Rapid Immunocytochemistry with Simple Heat-Induced Antigen Retrieval Technique for Improvement in the Quality of Cytological Diagnosis

Abstract

Rapid immunocytochemistry (ICC) can improve the accuracy of intraoperative cytological diagnoses; however, it is usually applied without heat-induced antigen retrieval (HIAR). We established a HIAR method for rapid ICC and evaluated its efficacy and reliability. Rapidly fixed smear samples were immunostained using 35 antibodies. We compared the results of HIAR by boiling in a pot or heating in an electric kettle. The smears were incubated for 3 min with each primary antibody and immuno-enzyme polymer reagent, and for 1 min with diaminobenzidine solution. HIAR for 1 min using the kettle method yielded the best cellular integrity. For 32 out of the 35 antibodies, results achieved using rapid ICC within 11 min were comparable to that achieved using standard ICC. HIAR was essential for 13 antibodies. For two of the antibodies, HIAR was not required when standard ICC was applied, but consistent staining with rapid ICC was obtained only with HIAR. In conclusion, we established a rapid ICC procedure using a simple HIAR method, which allowed efficient immunostaining of a panel of antigens, including nuclear antigens, within only 11 min. The combined use of this rapid ICC technique with other staining techniques could be useful for improving intraoperative cytological diagnoses.

Introduction

Rapid intraoperative diagnosis plays an important role in determining the most appropriate surgical strategies and assessment of the prognosis. The methods used for intraoperative diagnosis include frozen-section examinations and cytological analyses. Several comparative studies have demonstrated a similar diagnostic accuracy of intraoperative cytology to that of single frozen section examination, and that the accuracy may improve slightly when both approaches are applied in combination (Scucchi et al. 1997; Shidham et al. 2000). The potential advantages of intraoperative cytology over frozen-section examination include a reduced preparation time and the possibility of clear observation of the cellular details without any artifacts related to tissue freezing. In addition, intraoperative cytology is also useful for the evaluation of cavity fluids and of necrotic or spotty calcified tissues, from which it is difficult to prepare good frozen sections.

Immunocytochemistry (ICC) can be helpful in the differential diagnosis among carcinoma, sarcoma, lymphoma, melanoma, and benign tumors, to confirm tumor cell infiltration into resection margins, to detect metastases in the sentinel lymph nodes such as in patients with breast cancer or malignant melanoma, and to identify malignant cells in cavity fluids. However, standard ICC procedures usually take 2–4 h, precluding their use for intraoperative cytological diagnosis. Several rapid ICC methods have been introduced (Dabbs et al. 1995; Nomoto et al. 1995; Lambah et al. 2003; Salem et al. 2003; Fujishima et al. 2009; Francz et al. 2011); however, up until now, none has been introduced in routine clinical practice. Possible reasons for this may be that 1) these methods include longer procedural times of 15–25 min, or require an enhanced polymer one-step staining (EPOS) system, which is not commercially available at the present time; 2) these methods require three different fixatives (acetone, acetone/methanol, and ethanol); and 3) none of the methods has been established for nuclear antigens. Therefore, no ICC method that is rapid, sensitive, and simple, and is effective for a panel of antibodies on ethanol-fixed smears, has been described.

Heat-induced antigen retrieval (HIAR) is a key step for successful immunostaining (Shi et al. 2011). The efficacy of HIAR is influenced by the pH of the HIAR solution, the heating source, the temperature of heating, and the duration of heating (Shi et al. 1995; Taylor et al. 1996; Pileri et al. 1997). Current HIAR methods make use of 0.01 mol/L citrate buffer at pH 6.0 (CB6), 0.01 mol/L citrate buffer at pH 7.0 (CB7), 0.001 mol/L EDTA solution at pH 8.0, or 0.01 mol/L Tris buffer containing 0.001 mol/L EDTA at pH 9.0, with heating undertaken in a microwave (MW) oven (Shi et al. 1991; Cattoretti et al. 1993), autoclave (Bankfalvi et al. 1994), pressure cooker (Norton et al. 1994; Kamoshida et al. 2003), hot water bath (Kawai et al. 1994), or electric kettle (Namimatsu et al. 2005).

HIAR has been described to weaken or break the cross-linkages formed by formalin fixation, facilitating exposure of the epitopes to the antibodies (Ramos-Vara 2005; Yamashita 2007). The efficacy of HIAR is no longer restricted to formalin-fixed specimens, and we recently reported that HIAR is necessary to facilitate effective detection of all nuclear antigens and also some cytoplasmic and cell membrane antigens even in ethanol-fixed smears, with superior results obtained using the most popular HIAR solution, CB6, for most antibodies (Denda et al. 2012).

There have been no reports of rapid ICC using HIAR pretreatment. We conducted this study with the aim of establishing a rapid and reliable ICC procedure using a simple HIAR method for the detection of a panel of antigens on ethanol-fixed smears.

Materials and Methods

Cytology Specimens

We used freshly prepared ethanol-fixed smears from 33 surgical specimens, including 28 malignant tumors (4 gastric, 7 colorectal, 1 pancreatic, 4 breast, 1 prostatic adenocarcinoma, 3 testicular seminomas, 2 renal cell carcinomas, 2 glioblastomas, 1 astrocytoma, 1 basal cell carcinoma, and 2 malignant lymphomas), 4 benign tumors (1 gallbladder adenoma, 1 meningioma, 1 pheochromocytoma, and 1 uterine leiomyoma), and 2 lymph nodes extirpated for possible cancer metastasis. Cells were imprinted on aminopropyltriethoxysilane-coated slides by gently scraping the cut surface of the resected specimens. We also prepared smears of peritoneal fluid specimens obtained from four cancer patients (two ovarian, one gastric and one pancreatic adenocarcinomas) and three patients with reactive mesothelial cells. The peritoneal fluid samples were centrifuged at 1,500 rpm for 5 min at room temperature, and after discarding the supernatants, smears were prepared from the cell pellets.

The prepared ethanol-fixed smears were subjected to ICC using the immuno-enzyme polymer system, which involves a two-step procedure. This system has been widely used and reported to be a sensitive antigen detection method for routine immunostaining (Sabattini 1998; Denda et al. 2012).

Written informed consent, including consent for the research use of the specimens, was obtained from each patient.

Simple HIAR Method for Rapid ICC: a Preliminary Examination

First, we performed a preliminary examination to determine the optimal conditions for HIAR prior to rapid ICC. The parameters examined were the heating device and the duration of heating. In particular, we sought to identify the most efficient HIAR method for the rapid detection of standard nuclear antigens. Therefore, we compared the efficacy of two HIAR methods, namely the boiling method (boiling the buffer in a stainless steel pot that has been placed on an induction cooker) and the kettle method (heating in an electric kettle) for immunostaining of Ki-67 and p53 in imprint smears of gastric and colorectal adenocarcinomas.

We recently reported the efficacy of CB6, the most commonly used HIAR solution, as a useful HIAR solution for ethanol-fixed smears (Denda et al. 2012). Thus, we elected to use CB6 as the HIAR solution in this study for establishing a simple HIAR method. Imprint smears of gastric and colorectal adenocarcinomas were immediately fixed in 95% ethanol for 1 min, and then briefly rinsed in running water for 5 sec. The fixation time was determined based on the results of a previous study (Dabbs et al. 1995).

The HIAR procedure using the two aforementioned methods is described below.

1. Boiling method: Smears were placed in a stainless steel pot filled with 1 L of preheated CB6 for 1 or 3 min, so that the solution covered all the slides completely. The temperature of the CB6 solution was maintained by heating on an induction cooker (1,400 W, Model KZ-PH3; Panasonic, Kobe, Japan) set at high power (temperature variability, 98 ± 2C).

2. Kettle method: Smears were placed in an electric kettle (905 W, Model CD-XB22; Zojirushi, Osaka, Japan) filled with 800 mL of preheated CB6 for 1 or 3 min. The kettle was set at the “keep warm” temperature mode (98C), which maintained the temperature inside the kettle at a constant temperature of 95 ± 1C.

We simultaneously prepared control smears that were not subjected to any HIAR procedure, but processed for immunostaining in the same way as HIAR-treated smears. After the HIAR pretreatment, the smears were washed for 5 sec with PBS under continuous gentle agitation, and then immunostained using the rapid ICC procedure. The antibodies used for this preliminary examination were anti-Ki-67 monoclonal antibody (clone MIB-1, ready-to-use (RTU); Invitrogen, Carlsbad, CA, USA) and anti-p53 monoclonal antibody (clone DO-7, RTU; Nichirei Biosciences, Tokyo, Japan).

We evaluated the practical significance of the HIAR pretreatment by focusing on the quality and efficacy of the immunostaining and the degree of cellular adherence to the slides.

Rapid ICC Procedure

The smears were first fixed in 95% ethanol for 1 min. After rinsing in running water for 5 sec, the slides were subjected to the HIAR procedure conditioned from our preliminary examination. Control smears not subjected to HIAR treatment were processed for the rapid ICC procedure in the same way as the HIAR-treated smears. The incubation time and temperature of the primary antibodies, polymer reagents, and DAB solution were determined based on the results of previous studies (Dabbs et al. 1995; Francz et al. 2011). The smears were incubated with the primary antibody on a hot plate at 37C for 3 min, then washed with PBS for 5 sec and incubated with anti-mouse and anti-rabbit universal immunoperoxidase polymer (Histofine Simple Stain MAX-PO; Nichirei Biosciences) on the same hot plate for a further 3 min. After additional washing with PBS, the smears were incubated with the Liquid DAB + Substrate Chromogen System (Dako, Glostrup, Denmark) on the hot plate for 1 min, rinsed in running water for 5 sec, and counterstained with Mayer’s hematoxylin for 10 sec. Finally, the smears were rinsed for 10 sec in running water, dehydrated, and mounted. For the negative controls, the primary antibody was replaced with PBS containing 1% bovine serum albumin (BSA). The procedure for rapid ICC using a simple HIAR method is summarized in Table 1.

Table 1.

Procedure for Rapid Immunocytochemistry with Simple Heat-Induced Antigen Retrieval Method.

1.	Dip in 95% ethanol for 1 min	1 min
2.	Rinse in running water	5 sec
3.	Heat in preheated CB6 using an electric kettle for 1 min at 98C	1 min
4.	Wash briefly with PBS	5 sec
5.	Incubate with primary antibody for 3 min at 37C	3 min
6.	Wash briefly with PBS	5 sec
7.	Incubate with immune-enzyme polymer reagent for 3 min at 37C	3 min
8.	Wash briefly with PBS	5 sec
9.	Incubate with DAB solution for 1 min at 37C	1 min
10.	Rinse in running water	5 sec
11.	Immerse in Mayer’s hematoxylin solution for 10 sec	10 sec
12.	Rinse in running water	5 sec
13.	Dehydrate and mount	1 min
Total processing time		11 min

Abbreviations: CB6, 10 mM citrate buffer (pH 6.0); DAB, 3,3’-diaminobenzidine; PBS, phosphate buffered saline.

Application of the Rapid ICC Procedure Using a Panel of Antibodies

After defining the optimal HIAR conditions in the preliminary examination, we evaluated the efficacy of this HIAR approach for rapid ICC staining of additional diverse antigens. The types, sources, and dilutions of the 35 primary antibodies used are listed in Tables 2 –4. Of the 35 antibodies, 26 were used in their RTU form, whereas the remaining 9 concentrated (non-RTU) antibodies required 5-fold dilutions as compared with those for standard ICC with PBS containing 1% BSA. The immunostaining was performed at least twice for most of the antibodies.

Table 2.

Comparison between Rapid and Standard Immunocytochemistry for 10 Antibodies against Nuclear Antigens.

Antigen	Clone	Source	Types of observed cells	Rapid ICC			Standard ICC
				Dilution	NAR	HIAR1	Dilution	HIAR10
Androgen-R	F39.4.1	Biogenex	Prostatic AC	1:20	0	2+	1:100	3+
ER	1D5	Nichirei	Breast AC	RTU	1+	(2+, 3+)a	RTU	3+
Ki-67	MIB-1	Invitrogen	Gastric AC, Colonic AC, Glioblastoma	RTU	0	(2+, 3+)a	RTU	3+
Oct-3/4	C10	SCB	Testicular seminoma	1:20	1+	3+	1:100	3+
p53	DO-7	Nichirei	Gastric AC, Rectal AC	RTU	1+	3+	RTU	3+
PgR	A9621A	Nichirei	Breast AC	RTU	(1+, 2+)a	3+	RTU	3+
Topo-II	SWt3D1	Dako	Gastric AC, Colonic AC	1:100	(0, 1+)a	3+	1:500	3+
WT-1	WT49	Novocastra	Ovarian AC, Pancreatic AC	1:10	(0, 1+)a	(2+, 3+)a	1:50	3+
CDX-2	EP-P2764Y	Nichirei	Colonic AC, Rectal AC	RTU	0	1+	RTU	3+
MCM7	DCS-141	MBL	Gastric AC, Colonic AC	1:80	0	1+	1:400	3+

Abbreviations: AC, adenocarcinoma; Androgen-R, androgen receptor; AR, antigen retrieval; ER, estrogen receptor; HIAR1, heat-induced antigen retrieval for 1 min using an electric kettle; HIAR10, HIAR for 10 min using a pressure cooker; ICC, immunocytochemistry; MBL, Medical & Biological Laboratories; MCM7, minichromosome maintenance protein 7; NAR, no antigen retrieval; PgR, progesterone receptor; RTU, ready-to-use; SCB, Santa Cruz Biotechnology; Topo II, topoisomerase IIα; WT-1, Wilms’ tumor 1.

0, negative; 1+, weakly positive; 2+, moderately positive (weaker than that of standard ICC procedure but clearly distinguishable); 3+, strongly positive (equivalent to that of standard ICC procedure). ^a, slightly different staining intensity between samples.

Table 3.

Comparison between Rapid and Standard Immunocytochemistry for 15 Antibodies against Cytoplasmic Antigens.

Antigen	Type, Clone	Source	Types of observed cells	Rapid ICC			Standard ICC
				Dilution	NAR	HIAR1	Dilution	NAR
Calretinin	SP13	Nichirei	Ovarian AC, Mesothelial cells	RTU	3+	3+	RTU	3+
CK7	OV-TL12/30	Novocastra	Gastric AC, RCC	RTU	3+	3+	RTU	3+
CK20	Kg20.8	Novocastra	Colonic AC, Rectal AC	RTU	3+	3+	RTU	3+
Pan-CK	AE1/AE3	Nichirei	Gastric AC, Gallbladder AC	RTU	3+	3+	RTU	3+
HMWCK	34ßE12	Novocastra	BCC	RTU	3+	3+	RTU	3+
D2-40	D2-40	Nichirei	Testicular seminoma, Mesothelial cells	RTU	3+	3+	RTU	3+
GFAP	Polyclonal	Biogenex	Glioblastoma	RTU	3+	3+	RTU	3+
HGM	45M1	Novocastra	Gastric AC	1:10	3+	3+	1:50	3+
Muc-5AC	CLH2	Novocastra	Gastric AC	1:10	3+	3+	1:50	3+
SMA	1A4	Nichirei	Leiomyoma	RTU	2+	2+	RTU	3+
Synaptophysin	Polyclonal	Nichirei	Pheochromocytoma	RTU	2+	2+	RTU	3+
Vimentin	V9	Nichirei	RCC, Meningioma	RTU	3+	3+	RTU	3+
CK5/6	D5/16 B4	Dako	Mesothelial cells, BCC	1:10	(1+, 2+)a	(2+, 3+)a,b	1:50	3+b
bcl-2	124	Nichirei	B cell lymphoma	RTU	1+	2+c	RTU	3+c (HIAR10)
Chromogranin A	Polyclonal	Nichirei	Pheochromocytoma	RTU	1+	1+	RTU	3+

Abbreviations: AC, adenocarcinoma; AR, antigen retrieval; BCC, basal cell carcinoma; CK, cytokeratin; GFAP, glial fibrillary acidic protein; HGM, human gastric mucin; HIAR1, heat-induced antigen retrieval for 1 min using an electric kettle; HIAR10, HIAR for 10 min using a pressure cooker; HMWCK, high molecular weight CK; ICC, immunocytochemistry; NAR, no antigen retrieval; RCC, renal cell carcinoma; RTU, ready-to-use; SMA, smooth muscle actin.

1+, weakly positive; 2+, moderately positive (weaker than that of standard ICC procedure but clearly distinguishable); 3+, strongly positive (equivalent to that of standard ICC procedure). ^a, slightly different staining intensity between samples. ^b, HIAR is required for successful staining using the rapid ICC procedure, but not when using the standard ICC procedure. ^c, HIAR is exceptionally required for both rapid and standard ICC.

Table 4.

Comparison between Rapid and Standard Immunocytochemistry for 10 Antibodies against Cell Membrane Antigens.

Antigen	Type, Clone	Source	Types of observed cells	Rapid ICC			Standard ICC
				Dilution	NAR	HIAR1	Dilution	NAR
CA19-9	TA888	Zymed	Pancreatic AC	RTU	3+	3+	RTU	3+
CA125	TA347	Zymed	Ovarian AC	RTU	3+	3+	RTU	3+
CD20	L26	Nichirei	B cell lymphoma, Reactive LN	RTU	3+	3+	RTU	3+
CD45RB	PD7/26+2B11	Nichirei	B cell lymphoma, Reactive LN	RTU	3+	3+	RTU	3+
CD45RO	UCHL-1	Nichirei	B cell lymphoma, Reactive LN	RTU	2+	3+	RTU	3+
CEA	Polyclonal	KyowaMedex	Gastric AC, Colonic AC	1:20	3+	3+	1:100	3+
EMA	E29	Nichirei	Gallbladder AC, Rectal AC	RTU	3+	3+	RTU	3+
E-cadherin	36B5	Novocastra	Breast AC, Colonic AC	RTU	0	3+b	RTU	3+b
CD8	C8/144B	Nichirei	B cell lymphoma, Reactive LN	RTU	(1+, 2+)a	3+c	RTU	3+c (HIAR10)
CD79α	HM57	Nichirei	B cell lymphoma, Reactive LN	RTU	1+	2+c	RTU	3+c (HIAR10)

Abbreviations: AC, adenocarcinoma; AR, antigen retrieval; CA19-9, carbohydrate antigen 19-9; CA125, carbohydrate antigen 125; CD, cluster of differentiation; CEA, carcinoembryonic antigen; EMA, epithelial membrane antigen; HIAR1, heat-induced antigen retrieval for 1 min using an electric kettle; HIAR10, HIAR for 10 min using a pressure cooker; ICC, immunocytochemistry; LN, lymph node; NAR, no antigen retrieval; RTU, ready-to-use.

0, negative; 1+, weakly positive; 2+, moderately positive (weaker than that of standard ICC procedure but clearly distinguishable); 3+, strongly positive (equivalent to that of standard ICC procedure). ^a, slightly different staining intensity between samples. ^b, HIAR is required for successful staining using the rapid ICC procedure, but not when using the standard ICC procedure. ^c, HIAR is exceptionally required for both rapid and standard ICC.

Immunostaining was performed in accordance with a previously described rapid ICC procedure, and the results of immunostaining using the rapid ICC method were compared with those obtained using the standard ICC procedure.

Standard ICC was performed as described in our previous study (Denda et al. 2012). In brief, smears were fixed in 95% ethanol for 30 min and incubated in 0.03% hydrogen peroxide in methanol for 30 min to inactivate endogenous peroxidase. The smears were heated (if applicable) in CB6 solution in a pressure cooker at 120C for 10 min, and then cooled for 30 min to room temperature. The primary antibodies and polymer reagents were incubated at room temperature for 60 min, and the antigen detection processes used were the same as that described for the rapid ICC procedure.

Evaluation of Immunostaining

The immunostained smears was evaluated by three independent observers (TD, S Kamoshida, and JK), and graded from 0 to 3+ according to the intensity and clarity of the staining, in comparison with the results obtained from the standard ICC procedure. When the target cells were immunoreactive, the smear was interpreted as positive. The staining intensity was classified semi-quantitatively as follows: 0, negative; 1+, weakly positive; 2+, moderately positive (weaker than that obtained using the standard procedure, but still clearly distinguishable); and 3+, strongly positive (equivalent to that obtained using the standard procedure). The suitability of the intracellular antigen localization and presence/absence of background staining were also evaluated. Any differences in the scoring among the three investigators were settled by discussion, and each score was then provided as the consensus final score.

Results

Effect of HIAR Using the Boiling Method versus the Kettle Method

We evaluated the efficacy of HIAR using the boiling and the kettle methods for rapid ICC staining. In the boiling method, a greater degree of damage to the smears was observed when the boiling was carried out for 3 min as compared to 1 min, because of mechanical detachment of the cells from the slides with the longer duration of boiling. However, the cell loss was negligible when HIAR was performed using the kettle method, and clear and distinct positive staining was observed. For the kettle method, the results obtained with 1- and 3-min HIAR were comparable, although weak background staining was occasionally observed in smears subjected to 3-min HIAR. Based on these results, we concluded that 1-min HIAR using the kettle method was optimal for rapid ICC. Using the simple HIAR procedure, the immunostained smears were ready for interpretation within 11 min.

Rapid ICC Staining Using a Panel of Antibodies

The results of rapid ICC with the panel of antibodies are summarized in Tables 2–4, grouped according to the intracellular antigen localization. The cellular morphologies observed in the smears subjected to rapid ICC were almost identical to those observed in the smears treated by the standard ICC procedure, and no conspicuous artifacts were observed. To obtain optimal staining with the rapid ICC procedure, non-RTU primary antibodies had to be applied at 5-fold dilutions as compared with those for the standard ICC procedure; on the other hand, RTU primary antibodies were directly applicable without dilution.

Specific and appropriate immunostaining was seen with most antibodies applied in this rapid ICC procedure (Tables 2–4). Figures 1–3 illustrate several examples of immunostaining, showing a comparison of the immunostaining obtained using the rapid ICC procedure with or without HIAR, and the standard ICC procedure.

Figure 1.

Efficacy of heat-induced antigen retrieval (HIAR) for rapid immunocytochemistry of nuclear antigens. Colonic adenocarcinoma stained with anti-Ki-67 monoclonal antibody (A–C), gastric adenocarcinoma stained with anti-p53 monoclonal antibody (D–F), breast adenocarcinoma stained with anti-estrogen receptor monoclonal antibody (G–I), and ovarian adenocarcinoma stained with anti-WT-1 monoclonal antibody (J–L). Immunostaining was performed using a rapid staining procedure with (B, E, H, K) or without (A, D, G, J) HIAR or using the standard procedure with HIAR (C, F, I, L). Smears not subjected to HIAR exhibited no specific staining in the tumor cells (A, D, G, J), whereas false-positive staining in the neutrophils was caused by endogenous peroxidase activity (D; arrows). Application of HIAR for rapid immunocytochemistry yielded strongly positive staining in the nuclei of the tumor cells (B, E, H, K), which was equivalent to that obtained using the standard procedure with HIAR (C, F, I, L). HIAR also inactivated endogenous peroxidase (E, F). Scale bar = 50 μm.

Figure 2.

Examples of immunostaining of cytoplasmic antigens. Mesothelial cells stained with anti-calretinin monoclonal antibody (A–C), gallbladder adenoma stained with anti-pan-cytokeratin monoclonal antibody (D–F), pheochromocytoma stained with anti-synaptophysin polyclonal antibody (G–I), and renal cell carcinoma stained with anti-vimentin monoclonal antibody (J–L). Immunostaining was performed using a rapid staining procedure with (B, E, H, K) or without (A, D, G, J) heat-induced antigen retrieval (HIAR) or using the standard procedure without HIAR (C, F, I, L). Strongly positive staining was observed in the cytoplasm of the majority of the cells in all smears. In smears not subjected to HIAR, false-positive staining in the neutrophils was also caused by endogenous peroxidase activity (A; arrow). Scale bar = 50 μm.

Figure 3.

Examples of immunostaining of cell membrane antigens (A–F) and efficacy of heat-induced antigen retrieval (HIAR) for rapid immunocytochemistry (G–L). Ovarian adenocarcinoma stained with anti-CA125 monoclonal antibody (A–C), malignant lymphoma stained with anti-CD45RB monoclonal antibody (D–F), mesothelial cells stained with anti-cytokeratin 5/6 monoclonal antibody (G–I), and breast adenocarcinoma stained with anti-E-cadherin monoclonal antibody (J–L). Immunostaining was performed using a rapid staining procedure with (B, E, H, K) or without (A, D, G, J) HIAR or using the standard procedure without HIAR (C, F, I, L). HIAR for rapid immunocytochemistry yielded positive staining in the cytoplasm (H) or cell membrane (K) of the tumor cells. However, the standard procedure without HIAR also yielded adequate staining (I, L). In smears not subjected to HIAR, false-positive staining in the neutrophils was caused by endogenous peroxidase activity (A, G; arrows). The scale bar in part F refers to parts D-F. Scale bar = 50 μm.

For 27 of the 35 antibodies, the staining results obtained using our rapid ICC procedure were equivalent to those obtained using the standard ICC procedure (graded as 3+). For five of the antibodies, the staining intensity obtained using rapid ICC was slightly reduced as compared with that obtained using the standard procedure (graded as 2+). Several repeated immunostaining rounds yielded relatively consistent results, indicating that the rapid ICC procedure was reproducible. For five of the antibodies (antibodies against pan-cytokeratin, cytokeratins 7 and 20, human gastric mucin, and synaptophysin), the rapid ICC procedure yielded a weak background in comparison with that obtained using the standard ICC procedure. No immunostaining was observed in the negative controls (with omission of the primary antibody).

HIAR was found to be an important tool to uncover the epitopes in both the rapid and standard ICC procedures, and was shown to be essential for staining of 8 of the 10 nuclear antigens (androgen receptor, estrogen receptor, Ki-67, Oct-3/4, p53, progesterone receptor, topoisomerase IIα, and WT-1; Fig. 1), 1 of the 15 cytoplasmic antigens (bcl-2), and 2 of the 10 cell membrane antigens (CD8 and CD79α) (Tables 2–4). For 12 cytoplasmic and 7 cell membrane antigens, adequate staining (2+ or 3+) was obtained with both the rapid and standard ICC protocols, without HIAR treatment (Tables 3 and 4; Figs. 2 and 3A–F). However, application of HIAR did not deteriorate the staining intensity/quality of staining with these antibodies against cytoplasmic and cell membrane antigens.

Anti-cytokeratin 5/6 and anti–E-cadherin antibodies, which did not require HIAR pretreatment when standard ICC staining was used, stained efficiently using the rapid ICC procedure only after HIAR pretreatment (Tables 3 and 4; Fig. 3G–L). On the other hand, endogenous peroxidase was inactivated by HIAR, irrespective of the heating methods used or the durations of heating (Figs. 1D–F, 2A–C, 3A–C and 3G–I).

Successful immunostaining (2+ or higher) using the current rapid ICC procedure was not detectable for antibodies against CDX-2, minichromosome maintenance protein 7 (MCM7), and chromogranin A (Tables 2 and 3).

Discussion

ICC, applied in combination with Papanicolaou staining, can improve the accuracy of intraoperative cytological diagnosis by distinguishing benign cells from malignant cells and through the identification of malignant cells in resection margins, sentinel lymph nodes, and body fluids. Several studies have attempted various methods to speed up the ICC process to make it feasible for intraoperative cytological diagnoses. Dabbs et al. reported a rapid streptavidin-biotin method for ethanol-fixed scrape smears (Dabbs et al. 1995). This three-step method was based on a 2- or 3-min incubation time for the primary antibody and detection system, and could thus be completed within 15–20 min. However, whereas cytoplasmic antigens (cytokeratin, desmin, and vimentin) and cell membrane antigens (CEA, placental alkaline phosphatase, CD20, CD45RB, and CD45RO) could be detected, the method was not useful for the staining of nuclear antigens. In addition, the streptavidin-biotin method has the drawback of having a potentially high rate of false-positive staining caused by endogenous biotin.

The EPOS system, where primary antibodies are linked to a chemically inert polymer complex, has been shown to be effective in the rapid detection of cytokeratin in touch-imprint smears for the intraoperative evaluation of sentinel lymph node metastases in breast cancer (Salem et al. 2003; Fujishima et al. 2009). However, the EPOS immunostaining reagent is not currently commercially available. On the other hand, some studies have reported the usefulness of MW irradiation (not heating) for speeding up ICC staining. Nomoto et al. developed a rapid ICC method using MW irradiation for incubation of the primary antibody to allow effective staining of CEA and carbohydrate antigen 19-9 in acetone-fixed smears within 25 min (Nomoto et al. 1995). Furuhata et al. reported a rapid multiple ICC staining method also using MW irradiation, through which triple staining of MOC-31, BerEP4, and CEA could be completed in 19 min (Furuhata et al. 2010). This method, however, requires an expensive MW processor to provide controllable output and temperature. Recently, Francz et al. detected pan-cytokeratin (AE1/AE3) in acetone-fixed imprint smears using the immuno-enzyme polymer method, with incubation of antibodies at 37C for 3 min and 10 sec, with a total reaction time of 20 min (Francz et al. 2011). The immuno-enzyme polymer system has been shown to be highly sensitive; moreover, false-positive staining due to endogenous biotin is negligible in this system (Sabattini et al. 1998). In summary, these previously reported methods involve a procedure time of 15–25 min, are only useful for the detection of non-nuclear antigens without HIAR, and require specialized equipment or reagents.

HIAR is a key step in obtaining successful immunostaining results, because it facilitates the exposure of the epitopes to the antibodies (Shi et al. 2011). Because HIAR contributes to the disruption of formaldehyde-induced cross-linkages (Ramos-Vara 2005; Yamashita 2007), it has been considered to be unnecessary for ethanol-fixed smears. However, recent investigations, including our own, have indicated the usefulness of HIAR for detecting various markers in ethanol-fixed smears, including nuclear antigens, such as CDX-2, hormone receptors, Ki-67, MCM7, Oct-3/4, p53, p63, proliferating nuclear cell antigen, thyroid transcription factor-1, topoisomerase IIα, and WT-1 (Shtilbans et al. 2005; Kimura et al. 2010; Ikeda et al. 2011; Denda et al. 2012). In addition, we have also demonstrated that HIAR was effective for some cytoplasmic and cell membrane antigens (Denda et al. 2012). Nevertheless, to the best of our knowledge, HIAR has not been used in the rapid ICC staining of ethanol-fixed smears.

Different heating methods have been used for HIAR, including the MW oven (Shi et al. 1991; Cattoretti et al. 1993), autoclave (Bankfalvi et al. 1994), pressure cooker (Norton et al. 1994; Kamoshida et al. 2003), hot water-bath (Kawai et al. 1994), and electric kettle (Narimatsu et al. 2005). The MW heating method has several drawbacks, including the difficulty in controlling the heating temperature because of the uneven distribution of the heat and the limited number of sections that can be treated in a MW oven (Bankfalvi et al. 1994). Longer times are needed to attain the effective internal pressure in the autoclave and pressure cooker methods. Simple and timesaving HIAR pretreatment would be more suitable for rapid ICC. In this study, we compared two heating methods the boiling method (where slides were incubated in preheated CB6 in a stainless steel pot heated on an induction cooker set at high power) versus the kettle method (where slides were incubated in CB6 maintained at 98C in an electric kettle). In addition, we also compared the outcomes of two durations of heating (1 min versus 3 min). Cell loss was severe when the boiling method was used for HIAR but negligible using the kettle method. Results obtained with a 1-min HIAR using the kettle method were comparable with those obtained with a 3-min HIAR. Thus, incubation of the slides for 1 min in CB6 heated in an electric kettle at 98C was the most effective method for antigen retrieval in ethanol-fixed smears to maintain cellular integrity.

Table 1 summarizes the established procedure for rapid ICC staining using a simple HIAR method. This method shortened the procedure time but retained the detection sensitivity: (1) the smears were fixed in 95% ethanol to maintain both antigen integrity and morphology; (2) the endogenous peroxidase blocking step was omitted; (3) a simple HIAR method was applied; (4) the incubation time for each antibody was shortened to 3 min; (5) all incubations were carried out on a hot plate at 37C; and (6) washing was performed under continuous slide agitation. We found that various antigens in ethanol-fixed smears could be detected within 11 min using this rapid ICC procedure. HIAR was necessary for the rapid detection of eight nuclear antigens (three steroid hormone receptors, Ki-67, Oct-3/4, p53, topoisomerase IIα, and WT-1), two cytoplasmic antigens (cytokeratin 5/6 and bcl-2), and three cell membrane antigens (CD8, CD79α, and E-cadherin). Immunostaining with antibodies for cytokeratin 5/6 and E-cadherin did not require HIAR when the standard ICC procedure was applied, whereas, when rapid ICC was used, clear staining was obtained only after HIAR pretreatment. Furthermore, HIAR did not deteriorate the quality of immunostaining and did not damage the cell morphology. Fortunately, blocking of endogenous peroxidase was unnecessary, because the enzyme was inactivated during HIAR. These results suggest the usefulness of this simple HIAR method in rapid ICC.

Unfortunately, the current rapid ICC procedure was not suitable for immunostaining with the antibodies for CDX-2, MCM7, and chromogranin A, even though standard ICC yields satisfactory staining with all three antibodies. The reasons for this remain unknown, although it is conceivable that certain antibodies may require long incubation times and/or high concentrations to allow the antibodies to pass through the cell membrane and gain access to the epitopes. Further examinations to determine the optimal staining condition are therefore still necessary.

In conclusion, we developed a rapid ICC procedure using a simple HIAR method, which requires no special equipment. This rapid ICC method allowed immunostaining of a panel of antigens, including nuclear antigens, within only 11 min. When used in combination with Papanicolaou smear reading, this rapid ICC method may be useful to improve the accuracy of both intraoperative and routine cytological diagnoses.