A new reporting form for breast cytology
- First Online:
- Cite this article as:
- Tsuchiya, S., Akiyama, F., Moriya, T. et al. Breast Cancer (2009) 16: 202. doi:10.1007/s12282-009-0128-6
- 75 Views
The Cytology and Core Needle Biopsy Subcommittee, organized under the Rules Committee of the Japanese Breast Cancer Society, has prepared a new form for breast cytology reports. This reporting form consists of “diagnostic categories” and “recommendations.” The “diagnostic category” is either “specimen inadequacy” or “specimen adequacy.” The judgment on “specimen adequacy” is subdivided into four categories: “normal or benign,” “indeterminate,” “suspicious for malignancy,” and “malignant.” The “recommendation” indicates descriptions of cytological features and estimated histological type of tumor (these descriptions should be as detailed as possible). On the basis of an analysis of cytological data from 3,439 cases performed before preparing this form, the subcommittee has attached the following recommended goals to this form: (1) the percentage of “specimen inadequacy” should be 10% or less of all samples, (2) the percentage of “indeterminate” samples should be 10% or less of all “specimen adequacy” cases, and (3) 90% or more of “suspicious for malignancy” cases should be diagnosed as “malignant” in a subsequent histological examination. We hope that modification of this form, if it requires revision in the future, will be evidence-based, as was the process for compiling this set of rules.
KeywordsBreastFine-needle aspiration cytologyNew reporting form
In Japan, the Papanicolaou classification (Pap classification) has long been used in breast cytology reporting forms. However, the Journal of the International Academy of Cytology has for some time now refused papers using the Pap classification. Several new forms of breast cytology reports were prepared in the 1990s, primarily by Western investigators [1–4]. It has been suggested that a new form for this kind of report in Japan should be prepared to meet these demands.
The Japanese Breast Cancer Society recently organized the Cytology and Core Needle Biopsy Subcommittee within the Rules Committee and began to prepare a new breast cytology reporting form to be included in the 15th edition of the General Rules for Clinical and Pathological Recording of Breast Cancer . Cytological data from about 3,400 cases, supplied by its, were analyzed by the 10 subcommittee members and used as evidence for preparing a new reporting form. A draft reporting form was sent to several councilors of the Society (surgeons and pathologists) to ask their opinions and thus enable problems with the draft form to be identified and corrected.
This paper presents the new form, accompanied by a summarized presentation of the analysis of the data from the approximately 3,400 cases that served as the basis for its development.
Problems with the current breast cytology reporting form and countermeasures
Breast cytology is mostly based on specimens collected by fine-needle aspiration (FNA). The role of FNA differs from that of cervical exfoliative cytology in that the former is performed on individuals known to have a mass, while the latter is performed to check for presence/absence of lesions. The Pap classification, used for the current reporting form, is not suitable for FNA, which is capable of directly identifying the disease, i.e., FNA can categorize the tumor’s histological type. This is probably the major reason that the Journal of the International Academy of Cytology now ignores the Pap classification. Therefore, before preparing a new reporting form, the subcommittee listed problems with the Pap classification as follows: (1) it does not include any conclusions on sample appropriateness, (2) it provides no clear criteria that delineate cytological categories, and (3) it does not include the concept of histological type characterization. The subcommittee discussed these three problems as well as the need to dispatch patient information from clinicians to pathologists.
The first problem, deciding on the appropriateness of samples, is crucial for precision control of cytology. The Bethesda system for uterine cancer includes clear instructions on how to arrive at a decision . The Pap classification, however, does not include any. Medical facilities that use this classification sometimes add this concept at their own discretion. For example, numerous medical facilities place inappropriate samples in class I. Another problem that needs to be resolved is to demonstrate, with evidence, the maximum tolerable percentage of inappropriate samples among all cases undergoing cytological examination.
Regarding the second problem (lack of clear criteria for judgment), when the Pap classification I–V is reviewed in detail, the term “malignancy” is used for classes II to V, but no evidence is given for the finding of “malignancy.” The new reporting form, therefore, should list lesions that fall under each class. Class III in the existing classification includes several concepts such as “dysplasia” (which often leads to the detection of cancer on follow up), “borderline lesions” (in the sense used in histology and cytology), and “difficult to distinguish” (inability to distinguish between benign and malignant lesions under the current diagnostic capacity). The meaning of these concepts differs among examiners and medical facilities. The new reporting form replaces class III with the category of “indeterminate.”
Regarding the third problem (lack of a concept of histological type estimation), we must point out that “histological type” yields useful medical information only when its result is entered in a cytological report. Diagnostic imaging for breast diseases uses mammography, ultrasonography, CT, MRI, etc. Reports of the results of diagnostic imaging include identification of the histological types of lesions. Cytodiagnosis, which distinguishes between benign and malignant lesions, without referring to estimated histological types, is not sufficiently useful for surgeons when making a general diagnosis on the basis of various preoperative examinations. To what extent does cytology allow estimation of histological types? When the data collected by the subcommittee were analyzed, estimation of the histological type of benign lesions was possible in 93% of cases with fibroadenoma, 81% of fibrocystic change cases, and 92% of intraductal papillomas. Estimation of the histological type of cancers was possible in 85% of invasive ductal carcinomas, NOS, 91% of mucinous carcinomas, 54% of lobular carcinomas, and 23% of ductal carcinomas in situ. These results indicate that estimation of histological types on the basis of cytological data is possible at an accuracy comparable to or higher than that of diagnostic imaging.
The inadequacy of medical information supplied by clinicians to pathologists has been pointed out for many years. This problem is closely related to the view held by clinicians that cytology is merely an “adjuvant means” of making a definite diagnosis. However, as stated above, FNA, which is often used for breast cytology, is highly likely to provide data that can greatly contribute to making a definite diagnosis of a lesion, since it directly punctures the mass. When clinicians request a histological examination, they usually provide background information on individual patients to the pathologists, since histological examination is viewed as a means of making a definite diagnosis. Because FNA can also be viewed as a means of making a “definite diagnosis,” pathologists should demand related medical information (including information from diagnostic imaging) from clinicians requesting cytological examination. This will contribute to improved accuracy of cytodiagnosis. The Revised General Rules for Clinical and Pathological Recording on Breast Cancer include this point with the aim of facilitating better cytodiagnosis.
Details of the new reporting form
New reporting form for breast cytology
Normal or benign
Suspicious for malignancy
2. Findings: descriptions of cytological features and estimated histological type of lesion
Analytical results according to the new reporting form
Normal or benign
Of the council members, 98% affirmed the need to adopt this new reporting form.
This category refers to cases where diagnosis is difficult due to inappropriately prepared specimens or an inadequate number of cells present. In reporting forms used in Western countries, numerical criteria (e.g., the minimum number of cell clusters acceptable as a valid sample) are sometimes used . However, even when a sample contains only a few cells, the lesion can be suspected of being invasive ductal or lobular carcinoma if the cells assume a linear pattern (abnormal architecture). Therefore, our criterion does not refer to any particular number of cells or cell clusters.
Of the 3,439 cases analyzed by the subcommittee, 371 were deemed inadequate (10.8%; Table 2). We therefore attached a recommended goal to the criterion, stating that the percentage of inappropriate cases among all cytological cases should be 10% or less. This is because the significance of submitting samples to laboratories will be reduced if the percentage of inadequate cases is too high. Inadequate samples may be associated with puncture techniques, methods used for preparing samples, and other factors that may involve clinicians and pathologists. If this percentage exceeds 10%, the cause needs to be checked without delay by both clinical and pathological staff. Since 88 (23.7%) of these inadequate cases were histologically rated as malignant, efforts are needed to reduce inadequate cases as soon as possible (Table 2).
In the questionnaire survey, 98% of respondents supported the validity of the “inadequate” and “adequate” categories, and 90% affirmed the adoption of the four subcategories of “adequate.”
Normal or benign
Of the 3,068 cases analyzed, excluding “inadequate” cases, 1,264 (41%) were rated as “normal or benign.” Although there is still room for argument about what range of cytological features can be deemed as normal, we added the rating “normal” to this category, considering that normal mammary glands around the tumor may also be punctured by the needle. In addition to normal mammary gland cells, the following histological types are included in this category: most fibroadenomas, some fibrocystic changes, intraductal papilloma, and inflammatory lesions.
Of all cases rated as normal or benign, 85 (6.7%) were eventually found to have cancer (Table 2). In these 85 cases, cancer cells were not detected by repeated microscopy, and the inability to detect cancer appeared to be associated with methods of puncture or other skills, rather than careless microscopy. Ideally, clinicians should make efforts to acquire high-level puncturing skills, and to this end, adequate opportunities should be given to pre-graduate medical students and post-graduate doctors to enhance their puncture skills. Higher-level clinicians should also improve their teaching methods.
This category refers to cases where it is difficult to determine cytologically whether a lesion is benign or malignant. This category includes papillary lesions, epithelial hyperplasia, complex fibroadenoma, etc. Histological distinction of benign lesions from malignant lesions is often difficult with these lesions. The same can be said for cytological diagnosis.
A recommended goal has been attached to this criterion, stating that the percentage of cases rated as “indeterminate” should be 10% or less among all adequate cases (3,068 cases). This percentage was determined on the basis of the data that 236 samples (7.7%) of the 3,068 adequate cases were placed in this category when analyzed by the subcommittee (Table 2). In the questionnaire survey, 61% of respondents agreed with this percentage (10%), although some respondents said that this criterion was too severe. However, since the usefulness of cytodiagnosis will be weakened if this percentage is 20–30% or more, we adopted 10%. Every laboratory should ideally make efforts to reduce this percentage to below 10%. Of the 236 cases rated by the subcommittee as “indeterminate,” 133 (56.4%) were later rated as benign and 103 (43.6%) as malignant. Thus, benign lesions predominated over malignant ones at a ratio of about 6:4. Of the 133 benign lesions, 51 were fibroadenomas, 42 were fibrocystic changes, and 13 were intraductal papillomas. These three types accounted for 80% of benign lesions. Of the 103 malignant lesions, about half were invasive ductal carcinoma (48 cases), followed by ductal carcinoma in situ (18 cases), papillary carcinoma (13 cases), and invasive lobular carcinoma (9 cases).
Suspicious for malignancy
This category was defined as primarily including ductal carcinoma in situ, invasive lobular carcinoma, and others with a few atypical features. A recommended goal was attached to this criterion, stating that the percentage of cases rated as “malignant” in subsequent histological examinations should be 90% or more of cases initially rated as “suspicious for malignancy.” This percentage is based on the evidence that 87 (92.6%) of the 94 “suspicious for malignancy” cases were later found to have malignant lesions (Table 2). In the questionnaire survey, 64% of the respondents affirmed this percentage (90%) for use with this criteria. Benign lesions that were initially rated as “suspicious for malignancy” include three cases of fibrocystic change, two cases of fibroadenoma, one case of nipple adenoma, and one case of ductal adenoma. Nipple adenoma and ductal adenoma tend to be overestimated as malignant by clinical observation, diagnostic imaging, and histological examination.
Past reports on FNA of the breast using the conventional Pap classification have sometimes combined class IV with class V into a single category for statistical analysis. In recent years, some lawsuits have been brought against surgeons who performed mastectomy on class IV cases. This appears to be an outcome of regarding class IV as a synonym of class V. This new reporting form clearly states that 10% of “suspicious for malignancy” cases may actually be benign cases and that clinicians need to be aware of this fact when determining surgical treatment.
This category refers to malignant tumors, including breast carcinoma (primary and metastatic), nonepithelial malignant tumor, etc. We cannot deny that histological diagnosis inevitably involves misdiagnosis, although its frequency is very low. Of the 1,474 cases cytologically rated as “malignant,” 5 cases (0.3%; 2 cases of invasive ductal carcinoma, 2 cases of noninvasive ductal carcinoma, and 1 case of invasive lobular carcinoma) were later found to be benign. Although none of these cases suffered unfortunate outcomes, it is the duty of our pathologists to minimize this percentage.
Needless to say, the most basic purpose of cytology reports is to report “estimated histological types” and accompanying “findings.” If a good sample is available, it is possible to estimate the histological types of both benign and malignant lesions with relatively high accuracy, as described above in the section “Problems with the current breast cytology reporting form and countermeasures.” When making a cytodiagnosis, emphasis tends to be placed on determining whether or not a given tumor is benign or malignant. However, when a histological type is estimated from cytological features, it is always accompanied by a judgment as to its benign or malignant nature. Therefore, this new reporting form emphasizes that the estimated histological type needs to be included whenever possible.
A new reporting form for breast cytology has been presented in this paper. The new form is designed to be independent of the conventional Pap classification. It is composed of the following factors: (1) appropriateness of sample (adequate or inadequate), (2) evident diagnostic criteria, (3) estimated histological type, and (4) recommended goals for the percentage of inadequate cases, the percentage of “indeterminate” cases, and the percentage of truly positive (malignant) cases among “suspicious for malignancy” cases.
FNA is a major technique for breast cytodiagnosis. Although other breast cytodiagnosis techniques, e.g., nipple discharge, are likely to involve a slightly higher percentage of inadequate samples, this reporting form would also appear to be applicable to these techniques.
The 6th edition of the General Rules for the Description of Thyroid Cancer (The Japanese Society of Thyroid Surgery) includes a new reporting form for thyroid cytology, which is quite akin to our new reporting form for breast cytology . Although the details of manipulation and the diagnostic criteria differ between these two forms corresponding to the differences in the organs and lesions concerned, the overall design and purpose are the same. We anticipate that this kind of reporting form will be adopted for other organs for which FNA is applicable.
This new reporting form was adopted in the 15th edition of the General Rules for Clinical and Pathological Recording of Breast Cancer . It is, however, likely that this new form will need revision in the future. We hope that modification of this form, if it requires revision in the future, will be evidence-based.