Combined Core Needle Biopsy and Fine Needle Aspiration Cytology Using Returned Cell Block Method is Effective for Diagnosis of Cystic Lesions of the Breast

Taijiro Kosaka1, Shiho Azami2, Mizuki Iino2, Yuuji Aoki2, Kanako Ogura2, Toshiaki Kitabatake3, Kuniaki Kojima3 and Toshiharu Matsumoto2* 1Department of Breast Surgery, Sainokuni Higashiomiya Medical Center, Japan 2Department of Diagnostic Pathology, Juntendo University Nerima Hospital, 3-1-10 Takanodai, Nerima-ku, Tokyo 177-8521, Japan 3Department of Breast Surgery, Juntendo University Nerima Hospital, Tokyo, Japan


Introduction
Cystic lesions of the breast are common findings on mammography, ultrasonography, and magnetic resonance imaging. So, the evaluation of sonographic findings is important for diagnosis, treatment, and follow-up. The sonographic evaluation of such lesions has been reported based on the Breast Imaging Reporting and Data System (BI-RADS) and various types of sonographic findings of cystic lesions of the breast have been proposed [1][2][3][4][5]. An excellent review of the literature on breast cystic masses was published by Berg, et al., and they described the management of various types of cystic lesion with the adoption of aspiration cytology and biopsy [4]. Until now, the published articles concerning cystic lesions of the breast have described the necessity of aspiration cytology and biopsy in cases with suspicious of malignancy [1][2][3][4][5]. However, a detailed study comparing aspiration cytology and biopsy for cystic lesions of the breast has not been reported, and so a decision on the practical way to use both methods is needed for the diagnosis of such lesions.
The authors initially published an English article describing the returned cell block method (RCBM: cell block from a Papanicolaou staining smear on a glass slide), and also reported the usefulness of RCBM for the fine needle aspiration cytology (FNAC)-based evaluation of tumors of the breast in 2016 [6].
The present study compares the diagnostic accuracy between core needle biopsy (CNB) and FNAC with or without RCBM in patients with cystic lesions of the breast. According to the results of this examination, a practical method using CNB and FNAC with RCBM for the diagnosis of cystic lesions of the breast is determined.

Materials and Methods
Among the patients undergoing ultrasonographically-guided FNAC examination of mammary lesions from January 1, 2012 to December 31, 2013 in Juntendo University Nerima Hospital, 138 with FNAC findings of a cyst, papilloma, or papillary carcinoma were selected. All patients were examined with both ultrasonography and mammography, and so a detailed re-evaluation of the sonographic findings from both examinations was done. Subsequently, 47 patients with definite cystic lesions were selected and examined. All of the examined patients were female and their ages ranged from 30 to 87 years (average, 57 years).
The sonographic findings in the 47 patients were evaluated according to BI-RADS [5]: simple cyst ( Figure 1; 7 cases), complicated cyst ( Figure 2; 24 cases), complex cyst ( Figure 3; 15 cases), and unclassified cyst (1 case). These 4 types were compared with a benign or malignant group according to the cytological and histological diagnoses.
In all cases, the evaluation of FNAC by Papanicolaou staining smear was performed, and RCBM was added in cases that were difficult to evaluate with Papanicolaou staining smear. RCBM was described in our previous article, and it consists of making a cell block from a Papanicolaou staining smear on a glass slide, cutting sections of the block, and staining with conventional (HE stain and others) and immunocytochemistry stains [6]. According to the using of RCBM, easily recognition of morphology and application of immunocytochemistry can be made in the atypical cell clusters observed in a Papanicolaou staining smear [6]. In 6 patients undergoing repeated FNAC examinations, the findings of repeated FNAC were also evaluated.
For 26 patients, CNB examination was performed. For 15 patients, an operation, which consisted of tumorectomy (2 cases), partial resection of the breast (12 cases) and mastectomy (1 case), was performed. The data on the histological findings from CNB and surgical materials were reviewed. The pathological diagnosis by CNB and surgical materials was based on the newly published WHO classification of tumors of the breast [7]. The follow-up data on the examined cases were added to assess the benignity or malignancy of cases without histological examination. This study was approved by the Research Ethics Committee of Juntendo University Nerima Hospital. The surgeons obtained a written informed consent for the approval of cytological and histological examinations from all patients when they were examined with FNAC and CNB and they were operated. The approval of the opening of the data was not obtained from all of the patients, so personal information to pursue each of the patients was not included in the manuscript.

Results
Simple cysts according to BI-RADS were observed in the benign group only, and complicated and complex cysts were found in both groups of benignity and malignancy ( Table 1). The diagnostic accuracy increased with the addition of RCBM compared with Papanicolaou staining smear only (Table 2 and Figure 4).
Comparisons between FNAC and CNB in 11 patients diagnosed with a carcinoma based on the surgical materials are shown in Table  3. In 5 patients, a diagnosis of carcinoma was made on both CNB and FNAC from Papanicolaou staining smear only (case no. 7-11; Figure  5). In 2 patients, carcinoma was diagnosed by FNAC using RCBM, but in these cases the diagnosis of carcinoma could not be made by CNB (case no. 4 and 5; Figure 6). In 1 patient, carcinoma was diagnosed by CNB, but the diagnosis of carcinoma could not be made by FNAC, even with the use of RCBM (case no.2). A comparison between FNAC and CNB in 5 patients diagnosed with carcinoma from CNB without surgery is shown in Table 4. In 4 patients (case no. 13-16), the diagnosis of carcinoma was made based on FNAC from Papanicolaou staining smears, but in 1 patient (case no. 12; Figure 7), a diagnosis of malignancy could not be made by FNAC.  The summary of the diagnosis of carcinoma from FNAC and CNB is shown in Table 5. The diagnostic accuracy of FNAC from Papanicolaou staining smear only was 56.2%. The diagnostic accuracy increased to 75% on using RCBM, and the same accuracy was found with CNB. Moreover, the accuracy increased to 87.5% on combining CNB and FNAC using RCBM.

Discussion
The correlation between sonographic subtypes and the pathological diagnosis has been studied [1][2][3][4][5]8]. Subsequently, the simple cyst subtype is almost always benign and routine follow-up screening is recommended [8]. On the other hand, non-simple cyst subtypes are not defined as benign and if there are findings suggesting malignancy, aspiration cytology and biopsy are recommended [8]. Chang et al. examined 175 sonographic cystic lesions using aspiration cytology and biopsy, and demonstrated malignancy in 40 patients showing a complicated solid and cystic mass or cyst with a thick wall/septa or nodule [2]. Berg et al. examined 150 sonographic cystic lesions using aspiration cytology only (55 cysts) and CNB only (81 cysts) and both aspiration and CNB (14 cysts), and demonstrated malignancy in 18 cysts showing a mass with a thick wall and/or thick septum, or intracystic mass, mixed and solid, or a solid mass with eccentric cystic foci [1]. In both studies by Chang et al. and Berg et al. [1,2], a comparison between aspiration cytology and biopsy regarding the accuracy of the diagnosis of malignancy was not performed.
Daly et al. examined 243 patients with sonographic complicated cysts and reported that one of them (0.4%) was diagnosed with malignancy by aspiration cytology [3]. In 3 cases, atypical cells were      observed in aspiration cytology, but in all cases the lesions were benign on surgical excision [3]. From this study, biopsy may be superior to aspiration cytology. Consistent with this, in the present study, the accuracy rate of CNB (75%) was greater than that of FNAC using Papanicolaou staining smear only (56.2%). However, on using RCBM, the accuracy rate of FNAC increased to 75%, and 2 patients were diagnosed with carcinoma using this method, but these patients could not be diagnosed with carcinoma by CNB. On the other hand, in 2 patients, carcinoma was diagnosed by CNB only. Subsequently, the accuracy rate of diagnosing malignancy on combining CNB and FNAC using RCBM reached to 87.5%. The results of the present study indicate that a combination of CNB and FNAC using RCBM is effective for the diagnosis of cystic lesions of the breast.
Intracystic papilloma and intracystic papillary carcinoma are common neoplasms showing cystic changes among mammary tumors [9]. In the present study, such tumors were examined in detail. On the other hand, as a rare neoplasm presenting with cystic changes in a breast tumor, squamous cell carcinoma is considered [10]. Squamous cell carcinoma of the breast usually presents as a cystic lesion and the cystic cavity is lined by squamous cells with various degrees of nuclear atypia and pleomorphism [10]. The neoplastic cells infiltrate the adjacent stroma and an increase in size occurs on follow-up. Among the reported cases of squamous cell carcinoma of the breast, the diagnosis using FNAS was different: benignity or malignancy [11,12]. This difference was caused by the difficultly in the evaluation of the atypia of squamous epithelium, and so if RCBM is used for such tumors, it provides a precise evaluation of the atypical status of squamous epithelium and relationship between the atypical epithelium and stroma, leading to a precise diagnosis.

Conclusion
Based on the results of the present study, we devised the algorithm for the diagnosis by CNB and FNAC using RCBM for cystic lesions of the breast (Figure 8). In the algorithm, the division of simple cyst and non-simple cysts is initially performed according to the findings    of sonographic examinations. In a simple cyst, FNAC becomes a key point, so the presence of atypical cells in a Papanicolaou staining smear by FNAC is followed by the adoption of RCBM. On the other hand, in a non-simple cyst, both examinations of CNB and FNAC by Papanicolaou staining smear are conducted if sonographic findings indicate the possibility of malignancy. If atypical cells are present in Papanicolaou staining smear by FNAC and a diagnosis of malignancy cannot be made by FNAC and CNB, the adoption of RCBM is followed. Finally, we think that the proposed algorithm is optimal for diagnosis by CNB and FNAC for cystic lesions of the breast.