Monocyte distribution width in breast cancer: Results from a retrospective exploratory pilot study

Introduction

Breast cancer (BC) is the most common malignancy and the second cause of cancer-related deaths among women worldwide (1). By 2040, over 3 million of BC new cases and 1 million of related deaths are annually predicted as a result of multiple and interconnected factors, including: global population growth, unhealthy lifestyles, genetic predispositions and environmental exposures, among others (2-4). Taking into account tumor load and molecular markers, the therapeutic approach involves a multidisciplinary setting where standard management is lined with targeted and immune therapies (5). Despite the constant therapeutic advances, the 5-year survival rate for advanced BC patients is not greater than 30%, making prevention and early screening crucial for successful recovery (6). Mammography, ultrasonography and magnetic resonance imaging are considered the primary tools for detecting the presence of BC. Nevertheless, circulating biomarkers can offer an attractive alternative for BC screening because of their low cost and noninvasive nature (7). Several biomarkers have been explored over the years, and some of them have been approved in BC, such as carcinoembryonic antigen (CEA), carbohydrate antigen 15-3 (CA15-3), and carbohydrate antigen 27-29 (CA27-29) (8). Regrettably, these markers have shown low diagnostic specificity and sensitivity, while, conversely, they provide useful insights for monitoring disease recurrence or progression, following response to treatment, and even determining targetable mutations to direct therapy (9-11). Hence, there is still a need to discover non-invasive, straightforward and low-risk biomarkers for the screening, diagnosis and management of BC.

Monocyte distribution width (MDW) is a novel hematological parameter provided along with the complete blood count (CBC) by the last generation of DxH hematology analyzers (Beckman Coulter, Inc.) (12). As a direct measure of monocyte volume, MDW reflects the degree of similarity or dissimilarity (anisocytosis) within the cell population. Changes in MDW are emerging as an early marker of innate immune activation in different pathological conditions, including sepsis, viral infections, and other inflammatory-related diseases (13-16). The clinical usage of MDW is currently restricted to sepsis, where its values correlate with the severity of this life-threatening condition (17,18).

A recent study reported alterations in MDW across different stages of chronic liver disease, from hepatitis B to hepatocellular carcinoma, passing through cirrhosis (19). Dynamic upward changes in MDW levels were detected in HBeAg-positive patients, while a positive association was observed with the pathological progression of cirrhosis. Notably, MDW was positively correlated with occurrence and development of hepatocellular carcinoma, providing the first evidence of its role in malignant disorders.

In the light of the aforementioned state-of-the-art, herein we assessed MDW as a potential hematological parameter in BC. Using a retrospective approach, a comparative analysis was carried out between control and BC samples for both MDW and absolute monocyte number (MONO#). Additionally, MDW was also examined with respect to the histological stratification of BC patients.

Materials and methods

Study population and baseline characteristics

A total of fifty-six BC patients admitted at Betania Hospital (Naples, Italy) between October 2023 and February 2024 were retrospectively analyzed in this study at the time of diagnosis. The following exclusion criteria were applied to select the enrolled population: i) male; ii) under 18 years of age; iii) stage IV (metastatic); iv) presurgical treatment such as neoadjuvant chemotherapy. The baseline data required for this study was extrapolated from the medical records, and include sex, age, histological appearance, grade, lymph nodes involvement, lymphovascular invasion and molecular typing (Table I). Similarly, fifty-six whole blood samples matching the following stated exclusion criteria were selected as control: i) male; ii) under 18 years of age; iii) cancer; iv) infection; v) sepsis; iv) systemic inflammatory response syndrome. Only control subjects who matched BC participants in age were selected for this study. This study was approved by the local Ethics Review Board ‘Campania 1’ (Reference Number 19/24 OSS), and conducted in accordance with the local legislation and institutional requirements.

Table I Characteristics of patients with breast cancer.

Table I

Characteristics of patients with breast cancer.

Factors	No. (%)
Age
<66.2	26 (46.43)
≥66.2	30 (53.57)
Histology
Invasive ductal carcinoma	46 (82.14)
Invasive lobular carcinoma	6 (10.71)
Others	4 (7.14)
Surgery
Mastectomy	19 (33.93)
Lumpectomy	35 (62.50)
Unknown	2 (3.57)
Node
N0	29 (51.79)
N1-3	12 (21.43)
NX	7 (12.50)
Unknown	8 (14.29)
Histological grade
I-II	32 (57.14)
III	18 (32.14)
Unknown	6 (10.71)
LV invasion
Positive	36 (64.29)
Negative	3 (5.36)
Unknown	17 (30.36)
ER status
Positive	46 (82.14)
Negative	6 (10.71)
Unknown	4 (7.14)
PR status
Positive	35 (62.50)
Negative	15 (26.79)
Unknown	6 (10.71)
HER2 status
Positive	22 (39.29)
Negative	26 (46.43)
Unknown	8 (14.29)
Molecular subtype
Luminal A	20 (35.71)
Luminal B	22 (39.29)
HER2-enriched	3 (5.36)
TNBC	3 (5.36)
Unknown	8 (14.29)

[i] ER, estrogen receptor; LV, lymphovascular; PR, progesterone receptor; TNBC, triple-negative breast cancer.

Blood analysis and data collection

A complete blood count was performed for both BC patients and controls by DxH900 hematology analyzers (Beckman Coulter, Inc., Brea, California) in compliance with the manufacturer's guidelines. MDW and MONO# data were retrospectively collected and evaluated for statistical purposes in this study.

Statistical analysis

Experimental data were statistically analyzed using GraphPad Prism Software (Version 8.0.2). Box and Whisker Plot were chosen to graphically display the whole data distribution (from the lowest to the highest values passing through median and quartiles). Data distribution was assessed by Anderson-Darling test for both #MONO and MDW. Welch's t-test or Brown-Forsythe and Welch ANOVA were performed to estimate the mean differences between two or more than two pairs of experimental groups when the normality test was passed, respectively. Games-Howell correction was used for multiple comparisons as a post hoc test. Conversely, Mann-Whitney non-parametric test was applied for non-Gaussian distribution datasets. All tests were two tailed, and a P-value of less than 0.05 was considered statistically significant. The area under the curve (AUC) for each of the receiver operating characteristic (ROC) curve is annotated with the 95% confidence interval (CI) by Wilson/Brown. False Discovery Rate (FDR) was calculated by two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli, by setting a Q-value of 0.05 as significant.

Results

Pathological features of BC patients

This study included fifty-six women with BC undergoing either mastectomy (33.93%) or lumpectomy (62.50%). Histologically, the most common subtype was the invasive ductal carcinoma (IDC) (46 out of 56, equal to 82.14%), followed by the invasive lobular carcinoma (ILC) (6 out of 56, 10.71%) and others (4 out of 56, 7.14%). This latter group enclosed all the BC patients that could not be included in the previous two subtypes, such as in-situ and sarcoma. Lymph node spreading assessment displayed a percentage equal to 51.79% (29 out of 56) for the N0 cluster, 21.43% (12 out of 56) for N1-3 and 12.50% (7 out of 56) for NX. The most frequently occurring grade was grade I-II (32 out of 56, 57.14%), while 18 patients showed a higher grading (32.14%). Regarding the molecular typing, 82.14% (46 out of 56), 62.50% (35 out of 56), 39.29% (22 out of 56) of the cases were classified as estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER2) positive, respectively. Luminal A (35.71%) and Luminal B (39.29) were the most frequently occurring molecular subtype, followed by HER2-enriched (5.36%) and triple-negative breast cancer (TNBC) (5.36%) (Table I).

Breast cancer patients exhibit higher MDW than controls

To assess potential differences in monocyte number and volume between BC patients and controls, we retrospectively analyzed MONO# and MDW data from the complete blood count obtained by DxH900 hematology analyzers (Beckman Coulter, Inc., Brea, California).

Initially, a normality test was carried out to define #MONO and MDW distribution in either control or BC patients. Anderson-Darling statistic indicated a Gaussian distribution for both control (P=0.0633) and BC (P=0.2753) group with respect to MDW. Conversely, #MONO did not pass normality test in control (P=0.0013) as well as BC (P<0.0001) samples.

MONO# displayed a median of 0.555x103±0.186 per microliter in the BC patients, while a value of 0.507x103±0.136 per microliter was observed in controls (P=0.3068 by Mann-Whitney test; Q 0.1611) (Fig. 1A). Interestingly, MDW showed a value of 19.41±1.66 in BC patients and 18.14±1.41 in control group (P<0.0001 by Welch's t-test; Q 0.0001) (Fig. 1C). ROC curve analyses provided an AUC of 0.5548 (95% CI: 0.4480-0.6617) for MONO#, while for MDW the same value stood at 0.7136 (95% CI: 0.6195-0.8078) (Fig. 1B and D).

Figure 1 Evaluation of the MONO# and MDW levels in patients with BC. Box and whisker plots illustrating the data distribution of (A) MONO# and (C) MDW in control and BC samples. Related receiver operating curve analysis is shown for (B) MONO# and (D) MDW in patients with BC. ****P<0.0001 (Welch's t-test). The Mann-Whitney test was used for MONO# analysis in (A). The AUC, 95% CI and P-value were determined using the Wilson/Brown method. AUC, area under the curve; BC, breast cancer; MDW, monocyte distribution width; MONO#, absolute monocyte number.

These results indicate no significant difference in MONO# between BC and control groups, while, conversely, MDW levels were higher in BC patients than in controls.

MDW levels correlate with invasive ductal carcinoma expressing estrogen and progesterone receptors

Focusing on the MDW parameter, we subsequently analyzed its performance across the different BC histopathological features. Histological appearance, ER, PR and HER2 data were obtained from biopsy reports and subsequently examined in relation to the MDW levels.

Compared to the control group (18.14±1.41), MDW values of 18.77±2.07 (P=0.8815; Q 0.8696), 19.53±1.76 (P=0.0002, Q 0.0011), and 19.84±2.43 (P=0.5824; Q 0.8697) were recorded in ILC, IDC and others histological subtype, correspondingly (Fig. 2A). As the only significant subset, ROC curve revealed an AUC value of 0.7288 (95% CI: 0.6329-0.8246) for IDC (Fig. 3A).

Figure 2 Investigation of the MDW distribution in patients with BC with different histopathological features. Box and whisker plots displaying MDW levels in controls and patients with BC categorized by (A) histologic, (B) molecular subtype, (C) ER, (D) PR or (E) HER2 status. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001 (Brown-Forsythe and Welch ANOVA). BC, breast cancer; ER, estrogen receptor; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma; MDW, monocyte distribution width; PR, progesterone receptor; TNBC, triple-negative breast cancer.

Figure 3 MDW-related ROC curves in patients with BC with different histopathological features. ROC curve analysis reporting the AUC achieved in (A) IDC, (B) Luminal A and B, (C) ER-positive and (D) PR-positive BC samples. The AUC, 95% CI and P-value were determined using the Wilson/Brown method. BC, breast cancer; ER, estrogen receptor; IDC, invasive ductal carcinoma; MDW, monocyte distribution width; PR, progesterone receptor; ROC, receiver operating characteristic.

Moving to the hormone-related receptors, BC patients showed MDW levels of 19.68±1.76 (P<0.0001; Q 0.0002) when ER was expressed and 18.12±1.57 (P=0.9996; Q 0.6997) in the absence of the latter (Fig. 2C). Remarkably, the AUC obtained from ER-positive ROC curve stood at 0.7504 (95% CI: 0.6588-0.8419) (Fig. 3C). Analogues results were also obtained concerning the PR status because the expression of this receptor was associated with higher MDW (19.90±1.80; P<0.0001; Q 0.0001) with respect to the negative subset (18.35±1.23; P=0.8372; Q 0.2930) (Fig. 2D). ROC analysis indicated good specificity and sensitivity in predicting the PR status by MDW (AUC: 0.7761; 95% CI: 0.6828-0.8693) (Fig. 3D). Concerning HER2, MDW values were 19.51±1.87 (P=0.0077; Q 0.0040) in HER2-positive and 19.55±1.79 (P=0.0030; Q 0.0032) in HER2-negative BC patients (Fig. 2E).

Starting from the ER, PR and HER2 status, the molecular subtype was subsequently addressed in relation to the MDW levels. Luminal A (20.02±1.64) and TNBC (17.89±1.64) exhibited the highest and lowest MDW values, correspondingly (Fig. 2B). Intermediate measures were detected for Luminal B (19.31±1.45) and HER2-enriched (18.36±1.23) instead (Fig. 2B). Statistical analysis also revealed significant differences in mean by comparing both Luminal A (P=0.0010; Q 0.0042) and Luminal B (P=0.0275; Q 0.0578) with controls. Remarkably, ROC curves displayed and AUC of 0.8031 (95% CI: 0.6897-0.9076) and 0.7188 (95% CI: 0.6013-0.8332) in these two subsets (Fig. 3B).

Overall, these findings propose a possible association between MDW and certain histopathological features in BC. Specifically, higher MDW levels were typically found in IDC expressing ERs and PRs. Elevated MDW values were also noted in BC patients presenting either Luminal A or B subtype.

MDW exhibits higher values in BC patients presenting grade III, lymph nodes involvement, and lymphovascular invasion

The prognostic significance of histologic grading has extensively been studied in BC, representing one of two benchmarks for the outcome prediction along with disease stage (20). The TNM system defines disease stage taking into account tumor size, involvement of nearby lymph nodes and dissemination to other parts of the body (21). More recently, additional prognostic factors, such as the lymphovascular invasion (LV), have been included to provide a more comprehensive assessment on BC progression (22,23).

Using the biopsy reports, we later evaluated the MDW levels across the different BC grades and stages, as well as LV invasion. MDW showed value of 19.32±1.64 (P=0.0028; Q 0.0026) in grade I-II and 19.88±2.00 (P=0.0050; Q 0.0026) in grade III, with respect to 18.14±1.41 as median in controls (Fig. 4A). Using the N category as one of hallmarks of the disease stage, 19.35±2.01 (P=0.0234; Q 0.0143) and 19.69±1.63 (P=0.0273; Q 0.0143) were the MDW average obtained in N0 and N1-3 clusters, correspondingly (Fig. 4B). Higher MDW values were also recorded in LV-positive patients (19.62±1.63; P<0.0001; Q 0.0001), while no significant changes were documented in LV-negative with respect to control group (17.93±0.19; P=0.6056; Q 0.2120) (Fig. 4C).

Figure 4 Assessment of the MDW with respect to grade, lymph node involvement, LV invasion, and surgery in patients with BC. Box and whisker plots outlining MDW distribution in controls and patients with BC grouped by (A) grade, (B) lymph node involvement, (C) LV invasion and (D) surgery. *P<0.05, **P<0.01, ****P<0.0001 (Brown-Forsythe and Welch ANOVA). BC, breast cancer; LV, lymphovascular; MDW, monocyte distribution width.

Taken together, these findings suggest that MDW levels could correlate with the progression of both grades and stages in BC. The elevated MDW levels detected in LV-positive BC patients, as well in those undergoing mastectomy (19.55±2.11; P=0.0262; Q 0.0138) rather than lumpectomy (19.37±1.70; P=0.0016; Q 0.0017) (Fig. 4D), support its possible correlation with the BC severity.

Discussion

The prognostic assessment of BC has undergone a profound change over the years, moving this pathological condition from a life-threatening to a potentially curable disease. What made it all possible was the persistent advances achieved in therapeutic field, along with the implementation of screening programs, which have ensured an early detection of BC (24,25). Nevertheless, there is a significant number of BC patients still progressing to the advanced metastatic stage due to tumor recurrence and drug resistance (26,27). Additionally, early detection remains intricate for certain BC subtypes like the TNBC (28,29). Therefore, identifying non-invasive, straightforward and low-risk biomarkers is still pertinent and timely for early detection of BC.

As a novel hematological parameter, MDW has recently been recognized as an effective diagnostic and prognostic biomarker in different pathological conditions, including sepsis, viral infections and other inflammatory-related diseases (13-15,30).

The growing interest in MDW is further supported by the increasing number of available studies over time. Not surprisingly, PubMed database reveals that 483 out of 539 MDW-related items have only been released in the last ten years. Although the vast majority of these findings focus on infectious and inflammatory-related illnesses, additional medical applications are beginning to emerge, particularly in malignant disorders. On that note, a recent study reported alterations in MDW across different stages of chronic liver diseases, demonstrating for the first time a positive correlation between MDW levels and occurrence and development of hepatocellular carcinoma (19).

Herein, we observed no difference in MONO# between BC patients and control group, while conversely, MDW was higher in BC patients than in controls. Although our results suggest that MDW could serve as a potential marker in BC diagnosis, the obtained AUC values highlight a ‘fair leaning towards good’ performance of this test in distinguishing the presence or absence of the disease. Nevertheless, as widely known, labelling systems for AUC are quite arbitrary since strong discriminatory ability is not sufficient to claim a positive effect in clinical practice. The primary purpose of AUC values should be to compare discriminative aptitude of different biomarkers (31). Future investigations should aim to conduct AUC comparative analysis between MDW and other hematological parameters currently employed for BC diagnosis. After all, a combination of different biomarkers has already been recommended to provide a reliable diagnosis in cancer, and thus MDW could be included on the list to further enhance BC detection and stratification (8,32).

With respect to this latter concern, we observed a positive association between MDW levels and some specific BC features, such as histologic subtype and molecular typing. Specifically, higher MDW values were detected in IDC but not in ILC, as well as in BC expressing both ERs and PRs. Conversely, elevated MDW was recorded in both HER2-positive and HER2-negative BC patients compared with controls.

The hypothesis that MDW levels could be related to the ER and PR expression was also suggested by the subtype analysis. Not coincidentally, higher MDW values were detected in Luminal A/B but not in HER2-enriched/TNBC, which do not express or exhibit very low levels of hormone receptors.

The possibility of using a simple blood count test to obtain relevant BC histological features could offer a rapid and non-invasive option to the well-recognized biopsy procedure. Although MDW will never replace biopsy, its assessment could accelerate the therapeutic decision-making processes.

If supported by future data, MDW could also serve as a conceivable predictive/prognostic marker in BC. The obtained data propose, indeed, the existence of a positive correlation between MDW and the progression of both BC grades and stages. Elevated MDW levels were also detected in LV-positive BC patients, reinforcing the possible association with the severity of BC.

Tumor microenvironment (TME) has revealed an active involvement of host immune cells in BC (33). TAMs constitute the most prominent immune infiltrated component of TME in BC, reaching almost 50% of non-cancer cells in some cases (34). They are usually recruited from circulating monocytes, which differentiate in situ via a CCL2-CCR2 chemokine signaling pathway (35,36). M1 and M2 identify two distinct TAM subtypes with competing views on tumor growth and progression, as well as on TME inflammation (33,37). Nevertheless, M1/M2 distinction is somewhat ambiguous because intermediate states have also been recognized in TME (38). Assuming a correlation with TAMs recruitment and differentiation, MDW fluctuations could intuitively reflect remodeling in TME. However, further studies are needed to confirm the mechanisms linking MDW to TAMs in BC, as well as in other kinds of cancer.

However, it is also necessary to mention some limitations of the study. Among the others, special attention should be paid to the sample size of the proposed investigation, as well as to the stated conclusions. Pilot studies constitute a useful analytical approach to test both feasibility and acceptability of proposed trials (39). Due to its preliminary feature, pilot study does not rely on power calculations to determine the appropriate sample size. However, different rules of thumb have been proposed for addressing this issue (40). Typically, a sample size ranging from 12 to 35 per group is applied as regards the continuous outcomes. In accordance with this statement, the International Standard Randomized Controlled Trial Number (ISRCTN) registry showed a median sample size of 30 (Interquartile Range 20-43) across the 592 continuous outcome studies carried out between 2013 and 2020(41). Whilst rules of thumb offer valuable guidance, the optimal sample size should be determined by considering the main purpose of the pilot study. A higher number is demanded to estimate differences among the experimental groups in a pilot approach for instance. Benchmark metric indicates that a sample size between 60 and 100 might be required to assess event rates in an intervention group (42). Despite the selected sample number is slightly below the proposed range, it broadly covers the 9% of the main planned trial, as suggested by Cocks and Torgerson for pilot study (43).

By gathering both the Italian National Statistics Institute (ISTAT) and the Italian Association of Medical Oncology (AIOM) data, the prevalence of BC has been estimated to 3.07% in 2024 (44,45). Applying a confidence interval (CI) equal to 95%, large scale trials should enroll a number of patients between 1941 and 125 to achieve a precision ranging from 0.77 to 2.99%, namely below its prevalence as recommended when the occurrence is not higher than 10% (46). In light of these considerations, the number of samples enrolled in this study confers a degree of precision equal to 4.51%. Exactly the restricted number of enrolled patients, as well as the resulting precision, requires a certain caution in interpreting the study results. Whilst deviations are statistically significant across the experimental groups (as supported by both P-value and Q-value), overestimation or underestimation cannot be excluded precisely because of the limited number of samples.

Moreover, the impact of confounding variables has not been addressed in our analysis, and thus further caution is required in drawing the appropriate conclusions. A very recent study has examined comorbidity affecting MDW pattern in non-infectious (control) group (47). Aside from increasing in cancer patients, MDW elevation was detected in some chronic diseases, such as diabetes, cirrhosis and immune suppression (47). Whilst these findings emphasize the relevance of confounding factors in MDW interpretation, identifying potential biases remains intricate even just at physiological level. Reference interval is still being defined for MDW, and thus changes by either age or gender are not yet understood for instance (48).

In conclusion, we demonstrated for the first time that MDW can serve as a potential hematological biomarker in BC. Higher MDW levels were typically observed in BC patients presenting IDC subtype and expressing either ERs or PRs. A positive correlation between MDW and BC severity was also suggested based on grade, lymph nodes involvement and LV-invasion. However, considering the exploratory pilot nature of the study, large-scale trials like multi-center prospective studies are mandatory to increase statistical robusteness, generalizability, and reproducibility of MDW in BC.

Acknowledgements

Not applicable.

Funding

Funding: The present study was funded by the University of Campania ‘Luigi Vanvitelli’, iRESCUE project.

Availability of data and materials

The data generated in the present study may be requested from the corresponding author.

Authors' contributions

LS and SN were involved in the conceptualization and design of the study. RN, GC and RC collected data. GN and RG curated the data, and confirmed the authenticity of all the raw data. RG also implemented the research methodology. GN handled preparation, creation and presentation of the published data. AR, SK and GP performed statistical analyses. MC and AL contributed to data interpretation. LS wrote the original draft. MC, AL and SN reviewed and edited the manuscript. SN supervised the study. LS acquired funding. All authors have accepted responsibility for the entire content of this manuscript and approved its submission. All authors read and approved the final version of the manuscript.

Ethics approval and consent to participate

The present study complied with all relevant national regulations and institutional policies, and was performed in accordance with the tenets of the Helsinki Declaration. The study approval was granted by the local Ethics Review Board ‘Campania 1’ located at the National Cancer Institute ‘Fondazione G. Pascale’ of Naples (Naples, Italy; approval no. 19/24 OSS). Written informed consent was obtained from all individuals included in the present study.

Patient consent for publication

The patients provided written informed consent for the publication of any data and/or accompanying images.

Competing interests

The authors declare that they have no competing interests.

References