Materials and Methods

Animals—Twelve healthy adult Beagles were used in the study. To be included in the study, dogs were required to be free of pulmonary disease. The health status of each dog was assessed via the medical history and results of physical examination, a CBC, serum biochemical analysis, and thoracic radiography (lateral and ventrodorsal views). A board-certified veterinary radiologist (HD) evaluated the thoracic radiographs. Dogs did not receive any medications for at least 30 days before the study but regularly received prophylactic anthelminthic treatment. Dogs were cared for in accordance with the Canadian Council on Animal Care guidelines at a Canadian Council on Animal Care–accredited facility. The study protocol was approved by a university institutional animal care and use committee.

Anesthesia—Dogs were sedated with an IM injection of butorphanol tartrate (0.2 mg/kg) combined with acepromazine maleate (0.02 mg/kg). Anesthesia was induced by IV administration of propofol (4 mg/kg) and maintained with a constant-rate infusion of propofol (400 μg/kg/min, IV) and boluses of propofol (2 mg/kg, IV) as required to provide an adequate plane of anesthesia. All dogs received fluids IV, and supplemental oxygen was provided via a sterile soft urinary catheter^a placed orally into the trachea. Physical examination, a continuous ECG, and pulse oximetry were used to monitor dogs during anesthesia. Normal oxygen saturation was defined as ≥ 95%. Experimental procedures were discontinued and dogs were allowed to recover from anesthesia if oxygen saturation decreased to < 90% for 10 minutes, oxygen saturation decreased to < 85% for 5 minutes, or a dog developed bradycardia (< 60 beats/min) that was unresponsive to standard interventions.

Bronchoscopy and BAL—Dogs were positioned in sternal recumbency for bronchoscopy. A routine bronchoscopic examination was performed and recorded for each dog. The trachea and all lung lobes, up to the second- and third-generation bronchi, were visually examined.

Two BAL techniques were performed in each dog. The caudal lung lobes were lavaged with a weight-based volume (2 mL/kg, divided into 2 aliquots),⁹ with the second aliquot infused immediately after aspiration of the first aliquot. To perform SPA, the tip of the bronchoscope^b was wedged into a second-generation bronchus. Sterile saline solution was infused through the biopsy channel, which was then followed by infusion of 4 mL of air to empty the channel. Pulsatile aspiration with a maximum negative pressure of 5 kPa (37.5 mm Hg) was applied immediately after infusion; aspiration was applied with a suction pump^c connected directly to the suction valve of the bronchoscope via a disposable aspiration trap^d (Figure 1). The MAPT technique was performed on the opposite caudal lung lobe by wedging the bronchoscope into a second-generation bronchus, passing a sterile polyethylene tube^e through the biopsy channel of the bronchoscope, and advancing the tube until palpable resistance was met. Sterile saline solution was infused through the polyethylene tube, which was then followed by infusion of 2 mL of air to empty the tube. A 35-mL syringe was attached to the polyethylene tube via a 20-gauge, 1-inch hypodermic needle (Figure 2). Gentle pulsatile MA with a 35-mL syringe was performed immediately after infusion.¹

View Full Size

Photograph of the equipment used for BAL via the SPA technique. The disposable aspiration tube is connected directly to the suction valve of a bronchoscope (black arrowhead). Standard polyvinyl chloride tubing (black arrow) is used to connect the disposable aspiration tube to the stem valve of the suction pump (not shown).

Citation: American Journal of Veterinary Research 74, 4; 10.2460/ajvr.74.4.523

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Photograph of the equipment used for BAL via the MAPT technique. A 35-mL syringe is attached to sterile polyethylene tubing via a 20-gauge, 1-inch hypodermic needle (black arrowhead). The polyethylene tubing is inserted into the lungs through the biopsy channel of a bronchoscope (black arrow).

Citation: American Journal of Veterinary Research 74, 4; 10.2460/ajvr.74.4.523

• Download Figure
• Download figure as PowerPoint slide

Aspiration via each technique was continued until fluid could no longer be recovered. For MAPT, the syringe was emptied as necessary and then used to continue aspiration. The order of the aspiration techniques and the lung lobe lavaged with each technique were determined with a random number table. The bronchoscope was cleaned and sterilized in accordance with a standard cold sterilization method^f–h between dogs.

The amount of fluid retrieved, duration of BAL (time from infusion of saline solution until the end of aspiration for the second aliquot) and lowest oxygen saturation were recorded for each technique. Dwell time (time from infusion of saline solution to the first attempted aspiration) was < 20 seconds for each collection.

Examination of BALF—Macroscopic surfactant was indirectly assessed by measuring the volume of bubbles in each sample immediately after sample collection. A surfactant score of 0 was assigned if no bubbles were present, a score of 1 was assigned for 0.10 to 0.70 mL of bubbles, a score of 2 was assigned for 0.71 to 1.50 mL of bubbles, and a score of 3 was assigned for > 1.51 mL of bubbles. Each BALF sample was identified by a unique code of 3 letters, immediately placed on ice after collection, and processed for analysis within 120 minutes after collection. Samples were not filtered during processing. Total nucleated cell counts were determined with electrical impedance.ⁱ An aliquot (200 μL) of each sample was cytocentrifuged^j for 6 minutes at 180 × g, and additional slides were prepared from fluid centrifuged for 5 minutes at 500 × g. Slides were stained with Wright stain, and differential cell counts of 400 leukocytes were performed at 400× magnification on cytocentrifuge preparations by a board-certified veterinary clinical pathologist (DB) who was not aware of the technique used to obtain each sample.

Microscopic assessment—In addition to performing the differential cell counts, the board-certified veterinary clinical pathologist (DB) also semiquantitatively assessed 5 variables reflective of sample and slide quality (Appendix). Cellularity and number of clusters and sheets of epithelial cells were assessed at 100× magnification^k; cell preservation, number of bacteria, and differential counts were determined at 400× magnification.^k Scores of ≥ 2 for cellularity and cellular preservation were required for a BALF sample to be considered of diagnostic quality.¹³ Cells were evaluated for erythrophagocytosis and presence of intracellular pigment during differential counting. Scores > 2 for epithelial cells or RBCs were considered to indicate excessive suction or trauma and thus a BALF sample of lesser quality¹³

Statistical analysis—Statistical analyses were performed with a statistical software program.^l Nonparametric analyses were performed because none of the BALF data had normal distributions. Comparisons of the percentage fluid retrieval, total and differential cell counts, lowest oxygen saturation, duration of BAL, and sample quality were analyzed with Wilcoxon matched-pairs signed rank tests. For all analyses, values of P ≤ 0.05 were considered significant.

Results

Dogs—The study group consisted of 12 healthy Beagles (5 neutered males and 7 spayed females) that ranged from 5 to 8 years of age (mean ± SD, 7.2 ± 1.0 years). Body weight of the dogs ranged from 9.2 to 13.5 kg (mean, 11.2 ± 1.5 kg). Vital parameters of all dogs were within anticipated limits. No respiratory abnormalities were detected during thoracic auscultation. Results of biochemical analyses and CBCs were unremarkable. One of the dogs had a grade 2/6 left-sided systolic heart murmur and evidence of mild left atrial enlargement during examination of thoracic radiographs. Considering that pulmonary vessels and pulmonary tissues appeared normal and the dog did not have abnormal clinical signs, compensated left heart disease was suspected. No dogs were excluded from the study on the basis of the results of physical examination, hematologic analysis, or thoracic radiography. One dog developed a decrease in oxygen saturation to 72% for 5 minutes immediately after induction of anesthesia. In accordance with the study protocol, the procedure was discontinued and that dog was excluded from the study

Bronchoscopy and BAL—Bronchoscopic examination revealed no abnormalities in 9 of the 11 remaining dogs. An oral mass deep in the oropharynx and a generalized, nodular appearance of the bronchial mucosa was identified in 1 dog. In another dog, the bronchial mucosa of the right middle and right caudal lung lobes was thickened and irregular. The right middle and right caudal lung lobes in that dog were also sensitive, and bronchoscopy induced considerable coughing during the BAL procedure.

The duration of BAL by use of MAPT ranged from 106 to 372 seconds, and the duration of BAL by use of SPA ranged from 140 to 317 seconds; these values did not differ significantly (Table 1). There was sudden loss of negative pressure during aspiration in 2 of the 11 dogs during MAPT.

Adverse effects of BAL—Transient (< 1-minute) decreases in oxygen saturation were detected in some dogs during both the MAPT and SPA techniques. These decreases in oxygen saturation were self-limiting and resolved without intervention. The lowest oxygen saturation during MAPT was 87%, and the lowest oxygen saturation during SPA was 88%; these values did not differ significantly. Iatrogenic linear abrasions were identified along the bronchial walls in 2 dogs after wedging of the polyethylene tubing during MAPT.

BALF analysis—Surfactant was macroscopically present in all samples obtained with SPA but was absent in 2 samples obtained with MAPT. Surfactant scores for samples obtained via SPA were significantly (P = 0.004) higher than scores for samples obtained via MAPT (median difference, 1.0; IQR, 1.0). Percentage of retrieved infusate ranged from 18.5% to 67.7% for MAPT and from 38.0% to 76.3% for SPA. The percentage of BALF retrieved was significantly (P = 0.001) higher for the SPA technique, compared with the percentage of BALF retrieved for the MAPT technique (median difference, 16.2%; IQR, 8.8%). Median total nucleated cell count for samples obtained via MAPT was 600 cells/μL (range, 400 to 1,400 cells/μL), and median total nucleated cell count for samples obtained via SPA was 900 cells/μL (range, 400 to 2,000 cells/μL); these values did not differ significantly.

Microscopic assessment of BALF—All microscopic scores for cellular preservation and cellularity were 3 or 4. Analysis of sample-quality variables revealed identical values for cellular preservation scores for both MAPT and SPA in individual dogs. Cellularity and RBC scores were not significantly different between MAPT and SPA. Erythrophagocytosis and intracellular pigment in macrophages were not detected. A significantly (P = 0.050) higher epithelial cell score was identified in BALF obtained via MAPT (median difference, 1.0; IQR, 1.0). Evidence of oropharyngeal contamination (squamous epithelial cells) was identified in 3 of 22 BALF samples.

Bacteria observed on cytologic examination were extracellular and single or in small colonies. There was no significant difference in bacterial scores between BALF obtained with the MAPT and SPA techniques.

No significant differences were identified in percentages or absolute differential cell counts of eosinophils with round nuclei (round eosinophils), eosinophils with segmented nuclei (segmented eosinophils), mast cells, or lymphocytes. Samples obtained via SPA had a significantly higher percentage (median difference, 7%; IQR, 14%; P = 0.008) and absolute number (median difference, 74 cells/μL; IQR, 217 cells/μL; P = 0.014) of neutrophils, compared with values for samples obtained via MAPT. Samples obtained via SPA had a significantly lower percentage (median difference, 11%; IQR, 13%; P = 0.007), but not absolute number (median difference, 106 cells/μL; IQR, 260 cells/μL; P = 0.460), of macrophages, compared with values for samples obtained via MAPT. When data were excluded from the analysis for 1 dog because unilateral localized pulmonary disease was identified during bronchoscopy, differential cell counts still differed significantly.

Discussion

Studies^8,9,14 in human and veterinary medicine have indicated the manner in which BAL techniques may influence sample quality and interpretation. For example, filtering and centrifugation of human BALF can decrease sample cellularity by up to 34% and may alter the differential cell count.⁸ Analysis of results of a recent clinical trial performed in healthy Beagles suggested that use of weight-based aliquot volumes provided more consistent BALF samples than did use of fixed volume boluses.⁹ Protocols for BAL in horses were standardized a few years ago with widespread implementation,¹⁵ which allows for better comparisons among studies. Such a unified approach has been lacking in small animal medicine; hence, we performed the present study to critically assess 2 currently used methods for fluid retrieval.

We chose to assess the percentage of fluid retrieved, presence of surfactant, sample cellularity, cellular preservation, and presence of RBCs and epithelial cells as our main indicators of BALF quality. These criteria were selected on the basis of currently available evidence in the human and veterinary literature.^5,13,14,16

A higher percentage of retrieved infusate supports the contention that the sample represents fluid from the alveoli.^12,13 Small amounts of retrieved fluid in humans may only represent a sample from the large airways and therefore be considered bronchial wash and not BALF.^10,17 A minimum of 33% fluid retrieved has been suggested as an adequate BAL procedure in human medicine.¹² Recovery of 40% to 75% of the BAL infusate has been reported in small animal medicine.^4,5 Three of 11 samples obtained via MAPT in the present study comprised < 33% of the original lavage volume; therefore, these samples would have been considered nondiagnostic. In contrast, all samples obtained via SPA comprised > 33% of the infusate; these values differed significantly.

A sudden loss of negative pressure was reported with the MAPT technique for 1 dog, and the BALF sample for that attempted collection was less than the 33% retrieval criterion. In a clinical patient, a third BAL aliquot would have been used at this site, and it is possible that a third aliquot would have resulted in a diagnostic sample for MAPT for this dog. Decreased fluid retrieval for the MAPT technique may have been related to punctures in the polyethylene tubing made by the hypodermic needle, which would result in sudden loss of negative pressure. The polyethylene tubes of the 2 dogs in which MAPT was associated with sudden loss of negative pressure were visually inspected after the procedure, and no punctures were detected.

Decreased fluid retrieval could also result from the inability to wedge the polyethylene tubing in a bronchus. For example, if the diameter of the polyethylene tubing was too narrow to occlude a bronchus, the ability to achieve negative pressure during aspiration would rely on wedging of the bronchoscope in the bronchus. It is possible that the sudden loss of negative pressure when aspirating during MAPT was secondary to the bronchoscope being withdrawn from the bronchus with loss of occlusion. In addition, if the polyethylene tubing was not occluding a bronchus, some of the BAL aliquot could move proximal to the tube opening instead of distal into the alveoli. This would decrease infusate retrieval. Occlusion of the tube against the bronchus wall would also greatly hamper fluid retrieval. A final possibility for decreased infusate retrieval is kinking of the tubing, but we were careful to ensure that there was no kinking of the tubing proximal to the bronchoscope biopsy channel.

Retrieval of fluid with the SPA technique may not be dependent on occlusion created by the bronchus around the tip of the bronchoscope.⁴ If the bronchoscope tip is not wedged perfectly in a bronchus, the SPA technique may result in increased fluid retrieval as a result of increased suction, compared with the lower achievable negative pressure of the MAPT. The increased suction associated with SPA may result in collapse of the airways proximal to the bronchoscope tip, thus creating an occlusion at the area of airway collapse. Therefore, SPA has the potential to increase the diagnostic yield of BAL in larger canine patients where the bronchoscope tip cannot be completely wedged in a bronchus.

Surfactant is secreted by type II pneumocytes into the alveoli and reduces alveolar surface tension.¹⁸ The presence of surfactant, as assessed via the presence of foam, in BALF confirms that a sample has been obtained from the alveoli.^4,5 In the present study, the presence of surfactant was assessed indirectly by measuring the volume of bubbles present in BALF samples immediately after collection. It is currently unknown whether the method of fluid retrieval influences the amount or characteristics of foam and therefore the volume of bubbles in BALF. However, the presence of foam rising to the top of BALF has been reported as a consistent finding for both manual and SPA techniques.^4,5 On the basis of this assessment, all samples obtained via SPA had positive results for the presence of surfactant; however, 2 of 11 samples obtained via MAPT had negative results for the presence of surfactant and would have been considered nondiagnostic. It is possible that BALF obtained via MAPT contained surfactant but that it was not grossly visible as bubbles. Direct measurement of surfactant in BALF samples would have strengthened the analysis for the present study; however, after allocating BALF for other analyses, inadequate volumes of BALF were available to measure phospholipid via the Bligh and Dyer method.¹⁹

In the present study, limits for cellularity and the cellular preservation scores were extrapolated from values reported in the veterinary clinical pathology literature.^5,13,20 A differential count on a minimum of 200 cells is required for high reproducibility of the proportion of neutrophils, alveolar macrophages, and eosinophils in cytocentrifuged canine BALF samples.²⁰ A differential count on 500 cells is required for high reproducibility of the proportion of lymphocytes.²⁰ However, there is no information available in the veterinary literature on what constitutes a minimally acceptable quality of preparation. A score of 2 for cellularity and cell preservation would be marginal. However, given that most cells in such a preparation can still be classified despite less-than-perfect cell preservation, we considered such preparations adequate. The MAPT and SPA techniques yielded BALF samples of adequate cellularity and cellular preservation, and there was no significant difference in scores between the techniques. It has been hypothesized that SPA may result in increased cellular damage, compared with cellular damage for samples obtained via MA and MAPT, because of higher and more variable negative pressures.⁵ It is interesting that individual cellular preservation scores for samples obtained via MAPT and SPA were identical for this group of healthy Beagles. Analysis of these results suggested that SPA was an adequate method for recovery of BALF for cytologic analysis.

Epithelial cells and RBCs were considered markers of BALF samples with poor quality.¹³ Large numbers of epithelial cells (> 5% of nucleated cells)¹⁰ may be present in BALF because of excessive suction, abnormal exfoliation, or trauma. Predominance of cuboidal and columnar epithelial cells indicates samples collected from the trachea and bronchi, whereas superficial squamous epithelium cells and large bacteria (Simonsiella spp) indicate oropharyngeal contamination.¹³ Evidence of hemorrhage is rarely found in BALF unless iatrogenic trauma from the BAL procedure results in bleeding. Detection of erythrophagocytosis or hemosiderin during cytologic evaluation can be used to differentiate in vivo pulmonary bleeding from iatrogenic trauma.¹³ There was no difference between the 2 aspiration techniques with regard to the proportion of RBCs in the BALF. A higher proportion of epithelial cells was present in samples obtained via MAPT. This could have resulted from trauma secondary to the polyethylene tubing, considering that iatrogenic linear abrasions were identified along the bronchial walls in 2 dogs following wedging of the polyethylene tubing. It could also have been related to the lower amount of infusate retrieved via MAPT, considering that bronchial washings and low-volume BAL aliquots have a higher proportion of epithelial cells and neutrophils.^13,17

The percentage of neutrophils and alveolar macrophages was significantly different in the BALF obtained via MAPT and SPA. A higher proportion of neutrophils and a lower proportion of macrophages were identified in the BALF obtained via SPA. The absolute neutrophil count was also significantly increased in BALF obtained via SPA. Increased numbers of neutrophils may result from underlying inflammatory pulmonary disease¹³; however, if any dogs in the present study had pulmonary inflammation, we would have expected an increased number of neutrophils in BALF obtained via both MAPT and SPA because other studies^13,21 have revealed similar differential cell counts among different lung lobes. Therefore, we suspect that the difference in differential cell counts was related to the aspiration technique. Polymorphonuclear cells may adhere to tubing, thus decreasing the number of neutrophils in samples obtained via MAPT. Alternatively, evidence in humans indicates that bronchial washings contain a higher proportion of neutrophils and epithelial cells than does fluid from the alveoli.^10,14,17 It is possible that the higher proportion of neutrophils in samples obtained via SPA reflected samples predominantly from the bronchi; however, we believe that this was unlikely, given that a concurrent increase in epithelial cells was not identified in BALF obtained via SPA.

Strengths of the present study included randomization of sample sites and order of sample collection for the aspiration techniques. The same team of investigators (KSW and AMND) performed the BAL procedures to reduce error and enhance precision. Although the investigators were aware of the order of the BAL techniques and location from which the samples were obtained, the veterinary clinical pathologist (DB) was not aware of the origin of the samples when performing the analysis.

The 2 BAL aliquots for each aspiration technique were pooled for analysis in the present study. There is controversy in human and veterinary medicine regarding pooling of sequential BALF aliquots. In humans, the initial BALF aliquot is considered to represent the bronchial airways more than the alveoli; however, pooling the initial BALF aliquot with subsequent aliquots does not significantly change the overall results.¹⁴ In an equine study,¹⁵ investigators identified a higher proportion of mast cells in the initial 100-mL BALF aliquot; however, the clinical relevance of this difference is unknown, and the equine BAL standardization group continues to recommend pooling of samples for analysis.¹⁵ In a cytologic study²² of feline BAL samples, investigators found no difference in results of a pooled sample versus results for 3 aliquots. Furthermore, pooling of samples was recommended in a recent review on airway evaluation in small animals⁴; hence, we decided to pool sequential BAL aliquots in the present study.

Other limitations of the present study included the relatively small sample size and that the dogs were healthy and of similar age, body weight, and breed. We chose middle-aged dogs for the study because age affects composition of BALF.²³ This might limit applicability of the results to dogs with substantial differences in signalments and dogs with pulmonary parenchymal disease. For example, bronchoscopy and BAL in larger dogs may require the use of a longer gastrointestinal scope instead of a bronchoscope. In addition, an increased propensity for airway collapse is observed with many diseases (such as bronchomalacia and chronic bronchitis) of the lower portion of the respiratory tract.^1,5 Infusate retrieval may be hampered if excessive negative pressure associated with aspiration results in airway collapse. Indeed, decreased fluid retrieval has been reported in humans with obstructive lower respiratory tract disease.¹² Also, the maximum negative pressure was empirically set at 5 kPa (37.5 mm Hg) on the basis of the technique used in human patients,¹⁴ but it is possible that negative pressures of 10 to 25 mm Hg could yield better samples because of a decrease in the incidence of airway collapse.

Another potential limitation of the present study was the use of sterile polyethylene tubing as part of the MAPT technique. Most reports of clinical trials that involve BAL in small animals define MA as a syringe attached directly to a bronchoscope biopsy channel. However, both techniques for MA are described in a veterinary internal medicine textbook,¹ and investigators for a prospective clinical trial in humans reported that MAPT resulted in a greater amount of BALF retrieved and a higher diagnostic yield than did MA.²⁴ Further studies are warranted to compare the use of MA without sterile polyethylene tubing in healthy dogs and to evaluate SPA and MAPT in larger dogs and in dogs with respiratory tract disease.

Analysis of results for the present study indicated that in healthy dogs, SPA provided better retrieval of BALF and BALF of higher quality than did MAPT. The SPA technique may improve the rate of diagnostic success for BAL in dogs, compared with the rate of diagnostic success for samples obtained via MAPT; however, further evaluation of these aspiration techniques is required in dogs with respiratory tract disease. In addition, analysis of the results suggested that expected differential cell counts between samples obtained with the MAPT and SPA techniques differ, with a lower percentage of neutrophils for samples obtained with MAPT. These findings further support that BAL protocols should be standardized to obtain results comparable among studies.