Threshold Estimation and Superthreshold Behavior of Ultrasound-induced Lung Hemorrhage in Rats: Role of Age Dependency

Abstract

Age-dependent threshold and superthreshold behaviors of ultrasound-induced lung hemorrhage were investigated with one hundred ten 12.6±0.8-day-old rats, one hundred ten 22.9±0.8-day-old rats, and one hundred 57.7±3.9-day-old rats. Exposure conditions were: 2.8 MHz, 10-s exposure duration, 1-kHz pulse repetition frequency and 1.3-μs pulse duration. The in situ (at the pleural surface) peak rarefactional pressure (p_r(insitu)) ranged between 1.4 and 10.8 MPa for which there were either 9 or 10 acoustic pressure groups for each of the three rat ages (10 rats/exposure group). For each of the three rat ages there were also shams; there were no lesions in the shams. The p_r(insitu) levels were randomized within each age group; rat age was not randomized. Individuals involved in animal handling, exposure, and lesion scoring were blinded to the exposure condition. In addition, one hundred fifty six 72-day-old rats were included from three completed studies (same experimental conditions) to provide a fourth age group for the analysis. Probit regression analysis was used to examine the dependence of the occurrence of lesions on p_r(insitu) in the four age groups. Likewise, lesion depth and lesion root surface area were analyzed using Gaussian tobit regression analysis. Although p_r(insitu) was a significant variable, no significant age dependence of the p_r(insitu) effect was found. Furthermore, age had no significant effect on either the rate of occurrence or the depth of lesions. Given the occurrence of a lesion, a weak age dependence was found for the median surface area of the induced lesion (p-value = 0.037).

INTRODUCTION

The study reported herein evaluates one of the risk factors for ultrasound-induced lung hemorrhage, that is, age dependency because there have been several conflicting observations. These conflicting observations have been summarized previously (O’Brien et al. 2003a). Briefly, monkey studies suggested that younger animals had a greater likelihood for ultrasound-induced lung hemorrhage compared to older animals (Tarantal and Canfield 1994). Mice were reported to have an age-independent ultrasound-induced lung-hemorrhage threshold (Dalecki et al. 1997a). Also, one research group reported that neonate (1-day-old) (Baggs et al. 1996) and young (10-day-old) pigs had an age-independent lung-hemorrhage threshold (Dalecki et al. 1997b). However, when these younger (1- and 10-day-old) pig studies are compared to another pig study of older animals (10- to 12-week-old) (O’Brien and Zachary 1997), younger pigs were more sensitive than the older pigs, that is, an age-dependent observation between separately conducted studies. And, age-dependent lung-hemorrhage thresholds were reported with middle-aged (39-day-old) pigs being least susceptible to lung damage, oldest (58-day-old) pigs being most susceptible to lung damage, and neonate (4.9-day-old) pigs being somewhere in between susceptibility to lung damage (O’Brien et al. 2003a).

Three species (mouse, pig, monkey) have been investigated but the age dependencies of ultrasound-induced lung hemorrhage in rats have not. Therefore, the study reported herein estimated the in situ peak rarefactional pressure threshold levels and the sensitivity to lung damage at superthreshold exposure conditions in Sprague Dawley rats at three ages (12.6±0.8- days, 22.9±0.8-days, 57.7±3.9-days old). In addition, a fourth age group (72-days old) was included for the age-dependent ultrasound-induced lung hemorrhage analysis from a compilation of three completed rat studies (Zachary et al. 2001; O’Brien et al. 2001, 2003b) for which the same experimental conditions were used.

MATERIALS AND METHODS

Animals

The experimental protocol was approved by IACUC (Institutional Animal Care and Use Committee, University of Illinois) and satisfied all campus and National Institutes of Health rules for the humane use of laboratory animals. Animals were housed in an AAALAC (Association for Assessment and Accreditation of Laboratory Animal Care, Rockville, MD) -approved animal facility, and provided food and water ad libitum.

There were a total of one hundred ten 12.6±0.8-day-old (mean±SD) rats (weight: 26.7±3.6 g; chest wall thickness: 1.6±0.2 mm), one hundred ten 22.9±0.8-day-old rats (weight: 53.4±6.6 g; chest wall thickness: 1.8±0.1 mm), and one hundred 57.7±3.9-day-old rats (weight: 235±48 g; chest wall thickness: 3.8±0.5 mm). All Sprague Dawley rats were obtained from Harlan (Indianapolis, IN). These three age groups will also be referred to as 13-day, 23-day and 58-day rats for convenience. Animal weights were measured at the time of the experiment, and animal ages were determined from exact birth date. The selection of these age (weight) groups was somewhat arbitrary because the mechanisms by which ultrasound produces lung hemorrhage are not understood, and thus the selection could not have a specific biological basis. Nevertheless, there are maturation changes as the animal ages, particularly those based on recognized rat time points. These time points are linked to biological events. Such categories include preweaned rats (13-day group), postweaned rats (23-day old group) and adult rats (58-day group).

In each age group, rats were assigned randomly to 11 (13-day and 23-day) or 10 (58-day) acoustic pressure levels (Table 1), each with 10 rats/pressure level. Pulsed ultrasound exposure conditions were: 2.8 MHz, 10-s exposure duration, 1-kHz pulse repetition frequency and 1.3-μs pulse duration. No lesions were produced in the sham exposed rats. The individuals involved in rat handling, exposure, necropsy and lesion scoring were blinded to the exposure condition. The exposure condition for each rat was revealed only after the final results were tabulated.

TABLE 1.

Mean age, mean (SEM) values of the in situ (at the pleural surface) peak rarefactional pressure p_r(insitu) and in situ peak compressional pressure p_c(insitu) and their respective lesion occurrence and mean (SEM) values of the lesion depth and surface area. All rats were exposed to pulsed ultrasound (1-kHz PRF, 10-s ED, 1.3-μs PD). The sham exposure conditions used a pulse repetition frequency of 10 Hz. Mean values of the Mechanical Index (MI), as measured according to the applicable standard (ODS 1998), are provided because the MI is a regulated quantity of diagnostic ultrasound equipment.

Mean age (day)	pr(in situ) (SEM) (MPa)	pc(insitu) (SEM) (MPa)	Mechanical Index	% Lesion occurrence (uncertainty)	Mean (SEM) lesion depth (mm)	Mean (SEM) lesion area (mm2)
13-day rats (10 rats per group)
13.1 (sham)	0.37 (0.001)	0.45 (0.001)	0.18	0	--	--
12.9	1.4 (0.002)	1.3 (0.002)	0.68	0	--	--
12.5	2.2 (0.002)	2.2 (0.002)	1.1	0	--	--
12.9	2.9 (0.004)	2.9 (0.005)	1.4	0	--	--
12.7	3.7 (0.008)	3.9 (0.009)	1.8	30 (14)	0.50 (0.27)	1.08 (0.67)
12.3	4.4 (0.007)	4.7 (0.007)	2.2	0	--	--
12.8	5.3 (0.008)	5.6 (0.009)	2.6	30 (14)	0.18 (0.14)	0.30 (0.26)
12.5	6.5 (0.006)	7.5 (0.007)	3.2	30 (14)	0.11 (0.07)	0.11 (0.07)
12.3	8.4 (0.013)	9.7 (0.015)	4.1	60 (15)	0.37 (0.13)	0.75 (0.23)
12.2	9.2 (0.005)	11.3 (0.006)	4.7	60 (15)	0.37 (0.15)	0.16 (0.21)
12.7	10.3 (0.006)	12.9 (0.008)	5.2	60 (15)	0.36 (0.12)	0.48 (0.15)
23-day rats (10 rats per group)
22.6 (sham)	0.37 (0.000)	0.45 (0.000)	0.18	0	--	--
23.0	2.9 (0.003)	2.9 (0.003)	1.4	0	--	--
23.1	3.7 (0.002)	3.9 (0.002)	1.8	10 (9)	0.02 (0.02)	0.07 (0.07)
23.2	4.6 (0.004)	4.9 (0.004)	2.3	20 (13)	0.07 (0.06)	0.11 (0.09)
23.0	5.5 (0.005)	6.1 (0.005)	2.8	30 (14)	0.29 (0.15)	0.36 (0.19)
23.0	6.4 (0.005)	7.4 (0.006)	3.2	40 (15)	0.35 (0.18)	0.67 (0.35)
23.2	7.2 (0.005)	7.7 (0.005)	3.6	50 (16)	0.54 (0.19)	1.59 (0.55)
23.3	7.7 (0.007)	8.9 (0.008)	3.8	70 (14)	0.70 (0.17)	1.34 (0.37)
22.0	8.6 (0.007)	10.1 (0.008)	4.3	60 (15)	0.54 (0.20)	1.09 (0.40)
23.0	9.2 (0.005)	11.2 (0.007)	4.7	90 (9)	0.96 (0.18)	1.99 (0.45)
22.8	10.3 (0.009)	12.9 (0.011)	5.2	90 (9)	0.87 (0.21)	1.68 (0.49)
58-day rats (10 rats per group)
56.9 (sham)	0.34 (0.002)	0.42 (0.003)	0.18	0	--	--
59.9	4.9 (0.028)	5.2 (0.030)	2.6	0	--	--
53.8	5.1 (0.026)	5.7 (0.028)	2.8	20 (13)	0.37 (0.28)	0.89 (0.83)
58.9	5.6 (0.029)	6.0 (0.031)	3.1	30 (14)	0.34 (0.19)	0.70 (0.36)
58.2	6.7 (0.034)	7.2 (0.036)	3.6	30 (14)	0.52 (0.27)	1.23 (0.73)
58.7	7.8 (0.041)	9.0 (0.047)	4.1	60 (15)	0.37 (0.14)	0.75 (0.44)
57.9	8.7 (0.049)	10.6 (0.060)	4.7	50 (16)	0.50 (0.22)	0.81 (0.45)
59.9	9.7 (0.052)	12.2 (0.065)	5.2	50 (16)	0.60 (0.22)	2.03 (1.00)
53.7	10.3 (0.059)	13.9 (0.079)	5.5	90 (9)	1.6 (0.31)	5.33 (2.08)
59.2	10.8 (0.058)	15.1 (0.081)	5.7	80 (13)	1.4 (0.31)	5.26 (1.31)

In addition, a fourth age group (156 72-days-old rats) was included for the age-dependent ultrasound-induced lung hemorrhage analysis from a compilation of three completed studies (Zachary et al. 2001; O’Brien et al. 2001, 2003b) for which the same experimental conditions were used.

Rats were weighed and then anesthetized with ketamine hydrochloride (87.0 mg/kg) and xylazine (13.0 mg/kg) administered intraperitoneally. Hair of the left thorax was removed with an electric clipper, followed by a depilatory agent (Nair^® Carter-Wallace, Inc., New York, NY) to maximize sound transmission. A black dot was placed at approximately the sixth to ninth rib to guide the positioning of the ultrasonic beam. Anesthetized animals were placed in a specially designed transducer holder. A removable pointer, attached to the transducer, was used to position the ultrasonic beam perpendicular to the skin at the position of the black dot with the beam’s focal region approximately at the lung surface. The holder with animal and mounted transducer were placed in highly degassed, temperature-controlled (30°C) water. In pulse-echo mode, the low-power output signal (see shams in Table 1) from the RAM5000 (Ritec, Inc., Warwick, RI) was displayed on an oscilloscope and used to adjust the axial center of the focal region to within 1 mm of the lung surface. Following exposure, animals were removed from the water and holder, and then euthanized under anesthesia by cervical dislocation.

The thorax was opened and the thickness of each left thoracic wall (skin, rib cage, and parietal pleura) was measured using a digital micrometer (accuracy: 10 μm; Mitutoyo Corp., Kawasaki, Kanagawa, Japan) to calculate the in situ peak rarefactional pressure. The left lung lobe was scored for the presence or absence of hemorrhage and then fixed by immersion in 10% neutral-buffered formalin for a minimum of 24 hours. After fixation, the elliptical dimensions of each lung lesion at the visceral pleural surface were measured using a digital micrometer where “a” is the semi-major axis and “b” is the semi-minor axis. The lesion was then bisected and the depth “d” of the lesion within the pulmonary parenchyma was also measured. The surface area (πab) of the lesion was calculated for each animal. Each half of the bisected lesion was embedded in paraffin, sectioned at 5 μm, stained with hematoxylin and eosin, and evaluated microscopically.

Exposimetry

The exposimetry and calibration procedures have been described previously in detail (Zachary et al. 2001). Ultrasonic exposures were conducted using a focused f/1 19-mmdiameter, lithium niobate ultrasonic transducer (Valpey Fisher, Hopkinton, MA). Water-based (highly degassed water, 22°C) pulse-echo ultrasonic field distribution measurements were performed according to established procedures (Raum and O’Brien 1997) and yielded a center frequency of 2.8 MHz, a fractional bandwidth of 12%, a focal length of 19 mm, a -6-dB focal beamwidth of 470 μm, and a -6-dB depth of focus of 2.7 mm. Ultrasound-induced lesions at superthreshold levels have generally been larger in diameter than the reported -6-dB focal beamwidth (O’Brien et al. 2001) because (1) there is lung motion as the animal breaths and (2) as p_r(insitu) increases above threshold, more of the beam incident on the lung surface is involved in producing damage. The reported water-based pulse-echo -6-dB focal beamwidth is provided for transducer/beam characterization purposes only (e.g. -40-dB focal beamwidth is 1 mm).

An automated procedure, based on established standards (ODS 1998; AIUM/NEMA 1998), was used to routinely calibrate the ultrasound fields (Sempsrott and O’Brien 1999; Zachary et al. 2001). Briefly, the source transducer’s drive voltage was supplied by a RAM5000. A calibrated PVDF membrane hydrophone (Marconi Model Y-34-6543, Chelmsford, UK) was mounted to the computer-controlled micropositioning system (Daedal, Inc., Harrisburg, PA). The hydrophone’s signal was digitized (500 MS/s, LeCroy Model 9354TM Oscilloscope, Chestnut Ridge, NY), the output of which was fed to the same computer (Dell Pentium II, Dell Corporation, Round Rock, TX) that controlled the positioning system. Off-line processing (Matlab, The Mathworks, Natick, MA) yielded the water-based peak rarefactional pressure p_r(invitro) and the water-based peak compressional pressure p_c(invitro) (Table 1). The Mechanical Index (MI) was also determined from the measurement procedure (ODS 1998); the $\text{MI}={\text{p}}_{\text{r}.3}/\sqrt{{\text{f}}_{\text{c}}}$ where p_r.3 is the computationally derated (0.3 dB/cm-MHz) peak rarefactional pressure (in MPa) and f_c is the center frequency (in MHz). The MI is reported because it is a regulated quantity (FDA 1997) of diagnostic ultrasound systems, and its value is available to system operators. Thus, there is value to provide the MI for each exposure setting in order to give general guidance to manufacturers and operators as to the exposure levels used in this study. Further, as seen from the MI definition, it is a quantity that must be determined from the water-based measurement process, and cannot be determined from the in situ (at the pleural surface) peak rarefactional pressure, p_r(insitu). The study herein did not use the MI as a design variable; rather the p_r(insitu) was the principal design variable.

A total of fourteen independent water-based calibrations of the 2.8-MHz transducer were conducted before, during, and after the four-month period of the experiments. At the exposure levels used in this study, the relative standard deviation (SD×100/mean) was 6% for p_r(invitro) and 8% for p_c(invitro).

The in situ (at the pleural surface) peak rarefactional and compressional pressures (p_r(insitu) and p_c(insitu), respectively; Table 1) were estimated, as has been done in previous studies (Zachary et al. 2001) for each rat, using the mean attenuation coefficient at 2.8 MHz of 2.8 dB/cm (Teotico et al. 2001; Towa et al. 2002) and the individually measured intercostal tissue thicknesses.

Statistical analysis

The primary interest in this analysis was to assess the combined effect of an animal’s age and p_r(insitu) on the occurrence, depth and surface area of lesions in exposed lungs. Probit regression analysis and tobit regression analysis were used to assess the significance of the age and p_r(insitu) effects. The effective dose (ED) thresholds corresponding to 5% and 50% risk levels are termed, respectively, ED₀₅ and ED₅₀ (O’Brien et al. 2001, 2003a, 2003b, 2006; Simpson et al. 2004).

The lesion occurrence data were modeled by probit regression analysis, with age-dependent models of the form

$$Pr(\text{Y}>0)=\mathrm{\Phi }({\beta }_{\text{a}0}+{\beta }_{\text{a}1}\text{x}),\text{a}=1,2,3,4,$$ (1)

where x is p_r(insitu) in megapascals (MPa); the index a = 1, 2, 3 or 4 refers to one of the four age groups: 1) 12–14 days; 2) 22–24 days; 3) 53–62 days and 4) 72 days; Y denotes the size of a lesion so Y > 0 denotes that a lesion is detected whereas Y = 0 denotes that no lesion is detected; and $\mathrm{\Phi }(\text{u})={(2\pi )}^{-1/2}{\displaystyle \sum _{-\infty }^{\text{u}}}{\text{e}}^{-{\text{x}}^2/2}\text{dx}$, the standard normal distribution function. Maximum likelihood estimates and standard errors for the coefficients β_a0 and β_a1 were computed from the occurrence data using the “glm” function in the R statistical programming language (Ihaka and Gentleman 1996; Hosmer and Lemeshow 2000).

To evaluate age and p_r(insitu) effects on lesion occurrence, goodness-of-fit likelihood ratio statistics (e.g. Hosmer and Lemeshow 2000, p.146) were computed for the following hypotheses versus the general alternative with eight unconstrained parameters. Below the hypotheses are ordered from least restrictive to most restrictive:

• H₀ (No interaction between age and p_r(insitu)): β₁₁ = L = β₄₁

• H₁ (No effect of age): β₁₀ = L = β₄₀; β₁₁ = L = β₄₁

• H₂ (No age or p_r(insitu) effect): β₁₀ = L = β₄₀; β₁₁ = L = β₄₁ = 0

The hypotheses were tested for significance by comparing the corresponding likelihood ratio statistic for each hypothesis to the appropriate reference chi-square distribution with degrees of freedom as indicated in Table 1. Note that if the computed p-value < 0.05, then the hypothesis is rejected at significance level 0.05, otherwise it is accepted. If H₀ is rejected, for example, at the common significance level 0.05, then the conclusion is that there is a significant interaction between age and p_r(insitu). If H₀ is accepted, then the effects of age and p_r(insitu) are additive, in which case a likelihood ratio test is performed to determine if the main effect for age is significant. If H₁ is accepted, indicating the lack of any age effect, then testing hypothesis H₂ provides a check on whether the study has adequate power to detect the well-known p_r(insitu) effect, as the vast previous literature strongly indicates an acoustic pressure effect.

Data on lesion depth and lesion surface area were modeled by tobit regression analysis. If a lesion occurred, then the measured depth and surface area were recorded. If no lesion occurred, then the lesion depth and surface area were recorded as zero values. Tobit regression analysis models these nonnegative-valued responses by treating 0 as a left-censoring point for Gaussian response data, that is, the model for the response Y, either depth (in mm) or square root of surface area (in mm), has the form

$$\text{Y}=max(0,{\gamma }_{\text{a}0}+{\gamma }_{\text{a}1}\text{x}+\epsilon ),\text{a}=1,2,3,4,$$ (2)

where ε is a Gaussian noise variable with mean zero and standard deviation σ. Maximum likelihood estimates and standard errors for the parameters γ_a0, γ_a1 and log(σ) were computed using the “survreg” function in R (Ihaka and Gentleman 1996; Galfalvy and Simpson 1999).

To evaluate age and acoustic pressure effects on lesion depth and surface area, likelihood ratio statistics were computed for the following hypotheses:

• H₀* (No interaction between age and p_r(insitu)): γ₁₁ = L = γ₄₁

• H₁* (No effect of age): γ₁₀ = L = γ₄₀; γ₁₁ = L = γ₄₁

• H₂* (No age or p_r(insitu) effect): γ₁₀ = L = γ₄₀; γ₁₁ = L = γ₄₁ = 0

Goodness-of-fit likelihood ratio statistics were computed for these hypotheses to test for age and p_r(insitu) effects. The goodness-of-fit statistics test each of these hypotheses against the general alternative model in which both the intercept, γ_0a, and the p_r(insitu) coefficient, γ_1a, depend on the age group.

The estimates for Model 1 (Eqn 1) and Model 2 (Eqn 2) were used to compute ED₀₅ and ED₅₀ thresholds and confidence intervals for these thresholds, as previously described (O’Brien et al. 2003a, 2003b; Simpson et al. 2004).

RESULTS

Chest wall thickness

The chest wall thickness over the left lung was measured for each rat. The mean±SD (range) values of the chest wall thicknesses were: 1.6±0.2 (1.3–1.9) mm (12–14-day, n = 110), 1.8±0.1 (1.6–2.1) mm (22–24-day, n = 110), 3.8±0.5 (3.2–4.6) mm (53–62-day, n = 100), 4.6±0.7 (3.3–6.0) mm (72-day, n = 156).

Lesion occurrence, depth and area

The ultrasound-induced lung hemorrhage results of rats as a function of age are shown in Fig 1. The data from the three youngest age groups (12–14-day, 22–24-day and 53–62-day rats) were acquired for this study. The 72-day rat data were compiled from three separate studies (Zachary et al. 2001; O’Brien et al. 2001, 2003b) for which the experimental conditions were the same as used herein.

Figure 1.

(a) Lesion occurrence, (b) lesion depth and (c) lesion surface area as a function of the in situ peak rarefactional pressure. The dashed lines are straight lines connecting the mean values. The data from the three age groups (12–14-day, 22–24-day and 53–62- day rats) were acquired for this study. The 72-day rat data were compiled from three separate studies (Zachary et al. 2001; O’Brien et al. 2001, 2003b) for which the exposure conditions were the same as used herein, that is, 2.8-MHz center frequency, 1-kHz pulse repetition frequency, 10-s exposure duration and 470-μm beamwidth. Error bars are the standard errors of the mean (n = 10).

The statistical evidence for age and p_r(insitu) effects on lesion occurrence is summarized in Table 2. The table presents goodness-of-fit likelihood ratio statistics for the models implied by hypotheses H₀, H₁ and H_2. The results indicate that H₀ and H₁ are accepted at the significance level 0.05, so there is no evidence of an age effect on occurrence of lesions, while H₃ is rejected, with a p-value less than 10⁻⁶, so p_r(insitu) is a highly significant factor and the test demonstrates a strong power in detecting it.

TABLE 2.

Goodness-of-fit likelihood ratio statistics for age and p_r(insitu) effects in the occurrence of lesions based on probit regression model: in each case the alternative hypothesis is the full model with age-dependent intercepts and p_r(insitu) coefficients.

Hypothesis	−2Log-likelihood ratio	df	p-value
H0: No age by pr(insitu) interaction	2.11	3	0.555
H1: No age main effect	6.97	6	0.324
H2: No age or pr(insitu) effect	145.4	7	< 10−6

The statistical evidence for age and p_r(insitu) effects on lesion depth is summarized in Table 3. The table presents goodness-of-fit likelihood ratio statistics of the hypotheses H₀^*, H₁^* and H₂^* versus the general model of age-dependent baselines and p_r(insitu) effects, based on tobit analysis of lesion depth. The results indicate that H₀^* and H₁^* are accepted, so there is no evidence of an age effect on lesion depth. The p_r(insitu)effect was statistically significant, with a p-value less than 10⁻⁶.

TABLE 3.

Goodness-of-fit likelihood ratio statistics for age and p_r(insitu) effects in lesion depth based on tobit regression model: in each case the alternative hypothesis is the full model with age-dependent intercepts and p_r(insitu) coefficients.

Null Hypothesis	−2 Log-Likelihood Ratio	df	p-value
H0*: No age by pr(insitu) interaction	4.36	3	0.225
H1*: No age effect	10.2	6	0.117
H2*: No age or pr(insitu) effect	159.8	7	< 10−6

The statistical evidence for age and p_r(insitu) effects on lesion surface area is summarized in Table 4. The table presents likelihood ratio tests of the hypotheses H₀^*, H₁^* and H₂^* versus the general model of age-dependent baselines and p_r(insitu) effects on lesion surface area, based on tobit regression analysis of the root-surface area. The interaction between age and p_r(insitu) is not significant, so an additive model is appropriate. The near significance of the goodness-of-fit statistic for H₁^* suggests a possible weak age dependence on the baseline surface area of lesions, given that they occur. This is confirmed by a likelihood ratio test of the age main effect, which gives a value of 12.21–3.70=8.51 on 3 degrees of freedom and a p-value = 0.0365.

TABLE 4.

Goodness-of-fit likelihood ratio statistics for age and of p_r(insitu) effects in lesion surface area on tobit regression model for root-surface area: in each case the alternative hypothesis is the full model with age-dependent intercepts and p_r(insitu) coefficients.

Null Hypothesis	−2 Log-Likelihood Ratio	df	p-value
H0*: No age by pr(insitu) interaction	3.70	3	0.295
H0*: No age effect	12.21	6	0.057
H0*: No age or pr(insitu) effect	166.5	7	< 10−6

Figure 2 displays 95% confidence intervals for ED₀₅ and ED₅₀ thresholds. These were derived from the full models that included age-dependent intercepts and p_r(insitu) coefficients (Eqn 1; Eqn 2). ED₀₅ thresholds are relevant for safety; ED₅₀ thresholds are p_r(insitu) levels at which 50% of the animals are expected to exhibit lesions. The figure is consistent with the lack of significant age differences among the ED₀₅ (safety) thresholds, although there is some evidence for age variation among the ED₅₀ thresholds derived from the lesion depth and surface area models.

Figure 2.

Plots of 95% confidence intervals for the estimated ED₀₅ (solid horizontal lines with the ED₀₅ estimates located at the centers) and ED₅₀ (dotted horizontal lines with the ED₅₀ estimates located at the centers) based on (a) lesion occurrence, (b) lesion depth and (c) lesion area in rats. The x-axis indicates the in situ peak rarefactional pressure, and y-axis displays the rat age. The two vertical lines in each box represent the weighted averages of the ED₀₅ and ED₅₀ estimates, respectively, with the average values at the base of each vertical line. For each of the four studies, a common exposure regime was used: 2.8-MHz center frequency, 1-kHz pulse repetition frequency, 10-s exposure duration and 470-μm beamwidth.

DISCUSSION

The objective of this study was to assess the role of rat age on the threshold and superthreshold behaviors of ultrasound-induced lung hemorrhage because reports from previous findings (Tarantal and Canfield 1994; Baggs et al. 1996; Dalecki et al. 1997a, 1997b; O’Brien and Zachary 1997; O’Brien et al. 2003b) suggest that there was a lack of consistency observed relative to ultrasound-induced lung hemorrhage as a function of the animal’s age.

There were no statistically significant age effects on occurrence or depth of lesions induced by focused ultrasound. There was some evidence of an age dependence of the lesion surface area given that a lesion had occurred, but no evidence that the p_r(insitu) coefficient was dependent on age. Estimated thresholds were consistent with what has been found in other rat studies (Zachary et al. 2001; O’Brien et al. 2001, 2003b). It is noteworthy that the strong age-dependence of thresholds found previously in pigs (O’Brien et al. 2003a) did not manifest itself in the present age-dependent threshold study in rats.

In summary, the conflicting observations remain. Four species have been investigated (mice, rats, pigs, monkeys). No age-dependent effect was reported in mice (Dalecki et al. 1997a). Age-dependent effects were reported in monkeys (Tarantal and Canfield 1994). Conflicting age-dependent effects were reported in pigs (Baggs et al. 1996; Dalecki et al. 1997b; O’Brien et al. 2003a). And, this study reported no age-dependent effect in rats. It is interesting to observe that no age-dependent effects have occurred in the two rodent models investigated, whereas to some extent, age-dependent effects occurred in monkeys and pigs. It is thus likely that age dependency of lung lesions is species dependent, and the higher the animal is in the evolutionary chain, the more likely is such an effect. However, this latter comment is speculative and not able to be tested with the currently available data. Nevertheless, if it were true, then it is likely that ultrasound-induced lung hemorrage thresholds would be age-dependent in humans, although the data do not suggest whether younger or older humans would be more susceptible to damage.