Periimplant Breast Gas at High Altitudes: Prevalence, Significance, and Possible Associated Factors
Abstract
OBJECTIVE. The objective of our study was to confirm the suspected incidental nature of periimplant gas and characterize possible predisposing factors.
MATERIALS AND METHODS. Three hundred twenty-one chest CT examinations of patients with breast implants over a 2-year period were identified using a research database query. Scans were evaluated for the qualitative presence of gas, the implant type, and the presence of implant rupture, capsular calcifications, and axillary clips. Subjects' self-reported home state address was included as a surrogate for recent airline travel or travel from lower altitudes to our center located in Denver, CO, which is 1 mile (1.6 km) above sea level.
RESULTS. Of the 321 study subjects, 55 (17.1%) had periimplant gas present. No subject had CT signs or clinical evidence of chest wall infection. Periimplant breast gas was significantly associated with residence outside Colorado (odds ratio [OR], 28.32; 95% CI, 10.60–75.70), silicone implant type (OR, 14.56; 95% CI, 5.61–37.81), and implant rupture (OR, 4.21; 95% CI, 1.74–10.18). Capsular calcifications were associated with gas in backward elimination analysis only (OR, 2.15; 95% CI, 1.03–4.50). No association was found between periimplant gas and implant location, patient age, or the presence of axillary clips.
CONCLUSION. Periimplant breast gas was relatively common in our patient population, and no association with infection was found. Our results suggest that the development of gas is related to atmospheric pressure, implant filler, and implant integrity. Gas around breast implants after airline travel or an altitude change can be safely dismissed in the absence of other associated findings.
Patients with breast implants have long reported unusual sensations in their breasts during flying or travel to mountainous areas, including pain, swelling, crepitus, and “sloshing” noises [1]. Gas in or around breast implants (periimplant breast gas) on CT of the chest is a common incidental finding at our institution (Fig. 1), an outpatient pulmonary referral center located approximately 1 mile (1.6 km) above sea level. The clinical significance and cause of the gas have not yet been determined. As little as 2 mL of gas in breast implants is reported to produce noise, and we postulate that the periimplant breast gas identified on CT may help explain the odd sensations or noises that patients with breast implants feel with changes in altitude [2]. Previously published studies in the literature on periimplant gas formation is limited to in vitro implant compression-decompression experiments and case reports of patients flying and then undergoing imaging. The existing studies in the literature have asserted the incidental and benign nature of this finding based only on individual cases [1, 3]. The proposed mechanisms have included implant volume changes, expansion of trapped gases in or around the implant, and a mechanism analogous to decompression sickness (the “bends”) in which bubble formation is due to the diffusion of gases in and out of an implant [3].
Fig. 1A —52-year-old woman with breast implants.
A, CT images show periimplant gas around silicone implants.
Fig. 1B —52-year-old woman with breast implants.
B, CT images show periimplant gas around silicone implants.
This study was performed to confirm the existing hypothesis that periimplant breast gas is a common incidental finding at high altitudes and is not necessarily associated with infection. In addition, we analyzed patient and implant data in hopes of resolving which of the competing theories about the causes of periimplant breast gas formation is correct.
Materials and Methods
Institutional review board approval and an exception for informed consent were obtained for this retrospective study and chart review.
Subjects
Our study was performed as a retrospective case-control review of CT examinations of the chest obtained at an outpatient tertiary and quaternary pulmonary referral hospital located in Denver, CO. A query was performed of our internal research database using the following criteria: CT of the chest performed in the 2 years between August 20, 2011, and August 20, 2013, and a final report containing the word “implants.” This query yielded 325 subjects who had breast implants; the results that included the word “implants” in the context of metastatic disease or implant removal were excluded. Four subjects were excluded from analysis because they had both saline and silicone breast implants. Only the most recent CT examination was reviewed. A total of 321 unique patients were analyzed. All examinations were performed without IV contrast material and were performed for the assessment of pulmonary or pleural disease.
Self-reported home state mailing address was then obtained from the medical record as a surrogate record of recent airline travel or travel from lower altitudes to our center. We assumed that patients with an in-state zip code did not require air travel and were physiologically acclimated to the elevation of our medical center. Conversely, those with zip codes outside Colorado were assumed to have required commercial airline travel and to not have been acclimated to the elevation of our medical center.
Image Analysis and Criteria
The CT examinations were reviewed independently by a third-year radiology resident and a cardiopulmonary imaging fellow. The examinations were scored as qualitatively positive or negative for multiple criteria (Table 1).
| Criteria | Values |
|---|---|
| Gas in or around breast implant | Present or absent |
| Evidence of infection | Present or absent |
| Implant side | Left breast, right breast, or both breasts |
| Implant location | Retropectoral, retroglandular, or both |
| Implant composition | Silicone or saline |
| Implant rupture | Present or absent |
| Capsular calcifications | Present or absent |
| Axillary clips | Present or absent |
Periimplant gas was scored as present or absent on the basis of a reviewer's visual assessment. If gas was seen in the implant envelope or in the soft tissues immediately adjacent to the implant, this variable was scored as present (Figs. 2 and 3). For saline implants with only a small volume of gas in the filling valve apparatus, this variable was scored as negative, because this finding was thought to be more than likely related to procedural factors and not to be in continuity with the bulk of the implant filling volume.
Fig. 2A —57-year-old woman with breast implants.
A, CT images show small amount of periimplant gas posterior to silicone implants.
Fig. 2B —57-year-old woman with breast implants.
B, CT images show small amount of periimplant gas posterior to silicone implants.
Fig. 3A —65-year-old woman with silicone implants and capsular calcifications.
A, CT scan obtained at approximately 5280 ft (1609 m) above sea level shows periimplant gas.
Fig. 3B —65-year-old woman with silicone implants and capsular calcifications.
B, CT scan obtained 6 months before A at another institution at approximately 700 ft (213 m) above sea level shows no periimplant gas at similar level.
Evidence of infection around the implant was defined as the presence of any of the following findings: any fluid collection around the breast or adjacent soft tissue, soft-tissue edema and swelling, or localized fat stranding. In any subject with CT findings suggestive of infection, the electronic medical record was accessed to assess for the presence of any clinical findings of localized soft-tissue infection.
Determination of implant composition as either saline or silicone was performed by visual estimate of the implant density on images obtained with standard abdominal window and level settings. Equivocal cases were settled by an ROI density measurement in the center of the implant: Implants with a density value 50 HU or greater were classified as silicone and those less than 50 HU were classified as saline.
Implant location, implant rupture status, and the presence of capsular calcification were determined on the basis of previously described methods [4]. Implants were scored as either ruptured or unruptured, and no distinction between intracapsular rupture and extracapsular rupture was made in the final analysis.
Statistical Analysis
The presence or absence of gas in either breast implant was modeled using a generalized linear mixed model with a binary distribution (logit link). The following variables were predictors of interest: implant type (silicone or saline), location of implant (retroglandular or retropectoral), capsular calcifications (present or absent), axillary clips (present or absent), rupture (any or none), Colorado address or non-Colorado address, age of patient, and breast implant side (left or right). Repeated measures in space were accounted for with an unstructured covariance matrix. All variables were entered into the model, and insignificant variables were removed with a backward selection. Variables with p values of < 0.15 remained in the model. Separate models were run for the presence of gas on the left versus right side, and the results were similar enough to combine results for the left and right sides into a single model. A p value of < 0.05 was considered statistically significant in all models. All analyses were performed with statistics software (SAS, version 9.4, SAS Institute).
Results
A total of 55 of 321 (17.1%) examinations showed periimplant gas. No examinations were scored as positive for evidence of infection. Our study group included a majority of patients with silicone implants (56.4%), patients residing in Colorado (56.1%), and patients with unruptured implants (82.9%). Characteristics were otherwise similar between the patients with scans showing periimplant gas and patients with scans showing no perimplant gas. A summary of subject and implant characteristics and associations with periimplant gas is given in Table 2.
| Variable | Gas In or Around Breast Implant on CT | p | |
|---|---|---|---|
| Present (n = 55 Examinations) | Absent (n = 266 Examinations) | ||
| Implant type, no. (%) of patients | < 0.0001a | ||
| Silicone | 49 (27.1) | 132 (72.9) | |
| Saline | 6(4.3) | 134 (95.7) | |
| Implant location, no. (%) of patients | 0.08a | ||
| Retroglandular | 23 (22.1) | 81 (77.9) | |
| Retropectoral | 30 (14.2) | 181 (85.8) | |
| Both | 2(33.3) | 4 (66.7) | |
| Capsular calcifications, no. (%) of patients | 0.19a | ||
| Present | 20 (21.5) | 73 (78.5) | |
| Absent | 35 (15.4) | 193 (84.6) | |
| Axillary clips, no. (%) of patients | 0.56a | ||
| Present | 5 (22.7) | 17 (77.3) | |
| Absent | 50 (16.7) | 249 (83.3) | |
| Rupture, no. (%) of patients | 0.0001a | ||
| None | 35 (13.2) | 231 (86.8) | |
| Present (extracapsular or intracapsular) | 20 (36.4) | 35 (63.6) | |
| Home address, no. (%) of patients | < 0.0001a | ||
| Colorado zip code | 7(3.9) | 173 (96.1) | |
| Non-Colorado zip code | 48 (34.0) | 93 (66.0) | |
| Patient age (y) | 0.043b | ||
| Mean ± SD | 63.0 ± 11.1 | 59.6 ± 11.5 | |
a
Fisher exact test for categoric variables.
b
Two-sample t test assuming equal variance for age.
A mailing address outside Colorado was the strongest factor associated with periimplant gas formation (odds ratio [OR], 28.32; 95% CI, 10.60–75.70) (Table 3). Also, silicone implant composition (OR, 14.56; 95% CI, 5.61–37.81), implant rupture (OR, 4.21; 95% CI, 1.74–10.18), and the presence of capsular calcification (OR, 2.15; 95% CI, 1.03–4.50) were also significantly associated with periimplant gas formation. The presence of axillary clips (OR, 1.65; 95% CI, 0.37–7.41) and the implant location (OR, 1.26; 95% CI, 0.69–2.32) were not significantly associated with the presence of periimplant gas.
| Variables | Univariate Analysis | Backward Elimination Analysisa | ||
|---|---|---|---|---|
| Odds Ratio (95% CI) | pb | Odds Ratio (95% CI) | pb | |
| Implant type (silicone vs saline) | 10.44 (4.23-25.81) | < 0.0001 | 14.56 (5.61-37.81) | < 0.0001 |
| Implant location (retroglandular vs retropectoral) | 1.26 (0.69-2.32) | 0.45 | ||
| Capsular calcifications (absent vs present) | 0.67 (0.38-1.17) | 0.16 | 2.15 (1.03-4.50) | 0.04 |
| Axillary clips (absent vs present) | 1.65 (0.37-7.41) | 0.51 | ||
| Implant rupture (any vs none) | 3.24 (1.63-6.45) | 0.0008 | 4.21 (1.74-10.18) | 0.0014 |
| Home address (non-Colorado zip code vs Colorado zip code) | 16.03 (6.86-37.41) | < 0.0001 | 28.32 (10.60-75.70) | < 0.0001 |
| Patient age (10-year increase) | 1.35 (1.02-1.79) | 0.03 | ||
| Implant side (right vs left) | 1.11 (0.84-1.47) | 0.45 | ||
a
A p value cutoff of 0.15 was used for backward elimination analysis. Implant location, presence of axillary clips, patient age, and side of implant were insignificant and were removed from the model.
b
The p value was calculated using the type 3 test from the generalized linear model.
Discussion
Periimplant gas was a common finding in our study, seen in 17.1% of subjects. None of the study subjects had evidence of breast infection. The presence of periimplant gas was highly associated with an address outside Colorado, silicone implant composition, the presence of capsular calcification, and implant rupture. The high prevalence of incidentally identified periimplant gas strongly suggests that this finding is benign and of no direct clinical concern in isolation.
The higher prevalence of periimplant gas in patients residing outside Colorado is in agreement with case reports [3] and in vitro experiments [5] and suggests decreased atmospheric pressure as the most likely cause for periimplant gas formation. Colorado has an estimated mean altitude of approximately 6800 ft (2073 m), significantly higher than most of the rest of the United States, which has an estimated mean elevation of 2500 ft (762 m) [6]. The mean elevation in Colorado and at our center (5280 ft [1609 m]) is similar to the average pressurized altitude in airline cabins (6214 ft, 1894 m) [7]. As others have previously suggested [3], we theorize that periimplant gas develops in a process analogous to decompression sickness (the bends): As atmospheric pressure decreases, dissolved nitrogen from the blood in the breast or from the implant itself comes out of solution in or around the implant. Further supporting this theory is our finding of a much higher rate of periimplant gas in association with silicone as compared to saline implants. The solubility of nitrogen in silicone implants is greater than that in saline implants; therefore, silicone implants have more potential for nitrogenous gas precipitation assuming that at least some of the nitrogenous gas comes from the implant itself [5].
The significant association of capsular calcifications and periimplant gas is of uncertain cause, but we hypothesize that it may be secondary to impaired diffusion of gases through a thick fibrous breast implant capsule, which then results in a longer time to equilibration after ambient pressure changes. In a histologic study of explanted silicone breast implant capsules [8], the authors noted that calcifications were associated with thicker capsules, which supports our theory.
Implant rupture was also significantly associated with the presence of periimplant gas. Given the likely pressure-related mechanism, the rupture of breast implants may lead to localized areas of low pressure as the total breast volume is focally reduced, leading to focal areas of gas precipitation when combined with low ambient atmospheric pressure. Another possible mechanism may be that extracapsular material results in a localized inflammatory response and in fibrous tissue formation, which then impair the reversal of gas precipitation.
Our study is limited by its retrospective design. Moreover, the subjects' mailing address state was used as a proxy for recent plane travel rather than actual documentation of this variable. Home addresses may be misreported, and study subjects could have been residents of Colorado for any amount of time before undergoing CT. Documentation of recent travel was ultimately not pursued to protect patient privacy. Also, CT examinations were reviewed by only one scorer. However, given the conspicuousness of periimplant gas, confirmatory scoring was not deemed mandatory. Additionally, examinations were reviewed by trainees, but both trainees had multiple dedicated months of experience in cardiothoracic and breast imaging. Although the time since implantation was not captured in our study, gas trapped during filling or manufacture of the implants was shown to diffuse through the implant shell within 45 days in an in vitro study [9] and is unlikely to have been a major confounder in our study. Last, we made no distinction between gas inside the implant envelope and gas immediately adjacent to the implant because we assumed that the underlying mechanism was the same for both phenomena.
In summary, our study provides the first systematic characterization of periimplant breast gas to date. It is a common finding on chest CT examinations performed at a high altitude, especially after recent travel, and can be safely ignored by radiologists. Patients with ruptured, silicone, or peripherally calcified implants are particularly predisposed to this phenomenon. Patients and plastic surgeons can be assured that isolated crepitus, swelling, or other odd sensations around breast implants during air or mountain travel are likely explained by simple precipitation of periimplant gas and are of no clinical concern.
Footnote
Based on a presentation at the Society of Thoracic Radiology 2014 annual meeting, San Antonio, TX.
References
1.
Bachman JJ. Boyle's law and breast implants. N Engl J Med 1994; 331:483–484
2.
Eppley BL. Preventing postoperative noise in saline breast implants. Plast Reconstr Surg 1999; 103:1979–1981; discussion, 1982
3.
Lewin JM. Peri-implant gas following air travel to high altitude. Plast Reconstr Surg 2013; 131:465e–467e
4.
Juanpere S, Perez E, Huc O, Motos N, Pont J, Pedraza S. Imaging of breast implants: a pictorial review. Insights Imaging 2011; 2:653–670
5.
Vann RD, Riefkohl R, Georgiade GS, Georgiade NG. Mammary implants, diving, and altitude exposure. Plast Reconstr Surg 1988; 81:200–203
6.
U.S. Census Bureau website. The 2012 statistical abstract: PDF version. www.census.gov/compendia/statab/2012edition.html. Accessed 2014
7.
Cottrell JJ. Altitude exposures during aircraft flight: flying higher. Chest 1988; 93:81–84
8.
Siggelkow W, Faridi A, Spiritus K, Klinge U, Rath W, Klosterhalfen B. Histological analysis of silicone breast implant capsules and correlation with capsular contracture. Biomaterials 2003; 24:1101–1109
9.
Schanzer A, Wong GB. Air dissipation in saline breast implants. Plast Reconstr Surg 2002; 109:2238–2241
Information & Authors
Information
Published In
Copyright
© American Roentgen Ray Society.
History
Submitted: May 13, 2015
Accepted: June 7, 2015
First published: August 21, 2015
Keywords
Authors
Metrics & Citations
Metrics
Citations
Export Citations
To download the citation to this article, select your reference manager software.
There are no citations for this item