Using Breast MRI to Evaluate Complications of Autologous Fat Grafting

Li-Kuo Shen1, Shoei-Loong Lin2, Cheuk-Sing Choy3, Wing P Chan4, Mimmo Gasbarri5 and Ting-Kai Leung4,6* 1Department of Radiology, Shuang Ho Hospital, Taipei Medical University, New Taipei City 235, Taiwan 2Department of Surgery, Taipei Hospital, Ministry of health and Welfare, Taiwan., R.O.C and Department of Surgery, School of Medicine, College of Medicine, Taipei Medical University, Taiwan 3Department of Emergency and Intensive Care, Taipei Hospital, Department of Health, Taiwan 4Department of Radiology, School of Medicine, College of Medicine, Taipei Medical University and Department of Radiology, Wan Fang Hospital, Taipei Medical University, Taiwan 5Dept. of Natural Health Studies / London College of Training & Management, London, UK 6Diagnostic Radiology Department, Taipei Hospital, Ministry of Health and Welfare, Taiwan


Introduction
Autologous fat grafting is a technique that involves using the patient's own adipose tissue to be applied on a subcutaneous area of the body to increase total fat volume [1][2][3]. Breast augmentation using autologous fat grafting is not associated with implant related problems such as implant leakage, breast deflation, visible or palpable implants, or the development of breast capsular contracture. Regarding successful cases of autologous fat grafting, under the condition of suspected tumor growth, a usual biopsy procedure can be performed without risk of a stabbing rupture of the implanted materials. However, autologous fat grafting in the breast is not a simple procedure and should be performed by highly trained and skilled surgeons. The recipient of autologous fat grafting is at risk of intraoperative pneumothorax, calcification, multiple cyst formation [4], focal breast indurations, pain, infection, and abscesses [5], abnormal breast discharge, and reactive lymphadenopathy. In contrast to the aforementioned successful cases, these patients require life-long follow up for breast cancer screening because of the difficulty of cancer detection. These patients should also undergo long-term imaging analyses [6,7]. Autologous fat grafts are also believed to have the disadvantage of the unpredictability graft tissue survival because of fat necrosis and re-absorption [8]. The incidence of calcifications (mostly due to fat necrosis) after all types of breast operations varies, but has been reported in as many as 50 percent of patients within 2 years [9].
It was also reported that breast asymmetrical densities were present in approximately half of patients throughout the follow-up period of post-mammoplasty. In additions, fat necrosis occurred in approximately one tenth of the patients [9]. Many plastic surgeons have rejected fat grafting in the breast because of the speculation that transplanted fat might undergo necrosis or die off gradually, forming lumps or calcifications. Fat grafting raises the concern of interfering with breast cancer detection. In fact, injecting small volumes of fat produces microcalcifications, which can be indistinguishable from carcinomas in mammograms; whereas injecting larger volumes might result in the liponecrotic pseudocyst formation that can interfere with physical breast examination and cause a growing and painful mass [10,11]. Regarding breast image screening and diagnosis for such patients, mammography is favored by radiologists to differentiate cancer microcalcifications from benign calcifications of the breast. However, most cosmetic mammoplasty patients are among the young age group that should receive non-ionized radiation screening of breast tissues. Therefore, the first line screening for this age group is usually sonography, but the sensitivity and specificity of sonography to differentiate calcification and microcalcifications is not reliable [12]. Breast MRI should be considered a first line screening tool or Breast MRI has also been reported to be more capable than mammography of early detection of fat necrosis and clinically nondetectable masses and nodules, although MRI is not used to detect calcifications or microcalcifications [13].
During the past 2 years in Taiwan, there has been a rapid growth of autologous fat grafting procedures performed for cosmetic reasons. In this study, we demonstrate cases of autologous fat grafting in the breast, including complications, based on clinical presentations and breast MRI imaging. We also discuss the possible method of early detection of the Complications of Autologous Fat Grafting of the breast.

Cases collections
From March 2011 to February 2012, twenty-one ethnic Chinese (Taiwanese) women were transferred to the Breast Health Center (Taipei Medical University Hospital) for delicate breast MRI study. These patients exhibited different symptoms and signs after the procedure. Clinical and image information were collected between March 2011and February 2012 from the hospital database. The patients were between 25 and 62 years old, and none had undergone treatment in the period between mammoplasty and breast MRI study. Furthermore, information collected and recorded included clinical expression, laboratory data such as Immunoglobulin E (IgE), Antinuclear Antibodies (ANA), complement (C3&C4), C-Reactive Protein (CRP), antinuclear antibodies (ANA), monocyte and bacteria culture, as well as the possible pathological diagnosis and method of treatment.

Image processing using delicate breast MRI
Pre-enhanced and post-enhanced subtracted images were obtained, and an enhanced curve analysis was performed. All imaging studies were performed on a 1.5 T dedicated spiral breast MRI system (Aurora Systems, USA) [14][15][16][17]. The breasts were fitted and positioned using a breast transmission-receiving (TR) coil. The off-resonance spiral image mode was used. A gradient echo with 10 ms of TR and T2-weighted spin echo with 2.5 ms of TR were acquired. The echo time was 5 ms for RODEO and 120 ms for the T2 sequences. The volumetric matrix size was 360x360x128 mm. The slice thickness was 1.1 mm. The field of view of each station was 20-36 cm. Sequences were performed before and after the infusion of 0.2 mmol/kg adolinium-diethylenetriamine pentaacetic acid (Magnevist; Bayer-Schering Pharma AG, Berlin, Germany), administered as a bolus dose with a power injector, followed by a 20-mL saline flush. Subtraction images were also performed [14][15][16][17].

Results
According to our data on these 21 patients (Table 1), the complications may have developed 12 days to 24 months later after        [18][19][20]. The mean of duration that represent for MRI image diagnosis was 8.1 months. All of the patients suffered from breast tenderness, palpable lumps, pain or fever. The most severe complication was immediate post-operative abscess formation and impending sepsis. All of the cases were diagnosed with fat necrosis and granuloma formations caused by fail of graft tissue survival or graft rejection. Serum test for two of the patients were exhibited CRP elevation, and serum test for IgE was also found elevated in another 3 patients. Among these 21 cases, 19% required operation or incisional drainage; 23.8% received oral antibiotics for conservative treatment; and 57.1% of the cases were suggested for observation after image diagnosis. Based on the MRI images (Figures 1-21), all of the fat grafted tissue from the 21 cases could be identified, even Case 4, who received the procedure for Case 17 ( Figure 17)

Discussion
There are two issues to discuss. Firstly, is that complications of the fat graft can be detected and survived; secondly is breast MRI is capable to distinguish malignancy from benign lesions. In early 2007, Kaufman MR et al reported no evidence of the post autologous fat grafting fat survivability and the predictability of volume restoration. In conducting a quantitative study using large-scale clinical assessment and three-dimensional volumetric imaging to provide useful outcome data, they found that fat survival of autologous fat graft in that period is questionable [21]. By contrast, a rapid growth of new procedures used to collect larger amounts of transplantable fatty tissue by using liposuction (e.g., water-assisted liposuction); and the procedure of autologous lipo-transplantation after tissue separations. By conducting a series of procedures that focus on the superfluous water, fat purification, and centrifugation, the viable fat tissue could then be harvested [22,23]. Smith et al. using breast MRI to compare the image morphological change of fat grafted breasts before and after using this type of improved technique in operations [22]. They found that 41.1% of the studied patients exhibited a significant volume increase  of grafted fat within the treated areas. The average size of increase was approximately 100-150 ml. Approximately 2.4% patients developed palpable subcutaneous nodules which finally proved to be granulomas [22]. For these cases that developed complications, the pathologic results for the excised granulomas, included fat necrosis, calcification, hyalinization, and fibroplasia [24]. Our observations support the results of these previous studies, in that the complication rate of granuloma formation (a complex of chronic inflammation with fat necrosis), is ahead of infection and mastitis. Therefore, granuloma formations due    Granulomatous formation is caused by the failed attempts of the human immune system to eliminate foreign substances. Granulomas are seen in various diseases that relate to both infectious and noninfectious [25][26][27]. Granulomatous formation is a type of inflammation, with an important feature that it may contain necrosis. The identification of necrosis in granulomas is important because granulomas with necrosis tend to have infectious causes. However, in most of our cases with granulomatous formations that were associated with fat necrosis, infection did not occur. Fat necrosis has been reported in many different surgical procedures, such as reduction mammoplasty, mastectomy and post reconstructive procedures and others [28][29][30][31][32][33][34][35]. It was also shown that some over-obese animals are predisposed to fat necrosis without trauma and surgical inductions [36,37].
The mechanism of post-operative fat necrosis is still poorly understood but a variety of factors are thought to be some of the causes, such as direct trauma, infection, avascularity, hypoperfusion  and lipolytic enzymatic digestion [36][37][38][39]. To compare the image features of granulomas under mammography, sonography and breast MRI, it was found the typical mammographic pictures of granulomas as asymmetrical focal densities with no distinct margins, ill-defined masses with spiculated contours, and bilateral multiple ill-defined nodules were seen [40]. Under sonography, granulomas may appear as nonspecific discrete or heterogenous hypoechoic masses [38][39][40]. Thus, the mammography and sonography may not provide a conclusive result for determination of the necessity of surgical treatment, or to rule out malignancy [38]. It was believed that a major part of autologous fat injection for breast augmentation would not survive due to the associated physiological response of fat to succumb to necrosis and          then lead to scarring and calcification. Under mammography, the clusters of microcalcifications in the breast may be observed but cannot be distinguished from malignant calcifications. A common biopsy or a biopsy using mammographically guided needle localization is needed [39,40].
On our dynamic breast MRI, the survival of fat grafted tissues can be qualitatively observed and quantitatively measured. Also, the most frequent finding of complication was focal or diffusely asymmetrical signal intensity changes with fluid or semi-fluid content, without significant mass effect. Nodular lesions, mass-like enhancement, ringlike enhancement, and nodular enhancement were observed. Although, most of the dynamic curves of post-gadolinium enhancement did not have characteristic findings, but based on the morphology and subtracted image, malignancy could be ruled out. From our result, we found four cases that had received operation, drainage or aspiration, but only five cases had higher levels of one or more than one of the following items: CRP, IgE or complement (C4) /ANA. We found the laboratory data are not able to predict all the corresponding post fat grafting infection and rejection with granuloma formations. It supports our contention that breast MRI should acting as first line screening method for early diagnosis and for treatment.

Conclusion
This is a study of using breast MRI to evaluate the conditions of post autologous fat grafting complications. Chronic inflammation with granuloma formations and fat necrosis has the highest incidence rate; we also found acute infection with abscess is the most urgent condition. Overall, 19% of the cases suffering complications had need of an additional operative treatment. Serum test for laboratory data should not be independently used as monitoring parameter for early detection of fat necrosis, granulomatous mastitis or abscess. However, breast MRI  is a reliable imaging tool for precise diagnosis on the fat survival, fat necrosis, and other complications.