Introduction
Decompression illness is a term used to describe diseases caused by intravascular or extravascular bubbles that are formed as a result of reduction in environmental pressure (decompression). The term covers both arterial gas embolism and decompression sickness (DCS; Vann et al., 2011). Bubbles can have mechanical, embolic, and biochemical effects with manifestations ranging from trivial to fatal (Vann et al., 2011).
Bert noted that the presence of gas bubbles in the blood was not a necessary cause of death or even visible symptoms (Bert, 1878). However, he suggested that gas in the vascular system or in the tissues must cause pain and decompression symptoms. He further suggested that the risks of paralysis or death depends upon the size of the gas bubble and hence, the capability to block the circulation.
The presence and amount of intravascular circulating bubbles can be determined by ultrasound technology. This technology is used in scuba divers as a non-invasive system for detection of intravascular bubbles after diving schedules, as an indicator of decompression stress (Doolette and Mitchell, 2001; Pollock, 2007; Germonpre et al., 2014; Møllerløkken et al., 2016). The absence of detectable bubbles has been found to highly correlate with the absence of DCS symptoms (Sawatzky, 1991), and large quantities of intravascular bubbles predispose for the development of DCS (Evans et al., 1972; Neuman et al., 1976; Spencer, 1976; Gardette, 1979; Sawatzky, 1991). However, the relationship between intravascular bubbles detected by ultrasound and DCS is largely probabilistic, with no absolute threshold for number or size of bubbles below which there is no risk (Weathersby et al., 1984). DCS has occurred in divers with few bubbles detected by Doppler (Bayne et al., 1985). In contrast, no symptoms of DCS were recorded in individuals with large quantities of intravascular bubbles (Nishi et al., 2003). A recent study has shown that venous gas emboli detected by two-dimensional echocardiography are an imperfect surrogate endpoint for DCS (Doolette, 2016). There are no studies on detectable bubbles and lethal decompression in human controlled studies for obvious ethical reasons; hence the relationship between detectable bubbles by ultrasound and fatal decompression remains unknown.
The relationship between intravascular gas bubbles in a given localization determined by ultrasound and the amount of gas bubbles and the distribution of these intravascularly or extravascularly elsewhere in the organism is unknown. This relationship might contribute to understand better the pathophysiology of DCS. In the present study, we compared the presence and amount of intravascular circulatory gas bubbles at the pulmonary artery of New Zealand White (NZW) rabbits after decompression using ultrasound, to the presence and amount of gas bubbles in different vascular and extravascular body locations found grossly post mortem (PM) during necropsy.
Discussion
To the authors knowledge this is the first study describing the relationship between bubble grades detected by ultrasound in the pulmonary artery in vivo, and the presence of macroscopic bubbles elsewhere in the organism.
The present study showed a high inter-individual variability in bubble grades detected by ultrasound, despite having similar subjects (species, gender, weight) exposed to the same treatment. This result supports previous work by Eckenhoff et al. (1990) who reported a considerable inter and intra-individual variability in bubble formation at any level of exposure (Eckenhoff et al., 1990).
The main finding from this study was that EB bubble grade measured in vivo by ultrasound correlated to mortality and to gas score determined PM: animals with EB bubble grades 1–3 survived for 1 h after decompression until sacrificed according to protocol and presented with low gas score when examined PM, while animals with EB bubble grades 4–5 died within 5–35 min after decompression and presented with high gas score of gas bubbles widely dispersed in different veins, including peripheral veins, and in the adipose abdominal tissue when examined PM.
The mortality result was unexpected since large quantities of intravascular bubbles have been reported in asymptomatic apparently healthy divers (Nishi et al., 2003). PM examination showed only two treatment responses: large amount of gas bubbles within veins and tissues in animals dying shortly after decompression vs. absence or very few gas bubbles in animals which survived and were later sacrificed. Similar observations on survival and amount of bubbles detected PM have been previously reported in sparrows, mice, rats, pigs, cats, dogs, and rabbits (including NZW), that have been rapidly decompressed from high pressures (Bert, 1878; Eggleton et al., 1945; Lever et al., 1966; Shim et al., 1967). Most of these studies support our findings regardless species and exposure pressure levels. Bert's results were more varied but they also showed that animals dying rapidly after sudden decompression always presented with large amounts of gas bubbles in the venous side of the circulatory system (Bert, 1878).
It is interesting to note that regardless the different sensitivity to decompression (demonstrated by higher prevalence of DCS symptoms and/or morbidity) of different species (Bert, 1878; Eggleton et al., 1945), necropsy findings were similar to our results showing a dual response in dead vs. surviving animals (Eggleton et al., 1945; Shim et al., 1967): all animals dying shortly after decompression showed large amounts of gas bubbles in the veins regardless of animal species, or hyperbaric exposure.
Our animals that died shortly after decompression (EB bubble grades 4 and 5) also presented with gas bubbles within the abdominal adipose tissue. This finding has been described before and used as a symptom to diagnose DCS in rats (Hyldegaard and Madsen, 1989).
The dual response (presence and high amounts of gas bubbles vs. absence of gas bubbles) of the gas score was very different from the high individual variability in EB bubbles grades measured by ultrasound. Differences found between EB bubble grade measured by ultrasound and PM gas score examination could be due to different resolution limits of the two methods. Monitoring with high-resolution ultrasound renders the possibility of detecting smaller bubbles than is possible by visual examination. Additionally, small bubbles are more susceptible to be trapped and excreted in the pulmonary capillaries (Francis and Simon, 2003) or simply get diluted. According to LaPlace equation, smaller bubbles have larger inner pressure, thus dissolve more quickly than large bubbles (Hrncír, 1996). Large bubbles are more stable and easier to see macroscopically. Additional limitation of this method is the low number of cases studied. More studies using fresh animals should be carried out to reassure the results presented in this manuscript.
Future pathological studies comparing animals with EB bubble grades 0–3 and EB bubble grades 4–5 might help us to understand better the unexpected mortality rate described in this study, as well as the pathophysiological mechanism of DCS.
This study has focused in lethal decompression, which is rare among human divers except for the most severe accidents, thus the application of these results in human clinical medicine might be limited. On the other hand, studying diseases in their most extreme expression might be helpful to detect better pathophysiological mechanisms of the disease that might not be detected otherwise. The results from the present experiment should be considered carefully due to the low number of individuals studied, but the results suggest a strong relationship between fatal decompression and high loads of gas emboli in NZW rabbits. This relationship remains unclear in less severe decompression cases in human divers (Evans et al., 1972; Neuman et al., 1976; Spencer, 1976; Gardette, 1979; Bayne et al., 1985; Sawatzky, 1991; Nishi et al., 2003).
Another limitation to our data is that there are to our knowledge no studies available today that describes where decompression induced vascular bubbles do originate from, and whether there are specific tissues or organs that are more “bubble-producers” than others. Recent studies are attempting to detect small stationary bubbles in tissues using Dual-frequency ultrasound (Swan et al., 2011, 2014). In this regard, the total gas score estimated for each animal might underweight the pathophysiological importance of high bubble loads in a single tissue. The gas score for each tissue might be more informative for that respect.
In summary, in the present study we have reported a dual response of presence of gas bubbles in postmortem examinations (high amount of gas bubbles vs. absence of gas bubbles) in animals of the same species, gender, and similar weight, exposed to the same diving profile. In vivo ultrasound measurements at the pulmonary artery related to quantity of gas widely distributed in the venous system in decompression cases. High bubble grades were related to high gas score causing death within 35 min after decompression in NZW rabbits.