A 38-year-old woman actively undergoing chemotherapy postlumpectomy for (node-negative, incidentally detected) left-sided invasive ductal carcinoma had developed a painful lump in her left gluteal region. Subjectively, the mass felt firm and somewhat nonmobile. Pressure upon the area (eg, sitting, especially driving) exacerbated her discomfort. As an adjunct to chemotherapy, she was also receiving periodic Lupron (leuprorelin, GnRH agonist) injections in an effort to downregulate luteinizing hormone and follicle-stimulating hormone production and, subsequently, reduce systemic estradiol levels. To the patient’s recollection, injections were routinely administered along the right gluteal region. She denied any other type of trauma to this location.


A mammographic image of her primary malignancy reveals a focal cluster of pleomorphic calcifications embedded within the inferolateral aspect of the left breast. An ultrasound-guided core biopsy of the lesion revealed a hypoechoic mass with ill-defined, jagged margins, intermittent posterior shadowing, and internal calcifications (pathology: invasive and in situ ductal carcinoma).

A targeted ultrasound of the left gluteal mass revealed a subcutaneous hypoechoic lesion interposed between the subcutaneous fat and gluteal muscle bed. Moreover, the sonographic appearance bore a striking resemblance to the primary breast malignancy, eg, hypoechoic, ill-defined margins, internal hyperechoic foci (presumably calcifications) and some internal flow with Doppler applied. Shortly after the ultrasound was completed, a pelvic MRI was performed. Key images shown include short tau inversion recovery (STIR), T1, and fat-suppressed T1+C. Upon STIR, the 1.8-cm lesion demonstrated complex, heterogeneous signal and surrounding hyperintense reticulation. Upon precontrast T1 imaging, the lesion was isointense to skeletal muscle and markedly hypointense to surrounding subcutaneous fat. After receiving 7 cc of Gadovist, the lesion enhanced intensely and somewhat heterogeneously.

Surgical Biopsy & Diagnosis

Within 24 hours, the patient was referred to a general surgeon. An outpatient fine needle aspiration was performed. High-power microscopy revealed “fat necrosis,” associated inflammatory cells, etc. No tumor cells were present.


Fat necrosis is uncommon, but not rare. From an imaging standpoint, to state that fat necrosis can incite fear due to its myriad appearances would be an understatement. Dating back to an issue of AJR from 1978, Bassett et al published five cases of fat necrosis in the breast ranging from “suggestive” (eg, single/multiple lipid-filled cysts with or without calcified walls) to that indistinguishable from carcinoma (eg, branching, rodlike microcalcifications embedded within a focal mass). To complicate matters, the interval between a traumatic event, sensation, and/or detection of a palpable lump or mass is often drawn out; clinical history may be noncontributory or even misleading.

In this case, the overarching concern was the possibility of metastasis to the subcutaneous tissues, especially following the targeted ultrasound. At the time of imaging, the presumption that Lupron injections were exclusively given on the contralateral side was even more discouraging. At this juncture, tissue sampling was necessary. Just before the scheduled biopsy, a careful review of the patient’s chart revealed that nearly all Lupron injections had indeed been given on the right side. However, at least one had been administered on the left side approximately a month prior. Updated with this critical piece of the puzzle, the surgeon felt confident neoplasm was less likely.

Fat necrosis, as mentioned above, can have a multitude of appearances on imaging. MRI may reveal a spiculated lesion, with or without fat signal intensity within. Low T1 signal with mixed T2 signal (hyper- and hypointense) can be seen. Septations can develop, imparting a cystic or grapelike morphology. Although some have concluded that “lack of a discrete mass” is suggestive, there are reports of masslike fat necrosis that simulated malignancy. Enhancement may also be variable, generally a function of vascularized fibrous or granulation tissue. In the absence of a defining history, overlapping imaging features may be noted with primary malignant fatty lesions (eg, liposarcoma). Trickier yet is differentiating fat necrosis within a histologically benign lipoma and malignant degeneration.

Injection site fat necrosis is haphazardly reported in the literature. Cases include those that seem intuitive retrospectively, whereas others are extraordinary. Gluteal injection site granulomas are quite common, generally the result of intralipomatous delivery, as opposed to intramuscular delivery. Fat necrosis can complicate matters, albeit far less frequently, if an acute inflammatory response ensues (eg, influx of lymphocytes, foamy histiocytes). Regarding the more extreme cases, in a report published last year in Radiology Case Reports, a middle-aged woman with a lung mass and bilateral breast lesions noted upon CT underwent a follow-up PET scan for staging purposes. As it turned out, the FDG-avid breast lesions were not neoplastic but, conversely, zones of fat necrosis that developed following self-administered injections of heroin. To reiterate just how confounding imaging of fat necrosis can be, it’s important to remember that it can also demonstrate hypermetabolism on PET.

In conclusion, fat necrosis is an entity that can mimic neoplasm. History, review of systems, and physical exam can be helpful, but not always. In many cases, such as in this instance, its presentation may be a stomach-churning “red herring” for patients and their families. In order to establish the diagnosis (and rule out malignancy) with confidence, biopsy may be imperative.

This entry of “On the Case” serves to give fat necrosis the attention it deserves. Whereas radiologic imaging embodies the art of medicine, fat necrosis embodies the art of deception.

— Rahul V. Pawar, MD, Department of Radiology, Division of Neuroradiology, is Clinical Assistant Professor, UMDNJ—New Jersey Medical School.

Figure 1: Mammogram Figure 2: Breast Ultrasound
Figure 3: Gluteal Ultrasound Figure 4: Coronal STIR
Figure 5: Coronal T1 Figure 6: Axial fat-suppressed T1+C


  1. Bassett LW, Gold RH, Cove HC. Mammographic spectrum of traumatic fat necrosis: the fallibility of ‘pathognomonic’ signs of carcinoma. AJR Am J Roentgenol. 1978;130(1):119-122.
  2. Chan LP, Gee R, Keogh C, Munk PL. Imaging features of fat necrosis. AJR Am J Roentgenol. 2003;181(4):955-959.
  3. Toney LK, Lam DL, Rabhar H. Drug injection-related fat necrosis of the breast with FDG PET-CT uptake. Radiol Case Rep. 2015;10(3):12-17.

Submission Instructions

  1. Cases should have clinical relevance and clear radiological findings.
  2. Seconds should include a title, history and course of illness, findings, diagnosis, and discussion.
  3. Word count should not exceed 800. At least three references are recommended.
  4. Cases may be submitted from any radiological subspecialty and imaging modality.
  5. Figures must be high-quality JPEG or TIFF images and labeled for ease of reference. Please keep images in their native format, without the addition of arrows or other means of highlighting the key findings.

Submit cases via e-mail to Rahul V. Pawar, MD, at rvp325@gmail.com or to Radiology Today at jknaub@gvpub.com.

Section Editor: Rahul V. Pawar, MD, DABR
Department of Radiology, Division of Neuroradiology
Saint Barnabas Medical Center/Barnabas Ambulatory Care Center