Abstract
Rapid immunohistochemical investigation, in addition to staining with hematoxylin and eosin, would be useful during intraoperative frozen section diagnosis in some cases. This study was undertaken to investigate whether the recently described EnVision system, a highly sensitive two-step immunohistochemical technique, could be modified for rapid immunostaining of frozen sections. Forty-five primary antibodies were tested on frozen sections from various different tissues. After fixation in acetone for 1 min and air-drying, the sections were incubated for 3 min each with the primary antibody, the EnVision complex (a large number of secondary antibodies and horseradish peroxidase coupled to a dextran backbone), and the chromogen (3,3′diaminobenzidine or 3-amino-9-ethylcarbazole). All reactions were carried out at 37C. Specific staining was seen with 38 antibodies (including HMB-45 and antibodies against keratin, vimentin, leukocyte common antigen, smooth muscle actin, synaptophysin, CD34, CD3, CD20, and prostate-specific antigen). A modification of the EnVision method allows the detection of a broad spectrum of antigens in frozen sections in less than 13 min. This method could be a useful new tool in frozen section diagnosis and research.
(
S
A certain amount of information about rapid IHC techniques has been published (Lloyd et al. 1985; Leong and Milios 1986; Ichihara et al. 1989; Tabibzadeh and Shah 1989; Yokota et al. 1989; Chilosi et al. 1994; Dabbs et al. 1995; Tsutsumi et al. 1995; Mokry 1996; Rott and Velkavrh 1997; Richter et al. 1999; Viale et al. 1999). However, most of the methods described have the disadvantage of taking 20 min or more to perform or requiring directly labeled primary antibodies. Using an HRP method and microwave irradiation, Ichihara et al. (1989) were able to immunostain frozen sections for the intraoperative diagnosis of pancreatic cancer in 13 min. However, only four different antibodies were tested in this study. Therefore, no rapid, sensitive, and simple method for IHC that is effective with a broad range of primary antibodies and can be used for intraoperative frozen section has yet been described.
A newly developed immunohistochemical detection system, EnVision, has recently become available and has been described as a very sensitive detection method for routine IHC (Sabattini et al. 1998; Vyberg and Nielsen 1998). We have used a modification of this method to create a rapid, simple, and sensitive method that allows IHC to be performed on frozen sections in less than 13 min and thus represents a potentially useful new tool for intraoperative frozen section diagnosis.
Materials and Methods
The EnVision (DAKO; Hamburg, Germany) used in this study is a two-step method in which application of the primary antibody is followed by a polymeric conjugate consisting of a large number of secondary antibodies (goat anti-mouse or goat anti-rabbit) bound directly to a dextran backbone containing HRP. One such conjugate contains up to 100 HRP molecules and up to 15 antibodies.
In this study, the results of immunostaining with a modified rapid protocol using EnVision on frozen sections were compared with the results of the HRP- and LSAB (labeled streptavidin–biotin; DAKO) methods routinely used for the immunostaining of frozen sections in the laboratory.
Tissue Specimens and Preparation
Tissue samples from surgical specimens (carcinoma of the breast, lymph nodes, including one with metastatic carcinoma, small intestine, skin, malignant melanoma, B-cell lymphoma infiltrating the breast, prostate, thyroid gland, and early pregnancy decidua) were snap-frozen in liquid nitrogen for 30 sec immediately after removal and then transferred to a cryostat (Leica; Bensheim, Germany). Serial frozen sections were cut at 5 μm and placed on APES (3-aminopropyltriethoxy-silane; Roth, Karlsruhe, Germany)-coated slides, air-dried for 30 sec, fixed in acetone at room temperature (22C, RT) for 1 min, air-dried for 15 sec at RT, and subjected to the staining procedure. No attempt was made to block endogenous peroxidase.
Primary Antibodies
The primary antibodies used are listed in Table 1. For positive controls and the determination of the optimal dilution of the primary antibodies, tissue known to contain the antigen in question was used. For negative controls the primary antibody was replaced with mouse or rabbit IgG at the appropriate dilution.
Routine IHC Procedures
For the routine IHC procedures, frozen sections were stained by the HRP or LSAB method. The tissue sections (prepared as described above) were rehydrated in Tris-buffered saline (TBS; 25 mM Tris-HCl, pH 7.4, 137 mM NaCl, 2.7 mM KCl) for 5 min at RT. They were then incubated with the primary antibody in “antibody diluent” (DAKO) for 30 min at RT, rinsed three times in TBS, and subjected to the detection reaction. For the HRP method, the sections were incubated with the HRP-labeled goat anti-mouse antibody or goat anti-rabbit antibody (both DAKO; dilution 1:100) for 30 min at RT. The sections were then washed three times in TBS and incubated at RT with 3,3′-diaminobenzidine (DAB) (Sigma Fast DAB tablet dissolved in deionized water; Sigma, Deisenhofen, Germany) for 5 min or 3-amino-9-ethylcarbazol (AEC-substrate solution, ready to use; DAKO) for 10 min. For the LSAB method, the sections were incubated with the biotinylated secondary antibody and HRP-conjugated streptavidin (both “ready to use”; DAKO) for 10 min, being gently rinsed in TBS between the incubation steps. The same chromogens were used as for the HRP method.
Rapid EnVision IHC Procedure
For details of the procedure, see Figure 1. The sections were incubated with the primary antibody in “antibody diluent” (DAKO) and goat-anti-mouse or goat anti-rabbit EnVision-HRP–enzyme conjugate was performed for 3 min each. The “highly sensitive 3,3′diaminobenzidine plus” (DAB+) and the “3-amino-9-ethylcarbazol plus” (AEC+) chromogens (both from DAKO) were used as substrates for the EnVision–HRP–enzymes. Staining intensity was further enhanced by modifying the manufacturer's protocol in that all incubation steps (primary antibodies, EnVision, and substrate reactions) were performed on slides placed horizontally on a thermal plate at 37C. After each incubation, the slides were dipped in TBS or, after the substrate reaction, in tapwater at RT and waved at maximum speed for 10 sec. Excess liquid (buffer/water) was soaked up by a paper towel. After the last wash in tapwater, the slides were dipped quickly in distilled water before counterstaining with Meyer's hematoxylin (Sigma; 15 sec), followed by 30 sec in hand-hot (42C) tapwater. The sections were mounted in Aqua Tex (Merck; Darmstadt, Germany) and examined with a light microscope (Leica). Staining was evaluated by three independent observers (AMG, PR, and TR) and graded from 0 to 3+ according to its intensity and specificity, taking into account the results obtained with the standard IHC procedures (HRP, LSAB).
Results
The results are summarized in Table 1. The rapid EnVision procedure produced strong and specific staining that was clearly distinguishable from the weak background staining with most of the primary antibodies applied. To obtain optimal staining with the rapid EnVision protocol the primary antibodies had to be applied at concentrations four- to tenfold higher than in the standard IHC procedures. When the rapid protocol was performed with the same concentrations as in the standard procedures, staining was at best very weak and was often absent. Tissue morphology after staining with the modified protocol was almost identical to that observed by the HRP and LSAB methods, and no conspicuous freezing artifacts were noted.
a aAg, antigen; ALC, anaplastic lymphoma cells; Ca breast, carcinoma of the breast; LCA, leukocyte common antigen; LN, lymph node; met, metastasis; Merkel Ca, Merkel cell carcinoma; PSA, prostate-specific antigen; PSAP, prostate-specific acid phosphatase; R, receptor; SM, smooth muscle.
b1, ATTC, Rockville, MD; 2, DAKO, Hamburg, Germany; 3, BD Biosciences, Heidelberg, Germany; 4, Dianova, Hamburg, Germany; 5, Coulter, Hamburg, Germany; 6, Medac, Hamburg, Germany; 7, Calbiochem, Bad Soden, Germany.
co, no specific staining; +, evaluation possible with some minor difficulty; ++, evaluation possible without problem; + + +, evaluation very easy (even at low magnification), staining result as with standard immunohistochemical procedures or better.
Step-by-step schematic of the rapid EnVision procedure. RT, room temperature; TBS, Tris-buffered saline, pH 7.4; H, Meyer's hematoxylin.
Carrying out the incubations with the primary and secondary antibodies and both chromogens on a standard heating plate at 37C resulted in a marked increase in staining intensity, thus improving the sensitivity of the procedure. The higher temperature of the hand-hot tapwater accelerated the “blueing reaction” of the Meyer's hematoxylin in comparison with the use of cold water. Washing with TBS after the substrate reaction produced slightly more background than when tapwater was used, so we modified the EnVision manufacturer's protocol at this point also.
Because of the extremely short incubation times, humid chambers were not required to avoid evaporation of the immunoreagents. No nonspecific staining was observed in the control sections, except for staining of mast cells that were immunoreactive with a wide variety of different antibodies, a phenomenon that has been attributed to nonspecific binding of primary antibodies through ionic linkage (Ruck et al. 1990).
Specific, evaluable, and reproducible immunostaining (graded from 1+ to 3 + ; see Table 1) was seen with 38 of the 45 antibodies applied using the rapid EnVision protocol (Figures 2–7). When the results were inconsistent, the mean value of the three individual results was taken. With some antibodies (against CD14, CD45, CD56, cathepsin D, epithelial membrane antigen, and smooth muscle actin), EnVision produced even less background staining and a better signal-to-noise ratio than the routine HRP or LSAB procedures. As a substitute for the two clones for estrogen and progesterone receptors and the monoclonal anti-CD3 and c-erbB2 antibodies, which failed to stain with the rapid protocol, other antibodies were available that produced useful staining results. The polyclonal antibody against CD3 needed an extended incubation time of 5 min to produce sufficient staining. The two polyclonal antibodies against S100 protein did not produce specific staining on cryostat sections, either with the rapid EnVision protocol or with the standard IHC procedures.
Discussion
Because in some cases it would be useful to be able to perform rapid immunohistochemical investigations, in addition to staining with hematoxylin and eosin, during intraoperative frozen section diagnosis, several studies have tried to reduce the time for immunostaining using various different techniques (Ichihara et al. 1989; Tabibzadeh and Shah 1989; Chilosi et al. 1994). However, some of the methods described still require 20 min or more to perform (Davidson and So 1989; Viale et al. 1999). Dabbs et al. (1995) reported a rapid ABC method that took 20 min or less, but only cytological scraping, not frozen sections, were investigated. Many studies have described the use of directly labeled primary antibodies to reduce the time for immunostaining of frozen sections to 7–15 min (Yokota et al. 1989; Chilosi et al. 1994; Tsutsumi et al. 1995; Rott and Velkavrh 1997; Richter et al. 1999). However, directly labeled primary antibodies are usually more expensive and are available only for a limited range of antigens. Using microwave treatment and an ABC method, Ichihara et al. (1989) were able to reduce the time for immunostaining frozen sections to 13 min, but only four antigens, all in specimens from cases of pancreatic cancer, were investigated. The most rapid IHC method in frozen section diagnosis that has been reported is the quick LAB method (Tabibzadeh and Shah 1989). This three-step method is based on 30-sec incubation periods for the primary antibody and detection components. However, only strongly expressed antigens, such as cytokeratins, leukocyte common antigen (LCA), and vimentin, were investigated, and the method has definite limitations when immunostaining for other antigens is carried out (our unpublished findings).
In this study, we found that a broad spectrum of antigens could be detected in frozen sections within less than 13 min, using a modified protocol for the EnVision system. Therefore, this rapid, simple, and sensitive immunostaining method can be used to greatly enhance the diagnostic information obtained by standard frozen section examination during surgery. A limitation of the procedure is the number of slides that can be handled. However, in our experience, two slides with two sections each, allowing the simultaneous detection of up to four different markers, produce sufficient information for intraoperative histopathological diagnosis in most cases. The staining procedure presented here is simple and timesaving, requires no special equipment, and can easily be adapted to different applications in routine diagnostic work and research. There is no need for the microwave treatment used in some other rapid IHC protocols that have been described (Leong and Milios 1986; Ichihara et al. 1989; Nomoto et al. 1995), which is often difficult to standardize for reproducible results. EnVision has been shown to be highly sensitive in the routine staining of paraffin sections (Sabattini et al. 1998; Belling et al. 1999) and, likewise, the rapid procedure described here allows the detection of a broad spectrum of even weakly expressed antigens. False-positive staining due to endogenous biotin is not seen with this method (Vyberg and Nielsen 1998). In addition, a marked increase in detection sensitivity was achieved with the use of the DAB+ or AEC+ substrate solution and an increase in the incubation temperatures to 37C, modifications that might also be useful in routine IHC.
Immunoreactivity for pancytokeratin in metastatic carcinoma of the breast in a lymph node. Original magnification ×180.
Co-expression of pancytokeratin (
Immunoreactivity for CD34 of blood vessel endothelium in carcinoma of the breast. Original magnification ×180.
Immunostaining by Melan A in malignant melanoma. Original magnification ×180.
Immunoreactivity for the leukocyte common antigen in tumor-infiltrating lymphocytes in carcinoma of the breast. Original magnification ×180.
Immunoreactivity for pancytokeratin (
For most antibodies, staining results with the rapid modified EnVision were comparable to those obtained using the standard and time-consuming LSAB method (Guesdon et al. 1979). The two antibodies against S100 protein, two of the three antibodies against the progesterone receptor, and one of the two antibodies against the estrogen receptor, c-erb B2, and CD3 did not produce sufficient staining with the rapid EnVision protocol or with the standard IHC procedures in frozen sections, but stained well in paraffin sections (our unpublished observations). Like the EPOS system (Chilosi et al. 1994), which allows immunostaining of frozen sections in less than 10 min, EnVision uses a dextran polymer coupled to HRP molecules for detection. The EPOS system has been shown to be effective for the rapid detection of cytokeratin, leukocyte common antigen, and the melanoma-associated antigen detected by HMB-45 (Richter et al. 1999). However, in this one-step staining procedure the primary antibodies are labeled, so that its use is limited to the few directly labeled antibodies that are commercially available. In contrast, EnVision, as a two-step detection method, provides a further increase in sensitivity and can be used with any suitable primary antibody. The primary antibodies should preferably be monoclonal (in our hands, two of the four polyclonal antibodies tested did not produce satifactory results with this method), well established for use on cryostat sections, and available at higher concentrations than those used in standard IHC procedures (not “ready to use”). The need for higher concentrations will result in slightly higher costs, but if the procedure is limited to intraoperative frozen section diagnosis, the benefit will outweigh this financial disadvantage, especially in view of the time saved. The additional diagnostic information will often help to avoid a second operation and will thus make the slightly longer processing time than that for HE staining acceptable. The technique could be used in frozen section diagnosis for the detection of micrometastases in lymph nodes, e.g. in sentinel nodes, for the classification of poorly differentiated tumors, including the distinction between large-cell lymphoma and carcinoma, or for detection of deposits of signet cell carcinoma in lymph nodes or resection margins. In non-oncological surgery, detection of endocrine markers by the rapid EnVision protocol could help to distinguish parathyroid gland tissue from thyroid tissue or lymphoid tissue when this is not possible on morphological grounds alone. For research purposes, this technique could allow the unequivocal identification of certain cell populations for single cell picking and subsequent DNA or mRNA analysis by the polymerase chain reaction.
In summary, the rapid EnVision protocol presented here allows immunostaining of frozen sections in less than 13 min. This method could represent a useful new tool in surgical pathology and research, enabling more accurate frozen section diagnosis than staining with hematoxylin and eosin alone.
Footnotes
Acknowledgements
Acknowledgments
Supported by a grant from the Federal Ministry of Education and Research (01KS9603) and by the Interdisciplinary Center of Clinical Research Würzburg (IZKF).
This work is dedicated to Professor Edwin Kaiserling on the occasion of his 60th birthday.