INTRODUCTION
Fine needle aspiration cytology (FNAC) is a valuable first diagnostic procedure used in the diagnosis of a suspected tumor, as it is a simple, quick, non/minimally invasive, and cost-effective procedure. FNAC can also allow more specific diagnoses, such as subtyping of tumors as well as analysis of prognostic, predictive, and therapeutic markers, preoperatively by performing immunocytochemistry (ICC) on FNA materials, both on direct smears and cell blocks.[1,2]
In cases of clinically suspected malignancy or recurrence, some of the smears prepared are stained by May-Grünwald Giemsa (MGG) and the rest are simultaneously wet-fixed (immediately placed in 95% ethanol) for the purpose of ICC. However, there are several cases, where there is no clinical suspicion of malignancy, but microscopic examination suggests so; hence, the need to perform ICC arises. This requires patient recall and re-aspiration, a process that is inconvenient to patients as well as time-consuming for pathologists.
The present study was done to elucidate the use of MGG-stained smears for the purpose of ICC, after de-staining and saline rehydration.
Although few studies have been performed on the quality of ICC results of air dried and rehydrated smears, information regarding MGG pre-stained smears is lacking and needs to be evaluated against the routine method.[3,4]
MATERIALS AND METHODS
The study comprised of 40 cases: 25 cases of breast diseases (fibroadenomas, benign proliferative lesions, and carcinomas) and 15 cases of reactive lymphoid hyperplasia (RLH). These cases were chosen on the basis of known positivity for the antigens: pancytokeratin and leukocyte common antigen (LCA) for breast and lymph nodes, respectively.
The air-dried smears prepared in each case were stained by standard MGG stain and after the report was prepared for each case, one smear was selected from each batch having adequate cellularity and morphology. It was then placed in xylene to remove coverslip. Destaining was done by keeping the smear overnight in methanol (95%). This was followed by rehydration in normal saline. The smear was then post fixed in methanol–acetone 1:1 cold solution. This was followed by the steps followed for conventional ICC for wet-fixed slides. In each case, wet-fixed slides were prepared simultaneously and were kept aside for conventional ICC.
The results of ICC staining of the smears prepared under conventional and proposed protocols were then assessed by two independent pathologists. Each smear was observed using a binocular compound microscope under 100× oil immersion lens (magnification = 1000×). Each smear was assessed twice (once by each pathologist) to determine the overall result and grade each smear semi-quantitatively for staining-intensity, stain-localization, staining-uniformity, counter-staining, and background-staining.
All the ICC data were entered into IBM SPSS software (v23.0) and compiled separately for pancytokeratin and LCA. The inter-observer agreement stood at 100%. Comparison between the results of wet-fixed and air-dried smears was done using Fisher's exact test to determine the “P value” (two-sided). The “P value” here is the expression of the probability of obtaining such results with air-dried smears when known positives are used and the null hypothesis is that the performances of both methods of preparation are equivalent. The results of the wet fixed smears are taken to be “expected results.”
Observations
1) Pancytokeratin expression
Of a total of 25 cases, 19 air-dried smears (76%) showed expression of pancytokeratin as opposed to 20 wet-fixed smears (80%) (P = 0.289) [Figure 1a].
;)
Figure 1: Overall result. (a): Pancytokeratin. (b): Leucocyte common antigen (LCA)
Analysis of staining intensity showed that 8% of air-dried smears had optimal staining (as compared to 44% of wet-fixed smears). Most of the air-dried smears were over-stained (36%) and the remainder weakly stained (32%) or showed no stain (24%). Of the wet fixed smears, 20% did not show staining (P = 0.145) [Table 1].
Table 1: Staining intensity
The fraction of air-dried smears showing staining of incorrect compartment (36%) was nearly equal to that of correct compartment (40%). 72% of wet-fixed smears showed correct localization. Comparison of concurrent over-staining and incorrect localization of stain in individual air-dried smears showed that 9 of 9 over-stained smears (100%) also had incorrect localization of stain and 9 of 10 smears (90%) with incorrect localization were over-stained. The same was not true of wet-fixed smears where only one showed aforementioned concurrent results (P = 0.523) [Table 2].
Table 2: Localisation of stain
Staining was considered as non-uniform if there was an edge artifact or there was gradient staining. It was seen in 64% of air-dried smears, in contrast to 24% of wet-fixed smears (P = 0.558) [Table 3].
Table 3: Uniformity of stain
Counter-staining was optimal in 96% of air-dried smears and 92% of wet-fixed smears (P = 1.000).
Background staining was seen more frequently in air-dried smears (68%) than wet-fixed smears (36%). In all such cases, background staining comprised of RBCs; macrophages were stained in nearly three-fourths of the cases; and a small number of cases showed staining of collagen (P = 0.182) [Table 4].
Table 4: Background staining
2) LCA/CD45 expression
Out of 15 cases, one case was negative for LCA via both conventional and proposed protocols. The case was excluded from the study, resulting in 14 cases being analyzed.
Out of 14 cases analyzed, all the 14 wet-fixed smears were positive for LCA, whereas only 12 air-dried smears showed positivity (86%) [Figure 1b].
Analysis of staining intensity showed that only two air-dried smears were optimally stained (14.3%), most of the air-dried smears (42.8%) showed over-staining, and the rest weakly stained or showed no stain. In contrast, 85.7% of wet-fixed smears showed optimal staining (P = 1.000) [Table 1].
Analysis of localization of stain revealed that 8 out of 14 air-dried smears (57.1%) showed correct localization of stain while 13 wet- fixed smears (92.8%) did so. Amongst the air-dried smears, all smears showing incorrect localization (4) were also over-stained. These comprised 4 of 6 over-stained smears (66%) (P = 0.429) [Table 2].
Counter-staining was optimal in all wet-fixed and air-dried smears (100%).
Extent of background staining did not differ greatly amongst the two smear preparations. 42.8% of air-dried smears and 57.1% of wet-fixed smears lacked any background staining. The staining profile was similar to pancytokeratin and in all cases with background staining, RBCs were stained; macrophages were stained in nearly three-fourths of the cases; and a small number of cases showed staining of collagen (P = 0.627) [Table 4].
DISCUSSION
The ideal smear preparation technique for diagnostic ICC remains a contended topic, owing to the fact that it has to compete with immunohistochemistry in terms of rapidity, ease of procedure, and the diagnostic accuracy. The convention is to separately fix wet smears in ethanol without any staining being involved prior to ICC; our study attempted to simplify the procedure and reduce the sources of error by incorporating ICC in the routine smear preparation meant for reporting, after being stained by MGG stain.
Air-drying has benefits including eliminating the issue of drying artifacts and easier handling (air-dried smears can be conveniently stacked in trays and stored for long periods of time without much difficulty). ICC on Papanicolaou-stained materials have been shown to be possible.[4,5,6] However, it has turned out to be difficult to do ICC on MGG-stained smears.[7]
Pancytokeratin: Upon examination, the wet-fixed smears that gave negative results [20%; Figure 1a] or weak staining intensity [28%; Table 1] were attributed to be due to air-drying artifacts. 24% [Figure 1b] of air-dried smears were negative. Air-drying artifacts are not seen in air-dried smears; hence, errors due to fixation via air-drying are unlikely. Based on the available literature, this lack of staining is likely due to the denaturant effects of methanol.[8] The sample was subjected to methanol twice: during fixation prior to the slide being placed in MGG stain solution and during de-staining. The prolonged contact with methanol is likely to have changed epitope configuration, which also explains why another one-third of the air-dried smears could stain only weakly. Hence, although the overall fraction of smears with no stain and weak stain is nearly the same by both preparations [56% in air-dried and 48% in wet-fixed; Table 1], it could be significantly reduced in wet-fixed preparations if fixation was performed properly, whereas it can only be reduced in case of air-dried smears if methanol exposure is reduced or eliminated, which is a tough prospect while dealing with the Giemsa stain. An alternative is to skip the process of de-staining and attempt ICC on Giemsa pre-stained smears themselves, as proposed in the study by Beraki et al.;[7] this would retain the advantages of air-drying while decreasing false negatives. The results of the study were far better, with no false negatives. Staining intensity was also optimal but specific antigen retrieval techniques and higher antibody concentrations were utilized.
The large fraction of air-dried smears [36%; Table 1] that seem to be paradoxically over-stained was in fact due to loss of integrity of cells and leakage of cytoplasmic contents. This also explains why all of these smears had been interpreted as having incorrect localization of the stain [Table 2]. Since it is cytoplasm that has been stained, the staining is correct and specific. The reason for the cytolysis in such a large number of smears is unclear, but may be due to technical error. These findings were also seen in the study by Beraki et al.[7] and a similar conclusion had been derived, although in their study, background staining was present as an exception rather than norm, unlike ours.
Despite the results being statistically insignificant for both pancytokeratin and LCA (P values are 0.289 and 0.145, respectively; α = 0.05); the tiny fraction (8%) of air-dried smears that have optimal staining is inadequate for the preparation method to replace conventional technique. It is likely that the P value has been influenced by the large quantity of air-drying artifact in the wet-fixed smears.
Non-uniform staining and background staining were more prevalent in air-dried smears, but the difference was statistically insignificant [Table 3]. Counter-staining did not seem to be influenced much by the methods of smear preparation.
LCA/CD45: Overall, better results were seen by both methods. Although none of the wet-fixed smears were negative, 14.3% of air-dried smears showed no stain [Figure 1b].
Analysis of staining intensity revealed a similar picture as that of pancytokeratin. Most of the air-dried smears were over-stained, the rest weakly stained or showing no stain, and a small fraction (14.3%) showed optimal staining [Table 1]. This again was not satisfactory enough to allow the proposed protocol to replace the conventional protocol, though the P value was 1.000. The association between over-staining and incorrect localization was present amongst the air-dried smears but to a lesser extent.
Non-uniform staining and background staining were more prevalent in air-dried smears, but the difference was statistically insignificant [Tables 3 and 4]. Counter-staining did not seem to be influenced much by the methods of smear preparation.
Although none of the results were statistically significant, and there were a few instances of ideal staining in both the markers, it may be possible that MGG-stained smears can produce results that are equivalent, if not superior to conventional technique, provided the issue of cytolysis and methanol exposure is addressed. The results are similar to those in the studies utilizing the methanol-based fixative PreservCyt™, thus further underscoring the influence of methanol.[9,10]
CONCLUSION
The proposed protocol of de-stained MGG smears as an alternative to conventional technique for ICC was unsuccessful in giving satisfactory results in all the five parameters assessed in the proposed ICC. Although the difference in the results of proposed and conventional protocol was statistically insignificant, there is a precedent to the success of ICC on MGG-stained smears;[8] an attempt to modify the protocol by addressing the issues of methanol exposure and technical error is likely to produce good results.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
REFERENCES
1. Kirbis IS, Maxwell P, Flezar MS, Miller K, Ibrahim M. External quality control for immunocytochemistry on cytology samples: A review of UK NEQAS ICC (cytology module) results Cytopathology. 2011;22:230–7
2. Schmitt F, Cochand-Priollet B, Toetsch M, Davidson B, Bondi A, Vielh P. Immunocytochemistry in Europe: Results of the European Federation of Cytology Societies (EFCS) inquiry Cytopathology. 2011;22:238–42
3. Beraki E, Sauer T. Determination of HER-2 status on FNAC material from breast carcinomas using
in situ hybridization with dual chromogen visualization with silver enhancement (dual SISH) Cytojournal. 2010;7:21
4. Krishnamurthy S, Dimashkieh H, Patel S, Sneige N. Immunocytochemical evaluation of estrogen receptor on archival Papanicolaou-stained fine-needle aspirate smears Diagn Cytopathol. 2003;29:309–14
5. Kumar S, Reddy S. Immunocytochemistry study on pulmonary cytology samples in lung cancer Indian J Mednodent Allied Sci. 2018;6:71–93
6. Zhang L, Krausz T, DeMay RM. A pilot study of Galectin-3, HBME-1, and p27 triple immunostaining pattern for diagnosis of indeterminate thyroid nodules in cytology with correlation to histology Appl Immunohistochem Mol Morphol. 2015;23:481–490
7. Beraki E, Olsen TK, Sauer T. Establishing a protocol for immunocytochemical staining and chromogenic
in situ hybridization of Giemsa and Diff-Quick prestained cytological smears CytoJournal. 2012;9:8
8. Burry RW Fixation Theory In: Immunocytochemistry: A Practical Guide for Biomedical Research. Illustrated ed. 2009 New York Springer Science & Business Media
9. Gruchy JR, Barnes PJ, Dakin Haché KA. CytoLyt® fixation and decalcification pretreatments alter antigenicity in normal tissues compared with standard formalin fixation Appl Immunohistochem Mol Morphol. 2015;23:297–302
10. Lloyd IE, Zhou W, Witt BL, Chadwick BE. Characterization of PD-L1 immunohistochemical expression in cell blocks with different specimen fixation and processing methods Appl Immunohistochem Mol Morphol. 2017
