Abstract
OBJECTIVE To quantify the effect of time and recumbency on CT measurements of lung volume and attenuation in healthy cats under general anesthesia.
ANIMALS 8 healthy research cats.
PROCEDURES Anesthetized cats were positioned in sternal recumbency for 20 minutes and then in left, right, and left lateral recumbency (40 minutes/position). Expiratory helical CT scan of the thorax was performed at 0 and 20 minutes in sternal recumbency and at 0, 5, 10, 20, 30, and 40 minutes in each lateral recumbent position. For each lung, CT measurements of lung volume and attenuation and the extent of lung areas that were hyperaerated (−1,000 to −901 Hounsfield units [HU]), normoaerated (−900 to −501 HU), poorly aerated (−500 to −101 HU), or nonaerated (−100 to +100 HU [indicative of atelectasis]) were determined with a semiautomatic threshold-based technique. A restricted maximum likelihood analysis was performed.
RESULTS In lateral recumbency, the dependent lung had significantly greater attenuation and a lower volume than the nondependent lung. Within the dependent lung, there was a significantly higher percentage of poorly aerated lung tissue, compared with that in the nondependent lung. These changes were detected immediately after positioning the cats in lateral recumbency and remained static with no further significant time-related change.
CONCLUSIONS AND CLINICAL RELEVANCE Results indicated that once anesthetized healthy cats were positioned in lateral recumbency, the dependent lung lobes underwent a rapid reduction in lung volume and increase in lung attenuation that did not progress over time, predominantly attributable to an increase in poorly aerated lung tissue.
From results of a study in humans,1 it is known that the physical density of the lung is determined by 3 components: lung tissue, blood, and gas. The relative proportions of these 3 components are closely associated with changes in lung volume, whereby a reduction in lung volume leads to an increase in the overall physical density of the lung.1–4 Pronounced changes in lung density and volume occur commonly during anesthesia or recumbency, situations where there is a significant and rapid increase in lung density and a significant reduction in lung volume within the dependent lung regions.2–15 The most common and well-described result of these lung changes is the development of pulmonary atelectasis, with atelectasis being defined in CT images as pixel attenuation values of −100 to +100 HU.2–6,9 In CT examinations, pulmonary atelectasis was the most prevalent pulmonary abnormality for 144 of 352 (41%) cats without respiratory tract signs.16 Because veterinary patients are commonly anesthetized and positioned in LR, SR, or DR for procedures such as thoracic radiography and CT, these lung changes have clinically important implications. Pulmonary atelectasis can complicate the interpretation of diagnostic imaging in 2 ways. Pulmonary atelectasis is potentially an incidental finding, but can mimic pathological lesions leading to misdiagnosis of pulmonary abnormalities.17,18 Also, pulmonary atelectasis may mask evidence of intrathoracic disease and complicate interpretation of thoracic radiographic and CT findings.16,17 Although the lung changes related to recumbency may challenge the diagnosis of lung disease, they could also provide useful functional information. By application of the concept of gravitational differences in lung attenuation, lateral decubitus CT has been successfully used as an alternative method to obtain expiratory scans in humans who are unable to follow breath-holding instructions. By accentuating the differences in lung attenuation between diseased lung tissue and atelectasis, this technique has allowed detection of air-trapping in human patients for whom CT scans obtained in a supine position were inconclusive.19–21
Computed tomography is considered the gold standard for quantification of changes in lung density and volume. On the basis of differences in attenuation (expressed in HU), it is possible to distinguish areas of lung tissue that are hyperaerated (−1,000 to −901 HU), normoaerated (−900 to −501 HU), poorly aerated (−500 to −101 HU), or nonaerated (−100 to +100 HU [indicative of atelectasis]).6,14 In the past decade, the application of this quantitative CT analysis of lung volume and attenuation in veterinary medicine has continued to expand, thereby improving our understanding of changes in pulmonary attenuation.15,19 However, there is a lack of information in veterinary medicine regarding quantification of lung volume and attenuation in relation to different body positions and changes in those variables over time, particularly for cats. The purpose of the study reported here was to use CT to quantify lung volume and attenuation changes over time and to estimate the prevalence of atelectasis (attenuation, −100 to +100 HU) in anesthetized healthy cats in different body positions. Our hypotheses were that there would be a significant increase in lung attenuation with development of atelectasis within the dependent lung during LR, compared with findings during SR, and that this change would increase in severity over time.
Materials and Methods
Animals
Eight healthy adult domestic shorthair cats belonging to an experimental animal providera were selected and transported to the University of Sydney for the duration of the prospective experimental study. The cats were allowed a period of acclimation prior to commencement of the study and also a period of monitored recovery after completion of the study before being returned to the source. The study was approved by the Animal Ethics Committee of the University of Sydney.
Cats were considered healthy on the basis of no history of respiratory tract disease and results of a physical examination (cardiac and pulmonary auscultation, abdominal palpation, and measurements of rectal temperature, heart rate, respiratory rate, and pulse rate) that were within reference ranges. For each cat, no abnormalities were detected during thoracic CT without contrast agent administration. In addition, a CBC, serum biochemical analyses, and assessment of serum total thyroxine concentration to detect major abnormalities that would preclude any cat from undergoing general anesthesia were performed. Minor serum biochemical abnormalities (including mildly high activities of creatine kinase and alanine aminotransferase and concentrations of total protein, sodium, and potassium) were not considered important. Among the 8 cats, there were 5 neutered males, 1 sexually intact male, and 2 spayed females. The cats' median age was 4 years (range, 1 to 6 years) and median weight was 4.15 kg (range, 3.4 to 5.8 kg). All cats were considered to be in good body condition (5 to 7 as determined on a 9-point scale).
Study design
Each cat was anesthetized and positioned in SR for 20 minutes and then positioned in LR for three 40-minute periods (ie, LLR1, RLR, and then LLR2). Helical CT examination of the thorax was performed at 0 and 20 minutes when cats were in SR and at 0, 5, 10, 20, 30, and 40 minutes when cats were in each LR position. The 0-minute time point was defined as the time that cats were first placed in each position. Each scan was acquired in expiratory pause, and all cats were spontaneously breathing. Following the CT examination, the cats were allowed to recover from anesthesia routinely. The CT images were then sent to the hospital picture archiving system for analysis at a later time.
Anesthesia
Food but not water was withheld from each cat for 12 hours before anesthesia. Thirty minutes prior to the CT examination, each cat was sedated with alfaxaloneb (2 mg/kg), acepromazine maleatec (0.02 mg/kg), butorphanol tartrated (0.2 mg/kg), and medetomidine (10 μg/kg)e administered IM. Fifteen minutes later, a 22-gauge catheter was aseptically placed into a cephalic vein, and alfaxaloneb was administered IV to effect. Prior to endotracheal intubation, 0.1 mL of lidocaine hydrochloridef diluted to 1% was administered topically to the larynx of each cat to prevent laryngeal spasm and facilitate intubation. Following endotracheal intubation, anesthesia was maintained with isofluraneg (1.0% to 1.5%) in oxygen through a Bain nonrebreathing circuit (flow rate, 1 L/min). Throughout anesthesia, Hartman solutionh was administered IV at a rate of 5 mL/kg/h. During the experimental procedure, heart rate, respiratory rate, arterial blood pressure, peripheral hemoglobin oxygen saturation as measured by pulse oximetry, rectal temperature, and end-tidal CO2 concentration were monitored.
CT examination
All CT scans (regardless of the cats' position) were acquired with the same 16-slice multidetector CT scanner.i Transverse CT images of the thorax of each cat were acquired with the same settings as follows: 120 kVp; 150 mAs; collimation, 16 × 1.5 mm; gantry rotation speed, 0.5 seconds; field of view, 180 mm; matrix, 512 × 512 pixels; slice thickness, 2 mm (with 1-mm increments); and collimator pitch, 1.438. A high-resolution algorithm and lung reconstruction filter were used. Acquisition times varied between 3.045 and 3.494 seconds. Timing of each scan was coordinated by visualization of a mainstream capnography (no time delay) trace with the scan obtained during the expiratory pause. The CT scan was repeated immediately if there was notable motion artifact that resulted in a poor-quality scan.
CT lung segmentation
A single observer (TSF) performed all measurements and segmentation with an image analysis workstationj,k and software.l For each helical CT acquisition, the lungs were segmented from the rest of the thorax as an ROI and further separated into the right and left lung fields. This was achieved in 4 stages. First, the lungs were segmented from the surrounding structures with a 3-D growing ROI tool by setting a threshold attenuation range of −1,280 to −200 HU (Figure 1). Second, the trachea was then manually removed from the ROI down to the level of the mainstem bronchi. Third, the segmented lungs were then separated manually into the left lung and right lung fields. The accessory lung lobe was included in the right lung field. Fourth, within each lung field, regions between −199 and +100 HU were manually included into the ROI by use of the brush tool taking care to exclude major blood vessels to the level of the primary bronchi. The end result was separate ROIs of the left and right lung fields (excluding the trachea to the level of the mainstem bronchi and major blood vessels) with an attenuation range of −1,280 to +100 HU for each helical CT acquisition.
Illustration of the semiautomatic segmentation of both lung fields as an ROI (green) as determined for 1 of 8 cats used in a study to quantify the effect of time and recumbency on CT measurements of lung volume and attenuation in healthy anesthetized cats. In a representative CT image, the lungs are segmented from the surrounding structures with a 3-D growing ROI tool by setting a threshold attenuation range of −1,280 to −200 HU. The trachea was then manually removed from the ROI to the level of the mainstem bronchi. The segmented lungs were then separated manually into the left lung and right lung fields; the accessory lung lobe was included in the right lung field. Regions of −199 to +100 HU within each lung field were manually included into the ROI by use of the brush tool (excluding major blood vessels to the level of the primary bronchi). Thus, separate ROIs for the left and right lung fields (excluding the trachea to the level of the mainstem bronchi and major blood vessels) with an attenuation range of −1,280 to +100 HU for each helical CT acquisition were obtained. Lung volume (expressed as cm3) was determined by use of an image analysis workstation and software; segmented lung ROIs were analyzed to determine lung attenuation (HU). In the present study, a classification scheme12 was used for lung attenuation analysis as follows: hyperaerated (−1,000 to −901 HU), normoaerated (−900 to −501 HU), poorly aerated (−500 to −101 HU), and nonaerated (−100 to +100 HU) lung tissue.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.874
Quantitative CT analysis
Lung volume was calculated for the segmented right or left lung by use of the image analysis workstation and software. A compute-volume tool was used to calculate the lung volume (expressed as cm3). The segmented lung ROIs were exported into a multi-image analysis graphic user interface toolm for calculation of lung attenuation. By means of a generatehistogram tool, lung attenuation (expressed as HU) histograms of the segmented lung ROIs were generated. The raw data from each histogram were then exported into a spreadsheetn for statistical analysis. The following classification scheme12 was used for lung attenuation analysis: hyperaerated (−1,000 to −901 HU), normoaerated (−900 to −501 HU), poorly aerated (−500 to −101 HU), and nonaerated (−100 to +100 HU) lung tissue.
Statistical analysis
Mean lung volume and attenuation value of the dependent lungs were compared with those of the nondependent lungs across all lateral positions. The volume and attenuation value of the left lung were each compared across all positions. Similarly, the volume and attenuation value of the right lung were each compared across all positions. The lung volume and attenuation value of the dependent lung was also compared across all lateral positions. Data were analyzed with a restricted maximum likelihood procedure.o Fixed effects considered for inclusion were position, time, and dependent or nondependent status for analyses with volume (cm3) and attenuation (HU) as the outcomes and were position, time, dependent or nondependent status, attenuation, and right or left lung for lung attenuation analysis. Percentages were logarithmically (loge) transformed before analysis to maintain the assumption of normality. Cat was included as a random effect in each model. A value of P < 0.05 was considered significant for all analyses. Posthoc analyses involving least significant differences were conducted to determine the significance of pairwise differences.
Results
The cats' condition remained stable throughout the duration of the study as determined on the basis of heart rate, respiratory rate, arterial blood pressure, hemoglobin oxygen saturation, rectal temperature, and end-tidal CO2 concentration. These variables remained within reference limits, and no complications were recorded.
Overall attenuation and volume analysis
There was a significant (P = 0.014) difference in overall attenuation and volume between the dependent and nondependent lungs when all data for the dependent lung were compared with those for the nondependent lung for all lateral positions (ie, during LLR1, RLR, and LLR2). There was a significantly higher mean attenuation (−565.5 HU [SD, 41.3 HU]) and lower mean volume (49.8 cm3 [SD, 12.2 cm3]) for the dependent lung, compared with the mean attenuation (−641.7 HU [SD, 31.4 HU]) and mean volume (72.9 cm3 [SD, 14.6 cm3]) for the nondependent lung in each position. There were no significant differences in mean lung attenuation or mean volume (P = 0.978 and 0.998 respectively) among scans performed at 0, 5, 10, 20, 30, and 40 minutes in each position. These findings indicated that the maximal changes in attenuation and volume within the dependent and nondependent lungs occurred rapidly, and those variables did not significantly progress over time (Figure 2). Given that there were no significant differences in mean attenuation or mean volume over time when cats were in SR or during LLR1, RLR, or LLR2, further comparisons were performed with means calculated for each period of recumbency from data obtained at all time points during that period.
Mean lung volume (A) and attenuation (B) of the right (red line) and left (blue line) lungs of 8 healthy anesthetized cats that were positioned in SR for 20 minutes and then underwent LLR1, RLR, and LLR2 (40 min/position) during which expiratory helical CT scans of the thorax were performed (at 0 and 20 minutes when cats were in SR and at 0, 5, 10, 20, 30, and 40 minutes in each LR position). At the time that each cat's position was modified, there is a significant and rapid decrease in volume and increase in attenuation for dependent lungs with no further significant change over time (up to 40 minutes). See Figure 1 for remainder of key.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.874
Comparison of the left lung variables when cats were in SR and during LLR1, RLR, and LLR2 revealed that the left lung had a significant increase in attenuation during LLR1 (−545.9 HU) and LLR2 (−560.4 HU), compared with the value during SR (−639.7 HU). There was also a significant decrease in volume of the left lung during LLR1 (42.1 cm3) and LLR2 (44.4 cm3), compared with the value during SR (58.8 cm3). During RLR, the left lung had a significant decrease in attenuation (−636.1 HU) and increase in volume (57.7 cm3), compared with data obtained during LLR1 and LLR2. There was no significant decrease in attenuation or increase in volume of the left lung during RLR, compared with findings when cats were in SR, indicating recovery but no significant hyperinflation of the nondependent left lung. During LLR1 and LLR2, there were no significant differences in attenuation or volume of the left lung.
Comparison of the right lung variables when cats were in SR and during LLR1, RLR, and LLR2 revealed that the right lung had a significant increase in attenuation during RLR (−590.4 HU), compared with the value during SR (−635.2 HU). There was also a significant decrease in the volume of the right lung during RLR (62.9 cm3), compared with the value during SR (76.1 cm3). During LLR1 and LLR2, the right lung had a significant decrease in attenuation (LLR1, −644.6 HU; LLR2, −644.4 HU) and increase in volume (LLR1, 8 0.5 c m 3; LLR2, 80.6 cm3), compared with data obtained during RLR. There was no significant decrease in attenuation of the right lung during LLR1 and LLR2, compared with findings when cats were in SR. However, there was a small but significant increase in the volume of the right lung during LLR1 and LLR2, compared with the lung volume during SR, indicating recovery and also a degree of hyperinflation of the nondependent right lung. During LLR1 and LLR2, there were no significant differences in attenuation or volume of the right lung.
Data obtained for the dependent lung during LLR1, RLR, and LLR2 indicated that the right lung had a significantly lower attenuation during RLR (−590.4 HU), compared with findings for the left lung during LLR1 (−545.9 HU) and LLR2 (−560.4 HU). Thus, it appeared that the dependent right lung (during RLR) underwent a less pronounced decrease in attenuation, compared with the dependent left lung (during LLR1 or LLR2). The dependent right lung also had a significantly higher volume (62.9 cm3), compared with values for the dependent left lung during LLR1 and LLR2; however, this difference was likely attributable to the overall higher volume of the right lung.
Sternal recumbency resulted in the lowest overall lung attenuation value (−636.4 HU), compared with overall values during all other periods of recumbency (LLR1, −595.3 HU; RLR, −608.1 HU; and LLR2, −602.4 HU). The volume of the right (76.1 cm3) and left (58.8 cm3) lungs in SR differed significantly; however, there was no significant difference in attenuation (left lung, −639.7 HU; right lung, −635.2 HU). This difference in volume between the left and right lungs was again attributed to the naturally larger volume of the right lung.
Lung attenuation classification
Similar to comparisons of lung attenuation and volume, all attenuation classification analyses were performed with means calculated for each period of recumbency from data obtained at all time points during that period. During LLR1, the percentage of poorly aerated lung tissue (30.36%) was significantly increased, compared with that when cats were in SR (16.44% [Table 1]). This was visualized as a shift of the HU curve toward the right (Figure 3). During LLR2, the percentages of poorly aerated (27.52%) and nonaerated areas of the left lung (1.82%) were significantly increased, compared with the findings when cats were in SR (16.44% and 1.18%, respectively). During RLR, the left lung had a significantly higher percentage of hyperaerated tissue (0.42%), compared with that when cats were in SR (0.26%) or during LLR1 (0.26%), and a significantly lower percentage of poorly aerated tissue (17.01%), compared with that during LLR1 (30.36%) or LLR2 (27.52%). Other comparisons of left lung attenuation in relation to periods of recumbency yielded no significant differences.
Representative histograms to illustrate the distribution of HU values for the left lung of 1 of the 8 study cats in Figure 2 during SR (A), immediately after the start of LLR1 (B), and at 40 minutes of the LLR1 (C). Notice the significant shift of the curve toward the right (increase in HU) as soon as the cat is transitioned into LLR. At the 40-minute point of LLR1, there is no further significant change in HU. Notice that the degree of true atelectasis (attenuation of lung tissue, −100 to +100 HU) is not notable at any time point. See Figures 1 and 2 for remainder of key.
Citation: American Journal of Veterinary Research 79, 8; 10.2460/ajvr.79.8.874
Mean percentage of hyperaerated, poorly aerated, normoaerated, and nonaerated tissue of the right and left lungs of 8 healthy anesthetized cats that were positioned in SR for 20 minutes and then underwent LLR1, RLR, and LLR2 (40 min/position).
| Position | Aeration classification | Right lung (%) | Left lung (%) |
|---|---|---|---|
| SR | Hyperaerated | 0.24a | 0.26b |
| Poorly aerated | 16.64 | 16.44a,c | |
| Nonaerated | 1.19 | 1.18a | |
| Normoaerated | 81.45 | 81.86 | |
| LLR1 | Hyperaerated | 0.33 | 0.26d |
| Poorly aerated | 15.89d | 30.36d | |
| Nonaerated | 1.18 | 1.78 | |
| Normoaerated | 82.19 | 65.17 | |
| Poorly aerated | 25 | 17.01e | |
| Nonaerated | 1.5 | 1.38 | |
| Normoaerated | 72.6 | 80.56 | |
| LLR2 | Hyperaerated | 0.43 | 0.3 |
| Poorly aerated | 16.02 | 27.52 | |
| Nonaerated | 1.23 | 1.82 | |
| Normoaerated | 81.94 | 68.51 |
Expiratory helical CT scans of the thorax were performed at 0 and 20 minutes when cats were in SR and at 0, 5, 10, 20, 30, and 40 minutes in each LR position. For the CT images, the lungs were segmented from the surrounding structures with a 3-D growing ROI tool by setting a threshold attenuation range of −1,280 to −200 HU. The trachea was then manually removed from the ROI to the level of the mainstem bronchi. The segmented lungs were then separated manually into the left lung and right lung fields. The accessory lung lobe was included in the right lung field. Regions of −199 to +100 HU within each lung field were manually included into the ROI by use of the brush tool (excluding major blood vessels to the level of the primary bronchi). Thus, separate ROIs for the left and right lung fields (excluding the trachea to the level of the mainstem bronchi and major blood vessels) with an attenuation range of −1,280 to +100 HU for each helical CT acquisition were obtained. By use of an image analysis workstation and software, segmented lung ROIs were analyzed to determine lung attenuation (HU). All attenuation classification analyses were performed with means calculated for each period of recumbency from data obtained at all time points during that period. A classification scheme12 was used for lung attenuation analysis as follows: hyperaerated (−1,000 to −901 HU), normoaerated (−900 to −501 HU), poorly aerated (−500 to −101 HU), and nonaerated (−100 to +100 HU) lung tissue.
For a given lung field, the value obtained when the cats were in SR differed significantly from the value obtained during LLR2.
For a given lung field, the value obtained when the cats were in SR differed significantly from the value obtained during RLR.
For a given lung field, the value obtained when the cats were in SR differed significantly from the value obtained during LLR1.
For a given lung field, the value obtained during LLR1 differed significantly from the value obtained during RLR.
For a given lung field, the value obtained during RLR differed significantly from the value obtained during LLR2.
During RLR, the percentage of poorly aerated lung tissue in the right lung was significantly greater than that during LLR1 (15.89% [Table 1]). During LLR2, the percentage of hyperaerated lung tissue of the right lung (0.43%) was significantly greater than that identified when the cats were in SR (0.24%). Other comparisons of right lung attenuation in relation to periods of recumbency yielded no significant differences.
Discussion
Compared with findings when cats were in SR, the dependent lung lobes during LR underwent a rapid and significant reduction in overall lung volume and an increase in attenuation, changes that were compatible with a reduction in pulmonary aeration. The changes in dependent lung lobe volume and attenuation occurred immediately after positioning the cats in LR and then remained static with no further significant change in either variable throughout each of the three 40-minute periods of LR. It was previously assumed that prolonged LR results in a greater degree of recumbency-associated atelectasis.17,22 This was not proven in the present study, given that a reduction in pulmonary aeration within the dependent lung occurred immediately following repositioning and did not worsen over time. A recent study14 in dogs revealed a similar rapid reduction in aeration of the dependent lung, wherein a reduction in aeration of the dependent lung lobes was detected within 3 to 8 minutes after positioning in LR and was associated with little progression over time. From the results of the present study, it is unclear whether prolonged LR (> 40 minutes) would lead to a further reduction in pulmonary aeration and increase in the degree of atelectasis. Given the lack of progressive change in those variables in the study cats over time during LLR1, RLR, or LLR2, we think it is unlikely that the degree of pulmonary aeration would decrease further with LR of > 40 minutes' duration. Moreover, placement of research cats in LR for periods > 40 minutes may be neither necessary nor ethically justifiable.
The present study was designed to maximize the potential formation of atelectasis (nonaerated lung) in the study cats. This was achieved by use of inhalation anesthesia23 with 100% inspired oxygen concentration15 and acquisition of helical CT scans during the expiratory pause. Despite these factors, analysis of the lung attenuation classification revealed that the overall percentage of atelectasis formation during SR or LR in the present study was small (up to 1.82%). The most significant change in the dependent lung in LR was an increase in the percentage of poorly aerated lung tissue. Although there was a significant increase in nonaerated lung tissue within the dependent left lung lobe during LLR1, compared with the percentage when cats were in SR, this difference was small (1.82% vs 1.18%). Differences in the percentage of nonaerated lung tissue among other positions were not significant. These findings suggested that the rapid increase in lung attenuation observed in the dependent lungs during LR is primarily a result of an increase in the percentage of poorly aerated lung tissue, rather than a significant increase in the percentage of nonaerated lung tissue. Results of the present study are consistent with those of a recent investigation14 that used a cross-sectional quantitative evaluation method in dogs, which determined that atelectasis formation (−100 to +100 HU) did not occur in healthy anesthetized dogs maintained in LR. This finding was in contrast to results of previous studies that revealed a high prevalence of atelectasis in anesthetized cats placed in DR15 or SR.16 A reason for this discrepancy may be that atelectasis was measured subjectively in the previous studies15,16 in contrast to the present study wherein atelectasis was objectively measured and defined as lung attenuation between −100 to +100 HU.4,12 It is possible that subjective assessment of lung attenuation may have resulted in areas of poorly aerated lung tissue (≤ −101 HU) being interpreted as atelectasis. In another study15 of anesthetized cats, atelectasis was measured objectively, and results indicated that the degree of atelectasis formation was higher than that determined in the cats of the present study. Two important methodological differences may explain this discrepancy. First, the cats in that study15 underwent CT following routine ovariohysterectomy resulting in a significantly longer time before the scans were performed (mean interval, 62.2 minutes), compared with procedures in the present study. It is possible that the increased duration of anesthesia may have contributed to the higher degree of atelectasis in the cats that underwent surgery. However, given the lack of progression of the lung changes over time in the present study, we think that it is unlikely that the prolonged period of anesthesia contributed significantly to the difference in atelectasis formation. Second, the cats in that study15 were placed in DR for the duration of the procedure, unlike the protocol used in the present study in which cats were positioned in SR, LLR, and RLR. Although the effect of DR was not assessed in the present study, it is known that lung attenuation increases during DR, compared with the effects of placement in other recumbent positions.24,25
A recumbency-associated increase in lung attenuation or radiographic opacity can complicate the interpretation of CT or radiographic assessments of veterinary patients. It has always been assumed that this was a result of border effacement of areas of lung disease with atelectatic lung tissue.17 On the basis of the results of the present study, increased lung attenuation or radiographic opacity in diagnostic imaging studies may largely be attributable to an increase in the percentage of poorly aerated lung tissue, and not necessarily a result of increases in the degree of true atelectasis or percentage of nonaerated lung tissue. However, it is unclear whether the increase in lung attenuation found in the present study could still interfere with the identification of lung disease. Future studies to investigate lesion detection in association with the dependent-nondependent status of lungs may provide useful information in this regard.
Compared with the data obtained when the cats of the present study were in SR, the significant reduction in pulmonary aeration of the dependent right lung during RLR was less pronounced than the significant reduction in pulmonary aeration of the dependent left lung during LLR1 or LLR2. This finding supported that of a previous study,14 in which there was no decrease in aeration of the right lung lobes in dogs placed in RLR but patchy areas of abnormally increased attenuation were infrequently detected in the left cranial lung lobe in dogs placed in LLR. Another recent study24 revealed that ground-glass opacities were detected more frequently and were more severe in CT images from dogs in LLR, compared with findings for other positions. The reasons for this are still unclear and at this stage speculative. The presence of a cardiac notch may reduce compression of the right middle lung lobe by the heart, leading to a smaller reduction in pulmonary aeration. The phrenicopericardial ligament may also act as an anchor to the apex, thereby limiting displacement of the heart and subsequent lung compression in RLR, compared with changes associated with LLR.
In the present study, the lowest overall lung attenuation was detected when cats were in SR. This was consistent with results of previous studies involving dogs, in which positioning in SR resulted in the highest functional residual capacity26 and least amount of quantifiable atelectasis,24 compared with findings for dogs positioned in LR. Among the cats of the present study, mild compensatory hyperinflation of the nondependent right lung was noted during LLR1 and LLR2 as evidenced by a significant increase in volume of the nondependent right lung, compared with right lung volume when cats were in SR. A similar effect was not observed in the left lung during RLR. A reason for lack of compensatory hyperinflation in the left lung may be the study design, wherein the cats were transitioned from SR into LLR1 first, resulting in the left lung undergoing a considerable reduction in pulmonary aeration before cats were transitioned into RLR. Whether cats undergoing RLR first would have a different effect on pulmonary aeration warrants further investigation. It may also be that the larger overall volume of the right lung allows for greater overall hyperinflation. Inclusion of the accessory lung lobe in the ROI for the right lung may have also contributed to an overall greater nondependent right lung volume because the accessory lung lobe appears to be less commonly affected by pulmonary atelectasis or collapse,16 compared with the other lung lobes. Although there was a significant increase in the percentage of hyperaerated lung tissue in the left lung during RLR, compared with findings when cats were in SR and during LLR1, and in the right lung during LLR2, compared with the percentage when cats were in SR, the magnitude of these changes was small and a concurrent significant decrease in overall attenuation was not noted.
In the present study, acquisition of diagnostic-quality helical CT scans during the expiratory pause and without breath-hold techniques was crucial to maximize the potential for atelectasis formation in the study cats. To attain this goal, experimental requirements included short CT scan times, slow respiratory rates in the cats, and accurate coordination of the timing of the CT scan with the expiratory pause. To achieve the short scan times, a high collimator pitch (1.438) and slice thickness (2 mm) were used, which resulted in scan times of 3.045 to 3.494 seconds. Slower respiratory rates were achieved by maintaining the cats under a deep plane of anesthesia. Mainstream capnography was used to remotely monitor the respiratory rate and breathing phases of the cats to achieve accurate coordination of the CT scan with the expiratory pause. Control of these factors resulted in diagnostic-quality CT scans of the thorax in the expiratory pause without notable motion artifact.
Although potentially confusing, the quantitative HU classification scheme used in the present study defines nonaerated lung tissue as areas with attenuation of −100 to +100 HU, which implies that there is retained air within the lung tissue. However, previous studies27–37 that used the same classification scheme indicated that the voxels within the nonaerated lung compartment are not strictly gas free because they have a gas-to-tissue ratio of between 1:10 and 0. This takes into account that even in complete small airway collapse, some gas remains in the pulmonary unit behind the collapsed bronchioles.27 This method of quantitative classification has been used extensively in human medical research27–36 and some veterinary medical investigations.11,12 In humans, the classification scheme has been widely used in the assessment of pulmonary function and disease as a separate and sometimes complementary tool to morphological subjective assessment. It is this quantitative classification that derives the most functional information in pulmonary pathophysiologic studies.32–37
The changes in pulmonary volume and attenuation in the dependent lung of the cats of the present study may reflect normal pulmonary collapsibility in healthy animals. This has potential future applications, particularly in research of COPD and emphysema because these conditions result in a reduction of pulmonary collapsibility.30 Studies31–37 in humans have revealed that quantitative CT indices of lung collapsibility are correlated with pulmonary function test results in patients with COPD; in those studies, lower lung collapsibility suggested more severe COPD and higher lung collapsibility suggested less severe COPD. This field of research into airway collapsibility and atelectasis of lungs with quantitative CT indices and correlation with pulmonary function tests in humans with COPD is particularly interesting in its potential application to veterinary patients. In dogs and cats, COPD is a difficult syndrome to diagnose because its clinicopathologic features overlap with those of other lower airway diseases7 and, in general, detection of any lower airway disease with conventional radiography is difficult.38,39 Acquisition of expiratory thoracic CT scans from veterinary patients is challenging and could be considered similar to the challenges of acquisition of expiratory thoracic CT scans from young children.20,21 Quantitative CT analysis of lung volume and attenuation in veterinary patients placed in LR could potentially replace the need to acquire expiratory scans to obtain information regarding pulmonary function and air trapping.19–21,28–37 Results of the present study may provide a starting point for investigation of the value of quantitative thoracic CT scans obtained from patients in LR as a functional tool in cases of COPD and emphysema. Such functional information may prove useful for the diagnosis, prognosis, and treatment monitoring of feline lung diseases.
There were several limitations to the present study including the small number of cats, lack of evaluation of intra- or interobserver repeatability for lung segmentation, effect of positioning sequence, and a deep plane of anesthesia. Given the small number of animals used in the study, there was insufficient power in the statistical analysis to detect significant differences at some points where the raw data indicated differences. A repeated study with a larger number of animals may be able to elucidate these differences more clearly. Although it would have been interesting to extrapolate the degree of change in attenuation and volume of the lungs in each position, the small number of study cats yielded insufficient data from which to extend these results to the entire feline population. The CT lung segmentation was performed by only 1 observer and was not repeated. We attempted to minimize the effect of intraobserver discrepancies by the use of a semiautomated method for lung segmentation. However, a degree of manual segmentation by means of a brush tool was still required; therefore, possible intraobserver discrepancies could not be completely excluded. An additional phase of the study wherein the cats were transitioned to RLR following the initial SR positioning might have revealed some differences in pulmonary aeration, compared with findings after transitioning from the SR positioning to LLR first. It is unlikely that there would have been any difference in the prevalence of atelectasis formation, given the low prevalence detected in the present study. For each study cat, morphological pulmonary disease was ruled out by means of an initial diagnostic CT scan; however, functional pulmonary studies were not performed. As such, the presence of subclinical functional pulmonary disease that may influence pulmonary aeration cannot be completely excluded. Lastly, the cats were assessed in a deep plane of anesthesia, which may not reflect the routine handling of clinical patients undergoing diagnostic CT. Therefore, care should be taken when extrapolating the results of the present study to such cases.
Results of the present study indicated that there was a rapid and significant reduction in lung volume and an increase in lung attenuation of the dependent lung lobes in healthy cats anesthetized in LR. These changes occurred immediately (at 0 minutes) after each cat's position was modified and the subsequent scan was initiated. The lung volume and attenuation changes remained static over time in each position. The attenuation change was predominantly due to an increase in the percentage of poorly aerated lung tissue (areas of attenuation of −500 to −101 HU). Additionally, data obtained in the present study indicated that the percentage of true atelectasis in the lungs (areas of attenuation of −100 to +100 HU) of the healthy cats in the present study was minimal. Therefore, increased attenuation within the dependent lungs observed in thoracic radiographic and CT images of anesthetized animals in LR may be attributable predominantly to an increase in the percentage of poorly aerated lung tissue rather than to an increase in the percentage of truly atelectic areas of tissue.
Acknowledgments
This manuscript represents a portion of a thesis submitted by Dr. Foo to the Faculty of Veterinary Science, University of Sydney, as partial fulfillment of the requirements of the Master of Veterinary Clinical Studies degree.
The authors declare that there were no conflicts of interest.
Presented in abstract form at the 2017 Australia and New Zealand College of Veterinary Scientists (ANZCVS) Science Week, Surfers Paradise, QLD, Australia, July 2017.
The authors thank Helen Laurendet for technical assistance.
ABREVIATIONS
| COPD | Chronic obstructive pulmonary disease |
| DR | Dorsal recumbency |
| HU | Hounsfield units |
| LLR | Left lateral recumbency |
| LLR1 | First period of left lateral recumbency |
| LLR2 | Second period of left lateral recumbency |
| LR | Lateral recumbency |
| RLR | Period of right lateral recumbency |
| ROI | Region of interest |
| SR | Sternal recumbency |
Footnotes
Eurofins Agroscience Services Pty Ltd, Orange, NSW, Australia.
Alfaxan, Jurox Pty Ltd, Rutherford, NSW, Australia.
Acepromazine, Ceva Animal Health Pty Ltd, Glenorie, NSW, Australia.
Butorgesic, Troy Laboratories Pty Ltd, Smithfield, NSW, Australia.
Medetate Injection, Jurox Pty Ltd, Rutherford, NSW, Australia.
Lignocaine hydrochloride 2% Pfizer Australia Pty Ltd, West Ryde, NSW, Australia.
Isoflo, Abbott Australasia Pty Ltd, Botany, NSW, Australia.
Viaflex, Baxter Healthcare Pty Ltd, Toongabbie, NSW, Australia.
Phillips, 16-slice Brilliance CT V2.3, Phillips Medical Systems Netherlands, Eindohoven, the Netherlands.
ASUS PA238QR, Beitou District, Taipei, Taiwan.
Macbook Pro OS × 10.11, 1 Infinite Loop, Cupertino, Calif.
OsiriX v8.0.1 64-bit, Pixmeo SARL, Bernex, Switzerland.
Mango v4.0.1, Research Imaging Institute, University of Texas Health Science Center, San Antonio, Tex.
Microsoft Excel 2008 for Mac, Version 14.6.9, Microsoft Corp, Redmond, Wash.
GenStat, 17th edition, VSN International Ltd, Hemel Hempstead, England.
References
1. Verschakelen JA, van Fraeyenhoven L, Laureys G, et al. Differences in CT density between dependent and nondependent portions of the lung: influence of lung volume. AJR Am J Roentgenol 1993;161:713–717.
2. Ahlberg N, Hoppe F, Kelter U, et al. A computed tomographic study of volume and x-ray attenuation of the lungs of Beagles in various body positions. Vet Radiol 1985;22:43–47.
3. Duggan M, Kavanagh BO. Pulmonary atelectasis: a pathogenic perioperative entity. Anesthesiology 2005;102:838–854.
4. Magnusson L, Spahn DR. New concepts of atelectasis during general anesthesia. Br J Anaesth 2003;91:61–72.
5. Brismar B, Hedenstierna G, Lundquist H, et al. Pulmonary densities during anesthesia with muscular relaxation: a proposal of atelectasis. Anesthesiology 1985;62:422–428.
6. Woo SW, Berlin D, Hedley-White J. Surfactant function and anesthestic agents. J Appl Physiol 1969;26:571–577.
7. Trzil JE, Reinero CR. Update on feline asthma. Vet Clin North Am Small Anim Pract 2014;44:91–105.
8. Henao-Guerrero N, Ricco C, Jones JC, et al. Comparison of four ventilator protocols for computed tomography of the thorax in healthy cats. Am J Vet Res 2012;73:646–653.
9. Lundquist H, Hedenstierna G, Strandberg A, et al. CT-assessment of dependent lung densities in man during general anaesthesia. Acta Radiol 1995;36:626–632.
10. Joyce CJ, Baker AB. What is the role of absorption atelectasis in the genesis of perioperative pulmonary collapse? Anaesth Intensive Care 1995;23:691–696.
11. Staffieri F, De Monte V, De Marzo C, et al. Alveolar recruiting maneuver in dogs under general anesthesia: effects on alveolar ventilation, gas exchange, and respiratory mechanics. Vet Res Commun 2010;34:S131–S134.
12. Staffieri F, Franchini D, Carella GL, et al. Computed tomographic analysis of the effects of two inspired oxygen concentrations on pulmonary aeration in anesthetized and mechanically ventilated dogs. Am J Vet Res 2007;68:925–931.
13. Giglio RF, Winter MD, Reese DJ, et al. Radiographic characterization of presumed plate-like atelectasis in 75 non-anesthetized dogs and 15 cats. Vet Radiol Ultrasound 2013;54:326–331.
14. le Roux C, Cassel N, Fosgate GT, et al. Computed tomographic findings of pulmonary atelectasis in healthy anaesthetized Beagles. Am J Vet Res 2016;77:1082–1092.
15. Staffieri F, De Monte V, De Marzo C, et al. Effects of two fractions of inspired oxygen on lung aeration and gas exchange in cats under inhalant anaesthesia. Vet Anaesth Analg 2010;37:483–490.
16. Lamb CR, Jones ID. Associations between respiratory signs and abnormalities reported in thoracic CT scans of cats. J Small Anim Pract 2016;57:561–567.
17. Thrall DE. Principles of radiographic interpretation of the thorax. In: Textbook of veterinary diagnostic radiology. 6th ed. St Louis: Saunders Elsevier, 2013;474–488.
18. Morandi F, Mattoon JS, Lakritz J, et al. Correlation of helical and incremental high-resolution thin-section computed tomographic imaging with histomorphometric quantitative evaluation of lungs in dogs. Am J Vet Res 2003;64:935–944.
19. Franquet T, Stern EJ, Gimenez A, et al. Lateral decubitus CT: a useful adjunct to standard inspiratory-expiratory CT for the detection of air trapping. AJR Am J Roentgenol 2000;174:528–530.
20. Choi SJ, Choi BK, Kim HJ, et al. Lateral decubitus HRCT: a simple technique to replace expiratory CT in children with air trapping. Pediatr Radiol 2002;32:179–182.
21. Lucaya J, Garcia-Pena P, Herrera L, et al. Expiratory chest CT in children. AJR Am J Roentgenol 2000;174:235–241.
22. Morandi F, Mattoon JS, Kakritz J, et al. Correlation of helical and incremental high-resolution thin-section computed tomographic and histomorphometric quantitative evaluation of an acute inflammatory response of lungs in dogs. Am J Vet Res 2004;65:1114–1123.
23. Oliveira CR, Mitchell MA, O'Brien RT. Thoracic computed tomography in feline patients without use of chemical restraint. Vet Radiol Ultrasound 2011;52:368–376.
24. Lee SK, Park S, Cheon B, et al. Effect of position and time held in that position on ground-glass opacity in computed tomographic images of dogs. Am J Vet Res 2017;78:279–288.
25. Kim T, Chang D, Chang J, et al. Assessment of computed tomographic lung density in Beagle and Shih Tzu dogs. J Vet Clin 2010:273–283.
26. Rozanski EA, Bedenice D, Lofgren J, et al. The effect of body position, sedation, and thoracic bandaging on functional residual capacity in healthy deep-chested dogs. Can J Vet Res 2010;74:34–39.
27. Gattinoni L, Caironi P, Pelosi P, et al. What has computed tomography taught us about the acute respiratory distress syndrome? Am J Respir Crit Care Med 2001;164:1701–1711.
28. Vieira SR, Puybasset L, Richecoeur J, et al. A lung computed tomographic assessment of positive end-expiratory pressure-induced lung overdistension. Am J Respir Crit Care Med 1998;158:1571–1577.
29. Mets OM, de Jong PA, van Ginneken B, et al. Quantitative computed tomography in COPD: possibilities and limitations. Lung 2012;190:133–145.
30. Yamashiro T, Matsuoka S, Bartholmai B, et al. Collapsibility of lung volume by paired inspiratory and expiratory CT scans: correlations with lung function and mean lung density. Acad Radiol 2010;17:489–495.
31. Arakawa A, Yamashita Y, Nakayama T, et al. Assessment of lung volumes in pulmonary emphysema using multidetector helical CT: comparison with pulmonary function tests. Comput Med Imaging Graph 2001;25:399–404.
32. Xie X, de Jong PA, Oudkerk M, et al. Morphological measurements in computed tomography correlate with airflow obstruction in chronic obstructive pulmonary disease: systematic review and meta-analysis. Eur Radiol 2012;22:2085–2093.
33. Akira M, Toyokawa K, Inoue Y, et al. Quantitative CT in chronic obstructive pulmonary disease: inspiratory and expiratory assessment. AJR Am J Roentgenol 2009;192:267–272.
34. Lee YK, Oh YM, Lee JH, et al. Quantitative assessment of emphysema, air trapping, and airway thickening on computed tomography. Lung 2008;186:157–165.
35. Matsuoka S, Kurihara Y, Yagihashi K, et al. Quantitative assessment of peripheral airway obstruction on paired expiratory/inspiratory thin-section computed tomography in chronic obstructive pulmonary disease with emphysema. J Comput Assist Tomogr 2007;31:384–389.
36. Kauczor HU, Hast J, Heussel CP, et al. CT attenuation of paired HRCT scans obtained at full inspiratory/expiratory position: comparison with pulmonary function tests. Eur Radiol 2002;12:2757–2763.
37. Newman KB, Lynch DA, Newman LS, et al. Quantitative computed tomography detects air trapping due to asthma. Chest 1994;106:105–109.
38. Adamama-Moraitou KK, Patsikas MN, Koutinas AF. Feline lower airway disease: a retrospective study of 22 naturally occurring cases from Greece. J Feline Med Surg 2004;6:227–233.
39. Gadbois J, d'Anjou MA, Dunn M, et al. Radiographic abnormalities in cats with feline bronchial disease and intra- and inter-observer variability in radiographic interpretation: 40 cases (1999–2006). J Am Vet Med Assoc 2009;234:367–375.
