and interpretation.
ABBREVIATIONS
| AED | Antiepileptic drug |
| ICP | Intracranial pressure |
Footnotes
Gelfoam, Pharmacia & Upjohn Co, Kalamazoo, Mich.
R, version 3.2.3, R Foundation for Statistical Computing, Vienna, Austria. Available at: www.r-project.org/. Accessed Jun 8, 2016.
References
1. Bagley RS. Pathophysiology. In: Fundamentals of veterinary clinical neurology. Ames, Iowa: Blackwell Publishing, 2005;41–132.
2. Hu H, Barker A, Harcourt-Brown T, et al. Systematic review of brain tumor treatment in dogs. J Vet Intern Med 2015;29:1456–1463.
3. Dickinson PJ. Advances in diagnostic and treatment modalities for intracranial tumors. J Vet Intern Med 2014;28:1165–1185.
4. Stoica G, Levine J, Wolff J, et al. Canine astrocytic tumors: a comparative review. Vet Pathol 2011;48:266–275.
5. Rossmeisl JH Jr, Jones JC, Zimmerman KL, et al. Survival time following hospital discharge in dogs with palliatively treated primary brain tumors. J Am Vet Med Assoc 2013;242:193–198.
6. Van Meervenne S, Verhoeven PS, de Vos J, et al. Comparison between symptomatic treatment and lomustine supplementation in 71 dogs with intracranial, space-occupying lesions. Vet Comp Oncol 2014;12:67–77.
7. Axlund TW, McGlasson ML, Smith AN. Surgery alone or in combination with radiation therapy for treatment of intracranial meningiomas in dogs: 31 cases (1989–2002). J Am Vet Med Assoc 2002;221:1597–1600.
8. Niebauer GW, Dayrell-Hart BL, Speciale J. Evaluation of craniotomy in dogs and cats. J Am Vet Med Assoc 1991;198:89–95.
9. Klopp LS, Rao S. Endoscopic-assisted intracranial tumor removal in dogs and cats: long-term outcome of 39 cases. J Vet Intern Med 2009;23:108–115.
10. Greco JJ, Aiken SA, Berg JM, et al. Evaluation of intracranial meningioma resection with a surgical aspirator in dogs: 17 cases (1996–2004). J Am Vet Med Assoc 2006;229:394–400.
11. Ijiri A, Yoshiki K, Tsuboi S, et al. Surgical resection of twenty-three cases of brain meningioma. J Vet Med Sci 2014;76:331–338.
12. National Cancer Institute website. Treatment option overview for adult primary CNS tumors. Available at: www.cancer.gov/types/brain/hp/adult-brain-treatment-pdq#section/_69. Accessed Jun 19, 2016.
13. McPherson CM, Sawaya R. Technologic advances in surgery for brain tumors: tools of the trade in the modern neurosurgical operating room. J Natl Compr Cancer Netw 2005;3:705–710.
14. Mehdorn HM. Intracranial meningiomas: a 30-year experience and literature review. Adv Tech Stand Neurosurg 2016;43:139–184.
15. Karsy M, Guan J, Cohen A, et al. Medical management of meningiomas: current status, failed treatments, and promising horizons. Neurosurg Clin N Am 2016;27:249–260.
16. Albert FK, Forsting M, Sartor K, et al. Early postoperative magnetic resonance imaging after resection of malignant glioma: objective evaluation of residual tumor and its influence on regrowth and prognosis. Neurosurgery 1994;34:45–60.
17. Paldor I, Awad M, Sufaro Y, et al. Review of controversies in management of non-benign meningioma. J Clin Neurosci 2016;31:37–46.
18. Landriel Ibañez FA, Hem S, Ajler P, et al. A new classification of complications in neurosurgery. World Neurosurg 2011;75:709–715.
19. Brell M, Ibanez J, Caral L, et al. Factors influencing surgical complications of intra-axial brain tumours. Acta Neurochir (Wien) 2000;142:739–750.
20. Seicean A, Seicean S, Schiltz NK, et al. Short-term outcomes of craniotomy for malignant brain tumors in the elderly. Cancer 2013;119:1058–1064.
21. Sawaya R, Hammoud M, Schoppa D, et al. Neurosurgical outcomes in a modern series of 400 craniotomies for treatment of parenchymal tumors. Neurosurgery 1998;42:1044–1055.
22. Cabantog AM, Bernstein M. Complications of first craniotomy for intra-axial brain tumour. Can J Neurol Sci 1994;21:213–218.
23. Schiavolin S, Broggi M, Acerbi F, et al. The impact of neurosurgical complications on patients’ health status: a comparison between different grades and complications. World Neurosurg 2015;84:36–40.
24. Chang SM, Parney IF, McDermott M, et al. Perioperative complications and neurological outcomes of first and second craniotomies among patients enrolled in the Glioma Outcome Project. J Neurosurg 2003;98:1175–1181.
25. Motta L, Mandara MT, Skerritt GC. Canine and feline intracranial meningiomas: an updated review. Vet J 2012;192:153–165.
26. Fransson BA, Bagley RS, Gay JM. Pneumonia after intracranial surgery in dogs. Vet Surg 2001;30:432–439.
27. Cavanaugh RP, Aiken SW, Schatzberg SJ. Intraventricular tension pneumocephalus and cervical subarachnoid pneumorrhachis in a Bull Mastiff dog after craniotomy. J Small Anim Pract 2008;49:244–248.
28. Garosi LS, Penderis J, Brearley MJ, et al. Intraventricular tension pneumocephalus as a complication of transfrontal craniotomy: a case report. Vet Surg 2002;31:226–231.
29. Marino DJ, Dewey CW, Loughin CA, et al. Severe hyperthermia, hypernatremia, and early postoperative death after transethmoidal Cavitron ultrasonic surgical aspirator (CUSA)-assisted diencephalic mass removal in 4 dogs and 2 cats. Vet Surg 2014;43:888–894.
30. Bagley RS, Harrington ML, Pluhar GE, et al. Effect of craniectomy/durotomy alone and in combination with hyperventilation, diuretics, and corticosteroids on intracranial pressure in clinically normal dogs. Am J Vet Res 1996;57:116–119.
31. Bagley RS, Baszler TV, Harrington ML, et al. Clinical effects of longitudinal division of the corpus callosum in normal dogs. Vet Surg 1995;24:122–127.
32. Glass EN, Kapatkin A, Vite C, et al. A modified transfrontal sinus approach to the canine frontal lobe and olfactory bulb: surgical technique and 5 cases. J Am Anim Hosp Assoc 2000;36:43–50.
33. Rolston JD, Han SJ, Lau CY, et al. Frequency and predictors of complications in neurologic surgery: national trends from 2006 to 2011. J Neurosurg 2014;120:736–745.
34. De Lahunta A, Glass EN, Kent M. Upper motor neuron and cerebellum. In: Veterinary neuroanatomy and clinical neurology. 4th ed. St Louis: Saunders Co, 2014;222–236, 368–408.
35. Raisis AL, Musk GC. Anesthesia and analgesia. In: Platt SR, Olby NJ, eds. BSAVA manual of canine and feline neurology. 4th ed. Gloucester, England: British Small Animal Veterinary Association, 2014;412–414.
36. Hicks JA, Kennedy MJ, Patterson EE. Perianesthetic complications in dogs undergoing magnetic resonance imaging of the brain for suspected intracranial disease. J Am Vet Med Assoc 2013;243:1310–1315.
37. Snyder JM, Shofer FC, Van Winkle TJ, et al. Canine intracranial primary neoplasia: 173 cases (1986–2003). J Vet Intern Med 2006;20:669–675.
38. Klopp LS, Simpson ST, Sorjonen DA, et al. Ventral surgical approach to the caudal brain stem in dogs. Vet Surg 2000;29:533–542.
39. Sturges BK, Dickinson PJ, Bollen AW, et al. Magnetic resonance imaging and histological classification of intracranial meningiomas in 112 dogs. J Vet Intern Med 2008;22:586–595.
40. Bohn AA, Wills TB, West CL. el al. Cerebrospinal fluid analysis and magnetic resonance imaging in the diagnosis of neurologic disease in dogs: a retrospective study. Vet Clin Pathol 2006;35:315–320.
41. Cervera V, Mai W, Vite CH, et al. Comparative magnetic resonance imaging findings between gliomas and presumed cerebrovascular accidents in dogs. Vet Radiol Ultrasound 2011;52:33–40.
42. Leclerc MK, d'Anjou MA, Blond L, et al. Interobserver agreement and diagnostic accuracy of brain magnetic resonance imaging in dogs. J Am Vet Med Assoc 2013;242:1688–1695.
43. Young BD, Levine JM, Porter BF, et al. Magnetic resonance imaging features of intracranial astrocytomas and oligodendrogliomas in dogs. Vet Radiol Ultrasound 2011;52:132–141.
44. Keenihan EK, Summers BA, David FH, et al. Canine meningeal disease: associations between magnetic resonance imaging signs and histologic finding. Vet Radiol Ultrasound 2013;54:504–515.
45. Lamb CR, Croson PJ, Cappello R, et al. Magnetic resonance imaging finding in 25 dogs with inflammatory cerebrospinal fluid. Vet Radiol Ultrasound 2005;46:17–22.
46. Bentley RT, Ober CP, Anderson KL, et al. Canine intracranial gliomas: relationship between magnetic resonance imaging criteria and tumor type and grade. Vet J 2013;198:463–471.
47. Cherubini GB, Mantis P, Martinez TA, et al. Utility of magnetic resonance imaging for distinguishing neoplastic from non-neoplastic brain lesions in dogs and cats. Vet Radiol Ultrasound 2005;46:384–387.
48. Kraft SL, Gavin PR, DeHaan C, et al. Retrospective review of 50 canine intracranial tumors evaluated by magnetic resonance imaging. J Vet Intern Med 1997;11:218–225.
49. Ródenas S, Pumarola M, Gaitero L, et al. Magnetic resonance imaging findings in 40 dogs with histologically confirmed intracranial tumours. Vet J 2011;187:85–91.
50. Singh JB, Overmann A, Lang J, et al. Contrast media enhancement of intracranial lesions in magnetic resonance imaging does not reflect histopathologic findings consistently. Vet Radiol Ultrasound 2011;52:619–626.
51. Englot DJ, Magill ST, Han SJ, et al. Seizures in supratentorial meningioma: a systematic review and meta-analysis. J Neurosurg 2016;124:1552–1561.
52. Bagley RS, Gavin PR. Seizures as a complication of brain tumors in dogs. Clin Tech Small Anim Pract 1998;13:179–184.
53. Wu AS, Trinh VT, Suki D, et al. A prospective randomized trial of perioperative seizure prophylaxis in patients with intraparenchymal brain tumors. J Neurosurg 2013;118:873–883.
54. Raisis AL, Leece EA, Platt SR, et al. Evaluation of an anesthetic technique used in dogs undergoing craniectomy for tumor resection. Vet Anaesth Analg 2007;34:171–180.
55. Guilkey RE, Von Ah D, Carpenter JS, et al. Integrative review: postcraniotomy pain in the brain tumour patient. J Adv Nurs 2016;72:1221–1235.
56. Hassouneh B, Centofanti JE, Reddy K. Pain management in post-craniotomy patients: a survey of Canadian neurosurgeons. Can J Neurol Sci 2011;38:456–460.
57. Ovbey DH, Wilson DV, Bednarski RM, et al. Prevalence and risk factors for canine post-anesthetic aspiration pneumonia (1999–2009): a multicenter study. Vet Anaesth Analg 2014;41:127–136.
58. Reasoner DK, Todd MM, Scamman FL, et al. The incidence of pneumocephalus after supratentorial craniotomy. Anesthesiology 1994;80:1008–1012.
