Anesthetic Techniques Influence the Induction of Pulmonary Capillary Hemorrhage During Diagnostic Ultrasound Scanning in Rats

Abstract

Objectives

Puhnonary apillary hemorrhage can be induced by diagnostic ultrasonnd (US) during direct pulmonary US scanning in rats. The influence of specific anesthetic tedmiques on this bioeffect was examined.

Methods

Ketamine plus xylazine has been used previously. In this study, the influence of intraperitoneal injections of ketamine and pentobarbital, inhalational isoflurane, and the supplemental use of xylazine with ketamine and isollurane was tested. A diagnostic US machine with a7.6-MHz linear array was used to image the right lung of anesthetized rats in a warmed water bath at different mechanical index (MI) settings. Pulmonary capillary hemorrhage was assessed by measuring comet tail artifacts in the image and by morphometry of the hemorrhagic areas on excised lungs.

Results

Pulmonary capillary hemorrhage was greatest for pentobarbital, lower for inhalational isoflurane, and lowest for ketamine anesthesia, with occurrence thresholds at at Mis of about 0.44, 0.8, and 0.8, respectively. Addition of xylazine produced a substantial increaseinhemorrhageanda significant proportion of hemorrhage occurrence for ketamineat an MI of 0.7 (P < .01) and forisofluraneat an MI of 0.52 (P < .01).

Conclusions

Ketamine plus xylazine and pentobarbital yield lower thresholds than ketamine or isoflurane alone by nearly a factor of 2 in MI. These results suggest that the choice of the anesthetic agent substantially modifies the relative risks of pulmonary capillary hemorrhage from pulmonary US.

Pulsed ultrasound (US) can induce pulmonary capillary hemorrhage in mammals, which raises some concern for diagnostic US in medicine. This bioeffect was discovered more than 20 years ago¹ and was evahiated in subsequent research for dependence on various pulsed US parameters to assess the possible risk for diagnostic US. Reported exposure thresholds for pulmonary capillary hemorrhage found in studies of the response to varied peak rarefactional pressure amplitude were reviewed by the American Institute of Ultrasound in Medicine.² The results were interpreted in terms of the mechanical index(MI), which is an exposure index displayed on the screens of most US machines. The MI concept can be used for research by adjusting the peak rarefactional pressure amplitude measured in water for attenuation and dividing by the square root of the pulse center frequency. The thresholds, which were obtained mostly in anesthetized mice, corresponded roughly to an MI value of0.63.² For reference, the guideline upper limit for diagnostic US is an MI of 1.9.3 Ultrasound-induced pulmonary capillary hemorrhage was found at about the same derated peak rarefactional pressure amplitudes in mice, rats, rabbits, pigs, and monkeys. A meta-analysis ofl4 studies in mice and rats found that the pulmonary capillary hemorrhage effect had widely varying thresholds and appeared to depend on physical parameters, including pulse duration, pulse repetition frequency, and exposure duration, in addition to peak rarefaction al pressure amplitude and frequency.⁴ In another authoritative reviewby theAmerican InstituteofUltrasound in Medicine, patient pulmonary US exposure was expected to occur during incidental exposure to echocardiography, which was not thought to pose a consistently defined risk.⁵ However, this overview of the pubnonarycapillary hemorrhage problem has changed in recent years for two reasons.

First, most of the available data were obtained with pulsed US from laboratory exposure systems, and the applicability to actual diagnostic US was uncertain. Diagnostic US was used for studies in monkeys⁶and rats,⁷ but the results were mixed. Astudyofhuman hingsafterechocardiography indicated that 3.5-MHz diagnostic US up to 2.4 MPa (MI = 1.3),which was thoughtto interact with the lung as incidental exposure, did not cause pulmonary capillary hemorrhage.⁸ In a recent study in rats from our laboratory, diagnostic US from a 7.6-MHz linear array was used to directly scan the right lung ofanesthetized rats in a warmed water bath.⁹ The image displayed growing comet tail artifacts, which were indicative of pulmonary capillary hemorrhage on the surface of the lung. Pulmonary capillary hemorrhage was observed for several groups of rats scanned ata range of MI settingsfor 5 minutes, and a threshold for hemorrhage occurrence was indicated for an MI of about 0.44. This result appears to be lower than expected from the meta-analysis of physical parameters.⁴ The etiology of pulmonary capillary hemorrhage induced by diagnostic US remains poorly understood.

Second, direct pulmonary US examination has become a valuable diagnostic aid for pneumonia, pulmonary edema and effiision, pubnonary embolism, atelectasis, diffuse parenchymal disease, adult and newborn respiratory distress syndrome, and lung cancer. ¹⁰ Diagnostic US imaging has largely replaced chest radiography for pleural effusion, pneumothorax, hemothorax, and posttraumatic lung contusion. ¹¹– ¹² Pubnonary US is rapidly becoming routine in intensivecare, ¹3-¹5 emergencycare, ¹⁶and otherpoint-ofcare settings. ¹⁷ This adoption of routine pulmonary US completely changes the diagnostic US-induced pulmonary capillary hemorrhage problem of incidental exposure to one of direct, deliberate examination. The problem requires further examination from newperspectives to clarify potential patient risks and provide guidance for the safe use of pulmonary US.

One aspect of the problem in need of elaboration is the role ofbiological and physiologic factors. Age is one biological factor that has been explored. For mice and rats, pulmonary capillary hemorrhage thresholds appeared to have little dependence on age.^18,19 Neonatal swine also had pulmonary capillary hemorrhage thresholds comparable to those ofadult mice.²⁰ For older pigs, the experimentation is more difficult due to the size of the animals, but this animal model more closely simulates humans. O’Brien etal²¹found that older pigs were more sensitive than neonatal or middleaged pigs to 3.1-MHz pulsed US for 10-second exposures.

Other physiologic conditions may also be important. One key factor for research is the use of anesthetics. Forgeneral anesthesia in mice and rats, intraperitoneal injection of ketamine plusxylazine is typically used; for example, studies of rats and mice reviewed by Church and O’Brien⁴ all used ketamine plusxylazine. The doses were 87 mg/kg ketamine plus 13 mg/kgxylazine in rats or mice,²¹-²⁴ 100 mg/kg ketamine plus 10 mg/kg xylazine in rats,7 and 200 mg/kg ketamine phis 10 mg/kgxylazine in mice. ^1,17,25,26 The presently recommended anesthetic doses of 91 mg/kg ketamine and 9 mg/kgxylazine were used in our previous study.⁹

Ketamine is a dissociative anesthetic, which immobilizes with little respiratory depression, whereas xylazine is a sedative and analgesic, which produces substantial cardiovascular and respiratory depression.²⁷ Xylazine has been reported to induce pulmonary edema in rats at relatively high dosesof 21 to 45 mg/kg.^28,29 In a previous study using bronchoalveolar lavage albumin levels as representative of permeability injury, the normal doses of xylazine used for anesthesia did not produce a significant injury when compared with ketamine alone. However, the omission ofxylazine reduced the size of the comet tail artifacts produced during US scanning and significantly reduced the erythrocyte count in the bronchoalveolar lavage fluid found after scanning.³⁰ Statistical analysis using analysis of variance (ANOVA) suggested that a synergistic interaction occurred between the USscanningand the use of xylazine.

The goal of this study was to examine the influence of commonly used anesthetics and drug combinations on the induction of pulmonary capillary hemorrhage by diagnostic US scanning, using our previous methods.⁹ A common anesthetic for rodents is pentobarbital. This barbiturate yields prolonged anesthesia and can cause cardiovascular and respiratory system depression.²⁷ Inhalation anesthesia is typically used for larger animals but is also valuable for short-term anesthesia in rodents. Isoflurane typically is applied at 1% to 5% in oxygen or air from a vaporizing machine. Thresholds were determined using ketamine withoutxylazine, pentobarbital, and isoflurane. Pubnonary capillary hemorrhage results for ketamine with and without xylazine were compared. In addition, the effect of the addition of xylazine to is of huane was tested. The results show substantial influences of the specific anesthetics on US-induced pulmonary capillary hemorrhage.

Materials and Methods

Animal Preparation

All in vivo animal procedures were conducted with the approval and guidance of the University Committee on Use and Care of Animals. Female Sprague Dawley rats (CD IGS strain; Charles River Laboratories, Wilmington, MA) were used for this research. The study progressed in 3 parts involving ketamine, pentobarbital, or isoflurane anesthesia.

For the ketamine anesthesia, 36 rats weighing 224 ± 15g(mean ± SD) were used, with 2 animals dying shortly after administration of the agent. There were 6 scan groups of 5 ratseach and a shamgroup of 4 rats. Theketamine (Ketaved ketamine hydrochloride injection; Vedco, Inc, StJoseph, MO) was injected to deliver 100 mg/kg intraperitoneally for 3 scan groups and the sham group. In the additional 3 scan groups, ketamine 91 at mg/kg intraperitoneally was supplemented with 9 mg/kg intraperitoneal xylazine (AnaSed xylazine injection; Akorn, Inc, Decatur, IL), given either before the US scanning or immediately after scanning, representing a control for the potential effects of xylazine that are not dependent on scanning.

For pentobarbital anesthesia, 31 rats weighing 242±15 g were used, with 1 animal excluded for technical problems. There were 6 groups of 5 rats each, incruding 5 scan groups and 1 sham group. Pentobarbital (Nembutal sodium solution; Akorn, Inc) was injected at a dose of 50 mg/kg intraperitoneally.

Inhalation al use of isoflurane produces moderate cardiovascular and respiratory depression but can be rapidly modulated to yield the required level of anesthesia.²⁶ Oxygen was tried as the carrier gas in preliminary testing: however, this method seemed to produce spontaneous comet tail artifacts and areas of atelectasis on some rat's lungs. The use oflOO% oxygen can lead to problems with pulmonary function due to absorptive atelectasis, a serious potential problem with inhalational anesthesia.³¹ This problem can be mitigatedbyuse of lower oxygen concentrations or air (ie, 20% oxygen). For this study, air was theisofluranevehicle. Thirty rats weighing 227±10g were assigned to 6 scan groups with air as the carrier gas, mitigating the problems found with oxygen. The isoflurane (VET one isoflurane inhalation anesthetic; MWIVetOne, Meridian, ID) was delivered by a vaporizer (Surgivet Isotec 4; Smiths Medical ASD, Inc, St Paul, MN) at a concentration of 5% for induction. The concentration was subsequently reduced to as low as 1% for maintenance during handling and scanning. For 2 groups, xylazine was injected at anesthesia (before scanning) to assess the influence of this sedative on pulmonary capillary hemorrhage.

For all animals, the right thorax was shaved after anesthesia and depilated to allowgoodUS transmission into the body. The rats were mounted in dorsal recumbency on a plastic board, which then was mounted vertically in a 38°C vacuum-degassed water bath for US scanning. The water bath maintained the body temperature of the rats and provided reliable US transmission. After scanning and a 5-minute delay, the rat was euthanized under anesthesia by exsanguination of the inferiorvena cava. The trachea was tied off to maintain lung volume, and the lungs were removed for examination and photography using a stereo microscope with a digital camera (Spot Flex; Diagnostic Instruments, Inc, Sterling Heights, MI).

Ultrasound Scanning

An HDI 5000 diagnostic US machine (Philips Healthcare, Andover, MA) with a CL15-7 linear array was used for scanning, as described previously.⁹ The probe was positioned in partial contact with the right thorax and used to scan the right cranial or middle lobe. The machine was set for real-time imaging with B-mode optimization, a 2-cm depth, 1 focus at a 1-cm depth, and 39 frames per second. An initial Misetting of 0.21 was used to align the probe for a dear view of the lung. The primary variable for the thresh-old determinations was the on-screen MI setting, with a maximum of 0.9. The US field was characterized previously in the water bath with a calibrated hydrophone (model 805; Sonora Medical Systems, Inc, Longmont, CO).⁹ For this study, the measurements were rechecked and found to be essentially the same. Thecenterfrequency was approximately 7.6 MHz; the pulse repetition frequency was 10 kHz; and the pulse duration was 320 nanoseconds.⁹ The total length of the scan was approximately 2 cm, and the–6-dBw:idth was 1.0 mm. As before, scanning each rat at a single MI setting was used to assess the scan-response trends and the determination of thresh-old MIs. The water varues for peak rarefactional pressure amplitude were derated by an attenuation coefficient of 1.1 dB/cm/MHz3³² for the approximately 6-mm-thick rat chest wall to obtain 0.6, 0.8, 1.2, 1.6,and 1.9MPa, for MIs of 0.21, 0.37, 0.52, 0.7, and 0.9, respectively.⁹

Experimental Plan

The probe was aimed to obtain a dear bright image of the lung surface, which appeared as a bright horizontal line in the image with secondary A-lines atgreater depths. For a test, the US was quickly raised from an MI of 0.21 to the desired MI setting for 5 minutes of scanning. For the shamexposed animals, the same procedure was followed, except that the scanning at elevated MI settings was omitted. The images were used to obtain estimates of the width of the bright lung surface image, the time to the first clear appearance of a comet tail artifact, and the percentage of the bright lung surface involved in comet tail artifacts at the end of the scanning. The image of the scanned hing area was used to find the length, width, and area of the hemorrhagic region on the lung surface with the aid of image analysis software (Spotversion 5.1; Diagnostic Instruments, Inc).

When possible, physiologic data were collected for each rat on the heart rate and percent peripheral capillary oxygen saturation (SP_O2) using a pulse oximeter (SurgiVet V3395 TPR; Smiths Medical ASD, Inc). This procedure produced mixed results for both measurements due to difficulty in placing the sensor on the paw and to the upper limit of 350 beats per minute (bpm) for the instrument To obtain stable values, the sensor was set, and the readings were observed for up to 5 minutes. For the pentobarbital anesthesia, the percent SP_O2 could notbe measured due to high heart rates, which were determined from recordings from a laboratory electrocardiographic system (ECGA amplifier; Hugo Sachs Electronik, HarvardAppararus, March, Germany), digitizer (Powerlab 4/30; ADinstruments, Inc, Colorado Springs, CO), and electrocardiographic analysis software (Chart Pro 5 version 5.5.5; AD Instruments, Inc).

Statistical analysis was performed using SigmaPlot version 11.0 software for Windows (Systat Software, Inc, SanJose, CA). The Mann-Whitney rank sum test was used to compare means of the measured parameters, with statistical significance assumed at P < .05. The z test of proportions was used to assess the significance of the proportion of 5 rats that had pulmonary hemorrhage for each group scanned at a specific Ml The proportion test was used to locate the thresholdbetween the lowest MIwith significant hemorrhage occurrence and the next lower MI setting. In addition, 2-wayANOVA was used to gauge the influence of the anesthetic methods relative to the MI variation.

Results

For anesthesia by ketamine alone (n = 15), the heart rate and SP₀₂ averaged 308 ± 39bpm and 90% ± 4%, respectively, with no difference between results before and after scanning. For theketamine plusxylazine tests (n = 10),the heart rate and SP₀₂ were 327 ± 23 bpm and 92% ± 2% before addition of xylazine and 264 ± 18 bpm and 82%±4% 5 minutes after addition of xylazine. The latter results show the possible cardiopulmonary depression induced by the addition of xylazine with a reduced heart rate (P < .001) and SP₀₂ (P < .001). The use ofketamine alone as the anesthetic resulted in much less pulmonary capillary hemorrhage than was seen in the previous srudy.⁹ Scanning at an MI of 0.37 was omitted because hemorrhage was not seen even at the higher MI of 0.52. Results are listed in Table 1 for the range of the times to first appearance of comet tail artifacts, for the percentage of bright lung surface images with comet tail artifacts after scanning, for the hemorrhagic areas on the lung, and for the fractions of lungs with positive hemorrhage results. On the basis of the significant ocrurrence (5/5) at an MI of 0.9 and the lack of significance (2/5) atan Mlof0.7, the MI threshold was about 0.8. This finding compares to the loss of significance between MIs of 0.52 (4/5) and 0.37 (2/5) in the previous study, which indicated an MI threshold of about 0.44.⁹ The results with xylazine added before the scanning were comparable to the previous results, but adding thexylazine after scanning did not enhance the hemorrhage above that seen for ketamine alone. A 2-way ANOVA for the 0.7 and 0.9 Mis without and with xylazine showed that both the xylazine and MI had a significant influence on the results even after allowing for the infhience of the other variables; however, the interaction effect was only marginally significant (P= .052). The direct comparison for ketamine with xylazine given before or after scanning at an MI of 0.7 showed a dramatic increase in hemorrhage when xylazine was present during scanning (Table 1),with a significant difference in ocrurrence (P= .01). Figure 1 compares the stereo microscopic images for the 5 repetitions of each test Interestingly, the 2 shallow hemorrhagic regions for xylazine addition after scanning did not have any dear indication of comet tail artifacts in the US images, whereas the hemorrhage that penetrated deeper into the hing (daiker regions) had comet tail artifacts extending to the bottom of the images.

Table 1.

Results for Ketamine Anesthesia

Group, MI	PCH Start, s	CTA Width, %	PCH Area, mm2	PCH Fraction	z Test P
Sham	NA	NA	0	0/4	NA
0.52	NA	NA	0	0/5	NS
0.7	210–240	2.6 ± 4.0	0.1 ± 0.2	2/5	NS
0.7−X	300	0	0.2 ± 0.2	2/5	NS
0.7+X	2–60	91 ± 10	7.2 ± 2.6	5/5	<.01
0.9	10–300	26 ± 20	1.8 ± 2.1	5/5	<.01
0.9+X	5–10	94 ± 6	14.6 ± 5.0	5/5	<.01

Means ± SDs include 0 for rats with no discernible hemorrhage, so that the observed sizes in the positive results were approximately given by the means divided by the positive fractions. CTA indicates comet tail artifact; NA, not applicable; NS not significant; PCH, pulmonary capillary hemorrhage; −X, test with xylazine added after scanning; +X, test with xylazine added before scanning.

Figure 1.

Photomicrographs of the right lung cranial lobes from 10rats scanned at an MI of 0.7 with ketamine and without (top row) or with (bottom row) added xylazine during scanning (Table 1) All of the lobes scanned with added xylazine have substantial pulmonary capillary hemorrhage evident within the linear extent of the scan plane The lobes scanned without xylazine show no hemorrhage in 3 lobes and small surface evidence of hemorrhage in 2 (black arrows) The lungs hadseveral featuresthat were not dueto the US scanning a fluid-filled region due to impact with a rib (white arrow) and various wrinkles due to slight folding during removal. In addition. a few scanned or sham lungs hadvery small (possibly 1alveolus) red spots. which were not scored as pulmonary capillary hemorrhage. Scale bar indicates 5 mm

For pentobarbital anesthesia (n = 30), the heart rate averaged 377 ± 32 bpm. The relatively high heart rate precluded SP_O2 measurement. The use of pentobarbital as the anesthetic resulted in pulmonary capillary hemorrhage that was comparable to that seen in the previous srudy using ketamine plusxylazine.⁹ Results are listed in Table 2 forthe same parameters as in Table 1. On the basis of the significant occurrence (4/5) atan MI of 0.52 and insignificant occurrence (2/5) atan MI of 0.37, the MI threshold was about 0.44. This threshold was the same as that found for ketamine phisxylazine in the previous srudy.⁹ No additional tests were performed with added xylazine for pentobarbital anesthesia because the cardiovascular and pulmonary impact of such a combination could be detrimental.²⁷

Table 2.

Results for Pentobarbital Anesthesia

Group, MI	PCH Start, s	CTA Width, %	PCH Area, mm2	PCH Fraction	z Test P
Sham	NA	NA	0	0/5	NA
0.27	NA	NA	0	0/5	NA
0.37	7–240	32 ± 35	0.5 ± 0.9	2/5	.11
0.52	5–50	65 ± 40	4.9 ± 3.3	4/5	.01
0.7	10–90	71 ± 18	8.3 ± 4.2	5/5	<.01
0.9	1–15	100 ± 0	21.4 ± 10.5	5/5	<.01

Notations are as in Table 1.

For isoflurane anesthesia (n = 17), the heart rate and SP_O2 averaged 343 ± 11bpm and 91% ± 2%, respectively, with no observed difference between the results before and after scanning. For the 2 groupswithxylazine (n = 10),the heart rate and SP_O2 were 260 ± 17bpm and 90% ± 3%. Results are listed in Table 3 for the same parameters as in Tables 1 and 2. No tests were conducted for MIs of 0.27 and 0 (sham) because no effect was found for 5 rats scanned at an MI of 0.37. On the basis of the significant ocrurrence (5/5) at an MI of 0.9 and in significant ocrurrence (2/5) at an MI of 0.7, the MI threshold was about 0.8. This thresh-old was the same as that found forketamine alone (Table 1). However, when xylazine was added, pulmonary capillary hemorrhage was found in 5 of 5 rats atan MI of 0.52 (P= .01) butO of5 rats at an MI of 0.37. This finding indicated a lowering of the threshold for isoflurane plus xylazine to that found for pent obarbital anesthesia and ketamine phis xylazine anesthesia. A 2-way ANOVA for the 0.37 and 0.52 Mis without and withxylazine showed that both thexylazine and MI had significant influence on the results (P< .01),and that there was a significant interaction (P=.037).

Table 3.

Results for Isoflurane Anesthesia

Group, MI	PCH Start, s	CTA Width, %	PCH Area, mm2	PCH Fraction	z Test P
0.37	NA	NA	NA	0/5	NA
0.37+X	NA	NA	NA	0/5	NA
0.52	120–240	6.7 ± 9.3	1.3 ± 2.4	2/5	NS
0.52+X	10–60	68 ± 30	6.4 ± 4.4	5/5	<.01
0.7	2–20	23 ± 33	2.6 ± 3.9	2/5	NS
0.9	0.5–15	90 ± 14	17.5 ± 8.6	5/5	<.01

Notations are as in Table 1.

The comet tail artifact widths are presented in Figure 2 as the percentage of the hingsurface seen as a bright line in the image, shown with the means inchiding all 5 results (ie, 0 for the negative results). The bright line was indicative of perpendirular reflection, which also often generated repeated “A-lines” of multiple reflections at higher depths on the image. The comet tail artifacts developed within this bright-surface image region, which ranged from 7 to 18 mm in width. The comet tail artifact width correlated well with the actual width of the hemorrhagic area found in the lungs. This plot shows the variation of pulmonary capillary hemorrhage induced by identical scanning but with different anesthetic techniques, with pentobaibital giving the greatest hemorrhage and ketamine the least The hemorrhage with pentobaibital was comparable to that found for ketamine with xylazine.⁹ The addition of xylazine to ketamine during scanning raised the result at an MI of 0.7 to 90% (Table 1), which was higher than the result for pen tobarbital Likewise, the addition of xylazine for isoflurane anesthesia at an MI of 0.52 elevated the comet tail artifact width to 68% (Table 3), which was about the same as for pen tobarbital. Interestingly, the low result for isoflurane at an MI of 0.7 was not significantly different from the values obtained with ketamine but was significantly lower than the vahies obtained with pentobaibital However, the isoflurane result increased at an MI of 0.9 so that the reverse was true: significantly greater than theketamine result (P < .05) but not significantly different from the pentobaibital result.

Figure 2.

Means± SEs for the percentages of bright-line lung images which were involved with comet tail artifacts (CTAs) at the end of scanning (Tables 1-3) The occurrence of pulmonary capillary hemorrhage was statistically significant for MIs of 0.5 and higher for pentobarbital but only for an MI of 0.9 for isoflurane and ketamine-only anesthesia.

The areas showing pulmonary capillary hemorrhage on the hing surface, which maybe abetter gauge of the magnitude of the effect than the comet tail artifact, are plotted in Figure 3. The trends were essentially the same as in Figure 2 for comet tail artifacts. For pen tobaibital anesthesia, the increases in the areas were more pronounced with increases in the MI than for the comet tail artifact widths. This aspect of the 2-dimensional measurement likely reflects the changes in the width of the ultrasonic beam, over which the critical threshold level was exceeded. A 2-way ANO VA for the 0.7 and 0.9 Mis and the3 anesthetic methods (without xylazine) showed that the effect of the MI on the results was significantly influenced by the drug present (P = .047). At an MI of0.9, ketamine was significantly different from pentobarbital and isoflurane anesthesia.

Figure 3.

Means ± SEs for lung surfaces with evident pulmonary capillary hemorrhage for the 3 anesthetics. As in Figure 2, the results were statistically significantly different from shams at MIs of 0.5 and higher for pentobarbital but only for an MI of 0.9 for isoflurane and ketamine-only anesthesia

Discussion

Inthis study, the influence of different anesthetic techniques on pulmonary capillary hemorrhage induced by diagnostic US was investigated. The hemorrhage was greatest for pentobarbital anesthesia, lower for isoflurane inhalational anesthesia, and lowest for ketamine anesthesia (Figures 2 and 3), with thresholds at MIs of about 0.44, 0.8, and 0.8, respectively. Addition of xylazine produced a substantial increase in hemorrhage and a significant proportion of hemorrhage occurrence for ketamineatan MI of 0.7(Table 1) and for isoflurane at an MI of 0.52 (Table 3).

These results indicated that the cardiopulmonary effects of xylazine and pentobarbital yield lower thresholds than for ketamine or isoflurane alone by nearly a factor of 2 in MI. The time at which the comet tail artifacts were first seen on the images provides another interesting factor for analysis. The US interaction with the lung surface must begin at the initial pulse. However, the changes in the comet tail artifact start time may be viewed as changes in the time needed to induce substantial puhnonarycapillary hemorrhage; that is, xylazine and pentobarbital may reduce the time needed (Tables 1-3). If the experimental scan duration were reduced (eg, from 5 minutes to 1 minute), then the 1 of 5 and 2 of 5 results in Tables 1-3 might be eliminated. Unfortunately, the time-to-start parameter was somewhat subjective and variable, and the shallow hemorrhagic areas in Figure 1 did not correspond to any clear comet tail artifact indications. A more definitive appraisal, for example, in terms ofa temporal threshold at each MI, would require a different method.

The variation ofthesensitivityofthe lungs to different anesthetic techniques is likely due to variations in pulmonary physiologic responses. The moderate cardiopulmonary depression caused by xylazine or pentobarbital probably influences the depth of respiration. The degree oflung inflation has been shown to influence the acoustical impedance of the pleural surface, such that low inflation increases the impedance above that ofa simple air-water interface.^33,34 It seems possible that the respiratory depression may therefore result in enhanced energy transmission into the lungs and enhanced interaction with the pulmonary tissue.

The sensitivity to US-induced pulmonary capillary hemorrhage shown for anesthesia withxylazine orpen to barbital is likely also related to the physiologic response of the puhnonary microvasculature. The pulmonary capillary pressure is normally only about 2 to 15 mm Hg.³⁵ Higher pressure values distend the capillaries and apply some level of stress to the capillaryborder.³⁶ Pressure values above about 30 mm Hg are sufficient to induce edema.³⁷ Values higher than 40 to 50 mm Hg can cause capillary rupture.³⁵ Xylazine is an α2 receptor agonist and affects the pulmonary circulation, increasing the pulmonary capillary pressure. As noted above, relatively high doses of xylazine, in the range of21to45 mg/kg, cause pulmonaryedema in a few hours.^28,29 Studies in sheep show that α2 receptor agonists similar to xylazine produce elevated pulmonary capillary pressures in a dose-dependentmanner.^38,39 Pentobarbital causes cardiopulmonary depression with a relatively high heart rate, which may also lead to pulmonary microvascular perturbation, leading to increased pulmonary capillary hemorrhage. Both xylazine and pento barbital enhanced neurogenic pulmonary edema in rats.⁴⁰ The filling and distension of the capillaries are likely to reduce the amount of exogenous stress required to rupture the capillaries. Therefore, the additional stress required to cause pulmonary capillary hemorrhage may be reduced by xylazine or pentobarbital. The physiologic responses to these drugs likely prestress the microvasculature and make the capillaries vulnerable to the physical stress generated by the exposure to pulsed US.

Most studies of pulmonary capillary hemorrhage induced by US have used anesthetic techniques including xylazine, which may have resulted in relatively low thresholds. The observation that the milder anesthetic methods with ketarnine and isoflurane used here resulted in higher thresholds may have implications for human pulmonary diagnostic US. For example, ketamine alone produces a dissociated state, which is not far from wakefulness. It is entirely possible to extrapolate that awake healthy animals, orpatients, may be at relatively low risk of pulmonary capillary hemorrhage induced by pulmonary US. Conversely, pulmonary vulnerability may vary with different drugs and patient conditions. For example, drugs that are α2 receptor agonists such as clonidine (used for hypertension, attention deficit/hyperactivity disorder, and sedation),⁴¹ may have effects that are similar to those of xylazine. Various patient conditions can also produce pulmonary vasodilation with pulmonary venous hypertension. For example, left heart impairment due to heart disease can lead to increased pulmonary pressures and edema.⁴² Reduced air pressures due to upper airway obstruction or high altitudes⁴³ can also lead to vasodilation and edema. These drug effects and patient conditions may cause a relatively high risk of pulmonary capillary hemorrhage from pulmonary US. Interestingly, the use of xylazine for anesthesia in this and previous pulmonary capillary hemorrhage research may have fortuitously mimicked susceptible physiologic conditions and provided data relevant to high-risk patients. Further research on US-induced pulmonary capillary hemorrhage is needed to clarify the problem of physiologic variability for guiding sonographers in the safe use of pulmonary US for all patient conditions.

Abbreviations

ANOVA

analysis of variance

bpm

beats per minute

MI

mechanical index

SP_O2

peripheral capillary oxygen saturation

US

ultrasoud