Quality Assessment of Proteins and RNA Following Storage in Archival Formalin-Fixed Paraffin-Embedded Human Breast Cancer Tissue Microarray Sections

Abstract

Although the immunogenicity of formalin-fixed paraffin-embedded tissue sections can decrease during storage and transport, the exact mechanism of antigenic loss and how to prevent it are not clear. Herein, we investigated changes in the expression of estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER-2), E-cadherin, and Ki-67 in human breast tissue microarray (TMA) tissue sections stored for up to 3 months in dry and wet conditions. The positive rates of ER and PR expression were minimally changed after 3 months of storage, but the Allred scores of ER and PR stored in humid conditions decreased remarkably in comparison to fresh-cut tissue. The HER-2 antigenicity and RNA integrity of breast TMA sections stored in dry conditions diminished gradually with storage time, whereas the immunoreactivity and RNA quality of HER-2 in humid conditions decreased sharply as storage length increased. The area and intensity of E-cadherin staining in tissue sections stored in dry conditions did not change significantly and were minimally changed after 3 months, respectively. In contrast, the area and intensity of E-cadherin staining in tissue sections stored in humid conditions decreased significantly as storage length increased. Finally, the Ki-67 labeling index of tissue sections stored for 3 months in dry (9% decrease) and wet (31.9% decrease) conditions was decreased in comparison to fresh sections. In conclusion, these results indicate that water is a crucial factor for protein and RNA degradation in stored tissue sections, and detailed guidelines are required in the clinic.

Introduction

In an era of personalized therapy, various ancillary molecular tests are conducted with patient tissue specimens after histopathological examination to determine prognoses and design treatment strategies. In particular, transcriptional profiling analysis relies on surgical tissue specimen quality.^1,2 However, long-term preservation of high-quality specimens, such as frozen tissue, is not practical in a routine clinical care environment. The tissues obtained from patients are usually processed into formalin-fixed, paraffin-embedded (FFPE) blocks for diagnostic histopathology and archival storage.

Immunohistochemistry (IHC) is well established and commonly used in histopathology for diagnosis, prognosis, and biomarker identification, but its use is often plagued by poor reproducibility and inconsistent interpretation. Although there have been recent efforts toward improving IHC reagents, standardization, and automation,^3,4 there are several factors that can affect IHC staining quality. These include preanalytical conditions (e.g., cold ischemia, fixation, and processing time), section thickness, and paraffin block and slide storage conditions.^5–9 For the best staining quality, a fresh paraffin section should be used, but use of stored slides is inevitable. Prior studies have demonstrated that humidity and a long-term delay between cutting sections and IHC staining (section aging) can cause antigen decay, whereas cold temperature storage of paraffin blocks or sections prevents antigen degradation.^10–12

In contrast, there are reports that antigenicity in stored paraffin blocks is preserved for several decades with very slow and minimal antigen decay.^13,14 Although scientists believe that oxidation processes and archival conditions are involved in the loss of antigenicity in stored paraffin blocks and slides, the exact mechanism is unclear. To overcome these problems, several storage methods have been developed. Some methods recommend storage in a refrigerator or paraffin coating,^5,9,10 but these are controversial and cumbersome to use, which make them difficult to apply in the clinic.

The guidelines of the American Society of Clinical Oncology and the College of American Pathologists (ASCO/CAP) require both IHC and in situ hybridization (ISH) tests for the assessment of diagnostic and prognostic markers in cancer tissue specimens. ISH allows the detection of gene expression or genetic alterations at the cellular level within histological sections and has become an important diagnostic tool. It serves as an alternative to conventional IHC for many markers, including human epidermal growth factor receptor 2 (HER-2),¹⁵ phosphatase and tensin homolog,¹⁶ and programmed cell death ligand 1.¹⁷ In situ transcriptional profiling of single cells provides important insights into normal biological function and disease processes. In this context, RNA assays using FFPE tissues have become important for prognoses. Studies have shown that the quality of RNA from FFPE tissues relies on proper fixation, processing, tissue block storage, and age.^1,2,18–20

Breast cancer is a key example of how molecular and histological characterization play a fundamental role in treatment decisions. In most clinical pathology laboratories, IHC for estrogen receptor (ER), progesterone receptor (PR), and HER-2 is used for subtyping of breast cancer, and IHC examination for E-cadherin is also used when lobular carcinoma is suspected.^21,22 Ki-67 expression (cutoff value >25%) provides strong prognostic information in breast cancer patients.²³

In examining the expression of these major markers, slide quality management is important.^8,10,24 According to recently updated ASCO/CAP guidelines for breast biomarker reporting, variables that can affect test results include cold ischemia time, fixation time, type of fixative, any tissue treatment that could potentially alter immunoreactivity (decalcification), the status of internal and external controls, adequacy of sample for evaluation, primary antibody clone, and regulatory status (Food and Drug Administration cleared vs. laboratory-developed test). For HER-2 expression, ASCO/CAP guidelines recommend not to use unstained slides that have been cut more than 6 weeks before analysis.²⁵ However, there is no definite storage regulation for other subtype-determining or prognosis-related markers.

To expand our previous finding that water negatively impacts protein and RNA stability in FFPE tissue sections, we analyzed changes in the IHC staining of breast cancer subtype markers based on two storage methods at room temperature: vacuum packed with a desiccant or in a humidity chamber. In addition, we also investigated HER-2 messenger RNA (mRNA) level by assessing c-erbB2 expression in stored breast cancer tissue microarray (TMA) slides using RNA-ISH technology. IHC and RNA-ISH signal intensities were assessed by digital image analysis. Furthermore, we evaluated whether loss of bioanalytes from experimentally stored human breast TMA sections affected breast biomarkers accordant with the current ASCO/CAP clinical practice guidelines.

Materials and Methods

Specimens and TMA construction

This study included 100 consecutive breast cancer cases, which were resected surgically at Kangbuk Samsung Hospital between January and December 2010. Cases with neo-adjuvant chemotherapy were not included. Based on the histology, the total cohort consisted of 97 cases of invasive carcinoma and three cases of invasive lobular carcinoma. The patients' clinical data, including age at operation, sex, type of operation, and adjuvant chemotherapy, were obtained from electronic medical records. All slides of 100 breast cases were reviewed and pathologic findings, including histologic grade, tumor size, and lymph node metastasis, were recorded.

The molecular subtype was determined on the basis of IHC results (ER, PR, HER-2, and Ki-67), which were routinely performed with breast cancer tissue post-operation. TMA blocks were produced with 2-mm-diameter cores from representative paraffin blocks of each tumor, as described previously.²⁶ This study protocol was approved by the Institutional Review Board at Kangbuk Samsung Hospital (IRB No. 2020-08-029-001, Seoul, Republic of Korea).

Storage conditions of unstained slides

The unstained breast TMA slides were produced at 3 months, 1 month, 1 week, and 1 day before the scheduled date of the IHC and RNA-ISH tests. The 1-day-old slides were used as the controls for both tests. Sections were 5-μm thick, and slides were baked in an oven for 60 minutes. These slides were stored using either a vacuum packaging unit (Foodsaver; Jarden Corporation, Providence, RI) with a desiccant (Drierite, anhydrous calcium sulfate; W.A. Hammond Company, Xenia, OH) or a humidity chamber at room temperature. The humidity chamber was built using a plastic container (11 × 8 × 6 cm) containing 100 mL distilled water. The slides were placed on a rack above the water and sealed in the plastic container (humidity, 94%–96%). Then, they were collected and used for IHC staining or RNA-ISH at the same time to minimize intraexperimental variability.

IHC staining

Breast TMA sections from each condition were deparaffinized and rehydrated through xylenes and serial alcohol solutions, respectively. Slides were incubated for 10 minutes at 3% H₂O₂ to quench endogenous peroxidase activity after antigen retrieval. Detailed IHC conditions are described in Table 1. We used the IHC conditions that were established previously in our laboratory. The antigen-antibody reaction was detected with Dako Envision + Dual Link system-horseradish peroxidase (Dako, Carpinteria, CA) and visualized with 3,3′-diaminobenzidine (Dako). Mouse or rabbit immunoglobulin G and the omission of the primary antibody were used as a negative control. Slides were counterstained with hematoxylin and cover-slipped.

Table 1.

Antibodies Used for Immunohistochemistry

Antibody	Vendor	Clonality	Cat. no.	Incubation	Dilution	Antigen retrieval
ER	Dako	mAb Ms (1D5)	M7047	30 minutes at RT	1/500	20 minutes, pressure cooker, pH 9
PR	Dako	mAb Ms (PgR636)	M3569	30 minutes at RT	1/1000	20 minutes, pressure cooker, pH 9
HER-2	Dako	pAb Rb (c-erbB-2)	A0485	30 minutes at RT	1/2000	20 minutes, pressure cooker, pH 6
E-cadherin	Dako	mAb Ms (NCH-38)	M3612	30 minutes at RT	1/500	20 minutes, pressure cooker, pH 9
Ki-67	Cell marque	mAb Rb (SP6)	275R-15	30 minutes at RT	1/250	20 minutes, pressure cooker, pH 9

ER, estrogen receptor; HER-2, human epidermal growth factor receptor 2; PR, progesterone receptor; RT, room temperature.

The stained slides were scanned by an Aperio AT2 digital scanner with a 40 × objective (Leica Biosystems Inc., Buffalo Grove, IL). Representative images of each core were analyzed using Visiopharm software version 4.5.1.324 (Visiopharm, Hørsholm, Denmark). For ER and PR, nuclear staining was recorded separately as staining intensity (0 = negative, 1 = weak, 2 = moderate, and 3 = strong) and proportion; the scores were calculated using the Allred score system.²⁷ Membrane HER-2 and E-cadherin expression were quantified by staining area and intensity. HER-2 positivity was determined using the HER-2 scoring system from the ASCO/CAP clinical practice guideline-focused update, with scores of 0 and 1 considered negative, 2 equivocal, and 3 positive.²⁸ The Ki-67 labeling index was calculated using the fraction of tumor cell nuclei stained at moderate or higher intensity.

In situ hybridization

Chromogenic in situ detection of HER-2 mRNA was performed manually using the RNAScope in situ Hybridization Kit (Advanced Cell Diagnostics, Hayward, CA) according to the manufacturer's instructions. Briefly, the slides were deparaffinized, boiled with pretreatment reagent for 15 minutes, and then digested with protease at 40°C for 30 minutes, followed by hybridization for 2 hours at 40°C with probe-Hs-Erb-B2 receptor tyrosine kinase 2 (ERBB2) (Cat. No. 310081; Advanced Cell Diagnostics). In addition, peptidylprolyl isomerase B (Probe-Hs-PPIB; Cat. No. 313901; Advanced Cell Diagnostics) and Bacillus subtilis dihydrodipicolinate reductase (Probe-DapB; Cat. No. 310043; Advanced Cell Diagnostics) were used as positive and negative controls, respectively. Specific probe binding sites were visualized with the RNAScope 2.5 HD Reagent Kit (Brown) (Cat. No. 322370; Advanced Cell Diagnostics) and lightly counterstained with hematoxylin. Positive staining was detected as brown dots in the nucleus and/or cytoplasm.

Digital images were acquired using an Aperio AT2 digital scanner with a 40 × objective (Leica Biosystems Inc.). Captured digital images were imported into computer-assisted image analyzing software (Visiopharm v6.9.1) for quantification of ERBB2 mRNA expression. After training the system by digitally “painting” examples of the nucleus in the image, tumor nuclei were defined, and the cytoplasm was further defined by outlining the nucleus. Finally, the detection system was digitally trained to identify brown 3,3′-diaminobenzidine dots of HER-2. The mean value of brown dots per tumor cell in the representative areas of each TMA was calculated.

Statistical analyses

Variables were expressed as mean ± standard deviation (SD) or percentage and frequency. The differences between fresh and 1-week, 1-month, and 3-month samples from each group were analyzed using a paired t-test for continuous variables and using the McNemar's test for categorical variables, as appropriate. An interaction test to estimate the significant differences between groups over time was calculated using the linear mixed model for continuous variables and the generalized estimating equations method for categorical variables, as appropriate. A two-tailed p-value <0.05 was considered significant. All statistical analysis was performed using SPSS 24 (IBM SPSS Statistics 24; IBM Corporation, Armonk, NY).

Results

Patient characteristics

The patients' clinicopathological characteristics are summarized in Supplementary Table S1. All patients were female and the median age at operation was 50 years (range, 32–76 years). The mean tumor size was 2.2 cm (range, 0.7–5.7 cm). Twenty-five patients (25.0%) were the American Joint Committee on Cancer stage I, 43 (43.0%) were stage II, 28 (28.0%) were stage III, and 4 (4.0%) were stage IV. LN metastasis was detected in 48 (48.0%) patients. The biological subtypes of breast cancer included luminal A (n = 64, 64.0%), luminal B (n = 23, 23.0%), HER-2-enriched (n = 6, 6.0%), and triple-negative breast cancer (n = 7, 7.0%).

ER and PR expression

Of 100 patients, 89–94 interpretable samples were obtained for IHC and molecular assays. To assess the effects of storage period and humidity on the antigenicity of human breast FFPE TMA sections, ER and PR expression were evaluated by IHC staining (Fig. 1). Typical nuclear staining for ER and PR was detected in all tissue sections stored in dry conditions. In contrast, the ER and PR staining of tissue sections stored in the humidity chamber for 3 months were significantly weakened in intensity and number of stained cells (Fig. 1A, B). Using the Allred scoring system, we confirmed that the total score (TS) and intensity score (IS) for ER were significantly decreased after 3 months of storage in dry conditions, whereas there was no notable difference in ER proportion score (PS) among the dry conditions (Table 2).

FIG. 1.

Immunohistochemical assessment of ER and PR according to storage time in both dry and wet conditions. Representative images of ER (A) and PR (B) immunohistochemical staining in human breast tissue. High magnification images are shown in the inset. Scale bar, 100 μm. Quantitative analysis of ER (C) and PR (D) immunohistochemical staining. Data are presented as mean ± SD. ER, estrogen receptor; PR, progesterone receptor; SD, standard deviation.

Table 2.

Allred Scores for Estrogen Receptor and Progesterone Receptor Evaluation in Breast Cancer Tissues Stored Experimentally

		Dry			Wet
		n	Mean ± SD	p	n	Mean ± SD	p
ER
PS	Fresh	94	4.07 ± 1.56		94	4.07 ± 1.56
	1 Week	94	4.09 ± 1.54	0.829	94	4.01 ± 1.59	0.241
	1 Month	92	4.01 ± 1.73	0.277	94	3.72 ± 1.41	<0.001
	3 Months	89	4.12 ± 1.51	0.798	92	3.09 ± 1.64	<0.001
IS	Fresh	94	2.19 ± 0.88		94	2.19 ± 0.88
	1 Week	94	2.11 ± 0.93	0.103	94	1.98 ± 0.90	<0.001
	1 Month	92	2.09 ± 1.02	0.068	94	1.66 ± 0.74	<0.001
	3 Months	89	1.99 ± 0.85	<0.001	92	1.24 ± 0.60	<0.001
TS	Fresh	94	6.27 ± 2.37		94	6.27 ± 2.37
	1 Week	94	6.19 ± 2.38	0.373	94	5.99 ± 2.40	0.004
	1 Month	92	6.10 ± 2.67	0.104	94	5.38 ± 2.03	<0.001
	3 Months	89	6.11 ± 2.23	0.003	92	4.33 ± 2.11	<0.001
PR
PS	Fresh	93	3.76 ± 1.56		94	3.78 ± 1.55
	1 Week	93	3.55 ± 1.64	0.004	94	3.07 ± 1.56	<0.001
	1 Month	93	3.55 ± 1.58	0.005	93	3.04 ± 1.50	<0.001
	3 Months	91	3.44 ± 1.48	0.001	94	3.10 ± 1.31	<0.001
IS	Fresh	93	1.86 ± 0.87		94	1.86 ± 0.86
	1 Week	93	1.75 ± 0.92	0.032	94	1.40 ± 0.83	<0.001
	1 Month	93	1.66 ± 0.84	<0.001	93	1.20 ± 0.72	<0.001
	3 Months	91	1.49 ± 0.78	<0.001	94	1.22 ± 0.72	<0.001
TS	Fresh	93	5.62 ± 2.33		94	5.64 ± 2.32
	1 Week	93	5.30 ± 2.45	0.002	94	4.48 ± 2.28	<0.001
	1 Month	93	5.20 ± 2.31	<0.001	93	4.25 ± 2.09	<0.001
	3 Months	91	4.93 ± 2.12	<0.001	94	4.32 ± 1.90	<0.001

IS, intensity score; PS, proportion score; SD, standard deviation; TS, total score.

On the other hand, there were significant decreases for PS, IS, and TS for ER in sections stored in the humidity chamber. This remarkable decrease was detected at 1 week for IS and TS and 1 month for PS. The PS, IS, and TS scores for ER after 3 months in the humidity chamber were reduced by 24.1%, 43.4%, and 30.9%, respectively, in comparison to fresh-cut tissue sections.

Loss of PR antigenicity started earlier than ER (Fig. 1C, D), and there was a significant reduction in all three scores (PS, IS, and TS) for PR in all experimentally stored breast cancer tissues (Table 2). The PS, IS, and TS scores for PR were 91.5%, 80.1%, and 87.7% of the fresh-cut scores, respectively, after 3 months in the dry chamber. Meanwhile, these scores for PR after 3 months in the humidity chamber were 18%, 34.4%, and 23.4% respective reductions of the fresh-cut scores. Overall, the antigenicity of ER and PR was better preserved in dry chamber conditions in comparison to the humidity chamber (Fig. 1C, D).

Using the current ASCO/CAP clinical practice guidelines, the positive rates for ER and PR expression in fresh-cut breast cancer tissues were 86.2% and 83.9%, respectively. Fortunately, these positive rates were minimally changed during 3 months in either dry or humid conditions (Supplementary Table S2).

HER-2 expression

The membranous staining of HER-2 was well maintained for up to 1 week in both dry and wet conditions (Fig. 2). To objectively evaluate the change in antigenicity, we measured the area and intensity of HER-2 staining using digital image analysis. The stained area and intensity were maintained up to 1 week in dry condition, but the wet condition showed a significant decrease in stained area and intensity after 1 week (Table 3). The HER-2-stained area of tissue sections stored in dry conditions was relatively stable for up to 1 month (12.6% reduction compared to the fresh section), whereas the HER-2-stained area in humid conditions declined significantly after 1 week of storage (28.9% reduction, p < 0.001) compared to fresh tissue sections. Similarly, we confirmed that the intensity of HER-2 staining in stored tissue was more stable in dry conditions than in humid conditions (Table 3).

FIG. 2.

Evaluation of HER-2 immunochemical staining according to storage time in both dry and wet conditions. (A) Representative images of HER-2 immunohistochemical staining in human breast tissue. High magnification images are shown in the inset. Scale bar, 100 μm. (B) Quantitative analysis of HER-2 immunohistochemical staining. The data are presented as mean ± SD. (C) Change of HER-2-positive rate according to storage time in both dry and wet conditions. HER-2, human epidermal growth factor receptor 2.

Table 3.

Area and Intensity of Human Epidermal Growth Factor Receptor 2 Expression in Breast Cancer Tissues Stored Experimentally

	Dry			Wet
	n	Mean ± SD	p	n	Mean ± SD	p
Area
Fresh	96	26.85 ± 41.06		95	27.05 ± 41.24
1 Week	96	26.66 ± 38.97	0.902	95	19.24 ± 30.56	<0.001
1 Month	95	23.46 ± 35.60	0.047	96	17.30 ± 28.54	<0.001
3 Months	92	18.94 ± 33.70	0.004	93	8.71 ± 19.61	<0.001
Intensity
Fresh	96	1.11 ± 0.22		95	1.11 ± 0.22
1 Week	96	1.09 ± 0.23	0.064	95	1.01 ± 0.20	<0.001
1 Month	95	1.08 ± 0.22	0.022	96	0.97 ± 0.16	<0.001
3 Months	92	1.03 ± 0.21	<0.001	93	0.90 ± 0.18	<0.001

Based on the current ASCO/CAP clinical practice guidelines, positive, equivocal, and negative HER-2 expression values were 15.6%, 42.7%, and 41.7% in fresh sections, respectively. In dry conditions, these values changed at 1 month (positive, equivocal, and negative rates of 11.5%, 34.4%, and 54.2%) and significantly lost antigenicity at 3 months (positive, equivocal, and negative rates of 8.7%, 30.4%, and 60.9%) (Fig. 2C). In wet conditions, HER-2-positive, HER-2-equivocal, and HER-2-negative rates were 4.2%, 12.5%, and 83.3%, respectively, at 1 month, whereas they were 2.1%, 8.6%, and 89.3% at 3 months, and the results showed a remarkable decrease in HER-2 positivity (p < 0.001; Fig. 2C).

To appreciate the impacts of storage period and humidity on RNA integrity, we evaluated RNA quantity in breast cancer tissue sections, stored in dry or humid chambers for 3 months, using manual RNAScope ISH with ERBB2 probes (Fig. 3). HER-2 RNA expression showed a modest decrease after 1 week storage in both conditions (Fig. 3; Supplementary Table S3). In comparison to fresh-cut slides, HER-2 RNA expression gradually decreased in dry conditions as the storage length increased, with 15.3%, 16.2%, and 26.3% decrements at 1 week, 1 month, and 3 months. In contrast, there was a sharp decrease in HER-2 RNA expression after 1 month in humid conditions (1 week, 1 month, and 3 months; respective 21.5%, 76.2%, and 83.8% decrease in comparison to fresh-cut scores) (Supplementary Table S3).

FIG. 3.

RNA quality according to storage time in both dry and wet conditions. RNA quality was assessed by RNAScope in situ hybridization. (A) Representative RNAScope in situ hybridization images of ERBB2 in human breast tissue. High magnification images are shown in the inset. Scale bar, 100 μm. (B) Quantitative analysis of ERBB2 expression. Data are presented as mean ± SD. ERBB2, Erb-B2 receptor tyrosine kinase 2.

E-cadherin and Ki-67 expression

E-cadherin was stained in the cellular membrane of all cases of ductal carcinoma histology (93 of 96 cases, 96.9%) (Fig. 4A). In a detailed analysis of the stained area and intensity, the stained area of E-cadherin expression for the tissue sections stored in dry conditions was not changed significantly, while the intensity of staining changed slightly after 3 months (7.6% decrease in comparison to fresh cut) in the dry chamber (Fig. 4B).

FIG. 4.

E-cadherin immunohistochemical staining of human breast cancer tissue for slides stored over time and in dry and wet storage conditions. (A) Representative immunohistochemical images of E-cadherin before and after storage in both dry and wet conditions. High magnification images are shown in the inset. Scale bar, 100 μm. (B) Quantitative analysis of the immunohistochemical stained area and intensity of E-cadherin according to storage conditions. Data are presented as mean ± SD.

In contrast, the area and intensity of E-cadherin staining for the tissue sections stored in the humidity chamber decreased significantly as storage length increased. In comparison to fresh-cut sections, the stained area of E-cadherin was decreased by 25.3%, 38.9%, and 53.5% in the tissue sections stored in the humid conditions for 1 week, 1 month, and 3 months, respectively. The staining intensity for E-cadherin was also significantly reduced after 3 months in humid conditions (23.4% decrease in comparison to fresh cut) (Fig. 4B).

The Ki-67 immunostaining index showed a mean of 14.7% in fresh-cut breast cancer tissue sections, but this decreased gradually under both dry and wet conditions, with an especially prominent reduction under wet conditions (Fig. 5). In detail, the Ki-67 labeling index of tissue sections stored for 3 months (9% decrease in comparison to fresh cut) in the dry chamber was well maintained, while that after 1 week of storage in humid conditions started to decrease, with particular decrements after 3 months (31.9% decrease in comparison to fresh cut) (Supplementary Table S4).

FIG. 5.

Ki-67 labeling index of human breast cancer tissue for slides stored over time and in dry and wet storage conditions. (A) Representative immunohistochemical images of Ki-67 before and after storage in dry or wet conditions. High magnification images are shown in the inset. Scale bar, 100 μm. (B) Quantitative analysis of the Ki-67 labeling index according to storage conditions. Ki-67 labeling index was defined as the percentage of Ki-67 antigen-positive cells. Data are presented as mean ± SD.

Discussion

In this study, we used breast cancer TMA sections to compare changes in protein and RNA quality under dry and humid storage conditions. As expected, the intensity of IHC staining significantly decreased for all tested markers (ER, PR, HER-2, E-cadherin, and Ki-67) in TMA sections stored long term in humid conditions versus fresh tissue. Immunoreactivity was successfully detected with minimal reduction in staining intensity for all five markers in breast cancer TMA sections stored in dry conditions for 3 months. On the other hand, the RNA integrity of HER-2 was significantly reduced in both dry and humid conditions after 1 week of storage.

We have published previous data suggesting that endogenous and exogenous water play an important role in the loss of antigenicity in mouse FFPE tissues through hydrolysis. These results confirmed that RNA is substantially more sensitive than protein to humid storage conditions. Furthermore, we identified individual analyte specificity by IHC. The staining intensity decreased gradually in all markers as storage time increased, and the dry condition was more effective in preserving antigenicity than the humid condition.

Many factors, including storage time, temperature, humidity, photo-oxidation, exposure to sunlight, and drying, can affect IHC staining quality.^{12,13,29–33} Among these, we focused on storage duration and humidity, because clinical pathology laboratories routinely store unstained slides at room temperature and sometimes expose them to wet conditions. Several previous studies have reported refrigerator storage as a preventer of antigen degradation,^6,29,34 but storing slides at 4°C might be suboptimal due to hydrolysis from exogenous water.¹² Omilian et al reported recently that storage in a desiccator at room temperature was the most effective method to maintain antigenicity.⁹ Our study corroborated this finding, with the immunoreactivity of five markers (ER, PR, HER-2, E-cadherin, and Ki-67) being significantly protected in tissue sections stored in dry conditions in comparison to those in humid conditions. These results support our hypothesis that antigen degradation in FFPE tissue sections is due to hydrolysis during storage.

Because loss of ER and PR expression in tissue sections stored in various storage conditions has been reported, we analyzed the Allred scores and positivity for these two markers. In this study, a significant change of total Allred scores for ER was observed after 3 months of storage (mean ± SD; 6.11 ± 2.23) in dry conditions and after 1 week of storage (mean ± SD; 5.99 ± 2.40) in humid conditions. Total Allred scores for PR significantly changed after 1 week of storage in both conditions (Table 2). However, the positive percentage remained stable for up to 3 months in both storage conditions (Supplementary Table 1 ). This discrepancy between Allred score and positivity for ER and PR might be the consequence of a low threshold for positivity in the semiquantitative Allred score system and high positivity rates for ER (86.2%) and PR (84.0%) on fresh-cut slides.

A prior study reported that the ER-positive rate in breast TMA sections stored for 6 months at room temperature was similar to that of fresh sections, whereas the PR-positive rate was moderately disparate between the two conditions (stored vs. fresh tissue sections; 56.3 vs. 64.1).²⁴ In addition, Tabbara et al reported that ER and PR immunoreactivity were well preserved in air-dried ThinPreps stored for 56 days at room temperature.³⁵ These results suggest that ER and PR status can be evaluated in tissue sections stored at room temperature by IHC staining. Qualitative ER and PR analysis can be performed with specimens stored under dry conditions for a relatively long time. This will provide flexibility in performing ER and PR analysis for retrospective studies.

HER-2 is an invaluable indicator in the management of patients with breast cancer because it is the sole predictive marker for HER-2-targeted therapy. Thus, precise assessment of HER-2 status is an essential step for treatment of patients with breast cancer. However, long-term storage at high temperatures negatively impacts HER-2 antigen stability.^6,29,33 Furthermore, cold storage of tissue sections also results in significant antigenicity loss, despite some benefits in stability.^32,36 Our study shows that HER-2 staining is more stable in dry conditions than in humid conditions, which was confirmed using the ASCO/CAP guideline for HER-2 status evaluation. However, positive (15.6%), equivocal (42.7%), and negative (41.7%) HER-2 expression of fresh tissue sections remarkably changed in dry (11.5%, 34.4%, and 54.2%, respectively) as well as in humid (4.2%, 12.5%, and 83.3%) conditions even at 1 month (Fig. 2).

This significant decrease of positive or equivocal cases is clinically important because the candidates for HER-2-targeted therapy cannot be selected properly. Thus, we suggest that, for HER-2 IHC, it is necessary to establish more stringent storage guidelines, including duration and humidity. Equivocal cases of HER-2 (IHC 2+) require further assessment by fluorescence in situ hybridization. This study confirmed our previous findings that RNA is more susceptible to degradation during tissue processing and storage than is protein, suggesting that RNA-based assays can be more negatively affected in terms of data stability and reproducibility than protein-based methods in basic and clinical research.^1,37

The negative impact of water on human breast tissue section storage has not yet been acknowledged. We revealed previously that protein hydrolysis caused by inadequate dehydration during tissue processing is responsible for reduced immunoreactivity over time.¹² These findings confirm that exogenous water in stored sections plays a critical role in antigen instability. Loss of IHC staining intensity was accelerated in humid conditions compared to dry conditions for all five markers. DiVito et al reported that the antigenicity of cytokeratin, ER, and Ki-67 in stored tissue sections might be protected by combined nitrogen gas storage and paraffin coating storage because oxidation is blocked.¹³

However, paraffin dipping is impractical because it is too difficult to completely remove the paraffin from the slide, and it reduces staining intensity and positivity. Nitrogen gas storage may require specialized equipment as well. Interestingly, van den Broek and van den Vijver reported greater antigen degradation in membrane proteins than in cytoplasmic and nuclear proteins.³² Although loss of HER-2 membrane staining was detected in relatively short storage times (1 week) in dry conditions, E-cadherin membrane staining was stable up to 3 months in dry storage. These results suggest that the quality of IHC staining mainly depends on the characteristics of the target molecule and the IHC protocol, including antibody, detection system, and scoring method.

There are several limitations in this study. First, we did not consider storage temperature or special storage methods, such as paraffin dipping. This is because our initial focus was on storage of unstained slides in the clinical pathology laboratory, excluding complex methods of slide production. Second, because this study was performed using TMA blocks, the results of whole tumor sections were not included. In particular, the intratumoral heterogeneity of HER-2 IHC staining was reported in gastric cancer, so further research on whole tumor sections is needed.^38,39 Third, we stained stored and fresh sections simultaneously using an in-house standard IHC protocol, and digital image analysis with established algorithms was used to analyze the results. Thus, IHC staining results might differ in other laboratories depending on IHC protocol and scoring system.

Fourth, this study did not include detailed clinical or survival data on the patients, so we could not verify the clinical impact of antigenicity loss, such as choice of treatment. Finally, validation of HER-2 gene amplification was not performed in this study. In conclusion, we showed that TMA slides that were vacuum sealed and stored under dry conditions slowly lost antigenicity, but the degree of degradation differed depending on the antigen. Furthermore, RNA degradation is more susceptible compared with proteins even in short-term storage. Notably, the elimination of humidity in the storage method is beneficial in the protection of protein and RNA quantity during section storage. This storage methodology offers the advantage of delaying the loss of immunogenicity over time and can be easily incorporated into clinical practices.

Authors' Contributions

K.K. and J.Y.C. conceived the study and devised the experimental design. K.K., K.Y., C.P., and J.Y.C. performed the experiments. K.K., M.Y.L., J.W.K., J.Y.C., and S.M.H. contributed to interpretation of the results. K.K. and J.Y.C. wrote the article in consultation with S.M.H. All authors provided critical feedback, contributed to the research and writing, and reviewed the final article.

Author Disclosure Statement

The authors declare that there is no conflict of interest. The authors alone are responsible for the content and writing of the article.

Funding Information

This research was supported by the Intramural Research Program of the National Institutes of Health, National Cancer Institute, Center for Cancer Research.

Supplementary Material

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4