Breast cancer cryoablation with radiologic–pathologic correlation

Abstract

Breast cryoablation has become a viable option for the treatment of breast cancer in properly selected patient populations. With the increase in the use of cryoablation, post-treatment follow-up imaging and pathology has also gained increased importance. By using the proper imaging combination of diagnostic mammography, ultrasound and MRI, physicians are able to detect residual or recurrent malignancy and differentiate it from expected post-treatment findings. If suspicious imaging findings are seen, prompt biopsy and pathological diagnosis are essential. The pathologist must also be able to differentiate the expected post-procedural histological findings from those of recurrent or residual malignancy. These imaging and pathological findings must also be compared in order to ensure concordance and appropriate patient treatment.

Introduction

Percutaneous ablative modalities have been utilized as a medical treatment since the early 1990s and over the last decade have become a mainstay in the treatment of numerous benign and malignant lesions. ¹ The range of these lesions includes osseous lesions such as osteoid osteomas, solid organ lesions such as hepatocellular carcinoma or renal carcinoma, neurological diseases and recently, the techniques have been used with a variety of breast lesions. ² Percutaneous ablative techniques gained favor in the medical world due to the minimally invasive nature of the treatments, often lacking the need for full sedation, alternative options for poor surgical candidates, shorter recovery times and fewer cosmetic complaints. ²

There are several modalities available for use for ablation including radiofrequency ablation (RFA), cryoablation, microwave ablation, interstitial laser therapy, high-intensity focused ultrasound (HIFU), and irreversible electroporation (IRE). All of these different modalities have been used to treat a variety of lesions, yet cryoablation has been the most studied one in the treatment of breast lesions. Cryoablation utilizes percutaneous probes to introduce extreme cold to the target lesion to destroy the tissue. Several trials using cryoablation over the past decade have shown promising results in the treatment of fibroadenomas and early stage cancers, with clinical trials still ongoing. ³

The American College of Surgeons Oncology Group Z1072 Phase II treat and resect trial was undertaken to explore the effectiveness of cryoablation for treatment of breast cancer using pathologic examination of resected lesions after ablation. The trial demonstrated 92% success in ablation of invasive breast cancer (IBC) or ductal carcinoma in situ (DCIS) on pathological examination in lesions without multifocal disease and less than 2 cm in size. It also showed 100% success in lesions less than 1 cm in size. ⁴ The ongoing FROST (NCT01992250) and ICE3 (NCT02200705) trials investigate the role of cryoablation as a treatment for breast cancer instead of surgery in females 50 years or older with lesions less than 1.5 cm in diameter, with the 5 year local recurrence rates as the primary end point. ^5,6 3-year interim analysis of the ICE3 trial suggests cryoablation is a safe and effective treatment of low risk early stage breast cancer with local recurrence rates similar to those of the surgical standard. ⁶ Several other studies have focused on the use of multiple probes for treatment of larger lesions and the treatment of metastatic disease. One such study conducted by Littrup et al used multiple cryoablation probes to treat larger tumors, with a mean tumor diameter of 1.7 cm (range 0.5–5.8 cm) and an average of 3.1 probes producing 100% negative biopsy margins and no local recurrence at 18 months. ⁷ Pusceddu et al have also reported successful treatment of metastatic breast disease using CT cryoablation with complete tumor necrosis in 100% of cases at 6 month follow-up with minor side-effects. ⁸

With the growing success of cryoablation in the treatment of breast cancer, follow-up protocols have become increasingly important. Similar to surgical treatment, it is important to screen patient’s for recurrence or residual cancers. The preferred imaging modality for follow-up of cryoablation is the same as with surgery, including mammography, ultrasound and MRI. Recently, the role of positron emission tomography (PET)/CT has also been investigated in post-treatment follow-up. If a suspicious finding is found on any of these imaging modalities, prompt tissue biopsy should be obtained in order to obtain a pathological diagnosis. This combination of radiological findings and pathological diagnosis is essential in differentiating between the expected post-treatment changes within the breast tissue and residual or recurrent cancer. It is also important that the radiological and pathological findings are concordant to prevent malignant lesions from being missed and to assure additional biopsies are performed where necessary.

Implementing breast cryoablation in practice

With the success of the studies, breast cryoablation is becoming more prevalent as a viable option to treat early stage breast cancer and symptomatic fibroadenomas within the clinical setting. As such, multiple institutions are undergoing steps to implement the procedure into their practice. Considerations for implementing breast cryoablation include implementing strategic and operational plans, obtaining the proper equipment, creating the proper infrastructure and billing practices within the practice to perform cryoablation, overseeing the training and proper licensing of the practicing proceduralists, obtaining a referral base of physicians and collaborating with these physicians to select patients that would benefit from the procedure. ⁹ Due to the relatively recent adoption of cryoablation as a possible treatment for breast lesions, there may be a degree of institutional bias to overcome before the service is fully accepted into the line of treatment within the facility. Due to this hurdle, it is important to establish a well-organized strategic and operational plan that focuses on educating referring physician’s, hospital staff and patients on the procedure and its benefits as well as focusing on a team-based strategy that incorporates surgical, oncological and medicine services to get all stakeholders involved. ⁹

Once the initial plan for the program is established and a treatment team is created, facility infrastructure needs to be developed and equipment obtained to perform the procedure. There are several cryoablation devices available in today’s market, however, they all work based on one of two basic designs, which utilize the Joule-Thomson effect to cryoablate lesions. One is a closed loop circulation using liquid nitrogen which rapidly creates large ice balls when compared to the other system but is less customizable. ² The other system utilizes the Joule-Thomson effect caused by rapid decompression of argon gas and is more modifiable than the liquid nitrogen-based system since it can accommodate multiple cryoablation needles allowing for an overall greater size of the ice ball and variance in shape. ² Both units are similar in cost with argon devices being relatively more expensive in terms of the cost of the argon gas used to operate the ablation device. In terms of establishing billing for the procedure, the American Medical Association has already created CPT codes for both the ultrasound-guided cryoablation of fibroadenomas (Category I CPT code 19105) and breast cancer (Category III CPT code 0581T).

Training to perform breast cryoablation requires a strong skill set in ultrasound-guided procedures and breast imaging. In order to be successful in cryoablation, perhaps the most important step is precise placement of the ablation needle in the lesion, therefore the proceduralist must be well versed in breast ultrasound findings, ultrasound-guided breast procedures and experience with cryoablation techniques. As cryoablation is not often considered a core competency within residency, outside training is often necessary to obtain privileges to perform the procedure. ⁹ It is also important for the proceduralist as well as the medical staff to be versed on the pre- and post-procedural care of the patient and follow-up imaging requirements.

Performing breast cryoablation

Ultrasound-guided breast cryoablation involves techniques similar to ultrasound-guided core biopsy and typically takes less than 45 min to perform in an outpatient setting. Although there are CT and MRI compatible cryoablation devices, ultrasound is the most common imaging modality for intraprocedural monitoring. Under ultrasound guidance, the cryoablation needle is advanced into the targeted lesion with the needle remaining parallel to the long axis of the lesion.

After the cryoablation needle is properly placed within the lesion, saline hydrodissection using a spinal needle is often performed superficial to the lesion in order to protect the overlying skin from the cold temperature of the treatment while trying to maintain a distance of 2 mm between the skin and the growing ice ball. ¹⁰ Other protective techniques include warm pads on the skin, periodically lifting the cryoablation needle and utilizing deep saline hydrodissection with a Yueh catheter to prevent damage to the underlying pectoralis muscles. Decision to use a protective technique such as hydrodissection with a Yueh catheter is made before initiating the cryoablation, as during the treatment, the shadowing from the ice ball will obstruct visualization of the catheter on ultrasound. Therefore, proper placement of the catheter tip before ablation will allow the operator to intermittently inject saline during the procedure and be confident in the protection of the pectoralis muscle. Additionally, CT fluoroscopy can serve as an adjunct imaging modality for intraprocedural monitoring in order to directly visualize tissues deep to the ice ball that would otherwise be obscured on ultrasound.

A normal cryoablation treatment proceeds through several steps of freezing and thawing while maintaining positioning of the needle in the lesion. The initial freeze lasts approximately 10–20 s and helps ensure proper ice ball location relative to the lesion, as well as freezing the needle in place after confirming proper placement. ¹⁰ Although each cycle of the treatment is determined based on the size and type of lesion, it typically follows a pattern of freeze-passive thaw-freeze. When treating a 1 cm fibroadenoma, a treatment cycle of 2–2–2 min would suffice given the lack of a requirement for an ablative margin as with malignant lesions. ¹¹ For malignant lesions less than 1 cm, the pattern typically is 6–10–6 min respectively and for lesions up to 2 cm the pattern is usually 8–10–8 min. ¹⁰ During the freezing phase of the cryoablation, an echogenic ice ball is formed with the goal to create an increased ablative margin with the ice ball visible at least 1 cm past the tumor margin. ¹⁰ If multiple cryoprobes are used to treat larger lesions, the same steps are taken to ensure a greater than 1 cm ablative margin. ⁷ After the second freeze, an active thawing is performed either by electric heating of the needle or the use of helium gas through the needle. After the procedure, serosanguineous drainage from the site is likely to occur for 24–48 h with mild overlying skin bruising and palpable lump that will decrease over time.

Cryoablation mechanism of action

The mechanism of action of cryoablation is considered to be multifactorial in nature. The initial mechanism of action is thought to be secondary to coagulative necrosis of the cells closest to the cryoablation needle. ¹² This coagulative necrosis is thought to take place due to changes in osmotic gradients during the freezing and thawing phases, resulting in rupture of the targeted cells. ³ Furthermore, it is believed that ice crystal formation within cells, oxidative stress and surrounding endothelial dysfunction resulting in tumor cells ischemia also play a role in the coagulative necrosis process, as shown in studies investigating cryoablation of other neoplasms throughout the body including hepatic, renal and lung malignancies. ¹³ The secondary mechanism of action of cryoablation affects cells that are not exposed to lethal levels of hypothermia at the periphery of the cryoablation zone. This mechanism of action is thought to be due to damage to mitochondria secondary to the cold temperatures and resulting in delayed apoptosis-mediated cell death. ¹²

Recently, studies have explored the effect of coagulative necrosis on immune response within the body both locally and systemically. Coagulative necrosis of tumor cells releases cytokines and tumor antigens into the bloodstream which in turn recruit antigen-presenting cells (APCs) to the tumor and enhance the presentation of these tumor antigens to T cells to create an adaptive immune response. ¹⁴ It has been demonstrated in studies that cryoablation has increased tumor-specific T cells, increased systemic anti-tumor immune response and improved overall survival. ³ A recent trial by Kahn et al concluded that cryoablation induced a robust tumor-specific tumor-infiltrating lymphocytes (TILs) response compared with resection in the treatment of triple-negative breast cancer in animal models, suggesting an abscopal effect leading to the prevention of cancer recurrence and metastasis. ¹⁵ TIL scoring was also found to provide a viable biomarker for the further study of the abscopal effect of local therapies such as cryoablation, which may inform future human trials. ¹⁵ Furthermore, it has been shown that cryoablation creates a stronger adaptive immune response than other temperature-based modalities such as microwave ablation or RFA when treating other neoplasms within the body. ¹³

Given the increase of T-cell-mediated immune response secondary cryoablation, combination treatments with checkpoint inhibitors such as, Ipilimumab (anti–CTLA-4) and antibodies against programmed death-1/ligand-1 (PD-1/PD-L1), have been investigated to further increase this adaptive immune response. Checkpoint inhibitors release T-cell inhibition thereby promoting effector T-cell activation and proliferation creating greater destruction of tumor cells. ¹⁶ McArthur et al have recently conducted a pilot study comparing the use of Ipilimumab (anti–CTLA-4) and cryoablation in the treatment of early-stage breast cancers alone and in combination, with favorable intratumoral and systemic immune responses seen most predominantly with the combination treatment. This opens the possibility to not only treat breast cancer locally but possibly also affect distal metastatic disease due to the synergistic immune response created by the combination therapy. ¹⁶

Preferred candidates for cryoablation

Breast cryoablation is typically used for the treatment of invasive ductal carcinomas (IDCs) measuring less than 2 cm. However, trials have been undertaken in the treatment of other lesions. Other lesions such as DCIS, IDC with extensive intraductal component or invasive lobar carcinoma (ILC) are challenging to visualize and estimate extent of disease on ultrasound thus making treatment more difficult. ¹⁰ The ideal lesion for cryoablation treatment as a low-grade IDC less than 1.5 cm, hormone receptor-positive and HER2-negative, and located at least 5 mm from the skin and easily visualized on ultrasound imaging. ⁵

Imaging findings post-cryoablation

Imaging is often performed at 6-month intervals for at least 2–3 years following cryoablation and annually thereafter. An optional earlier 1- to 3 month post-cryoablation imaging follow-up can be performed to confirm appropriate targeting if there was any doubt regarding optimal targeting during the procedure.

Mammography

Immediately after cryoablation, mammography demonstrates regional edema, due to both expected ablation change as well as saline used for hydrodissection for skin protection. A well-defined ablation zone is demonstrated by 1 month after cryoablation, recapitulating the shape of the ice ball at the time of the procedure. This ablation zone has a radiodense rim with internal lucency. Within the first 6 months after the procedure, a residual mass of the ablated tumor is often still seen. However, after 12 months, any residual mass has typically resolved, being replaced by fat-density. The biopsy clip placed at the time of tissue diagnosis, before cryoablation, serves as an important anatomic landmark, helping to confirm appropriate targeting of cryoablation during mammographic follow-up. The ablation zone progressively decreases in size and becomes less conspicuous over time (Figure 1). Suspicious findings on mammography include a new asymmetry, focal asymmetry, mass, architectural distortion, or suspicious calcifications located within the region of the ablation, including within the ablation zone, at the margin of the ablation zone, or immediately outside of the ablation zone. ⁶ If suspicious findings are encountered on mammography, further work-up is warranted with ultrasound. Percutaneous biopsy would then be performed using the imaging modality best demonstrating the suspicious finding.

Figure 1.

73-year-old female with biopsy-proven IDC, Nottingham Grade II of III, ER/PR+ HER2-, demonstrating TP imaging at 12 months status post-cryoablation, consistent with tumor recurrence at the margin of the ablation zone. Paired images from mammography, axial subtracted contrast-enhanced MRI, and hematoxylin and eosin stained core specimen from percutaneous biopsy are provided before ablation, 6 months post-ablation, and 12 months post-ablation. 1A: 1: Pre-ablation. Mammography and MRI demonstrate the biopsy clip (yellow arrows) within the center of the mass. 2: High powered hematoxylin and eosin staining demonstrates invasive ductal carcinoma, Nottingham Grade II of III from initial percutaneous biopsy. 3: 6-month post-ablation. Mammography demonstrates the clip well centered within the ablation zone with near-complete resolution of the mammographically visible mass, an expected finding. On MRI, there is no suspicious enhancement (red arrow) within the ablation zone, however, in retrospect, there is asymmetric thickening at the posterolateral margin of the ablation zone. 4: 12-month post-ablation. Mammography and MRI demonstrate a new mass at the lateral margin of the ablation zone (blue arrow). 1B: Expected post-ablative histological changes seen status post ultrasound-guided 9-gauge core biopsy of the ablation zone at 6 months post-ablation. 1. Medium powered hematoxylin and eosin staining shows necrotic debris initially after ablation therapy. 2. Medium powered hematoxylin and eosin staining shows fat necrosis, histocytes, chronic inflammation, and giant cells 6 months post-ablation. 3. Medium powered hematoxylin and eosin staining shows fat necrosis and histocytes 6 months post ablation. 4. Medium powered hematoxylin and eosin staining shows fibrosis, myofibroblastic proliferation, histocytes, and giant cells 12 months post-ablation. 1C: Status post-MRI-guided 9-gauge core biopsy of the mass at the margin of the ablation zone 12 months after cryoablation. Low powered hematoxylin and eosin staining of the core biopsy specimen shows post-treatment fibrosis (blue arrows), post-treatment chronic inflammation (yellow arrow), post-treatment fat necrosis (black arrow), and recurrent invasive carcinoma (white arrow). 1D: Status post-MRI-guided 9-gauge core biopsy of the mass at the margin of the ablation zone 12 months after cryoablation. Medium powered hematoxylin and eosin staining shows post-treatment fibrosis and fibroblastic reaction (blue arrow) and recurrent invasive carcinoma (yellow arrow). 1E: Status post-MRI-guided 9-gauge core biopsy of the mass at the margin of the ablation zone 12 months after cryoablation. High powered hematoxylin and eosin staining shows viable invasive carcinoma with mitotic figures (blue arrows). This malignant pathology is concordant with the imaging findings. 1F: Status post-surgical excision of the biopsy-proven recurrent cancer 12 months after cryoablation. Hematoxylin and eosin staining shows invasive carcinoma at the margin of the ablation zone, better demonstrating radiologic–pathologic correlation with 12 month follow-up mammography and MRI (1A). IDC, invasive ductal carcinoma; TP, true positive.

Ultrasound

Post-procedural ultrasound findings follow those seen on mammography. Immediately after the ice melts on the day of the procedure, there is regional edema and there may be an imperceptible echogenic boundary, which corresponds to expected post-ablation coagulative necrosis and inflammation. ¹⁰ As in mammography, a sonographic ablation zone becomes well-defined approximately 1 month after cryoablation, appearing as a mass of mixed echogenicity. Within the first several months, the residual mass of the ablated tumor often remains visualized within this ablation zone, but after 6–12 months, that residual mass is no longer seen, becoming indistinguishable from the increasingly heterogeneous ablation zone. ² A persistent or new mass within, at the margin, or immediately outside of the ablation zone would be considered suspicious and should prompt biopsy.

Magnetic resonance imaging

MRI of the breast is an important complementary imaging modality for evaluating cryoablation success. Because viable tumor tends to enhance more avidly and rapidly than normal breast tissue, subtracted post-contrast MRI sequences provide both anatomic and physiologic information, whereas mammography and ultrasound provide only anatomic detail. 6 months after cryoablation, MRI will demonstrate the ablation zone with resolution of tumoral enhancement seen prior to the ablation procedure. There may be absent or minimal rim enhancement related to post-ablation inflammatory changes and fat necrosis. With subsequent imaging follow-up, the ablation zone will decrease in size and any mild enhancement will resolve. Likely, the best predictor of complete tumor ablation is the continued lack of enhancement within the ablation zone on follow-up MRI. ² It is important to note that false positives can occur due to enhancing fat necrosis. While fat necrosis tends to present as uniform rim enhancement or mild indistinct enhancement within the ablation zone, it can also present focally as an enhancing mass, which is suspicious, warranting biopsy. ¹⁷ It has been shown that MRI has a negative predictive value of 81–83% for recurrence on post-treatment imaging. ⁴ Due to the relatively low negative-predictive value, it is important to combine ultrasound and mammography with MRI to ensure maximum detection. Suspicious findings for recurrence on MRI include an enhancing mass in the ablation zone, new areas of nonmass enhancement, or nodular enhancement along the periphery of the ablation zone as seen on ]. ¹⁸ Suspicious imaging findings should prompt biopsy.

Positron emission tomography/CT (PET/CT)

PET/CT is another imaging modality that combines anatomic detail through CT cross-sectional imaging with functional imaging through PET. The most common PET radiopharmaceutical is ¹⁸F-fludeoxyglucose (FDG), which is a glucose analog that serves as a surrogate measure of metabolic activity. ¹⁹ Because cancers tend to be more metabolically active than the surrounding normal tissues, ^F18-FDG PET highlights these cancers on PET imaging.

PET/CT is most commonly used in breast cancer for staging and for the evaluation of response to treatment. In patients with certain forms of breast cancer, PET/CT can be used to find distant metastases or show decreased metabolic activity in lesions that are responsive to current oncological therapy. ²⁰ Currently, its role within the follow-up of postablative imaging and breast cancer are being studied. Adachi et al have recently investigated ^F18-FDG PET/CT findings in post-cryoablation treated breast cancer and have described two major findings. Initially after cryoablation, it is thought that there will be increased metabolic activity within the region of treatment due to an inflammatory reaction. After approximately 6 months to 1 year, findings on PET/CT are usually classified as either a fatty mass or a non-fatty mass in the region of ablation that correspond to fat necrosis and post-necrotizing changes such as fibrosis, granulation tissue and scarring, respectively. ¹⁹ The fatty mass often shows less metabolic activity than the non-fatty masses; however, this may be secondary to inflammatory infiltrate in the non-fatty mass as opposed to recurrence, as the study found no instances of recurrence in these lesions. PET/CT has been used successfully in the detection of metastatic disease, for which trials of cryoablation in the treatment of metastatic lesions have demonstrated positive results. ^8,19 As the investigation into the correlation between PET/CT and post-ablative recurrence continues, there are hopes that it can be used in combination with the other imaging modalities to further increase detection rates.

Another potentially promising PET radiopharmaceutical is ¹⁸F-fluestradiol (FES), which is an estrogen analog that binds to hormone receptors on ER+ breast tumors, thus highlighting a more specific functional characteristic on imaging than metabolic activity seen in ¹⁸F-FDG PET. ²¹ More research is needed in this area.

Pathologic findings

Breast cancer

IDC is the most common type of breast cancer arising from the terminal duct lobular unit (TDLU). It is thought that IDC initially begins as dDCIS and is later classified as IDC after invasion through the duct wall and into the breast stroma. Invasive ductal carcinoma differs from ILC in that it arises from the epithelial cells lining the terminal ducts. ²² IDC is further classified into several subtypes based on cell type, location of mucus secretion, architectural features and immunohistochemical profile. For the purpose of cryoablation, the most commonly treated variety is IDC no specific type (NST) which is also the most common type of IDC, constituting 40–75% of all invasive breast carcinoma. IDC NST often presents with a heterogeneous morphology and growth pattern with a variable amount of ductal differentiation. ²³ The tumor cells often show a high degree of pleomorphism with multiple areas of mitosis and prominent nucleoli with commonly seen areas of necrosis and calcifications which help with detection on mammography. ²⁴ These features of IDC NST are used to determine the grade of the carcinoma through the Nottingham combined histologic grade system. This grading system, which is recommended by the College of American Pathologists, utilizes tubule formation, nuclear pleomorphism, and mitotic activity to provide a grade of the tumor’s differentiation and is used to determine the tumor’s prognosis. ²⁵

IDC NST also has a variable immunohistochemical profile based on the expression of ER, PR, HER2/neu, cytokeratin 5/6, EGFR, and Ki-67 which correspond to the tumor’s molecular classification. The molecular classification of IBCs is divided into four subtypes: Luminal A, luminal B, HER-2 overexpression and basal like. Luminal A is the most common variety and are estrogen and progesterone receptor (ER/PR) positive and HER2/neu negative with lower grade than the other varieties. ²³ Luminal B is the next most common variety, with ER/PR positive and variable HER2/neu expression, and a higher grade than luminal A showing high Ki-67 expression. HER-2 overexpression like the name suggests has high expression of HER2/neu and is usually ER/PR negative. This variety is also considered high-grade showing high Ki-67 expression. ²³ The basal cell classification is named due to its similar pattern of expression to that of basal epithelial cells of the mammary tissue. This variety is thought of as triple negative and high-grade with an immunohistochemical profile showing CK5/6 and/or EGFR positive, ER/PR negative, HER2 negative, and a high expression of Ki-67. ²⁴ This immunohistochemical profile is then used to determine treatment and prognosis for the carcinoma.

Histology findings post-cryoablation

Post-cryoablation histological changes in breast malignancies are similar to those found with ablated fibroadenomas. Post-treatment pathological examination is expected to show fat necrosis with inflammatory changes and necrotic debris. These changes are expected to evolve over time to eventually show chronic inflammation, fibrosis, hemosiderin laden macrophages and evolving fat necrosis (Figure 2). Post-procedural biopsies performed during clinical trials on cryoablation of breast malignancy have showed several common histological findings. Biopsies obtained initially after treatment will show necrotic debris and fluid from the destroyed cells, shadow of destroyed tumor cells and avital tumor cells. ²⁶ Dilated necrotic blood vessels are often present within the edematous stroma with a lack of inflammatory cell reaction at this time. ⁴ There may also be biopsy site or cryoprobe related material introduced during the procedure seen within the ablation zone.

Figure 2.

78-year-old female with biopsy-proven invasive ductal carcinoma, Nottingham Grade I of III, ER/PR+ HER2-, demonstrating late FP imaging at 24 months status post-cryoablation, consistent with fat necrosis within the ablation zone. Contrast-enhanced breast MRI showing FP for recurrence 24 month post-cryoablation. Follow-up ultrasound-guided biopsy (not shown) demonstrated fat necrosis on histological review. Paired images from MIP subtracted contrast-enhanced MRI and hematoxylin and eosin stained core specimen from percutaneous biopsy are provided before ablation, 6 months post-ablation, and 24 months post-ablation. 2A: 1: Pre-ablation. MRI demonstrates an enhancing mass (yellow circle). 2: 6-month post-ablation. MRI demonstrates resolution of previously seen mass enhancement, consistent with successful ablation of the previously seen cancer (blue circle). 3: 24-month post-ablation. MRI demonstrates a new irregular enhancing mass at the site of the previously ablated breast cancer (yellow arrow). 2B: Status post pre-ablation ultrasound-guided core biopsy of the mass. Hematoxylin and eosin staining demonstrates invasive ductal carcinoma, Nottingham Grade I of III. 2C: Status post-ultrasound-guided core biopsy of the ablation zone 6 months after cryoablation. Hematoxylin and eosin staining demonstrates fibrosis (1) and fat necrosis (2). This benign pathology is concordant with the imaging findings. 2D: Status post-MRI-guided core biopsy of this new mass 24 months after cryoablation. Hematoxylin and eosin staining demonstrates fibrosis (1), fat necrosis (1), histiocytes (2), giant cells (2), and chronic inflammation (2). This benign pathology is concordant with the imaging findings. FP, false positive; MIP, maximal intensity projection.

After 6 months post-ablation, chronic inflammatory changes are seen with the tissues with lymphocytes and histiocytes infiltrating the treated area. At this time, you also see fat necrosis and the formation of giant cells.The earliest signs of recurrent malignant cells at the border of the ablation zone also can occur during this time frame and may need staining for tumor antigens for accurate diagnosis to be made. At 12–24 months post-ablation, you continue to see a chronic inflammatory reaction and fat necrosis, however, you also begin to the formation of fibrosis. ⁴ Staining will show fibroblasts with negative estrogen receptor staining in the fibrotic tissues. At this time, recurrence is also the easiest to visualize, with tumor cells along the ablative border and multiple mitotic figures often visualized. After several years, the ablation zone should form an area of fibrohyaline scarring without significant breast or malignant tissue remaining.

Radiologic–pathologic correlation and concordance

Imaging and pathology correlation is critical to the practice of breast imaging, as the assessment of concordance drives clinical management. The imaging and pathology findings are considered concordant when the pathology result provides an acceptable explanation for the imaging features. For example, fat necrosis is considered the great mimicker and can appear as characteristically benign or as a suspicious mass. When the imaging features are equivocal or suspicious, a biopsy should be performed to confirm the true nature of the imaging finding. An enhancing mass at the margin of the ablation zone could represent cancer or fat necrosis, and either diagnosis could be considered concordant in that scenario. In the case of fat necrosis, routine or short interval follow-up would be suggested. In the case of cancer, the patient would require further oncology management, which could include surgical excision vs repeat cryoablation. If the histologic findings were that of non-specific fibroglandular tissue and fat, on the other hand, that would be considered discordant as the non-specific histology does not adequately account for an enhancing mass on imaging. This would prompt repeat biopsy or referral for surgical excision.

Unlike surgical excision, where the specimen is removed and submitted for histologic analysis to determine whether tumor extends to the margin of the excised tissue, suggesting tumor was left behind, cryoablation relies on imaging features to determine whether the entire tumor was treated adequately. Imaging findings can be divided into different categories: true negative (TN), true positive (TP), false positive (FP), and false negative (FN). TN imaging findings are those demonstrating expected postablation change with future imaging follow-up or histology confirming the absence of residual or recurrent malignancy (Figure 1B, Figure 2C and D, Figure 3D and E, Figure 4A and B). TP imaging findings are those demonstrating suspicious imaging features with histology confirming residual or recurrent malignancy (Figure 1A and C–E). FP imaging findings are those demonstrating suspicious imaging features with future imaging follow-up and/or histology showing benign findings, such as fat necrosis that could conceivably account for the suspicious appearance on imaging, and thus, is considered benign and concordant (Figure 2, Figure 3, Figure 4). FN imaging findings are those where imaging is interpreted as normal but subsequently found to represent residual or recurrent malignancy on future imaging follow-up and histology. In several studies, tumor recurrence was found to be rare after treatment; with recurring malignancy most often demonstrating the same immunohistochemical profile as the treated malignancy. ⁶

Figure 3.

91-year-old female with biopsy-proven invasive ductal carcinoma, Nottingham Grade II of III, ER/PR + HER2-, demonstrating early false-positive imaging at 6 months status post-cryoablation, consistent with fat necrosis at the margin of the ablation zone. Paired images from mammography and hematoxylin and eosin stained core specimen from percutaneous biopsy are provided before ablation, 3 months post-ablation, and 6 months post-ablation. 3A: Pre-ablation. Mammography demonstrates a spiculated mass, before and after biopsy with satisfactory placement of a ribbon clip (yellow arrow); note mild post-biopsy changes, including a small amount of bleeding along the biopsy tract. Ultrasound-guided core biopsy of the mass (not shown) with hematoxylin and eosin staining demonstrates invasive ductal carcinoma, Nottingham Grade II of III. 3B: 1: 3-month post-ablation. Mammography demonstrates a well-centered ribbon clip within the ablation zone as well as an asymmetry at the anterior margin of the ablation zone. The asymmetry was thought to be probably benign, and therefore, short interval follow-up was recommended rather than biopsy. 2: 6-month post-ablation. Mammography demonstrates interval enlargement of the asymmetry at the anterior margin of the ablation zone, warranting tissue sampling. 3C: Tomosynthesis-guided core biopsy was performed of the asymmetry with placement of a cylinder clip (yellow arrow). 3D: Hematoxylin and eosin staining of the core specimen demonstrates fat necrosis, histiocytes and giant cells on a background of chronic inflammation and fibrosis. This benign pathology is concordant with the imaging findings. 3E: Hematoxylin and eosin staining (1-4) and ER staining (5) of the core specimen of the false- positive mammography showing expected post-ablative histological changes at 6 months post-cryoablation. 1. Fat necrosis, histocytes, and giant cells 2. Fibrosis 3. Fibrosis and chronic inflammation 4. Fat necrosis, histocytes, and giant cells 5. ER staining is negative in fibrosis (yellow arrow) and positive in benign breast lobule (black arrow). ER, estrogen receptor.

Figure 4.

74-year-old female with biopsy-proven invasive lobular carcinoma, Nottingham Grade II of III, ER/PR + HER2-, demonstrating late false-positive imaging at 48 months status post-cryoablation, consistent with fat necrosis at the margin of the ablation zone. Note, the patient had a follow-up MRI 6 months later (not shown), which demonstrated enlargement of biopsied area, subsequently undergoing repeat biopsy and again demonstrating benign fat necrosis. The following images demonstrate radiologic–pathologic correlation at various time points. 4A: Pre-ablation: MRI demonstrates an irregular enhancing mass (orange circle) on the subtracted contrast-enhanced MRI MIP image. 6-month post-ablation: MRI demonstrates minimal circumferential enhancement at the margin of the ablation zone (blue circle), consistent with expected post-ablation change. 4B: Ultrasound-guided biopsy of the initial mass (not shown) with hematoxylin and eosin staining demonstrates discohesive tumor cells infiltrating the stroma in a single file pattern (yellow arrow). The biopsy is negative for E-cadherin on immunohistochemistry (positive foci in the image are entrapped normal breast lobules [black arrow]), consistent with invasive lobular carcinoma. 4C: CT (red circle), PET (orange circle), and fused PET/CT images (not shown), obtained for evaluation of the patient’s known non-small cell lung cancer, demonstrates corresponding mild circumferential FDG-avidity in the area of the fat necrosis seen on contrast-enhanced MRI from 6-month post-ablation (blue circle). PET/CT findings are consistent with expected post-ablation change. 4D: 48-month post-ablation: PET/CT and MRI demonstrate FDG-avidity (blue circle) and focal enhancement (orange arrow), respectively, at the anterior margin of the ablation zone, which is suspicious for recurrence and warranting tissue sampling. 4E: MRI-guided core biopsy of new area of enhancement (not shown) with hematoxylin and eosin staining demonstrates with fibrosis, fat necrosis, foreign materials and giant cells, hemosiderin deposition/laden macrophages, and chronic lymphohistiocytic inflammation. The foreign material (black arrows) in this case is very impressive and surrounded by giant cells. No evidence of recurrent invasive lobular carcinoma is visualized. 4F: Lobular tumor can be very easy to miss on H&E. Therefore, immunohistochemistry for cytokeratin cocktail and estrogen receptor was performed. Any residual or recurrent tumor should be highlighted as positive and stain brown. Other than the foreign material (black arrows), all the cells in the background are light blue, indicating reactive fibroblasts and no recurrent tumor cells. FDG, fludeoxyglucose; MIP, maximal intensity projection; PET, positron emission tomography

Conclusions

Cryoablation is a minimally invasive office-based procedure that has become more commonplace in treating early-stage breast cancer. Because there is no tissue specimen to submit to pathology, as there would be after surgical management, treating breast cancer with cryoablation highlights the importance of multimodality post-procedure imaging follow-up in order to identify suspicious findings that might represent residual or recurrent malignancy, and thus, warrant biopsy and histologic correlation. Mammography, ultrasound, and MRI are complementary imaging modalities that are routinely used in breast imaging practice. For breast cryoablation: ultrasound is used for cryoablation needle positioning, intraprocedural monitoring, and is often the biopsy modality of choice if suspicious imaging findings arise during follow-up; mammography best demonstrates the relationship of the biopsy clip within the ablated tumor relative to the ablation zone, thus serving as a critical anatomic landmark to determine successful cryoablation targeting; MRI provides a physiologic assessment of the breast tissue where resolution of enhancement on follow-up MRI suggests absence of viable tumor after cryoablation. PET/CT is another imaging modality that is currently being studied in the detection of initial breast cancer as well as recurrent or residual tumor. As research into the modality continues, PET/CT may eventually become another adjunctive imaging modality used to improve detection, treatment and post-procedural follow-up. Regardless of the imaging modality utilized, any suspicious imaging finding should undergo biopsy to distinguish residual or recurrent tumor from benign post-ablation change, such as fat necrosis. Careful attention to radiologic–pathologic correlation and determination of concordance is required to ensure optimal long-term treatment and outcomes. More study is needed to determine the optimal imaging follow-up for patients undergoing cryoablation for the treatment of breast cancer.