Cytological Diagnostic and Prognostic Methods using Immunocytochemistry (Estrogen Receptor) for Surgical Management of Breast Cancer

The purpose of this study was to assess the utility of fine needle aspiration cytology (FNAC), immunocytochemistry(ICC) using estrogen receptor(ER) in diagnosing breast lesions. This was done by comparing it to histopathology with immunohistochemistry(IHC), which serves as the gold standard for diagnosing these lesions. To compare these modalities of investigation,50 samples were collected using FNAC and were compared to the same samples obtained by histopathology. For FNAC the results were as follows, Sensitivity=100%, Specificity=100%,Diagnostic Accuracy=100% Positive Predictive Value(PPV)=100% and Negative Predictive Value(NPV)=100% . For ICC using ER the results were as follows, Sensitivity=100%,Accuracy=100%,Positive Predictive Value=100%,Negative Predictive Value=100%. This indicates that FNAC and ICC using ER can be used as a reliable alternative to gold-standard diagnostic tests when the latter cannot be done due to a lack of resources or in circumstances where there is a need to perform a painless, minimally invasive procedure such as in inoperable breast carcinoma. This study also involved using text data analysis on FNAC reports. On analysis, it was found that the useful words were 11.35% of the data set, implying that the process of normalization, will result in the formation of condensed data, which can then be utilized for assisting clinical chart reviews and clinical decision support systems.

Breast Cancer is the most common cancer among Indian females with a mortality of 12.7 per 100,000 women 1 .
Fine needle aspiration cytology (FNAC) of breast lumps is an important component of the triple assessment of palpable breast lumps 2 . The advantages of utilizing cytological examination over traditional tissue samples are that it is a safe, simple, quick, and cost-effective diagnostic modality 3 . However, not all the diagnostic criteria are present every time and in such circumstances, a second diagnostic approach is needed 3

.
As early diagnosis is of great significance for the patient, any new diagnostic modality in addition to FNAC will enhance its accuracy. In view of this immunocytochemistry (ICC) can complement routine investigations 4 .
ICC is of great utility in such circumstances as its practical utilization includes characterization of poorly differentiated neoplasms, differentiation of primary from metastatic tumors, determination of sites of origin of metastatic lesions, and prognostic assessment 5 . An immunocytochemical assay (ICA) on estrogen receptor (ER) improves the accuracy of diagnosis and gives an admirable prognostic indication 6 . Usually, malignancy on FNAC is diagnosed by Papanicolaou-stained(Pap) smears, following which ICC using ER is done for confirmation and subsequent treatment. So, the immunocytochemical results are obtained in the later stages of patient management. However, they have a great clinical scope if reported within a short time 7 .
If ICC is used as an adjunct to FNAC, then appropriate information can be provided about the prognosis of breast cancer, helping the clinicians in selecting the right protocol for treatment, avoiding immediate surgical intervention, and essentially saving time and additional expenses 8 .
This study was done by collecting FNAC samples and subjecting some of these samples to immunocytochemistry using ER. Once this was done, we obtained the histopathology and immunohistochemistry(IHC) reports of the same samples. We then compared FNAC with histopathology and ICC with IHC to determine the diagnostic indices. Using the results of this comparison, we found that ICC on FNAC can be used as an alternative to IHC, to give reliable diagnostic information regarding breast lesions.
The most widely available form of medical data is in the form of text or literature.
There is a need to process the medical literature that has been growing rapidly to locate and extract clinically useful information. To achieve this, Natural Language Processing(NLP) and Machine Learning(ML) approaches can be used. One such huge potential for the application of NLP and ML can be seen in the interpretation of breast cancer Electronic Health Records containing text data.
We can use the information obtained by this analysis for assisting clinical chart reviews, the application of which is currently limited by the lack of information in the form of records 9 . All the subjects, who were available and satisfied the inclusion and exclusion criteria, during the given period were selected for the study and subjected to further analysis. • The extra slide used was the electronically charged slide to perform immunocytochemistry(ICC) for estrogen marker(ER).FNAC smears were prepared on the charged slides directly. The smears thus obtained were fixed in 80% alcohol immediately.

• Study
The smear was then washed and processed with the antigen retrieval technique. The slides were observed, diagnosed and a standard report according to the format was given.
• This was then compared to the histopathology(HPE) processing from the biopsy of breast lumps, following which Sensitivity, Specificity, Accuracy, Positive Predictive Value, and Negative Predictive Value were determined.
• Finally, the analysis was completed by using text data analysis.

statistical analysis
The demographic characteristic i.e., Age was represented by Arithmetic Mean and Standard Deviation.
• The results of FNAC have been compared with histopathology obtained and that of ICC has been compared with IHC, using Sensitivity, Specificity, Positive Predictive Value, Negative Predictive Value, and Accuracy, along with the corresponding 95% Confidence Intervals.
• FNAC compared to histopathology True Positive = FNAC diagnosis and histopathology of False Positive = FNAC diagnosis was suggestive the lesion were suggestive of malignancy(a) of malignancy but on subsequent histopathology, the lesion was found to be benign(b) False Negative =FNAC diagnosis was suggestive of True Negative =both FNAC and histopathology a benign breast disease but on histopathology, of the lesion were suggestive of benign breast the lesions were found to be malignant(c) disease(d) • Similarly, ICC was compared to IHC

results and disCussion
The patient population consisted of all female patients in the age range of 18-74 years with a mean age of 46.3±13.782 years.
A total of 50 cases of FNAC breast were obtained during the given period of study.
Using FNAC, we were able to make correct diagnoses in all 50 cases. Out of those 28 were diagnosed as Malignant and 22 were diagnosed as benign. which were subsequently confirmed using the histopathological examination.
In the data we collected, there was no case of malignancy that was diagnosed as benign by FNAC.
It was observed that benign breast lesions were most common in the 4 th decade of life and the most common lesion was fibroadenoma. It was also observed that malignant lesions were more common in the 5 th and 6 th decades of life, wherein invasive ductal carcinoma was the most common lesion. Out of the 50 samples collected, ICC was performed on 25 samples. Out of these,16 cases were found to be malignant and 9 were benign. The cases which were diagnosed as malignant also underwent IHC and only those samples which had a gold standard (IHC) to be compared to, have been included in analyzing the results We calculated the Sensitivity, Specificity, Accuracy, Positive Predictive Value, and Negative  The reduction in the number of words is as shown in Figure 9 below disCussion From our study, we found that the age distribution of breast lesions was such that benign breast disease was most common in the 4 th decade of life and malignant lesions were most common in the 5 th and 6 th decades of life, this is in agreement with other studies in which the same trends can be observed 10,11 .
In our study 100% cytohistological correlation was seen in both benign and malignant cases, the Sensitivity of FNAC is 100%, the Specificity is 100%, the Accuracy is 100%, the The sensitivity and specificity of our study are found to be greater than the results seen in the meta-analyses by Wang et al., (12)This could be because, the sample size of our study is limited to 50, whereas the study performed by Wang et al., being a meta-analysis can be generalized to the population at large. The Accuracy of our study is in agreement with the literature available wherein Accuracy rates of 84%-99.5% have been reported 13,14 .
The study by Panjvani et al., has resulted in a value of 100% for Positive Predictive Value(PPV) and 97.85% for Negative Predictive Value(NPV), the study by Miškoviae et al., suggests a PPV of 95.5%, and NPV of 94.2%, and the study done by de Cursi et al., has resulted in a PPV of 99.6% and NPV of 97.6%, all of which are close to the values found in our study (13)(14)(15)(16). While histopathology is considered a gold standard, the presence of FNAC as an alternative for obtaining results of breast lesions, serves as a boon in resource-deficient areas because of its cost-effectiveness, rapid turnaround time, and repeatability 17,18 .
These results of Specificity and PPV are in agreement with the literature, however, the Sensitivity and NPV obtained in our study are greater than what is reported by Geethamala et al., wherein Sensitivity of 96.3% and NPV of 95.7% have been reported 4 .
This diagnostic modality can be preferred over histopathology in cases of advanced inoperable tumors, ICC using ER can serve as a reliable test to assess the ER status of such a breast lesion following which the decision of selecting hormone therapy can be done. In a patient with advanced breast cancer, by avoiding histopathology we spare the patient the additional distress of an invasive procedure and instead get the necessary results by using a relatively painless, minimally invasive test like FNAC on which ICC can be performed directly. This is also in accordance with the guiding principle of improving the quality of life in patients with advanced breast carcinoma 19 .  When we performed ICC using ER, we found that better staining of the samples and hence better results were obtained when the FNAC sample was placed on the slide in the form of a button as opposed to a smear, before staining with ER ( Figure 6). The sample also gave good results when there was little to no blood in it, so it is of great importance to obtain a bloodless sample for the same. It is also to be noted that a clinical correlation of the obtained FNAC sample will be helpful because non-neoplastic lesions such as lipoma, and mastitis need not undergo additional ICC using ER. As for the malignant cases, the results of ICC using ER were either reported as 'POSITIVE' or 'NEGATIVE'. However, in the gold standard method of IHC using ER, the Allred Score is used to give a detailed assessment of the Hormone Receptor Status, which cannot be done using ICC. text data analysis In the proposed work, medical text data used for the description of breast cancer were pre-processed to remove all types of noise and outliers. Normalization is a process by which the randomness and redundancy of text are reduced.

ConClusion
Fine needle aspiration cytology(FNAC) is a quick, cost-effective, minimally invasive, and accurate diagnostic method for the assessment of breast lumps, which can be done on an outpatient basis. FNAC provides good sensitivity, specificity, and positive and negative predictive values that are comparable to the gold standard, histopathology, especially in the hands of an experienced cytopathologist.
Immunocytochemistry (ICC) using ER done on FNAC samples is of high diagnostic accuracy as well and can be used as an adjunct to assess the estrogen receptor(ER) status, which is of vital prognostic significance, without the invasiveness of a surgical biopsy. Therefore, in resource-limited regions and with the aid of triple assessment, ICC on FNAC samples can be used instead of immunohistochemistry(IHC), especially in case of advanced inoperable carcinomas. Further studies on larger samples are needed to substantiate our findings. text data analysis The normalization process reduces the number of words in reports of FNAC diagnoses. This condensed data can serve as an important component in the development of a large-scale clinical decision support system.