Mammographic Findings After Breast Augmentation With Autologous Fat Injection

Abstract

Background: Conventional film-screen mammography is a highly effective tool for the early diagnosis of breast cancer. Although the mammographic spectrum of fat necrosis has been well documented, and many postsurgical findings mimic carcinoma in clinical examination or imaging studies, the evolution of the mammographic appearance has not previously been reported in patients with a history of breast lipoinjection.

Objective: The purpose of our study was to evaluate the mammographic findings of fat necrosis in patients who had undergone breast lipoinjection and to determine whether there are any specific features that help to distinguish fat necrosis caused by fat injection from more worrisome findings.

Methods: Bilateral mammography was performed on 20 patients who had received autologous fat injection for breast augmentation between February 1999 and June 2006. The time elapsed between surgery and the postoperative mammograms ranged from 6 months to 7 years, an average of 34.5 months. The mammographic findings of fat necrosis were divided into six categories: 1, radiolucent oil cysts; 2, microcalcifications; 3, coarse calcifications; 4, focal masses; 5, spiculated areas of increased opacity; 6, negative. The Breast Imaging Reporting and Data System (BI-RADS) was used to classify the lesions in the mammograms.

Results: The most common mammographic findings were benign bilateral scattered microcalcifications, followed by dispersed radiolucent oil cysts in the breast tissue. Microcalcifications were found on the mammogram of one patient as early as 11 months after lipoinjection. Only 3 patients showed clustered microcalcifications on their mammograms and were classified as BI-RADS III. These patients were later available for further digital mammography and reclassified as BI-RADS II.

Conclusions: Knowledge of the mammographic appearance and evolution of patterns of fat necrosis in patients who have undergone breast fat injection may enable imaging follow-up of these lesions, reducing the number of unnecessary biopsies or additional examinations and avoiding possible delays in the diagnosis of breast cancer. Because calcifications in breast parenchyma can be expected after breast fat injection, in our opinion this technique for breast augmentation should not be performed in patients with a family history of breast cancer.

To date, breast fat injection has not been proven to be a risk factor for development of breast cancer, nor has it been shown to modify the incidence of breast cancer in patients who have undergone this procedure. Mammography is still considered to be the “gold standard” for breast cancer screening because it has been shown to decrease mortality rates through early detection.¹–⁶ The sensitivity of mammography when detecting breast cancer depends on the density of a woman's breasts, which is affected by her genetic predisposition, hormonal status, and age. Other factors that influence the accuracy of the mammographic readings are as follows: the size and conspicuity of the lesion, the quality of the overall image, and the radiologist's skills and experience in interpreting mammograms.³,⁷

Little information has been reported on specific mammographic findings in patients who have undergone lipoinjection for breast augmentation. The only articles that have been published refer to patients who have had complications after undergoing this procedure and whose mammograms showed findings of fat necrosis, sometimes with calcifications that imitate malignity and require biopsy to clear the diagnosis.¹,⁸,⁹

Architectural distortion of tissue, masses, calcifications, fat necrosis, edema, or skin thickening may be radiologically visible after any kind of breast surgery. Generally, postsurgical changes in the breast decrease quickly after the first 6 months and finally stabilize mammographically about 2 years after surgery.¹⁰

Fat necrosis is a nonspecific histologic finding, and multiple processes are involved in its etiopathogenesis. In addition to surgery, the most common causes of fat necrosis are ischemia, radiation therapy and trauma.¹,⁹–¹⁵ Other rare incidents have been reported of fat necrosis in the breast caused by anticoagulant therapy with warfarin sodium (Coumadin) and enoxaparin sodium. Calciphylaxis (hypersensitivity to local calcinosis associated with secondary hyperparathyroidism in kidney failure) has also been reported.¹⁶–²⁰

Autologous fat injection for breast augmentation has become a controversial procedure. Fat necrosis, which generally evolves after this type of procedure, is usually occult in clinical follow-up and is generally evident through mammography. Even experienced radiologists encounter difficulties in assessing fat necrosis because of its wide spectrum of mammographic appearances, which may interfere with early cancer detection; this is seen as the procedure's most critical drawback.¹,⁹–¹¹,¹⁷,¹⁸,²¹–²⁴ The mammographic images of fat necrosis range from lipid cysts to findings that are suspicious for malignancy, such as clustered microcalcifications, spiculated areas of increased opacity, and focal masses.¹⁰,¹¹,¹⁸,²⁵–³¹ The purpose of our study was to investigate and identify the different mammographic appearances of fat necrosis caused by breast lipoinjection, particularly calcifications in symptom-free patients, as well as to determine whether there are any specific features that help radiologists and other specialists to distinguish the different aspects of fat necrosis caused by fat injection from other threatening findings.

Materials and Methods

The study began with the performance of a retrospective review of the mammographic images of some patients who had undergone breast lipoinjection. Later, we contacted additional patients who had undergone this type of procedure, and we obtained mammograms for our study. Between February 1999 and June 2006, these patients received treatment from several different surgeons, all of whom stated that they had used a standard microlipoinjection technique³²,³³ in which the fat for the transfer was usually obtained from several donor sites, such as the abdomen, back, thighs, and arms, through the conventional method of lipoplasty. Next, small skin incisions were made around the breast, and the fat was generally injected in small quantities. These patients said that their doctors informed them that the obtained fat (ranging between 150 and 300 cc, an average of 235 cc) was injected into and under the breast parenchyma and the pectoral muscle.

We reviewed conventional and digital bilateral mammograms with craniocaudal and mediolateral oblique (MLO) views of 20 patients ranging between 31 and 46 years of age, with an average of 36.9 years of age. None of them had a family history of breast cancer. Only three patients who were older than age 40 at the time of the breast augmentation had undergone preoperative mammography, with results reported as normal. The elapsed time between surgery and postoperative mammography ranged from 6 months to 7 years (average, 34.5 months). All of the patients were symptom free at the time of their mammography (Table 1).

Table 1

Patient data: age, time of postoperative mammogram, mammographic findings, and BI-RADS classification

No.	Age	PO Time (mos)	Volume of fat (cc) injected (in each breast)	Mammographic findings	BI-RADS category	Preoperative mammography result
1	45	6	250	Dense fibroglandular parenchyma.	II	Normal
				Bilateral heterogeneity of muscle density, presenting low-density strips probably caused by fat injection.
				Right axillary lymph node.
2	35	6	250	Mixed fibroglandular parenchyma.	II	No
				Bilateral heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection.
3	46	11	200	Mixed fibroglandular parenchyma.	II	Normal
				Punctate microcalcification in the right breast.
				Left intramammary lymph node.
				Right axillary lymph node.
4	36	13	300	Mixed fibroglandular parenchyma.	II	No
				Bilateral punctate microcalcifications.
				Bilateral axillary lymph nodes.
5	33	15	250	Mixed fibroglandular parenchyma.	II	No
				Bilateral scattered calcified and uncalcified lipid cysts of varied sizes.
6	34	15	250	Mixed fibroglandular parenchyma.	II	No
				Bilateral scattered calcified and uncalcified lipid cysts of varied sizes.
				Right axillary lymph node.
				Bilateral heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection.
7	37	16	150 cc	Mixed fibroglandular parenchyma.	II	No
				Calcified lipid cyst on the left breast.
				Bilateral heterogeneity of muscle density, presenting low-density strips probably caused by fat injection.
				Right axillary lymph node.
8	35	16	200 cc	Mixed fibroglandular parenchyma.	II	No
				Bilateral lipid cysts.
				Bilateral heterogeneity of muscle density, presenting low-density strips probably caused by fat injection.
				Bilateral axillary lymph node.
9	46	19	200 cc	Mixed fibroglandular parenchyma.	II	Normal
				Bilateral scattered spherical microcalcifications.
				Bilateral heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection.
				Bilateral axillary lymph nodes.
10	35	19	250 cc	Mixed fibroglandular parenchyma.	II	No
				Bilateral scattered spherical microcalcifications.
				Bilateral lipid cysts.
				Bilateral heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection.
				Right axillary lymph node.
11	39	19	300 cc	Mixed fibroglandular parenchyma.	III	No
				Bilateral scattered dystrophic calcifications.
				One clustered microcalcification in the lower inner quadrant and 2 clustered microcalcifications in the upper outer quadrant of the left breast.
				Bilateral heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection.
				Right axillary lymph node.
12	36	25	250	Mixed fibroglandular parenchyma.	II	No
				Bilateral scattered dystrophic calcifications.
				Bilateral heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection.
				Left axillary lymph node.
13	39	41	200	Mixed fibroglandular parenchyma with asymmetry in the upper right breast without suspicious findings.	III	No
				Bilateral scattered round microcalcifications.
				One clustered microcalcification in the upper outer quadrant of the left breast.
				Bilateral heterogeneity of muscle density, presenting low-density strips probably caused by fat injection.
				Right axillary lymph node.
14	35	48	250	Mixed fibroglandular parenchyma.	II	No
				Bilateral scattered round microcalcifications.
				Bilateral heterogeneity of muscle density, presenting low-density areas probably caused by fat injection.
				Bilateral axillary lymph nodes.
15	35	58	300	Mixed fibroglandular parenchyma.	II	No
				Bilateral scattered spherical microcalcifications.
				Bilateral lipid cysts.
				Heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection.
16	31	62	200	Mixed fibroglandular parenchyma.	III	No
				Bilateral scattered spherical microcalcifications.
				Clustered microcalcifications of the left breast in the upper outer quadrant.
				Small lipid cysts in the right breast.
				Lipid cyst in the left breast.
				Heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection
17	39	67	200	Dense fibroglandular parenchyma.	II	No
				Bilateral heterogeneity of muscle density, presenting low-density strips probably caused by fat injection.
18	33	73	300	Mixed fibroglandular parenchyma.	II	No
				Bilateral scattered round microcalcifications.
				Bilateral heterogeneity of muscle density, presenting low-density strips probably caused by fat injection.
				Bilateral lymph nodes.
19	34	77	200	Mixed fibroglandular parenchyma.	II	No
				Bilateral scattered round microcalcifications.
				Bilateral partially calcified lipid cysts.
				Heterogeneity of muscle density, presenting low-density strips probably caused by fat injection.
				Bilateral axillary lymph nodes.
20	35	83	200	Mixed fibroglandular parenchyma.	II	No
				Bilateral scattered round microcalcifications.
				Left axillary lymph node.

The mammographic findings of fat necrosis in our patients were divided into 6 categories: 1, radiolucent oil cysts; 2, microcalcifications; 3, coarse calcifications; 4, focal masses; 5, spiculated areas of increased opacity; and 6, negative (no abnormality seen)¹⁵ (Table 2). The Breast Imaging Reporting and Data System (BI-RADS) was used to classify the lesions: 0, “incomplete”: Additional imaging or comparison with outside films is required; I, “negative”: routine screening is needed; II, “benign finding”: nothing further is required other than routine screening; III, “probably benign findings”: for which short-interval mammographic follow-up is suggested to observe the stability of the lesion; IV, “suspicious finding”: for which biopsy is recommended; V, “highly suggestive of malignancy”: requiring biopsy.⁴,⁷,¹²,¹⁸,³⁴–³⁶ The mammographic findings were analyzed in terms of, bilaterality, size, and morphological appearance, such as shape of the lesion, margin characteristics, radiographic density, and location (Tables 1 to 3).

Table 2

Mammographic features of patients' lesions

Radiolucent oil cysts	4
Microcalcifications	9
Radiolucent oil cyst + microcalcifications	4
Coarse calcifications	0
Focal masses	0
Spiculated areas of increased opacity	0
Axillary lymph nodes	14
Axillary lymph node + Intramammary lymph node	1
Heterogeneity of pectoral muscle density	16
Negative (no abnormality seen)	0

Table 3

Patients with different types of microcalcifications

Round	5
Spherical	4
Punctuate	2
Dystrophic	2
Coarse	0
Cluster	3

Results

All of the postoperative mammograms were classified as BI-RADS II (85%) and III (15%). The predominant mammographic findings in the breast parenchyma were as follow. Nine mammograms (45%) showed microcalcifications with a wide variety of appearances, including round, spherical, punctate, dystrophic, and cluster forms (Figure 1). Four of the 20 patients (20%) had typical oil cysts, defined as lucent lesions surrounded by smooth rims that may or may not be calcified, and were considered pathognomonic of breast fat necrosis (Figure 2). Four patients (20%) had lipid cysts associated with round, spherical, and cluster microcalcifications (Figure 3). Sixteen patients (80%) had axillary lymph nodes; in one of these cases an intramammary lymph node was also noted.

Figure 1

Close-up digital view of left breast of patient 13 shows scattered calcifications (arrows and circle) with a wide variety of appearances.

Figure 2

Close-up MLO view of patient 6 shows cystic lesions with calcified capsules (arrows) of varied sizes.

Figure 3

Close-up of patient 16 shows oil cyst (arrow) and microcalcifications (circle).

Normally, the pectoral muscle represents a predominant area that is homogeneous in density in MLO views of mammograms.³⁷ However, the images of 16 patients in the study (80%) showed findings that are worth highlighting because the pectoral muscle appeared heterogeneous in density and presented low-density strips, with some of them showing microcalcifications probably caused by fat injection (Figure 4). This finding was observed in mammograms throughout our study.

Figure 4

Close-up of digital mammogram of patient 16 shows where the pectoral muscle appears heterogeneous in density (circled area) and presents low-density strips with some microcalcifications, probably caused by fat injection. Arrows indicate small lipid cysts.

The most frequent mammographic findings in the breast parenchyma were bilateral scattered microcalcifications, followed by radiolucent oil cysts by themselves or associated with microcalcifications. Three of the patients with microcalcifications showed cluster patterns. None of the patients in the study group demonstrated any other type of mammographic lesion related to any other illness.

Microcalcifications were found on mammography as early as 11 months into the study period; nevertheless, the main finding within the first year after surgery was the heterogeneity in pectoral muscle density (Figure 5). Patients who underwent mammography during the second postoperative year showed typical oil cysts, sometimes associated with microcalcifications as the main finding (Figure 6). The most significant mammographic findings after a 24-month postoperative period were diffuse microcalcifications in the breast tissue, which looked similar to lightly falling snow, sometimes associated with oil cysts (Figure 7). Three of the patients with microcalcifications showed cluster patterns and were classified as BI-RADS III. We were able to perform digital follow-up on two of these patients 6 months later and the other patient 1 year later (Table 4). These digital mammograms showed multiple bilateral calcifications related to fat necrosis, which had persisted, bilateral lipid cysts, and the clusters in the upper outer quadrant of the left breast evolved into non-suspicious small, round, and more regularly shaped calcifications. The patients were reclassified as BI-RADS II (Figure 8).

Table 4

Mammographic follow-up In Bi-RADS II I patients

No.	Age of patient	PO time (mos)	New mammographic findings	Previous mammographic findings	BI-RADS category
16	31	68	Small lipid cysts mainly in the lower right breast. The clustered microcalcifications in the upper outer quadrant of the left breast have changed. Now, these calcifications are round homogeneous, some of them with lucent center which is characteristic of benign lesions.	Mixed fibroglandular parenchyma. Bilateral scattered spherical microcalcifications. Clustered microcalcifications of the left breast in the upper outer quadrant. Small lipid cysts in the right breast. Lipid cyst in the left breast. Heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection	Re-classified BI-RADS I I
13	39	47	Mixed fibroglandular parenchyma with slight asymmetry in the junction between the upper quadrants of the right breast, without suspicious findings. Bilateral scattered spherical calcifications related to breast fat injection. The clustered microcalcifications in the upper outer quadrant of the left breast were stable without changes that suggest malignancy. Bilateral heterogeneity of muscle density, presenting low-density strips probably caused by fat injection.	Mixed fibroglandular parenchyma with asymmetry in the upper right breast without suspicious findings. Bilateral scattered round microcalcifications. One clustered microcalcification in the upper outer quadrant of the left breast. Bilateral heterogeneity of muscle density, presenting low-density strips probably caused by fat injection. Right axillary lymph node.	Re-classified BI-RADS I I
11	40	30	Bilateral retropectoral breast implants. Bilateral scattered spherical calcifications related to breast fat injection. The clustered microcalcifications evolved into round homogeneous calcifications with no changes that suggest malignancy.	Mixed fibroglandular parenchyma. Bilateral scattered dystrophic calcifications. One clustered microcalcification in the lower inner quadrant and 2 clustered microcalcifications in the upper outer quadrant of the left breast. Bilateral heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection. Right axillary lymph node.	Re-classified BI-RADS I I

Figure 5

A, B, The main finding within the first year after surgery was heterogeneity in pectoral muscle density, showing low-density strips as the arrows indicate in this digital mammogram of patient 2.

Figure 6

During the second postoperative year, patients showed typical oil cysts sometimes associated with microcalcifications as the main finding, seen in this MLO view of patient 5; arrows show lipid cysts of varied sizes.

Figure 7

MLO view of patient 14 shows diffuse microcalcifications in the breast tissue that look like lightly falling snow.

Figure 8

A, MLO and B, close-up views of patient 16. Mixed fibroglandular parenchyma, bilateral scattered spherical microcalcifications, and clustered microcalcifications (circled) can be seen in the left breast in the upper outer quadrant. Lipid cysts (arrows) are apparent in the breasts. Bilateral heterogeneity of pectoral muscle density, presenting low-density strips probably caused by fat injection, is also seen. C, In a digital follow-up view of the same patient, small lipid cysts are seen, mainly in the lower right breast. The clustered microcalcifications in the upper outer quadrant of the left breast have changed. Now, these calcifications are round and homogeneous, and some of them have a lucent center that is characteristic of benign lesions. D, Close-up view of calcification circled in Part A.

Discussion

Fat necrosis after breast fat injection may occur because of the mechanical trauma caused by the blunt cannula used for injection, no matter how small the cannula may be in diameter, and because the fat injected into the breasts is more prone to liponecrosis.¹,²⁵ The pathophysiology of fat necrosis explains its mammographic spectrum. It begins with an inflammatory process: necrotic fat and cellular debris are enclosed by peripherally developing fibrosis which eventually contracts into a scar. The mammographic findings depend on the degree of fibrosis. Initially, less extensive fibrosis is associated with a lipid cyst with a thin, fibrous capsule. With more extensive fibrosis, a spiculated mass may develop, which can be confused with a neoplasm. Early calcification of the fibrotic rim of a lipid cyst or collapse of a partially calcified lipid cyst may have an undefined mammographic appearance. With severe fibrosis, calcifications tend to be found and represent a relatively late finding that may be seen months or even years later.⁹,¹⁰,¹²,²¹,³⁸–⁴¹

In 1988 Bircoll⁴²,⁴³ reported that microcalcifications can occur as early as 3 months after breast lipoinjection in 1.4% of the patients. In our research, the earliest microcalcifications that we observed on the mammograms were 11 months after breast lipoinjection and were found in 65% of our patients. This was our most significant mammographic finding.

Microcalcifications are also the most important findings in the identification of carcinomas through mammography, which is the diagnostic imaging tool of choice.¹⁵,³⁶,⁴⁴ This is why a complete mammographic evaluation is critical in determining the morphology and extent of calcifications. It is important for radiologists and clinicians to be able to identify and classify typical benign breast calcifications as such in these patients, to decrease the number of unnecessary biopsies and reduce patients' anxiety.¹²,¹⁵ Because calcifications in breast parenchyma can be expected after breast fat injection, we believe that this technique for breast augmentation should not be performed on patients with a family history of breast cancer.

The patients who opt for this procedure should undergo a complete preoperative evaluation. There has been controversy about screening women younger than age 35 with mammography because they have higher breast density, which has been directly correlated with the rate of mammograms with “false-positive” results. Furthermore, mammograms for young patients are not recommended because of the risk for development of breast cancer induced by early exposure to radiation. It is also important to keep in mind that the efficacy of ultrasonography or other diagnostic tools for breast cancer screening has not yet been proven.⁴,⁶,¹²,⁴⁵ Consequently, we should consider the idea of preoperative mammography for patients over 35 as “clearance mammograms” as part of their workup.

For postoperative screening after breast augmentation with lipoinjection, we suggest that patients older than 35 who had negative preoperative mammography results undergo a second postoperative mammogram 12 months after the surgery. If a patient undergoes mammography 12 months after the procedure, it may show scars and calcifications resulting from surgery that should not be mistaken for malignancy further along in the follow-up, therefore providing a new baseline mammogram for future comparison.¹ If their mammograms have been categorized as BI-RADS I or II, the patients should continue with annual mammographic screening and clinical follow-up.

Patients younger than 35 who are symptom free after surgery should have the same clinical follow-up as the rest of the general population and wait until they are 35 to begin mammographic screening. If these patients present a palpable abnormality, ultrasound imaging rather than mammography should be performed. If a cyst or characteristically benign lesion is found through ultrasound scanning, no mammographic imaging is necessary.¹²

All patients whose mammograms show lesions that have been assessed as BI-RADS III should have short-term imaging (defined as 2 consecutive 6-month mammograms and then annual mammograms for 2 to 3 years) to establish their stability. Because there is only a 2% chance of the lesions being malignant, and abnormalities are not expected to change over the follow-up period, short-term imaging follow-up rather than immediate biopsy is acceptable and reduces patient anxiety. When no changes occur during the short-term follow up, the lesions are considered to be benign and re-classified as BI-RADS II and follow-up should continue as described for this category. If a patient shows any suspicious changes in the lesion during short-interval follow-up, a biopsy is required.⁴,¹²,⁴⁶

Patients whose mammograms have been categorized as BI-RADS IV and V should follow the Breast Imaging Reporting and Data System recommendations.⁶ With advances in imaging techniques and technology, the possibility of digital mammographic follow-up should be considered for young patients (under 35) and patients who have been classified as BI-RADS III. Research has suggested that digital mammography can reduce of the amount of radiation that a patient is exposed to by as much as 50% compared with conventional mammography.⁴–⁶,¹²

It is sometimes difficult to obtain a high-quality view with conventional mammography. Digital imaging is more sensitive to abnormalities and allows more flexible mapping and display of radiographic densities in a dense breast. Additionally, digital mammography allows the image to be manipulated to magnify the morphologic features of individual calcifications and to change the contrast to ensure the best image. It also makes it easier to identify and detect additional calcifications in a given cluster and other unsuspected clusters of calcifications, which are usually difficult to see with conventional screening methods.⁵,¹⁵,⁴⁷

Conclusion

An understanding of the appearance and evolution of specific mammographic findings helps radiologists and other specialists in follow-up of patients who have undergone breast lipoinjection, minimizing unnecessary biopsies and additional examinations while avoiding delays in diagnosing breast cancer and alleviating patient anxiety. Given that calcifications in breast parenchyma can be expected after breast fat injection and that this procedure has not been shown to modify the incidence of breast cancer, we believe that when a mammographic finding has been categorized as BI-RADS III, this finding has a high probability of being correlated with fat injection. Nevertheless, these patients will probably undergo additional mammography during their follow-up, and it is important to keep digital mammography in mind. A biopsy may be required in some cases to clear the diagnosis. In any case, additional clinical and radiological studies in a larger series of patients will be necessary to gain more knowledge on this topic.

Acknowledgement

The authors would like to thank Dr. Pablo Patiño for his assistance in the investigation and Donna Gerrish for her help in translating this manuscript.

Disclosures

The authors have no disclosures with respect to the contents of this article.