Verification of the Storage Stability of Fecal Hemoglobin in Stool Sampling Bottle Using Improved Buffer - A Pilot Study -

Abstract

Objective:

The temperature and duration of sample storage is known to affect the accuracy of the fecal immunochemical test (FIT). This study compared an improved stool collection buffer with the conventional buffer on the fecal hemoglobin (Hb) correlation and high-temperature stability and simultaneously evaluated the stability of fecal transferrin (Tf) and fecal calprotectin (FC).

Methods:

From June 2021 to December 2024, 130 stool samples collected from patients with informed consent were analyzed. Fecal Hb, Tf, and FC were measured in samples stored in both buffers, followed by correlation and stability assessments. High-temperature stability was evaluated at 37°C for 1 week.

Results:

The improved buffer showed strong correlations with the conventional buffer for three markers (r ≥ 0.95; slope 1.00 ± 0.10). In the high-temperature stability test, residual reactivities were assessed for Hb ≥10 μg/g, Tf ≥10 μg/g, and FC ≥50 μg/g. Hb residual reactivity was significantly higher with the improved buffer (p < 0.01), while Tf maintained high stability (> 90%) in both buffers. FC showed reduced stability in the conventional buffer under prolonged high-temperature exposure, whereas the improved buffer demonstrated better preservation. The improved buffer also maintained higher Hb stability at elevated concentrations (≥30 μg/g).

Conclusions:

The improved stool collection buffer effectively addressed the limitations of the conventional buffer by providing better Hb stability under high-temperature and high-concentration conditions, while preserving excellent stability for Tf and FC. These findings support its potential utility for postal stool collection in FIT-based colorectal cancer screening.

Introduction

Colorectal cancer (CRC) is the third most common cancer worldwide, accounting for approximately 1.9 million new cases and 0.9 million deaths annually[1]. To reduce CRC-associated mortality, population-based screening using the fecal immunochemical test (FIT) has been widely implemented[2-4]. FIT is a test that measures hemoglobin (Hb) concentration in stool. It is noninvasive, simple, and capable of detecting early-stage cancer and precancerous lesions[5-7], thereby contributing to improved prognosis and reduced mortality[8-10]. Although FIT has demonstrated high utility in population screening, its improvement in both diagnostic accuracy and participation rates remains a challenge[11]. In fact, global participation rates remain suboptimal, with less than 50% of eligible individuals undergoing screening in many countries[5-7]. In terms of diagnostic accuracy, several studies have reported that the storage and transport conditions of stool specimens after collection can affect measured values, potentially compromising the reliability of test results[12]. When Hb in stool is stored at room temperature, it deteriorates in a buffer solution used in current stool sampling bottle, resulting in a decrease in measured values, so refrigeration is recommended[13,14]. Moreover, the complexity of stool collection, storage, and sample submission procedures has been shown to directly affect participation rates[15], and simplifying these processes is considered to be crucial for increasing screening uptake.

In recent years, for reasons of convenience, there has been a growing demand for the direct mailing of stool samples to testing laboratories after their collection using stool sampling bottle. With the growing use of FIT globally, stool samples are increasingly being exposed to varying temperatures and durations during transportation. Previous studies have reported that individuals with high fecal Hb concentrations have an increased risk of CRC[16]. Therefore, maintaining the stability of Hb even when present at high concentrations in stool specimens is essential to prevent missed CRC detection in high-risk populations. If a buffer composition can be established that maintains the stability of fecal specimens for extended periods even under high-temperature conditions, it may help to prevent false-negative results owing to Hb degradation, ultimately reducing the global risk of missed CRC diagnoses.

Furthermore, stool specimens collected using the sampling bottles are widely used not only for FIT but also for other clinical tests. Specifically, fecal transferrin (Tf) measurement has been reported to have clinical significance as a complementary marker to FIT, whereas fecal calprotectin (FC) measurement has been clinically applied as a diagnostic and disease activity marker for inflammatory bowel disease (IBD). Because the levels of these markers are also affected by the stability of fecal proteins, they are greatly affected by storage and transportation conditions. Therefore, the use of an appropriate collection buffer is essential for obtaining reliable test results.

The aim of this study was to evaluate the correlation and stability of fecal Hb, Tf, and FC measurements between the conventional buffer and an improved buffer under high-temperature storage conditions using clinical stool specimens.

Methods

Patients and Study Design

From June 2021 to December 2024, stool samples were collected from 156 participants who provided informed consent for the use of their specimens. The inclusion criteria were outpatients and inpatients at Tokyo Medical University Hospital who were scheduled to undergo colonoscopy and were aged 20 years or older, regardless of sex. The exclusion criterion was any patient whom the principal investigator judged to be unsuitable for participation in this study. A stool collection cup containing HEPES-based buffer solution were distributed, and participants collected samples either on the day before or the day of their hospital visit. This HEPES-based buffer is composed of common buffer make-up shared between the conventional and improved buffer and that it does not interfere with either buffer. A total of 142 samples were submitted. After collection, samples were immediately placed in a cooler bag with refrigerants and stored under refrigeration. Upon receipt, the samples were frozen at −80°C, and subsequently sent to Eiken Chemical Co., Ltd. for analysis. Samples were stored at −40°C until testing, and after testing, specimens for research use were securely stored in a locked −80°C freezer.

The fecal concentration of each sample was calculated based on the measured weight of the collection container. The median fecal concentration among the 142 samples was 7.7% (range: 0.2%-21.7%). Twelve samples were excluded owing to insufficient buffer volume (e.g., spillage during sampling) as decreased container weight before sampling, leaving 130 valid samples for analysis (Figure 1). Using the conventional stool collection buffer as a control, fecal markers were measured with both the conventional and improved buffers, and correlation and stability tests were conducted. The improved buffer was designed by Eiken Chemical. The components that contribute to the stabilization of fecal protein markers were replaced with alternative components that exhibit a higher stabilizing effect. Other functional properties remain equivalent to those of the conventional buffer. The buffer composition is proprietary information of Eiken Chemical.

Figure 1.

The diagram of this study-flow.

Measurement samples were prepared by suspending each clinical stool specimen stored in HEPES-based buffer in the conventional or improved buffer to achieve a final fecal concentration of 0.5% of stool. Measurements were performed using the OC-Sensor PLEDIA^Ⓡ analyzer (Eiken Chemical Co., Ltd.) for Hb, Tf, and FC, using the reagents OC-SENSOR FIT Latex Reagent (Eiken), OC-Transferrin Auto^Ⓡ (Eiken), and OC-FCa Reagent (Eiken), respectively.

Correlation Test

Correlation analyses of Hb, Tf, and FC levels were performed between the improved buffer and the conventional buffer as the reference. For correlation analysis, the following cutoff values were used: Hb = 20 μg/g (100 ng/mL), Tf = 10 μg/g (50 ng/mL), and FC = 50 μg/g. Statistical analyses were performed using StatFlex software ver. 7 (Artech Co., Ltd., Osaka, Japan), and standard major axis regression was applied to calculate correlation coefficients, slopes, and intercepts.

Stability Test

Samples prepared with the conventional and improved buffers were incubated at 37°C for 1 week, and the residual reactivity of each fecal marker were calculated as the percentage of the initial value retained after the incubation.

Although the cutoff value for fecal Hb varies among institutions in Japan, thresholds of 20 to 30 μg/g (100-150 ng/mL) are commonly used. Therefore, an additional stability evaluation was performed for samples with a Hb concentration of 30 μg/g or more to assess the stability of Hb at high concentrations.

Statistical Analyses

All statistical analyses were performed using StatFlex software version 7.

For the correlation test, the correlation between the improved and conventional stool buffers was evaluated for fecal Hb, Tf, and FC concentrations using standard major axis regression analysis. The regression slope, intercept, and correlation coefficient (r) were calculated for each marker. Based on clinical cutoff values (Hb: 20 μg/g; Tf: 10 μg/g; FC: 50 μg/g), 2×2 contingency tables were created, and negative, positive, and overall concordance rates were determined.

For the stability test, the residual reactivity of each marker after a 1-week incubation at 37°C was calculated as the percentage of the initial value retained after incubation. Quantitative data were expressed as the median and range, and comparisons between the improved and conventional buffers were performed using the paired t-test. A p-value of less than 0.05 was considered to indicate a statistically significant difference between groups.

Results

Correlation Test

For fecal Hb, using a cutoff value of 20 μg/g, 23 samples were positive using the conventional buffer (including 11 samples ≥ 200 μg/g), whereas 24 samples were positive using the improved buffer. Only 3 samples showed discrepant results (Table 1). The results of the improved buffer demonstrated a strong correlation with those of the conventional buffer (slope = 0.944, intercept = 0.359, correlation coefficient = 0.993). The negative, positive, and overall concordance rates were 98.1%, 95.7%, and 97.7%, respectively (Figure 2).

Table 1.

Correlations between the Conventional Stool Buffer and the Improved Stool Buffer for Positive/Negative Determinations of Hb (Cutoff Value: 20 μg/g).

Hb		Conventional buffer
		(-) ≦20	(+) >20	(μg/g)
Improved buffer	(-)	105	1	106
	(+)	2	22	24
		107	23	130

Figure 2.

Scatter plot of the correlation between the conventional stool buffer and the improved stool buffer for Hb measurements.

For fecal Tf, using a cutoff value of 10 μg/g, 14 samples were positive using the conventional buffer and 15 using the improved buffer (Table 2). The results of the improved buffer showed a strong correlation with those of the conventional buffer (slope = 0.979, intercept = 3.600, correlation coefficient = 0.998), with negative, positive, and overall concordance rates of 99.1%, 100%, and 99.2%, respectively (Figure 3).

Table 2.

Correlations between the Conventional Stool Buffer and the Improved Stool Buffer for Positive/Negative Determinations of Tf (Cutoff Value: 10 μg/g).

Tf		Conventional buffer
		(-) ≦10	(+) >10	(μg/g)
Improved buffer	(-)	115	0	115
	(+)	1	14	15
		116	14	130

Figure 3.

Scatter plot of the correlation between the conventional stool buffer and the improved stool buffer for Tf measurements.

For FC, a cutoff of 50 μg/g was used to distinguish IBD from functional disorders. Seventy-four samples were positive using the conventional buffer, and 75 were positive using the improved buffer, with only 3 discrepant cases (Table 3). When using a higher cutoff of 300 μg/g for activity assessment in IBD, both buffers identified 24 positive samples (Table 3). The results of the improved buffer demonstrated a strong correlation with those of the conventional buffer (slope = 0.917, intercept = 11.618, correlation coefficient = 0.983). The negative, positive, and overall concordance rates were 96.4%, 98.6%, and 97.7%, respectively (Figure 4).

Table 3.

Correlations between the Conventional Stool Buffer and the Improved Stool Buffer for Positive/Negative Determinations of FC (a; Cutoff Value: 50 μg/g, b; Cutoff Value: 300 μg/g).

a.		Conventional buffer
		(-) ≦50	(+) >50	(μg/g)
Improved buffer	(-)	54	1	55
	(+)	2	73	75
		56	74	130

b.		Conventional buffer
		(-) ≦300	(+) >300	(μg/g)
Improved buffer	(-)	106	0	106
	(+)	0	24	24
		106	24	130

Figure 4.

Scatter plot of the correlations between the conventional stool buffer and the improved stool buffer for FC measurements.

Stability Test

The stability of each marker was assessed after 1 week of storage at 37°C for samples exceeding the following thresholds: Hb ≥ 10 μg/g (n = 18), Tf ≥ 10 μg/g (n = 14), and FC ≥ 50 μg/g (n = 73). Over-range samples, which measured value exceeded the measurement range for each marker (Hb: > 200 μg/g, Tf: > 250 μg/g and FC: > 2720 μg/g) were excluded.

For Hb, the residual reactivity after 1 week of storage was significantly higher using the improved buffer than using the conventional buffer (86% vs. 64%, p < 0.01), confirming the enhanced stability under high temperature condition using the improved buffer (Figures 5 and 6a).

Figure 5.

Comparison of the stability of Hb, Tf, and FC in stool samples stored at 37˚C for 1 week between the conventional and improved buffers. (* p<0.01)

Figure 6a, b, c.

Detailed analysis of Hb, Tf and FC stability in stool stored at 37˚C for 1 week.

For Tf, no significant difference in residual reactivity was observed between the 2 buffers; however, both maintained high stability (> 90%) after incubation (Figures 5 and 6b).

For FC, the residual reactivity was markedly improved using the improved buffer compared with using the conventional buffer (105% vs. 58%, p < 0.01), again demonstrating enhanced stability using the improved buffer under high-temperature conditions (Figures 5 and 6c).

To further assess Hb stability, positive samples were divided into the following 2 groups: < 30 μg/g (n = 8) and ≥ 30 μg/g (n = 10). After 1 week at 37°C, there was no significant difference between buffers in the < 30 μg/g group. However, regarding buffers in the ≥ 30 μg/g group, the improved buffer showed a higher residual reactivity than the conventional buffer, indicating enhanced Hb stability in the medium- to high-concentration ranges (Figure 7).

Figure 7.

Comparison of the stability of Hb in stool with varying Hb concentrations, stored at 37˚C for 1 week between the conventional and improved buffers. (* p<0.01)

Discussion

The enhancement of cancer screening programs is regarded as a key component of cancer control. Among the available modalities for CRC screening, FIT has well-established efficacy in reducing CRC mortality. However, persistently low participation and follow-up rates remain major public health challenges. Introducing a year-round postal turn in system for FIT specimens could be a crucial step toward improving participation rates in CRC screening programs.

Postal-based FIT in CRC screening has already been utilized in various European countries and regions[17]. In the United States, the Centers for Disease Control and Prevention has also published detailed guidelines for designing, implementing, and evaluating mailing programs aimed at increasing screening uptake[18]. In Japan, however, the Ministry of Health, Labour and Welfare has issued a notice stating that “mailing of specimens from screening participants to implementing institutions is, in principle, not allowed, as maintaining appropriate temperature control is difficult and test accuracy may be compromised”[19]. Nevertheless, the Colorectal Cancer Screening Manual (Revised 2021 Edition), published in May 2022[20], acknowledges that “Improvements to stool collection buffer solutions have been reported to enhance the stability of Hb in stool samples at room temperature. Therefore, the feasibility of seasonal mailing collection (excluding summer) as a means to increase screening participation requires continued discussion.” Considering these developments, enabling year-round postal submission―including during the summer―could represent an effective strategy to overcome a major barrier to CRC screening participation[21].

In the present study, the improved stool collection buffer demonstrated strong correlations (r ≥ 0.95) and regression slopes within 1.00 ± 0.10 for Hb, Tf, and FC compared with the conventional buffer. Although a few discrepancies were observed (3 samples for both Hb and FC), all occurred near the respective cutoff thresholds (Hb 20 μg/g; FC 50 μg/g) and within approximately 10% of the measured values, suggesting that these may have been owing to sampling variability or minor analytical deviations in the low-concentration ranges.

After storage at 37°C for 7 days, no stool samples had Hb concentrations below the cutoff value of 20 μg/g in either buffer. For Tf, 1 sample from each buffer became negative, whereas for FC, 25 samples became negative with the conventional buffer versus only 4 with the improved buffer, suggesting enhanced stability of FC in the improved buffer. Importantly, the improved buffer showed significantly higher stability for Hb across all concentrations under high-temperature conditions, overcoming a long-standing limitation of FIT. These results indicate that the improved buffer may be useful for year-round postal handling of FIT specimens.

Storage temperature and storage time have long been recognized as key determinants of FIT accuracy. Previous studies have reported that short-term refrigeration preserves Hb most effectively[22]. Refrigerated FIT specimens can maintain Hb levels equivalent to fresh samples for more than 2 weeks; however, compliance with refrigeration recommendations is limited, with only about two-thirds of participants adhering to proper storage practices[23]. This underscores the need for more robust buffer systems capable of maintaining sample stability without strict temperature control.

A large study of 20,371 participants in the French national CRC screening program found no significant associations between mailing time, season (including summer), and clinical performance of the FIT, such as positivity rate, detection rate, and positive predictive value. However, in a stability test using 247 FIT-positive specimens, fecal Hb levels decreased by 5.1% after 7 days at 20°C, and by 20.5% after 7 days at 30°C. The proportion of samples turning negative increased from 10.0% at 20°C to 23.8% at 30°C. Notably, whereas samples remained relatively stable during the first 4 days at 30°C (−4.5%), a marked additional decline (−18.5%) occurred between days 5 and 7. Even among CRC-positive samples, 22.0% (2/9) turned negative after 7 days at 30°C. These findings suggest that whereas mild mailing delays may have minimal clinical impact, high-temperature environments (≥ 30°C) significantly accelerate Hb degradation and increase the risk of false negative FIT results[24]. Similar temperature-dependent reductions in Hb concentration have been consistently reported at 20°C or greater[25-27].

In this study, Hb concentrations in samples using the conventional buffer decreased by approximately 20% after 7 days at 37°C, which was consistent with previous findings. By contrast, samples in the improved buffer maintained more than 80% of their Hb concentrations in both the < 30 μg/g Hb and the ≥ 30 μg/g Hb groups, demonstrating its significantly superior stability under high-temperature and high-Hb concentration conditions.

FC is a calcium-binding protein abundant in neutrophils, and its concentration increases in proportion to neutrophil migration into the intestinal lumen during inflammation. Although FC is considered relatively stable at room temperature[28], prior studies have reported a decrease in FC concentrations over time, with significant reductions observed after 2 days in extracted stool stored at ambient conditions[29]. Tf, in our study, showed no significant difference in stability between the 2 buffers, however, for FC, consistent with previous reports, a decrease in stability was observed in the conventional buffer after prolonged exposure to high temperatures. In contrast, a significant improvement in stability was confirmed with the improved buffer.

Collectively, these results suggested that reliable diagnostic measurements of Hb, Tf, and FC may be achievable even in samples stored at 37°C for 7 days.

An additional advantage of stabilizing all 3 markers (Hb, Tf, and FC) within a single buffer system is that it eliminates the need to divide specimens, enabling multiple assays to be performed using a limited stool volume. This approach can reduce participant burden by minimizing the number of collections and submissions required, while also decreasing sampling error inherent to multi-sample collection. Consequently, this may lower the risk of inconsistent or erroneous test results while simultaneously simplifying logistics and reducing costs.

Several limitations of this study should be acknowledged. First, this was a pilot study with a relatively small sample size, and hence larger-scale studies are required to validate generalizability. Second, sampling variability associated with stool consistency could not be completely eliminated, and may have contributed to some discordant cases. Third, as this was a single-center study, reproducibility under different institutional or environmental conditions remains unverified. Fourth, this study focused on analytical stability rather than diagnostic sensitivity, specificity, or clinical outcomes. In addition, the detailed composition of the improved buffer was not disclosed because it is proprietary

Conclusion

The improved stool collection buffer successfully addressed the key limitations of the conventional buffer, demonstrating superior Hb stability under high-temperature and high-Hb concentration conditions, as well as favorable stability and correlation for Tf and FC. The ability to stably measure multiple markers in a single specimen under increased temperatures has substantial clinical value. Overcoming the temperature-dependent instability of fecal Hb not only facilitates the feasibility of postal CRC screening throughout all seasons, but also represents a significant advancement toward expanding access and improving participation in CRC screening. Future multicenter and large-scale studies are warranted to further substantiate these findings and establish a robust scientific foundation for the nationwide implementation of postal FIT-based CRC screening.

Author Contributions

Substantial contributions to conception and design, acquisition of data: Fukuzawa, Enomoto, Takehara, Hosaka, Itoi.

Acquisition of data: Naito, Kono, Yamauchi, Morise, Uchida, Kagawa, Muramatsu.

Analysis and interpretation of data: Fukuzawa, Enomoto, Takehara.

Drafting of manuscript: Fukuzawa, Enomoto, Takehara.

Critical revision: Hosaka, Itoi.

Ethics Approval

This study was conducted as a collaborative research project between Tokyo Medical University Hospital and Eiken Chemical Co., Ltd. The study protocol was reviewed and approved by the institutional ethics committees of both organizations (Tokyo Medical University Hospital: study approval no. T2020-0114; and Eiken Chemical Co., Ltd.: study approval no.: 87-028).

Declaration of Conflicts of Interest

The authors declare that they have no conflicts of interest associated with this study.

Informed Consent

Written informed consent was obtained from all participants included in this study.

Funding Statement

This research was conducted without any specific grants from funding agencies in the public, commercial, or not-for-profit sectors.