Molecular profiling of breast cancer pleural effusions using cytology specimens: Impact of sample source and cytological features on next‐generation sequencing performance

ABSTRACT

Background

Malignant effusions in breast cancer are a minimally invasive source for molecular profiling. Effusion‐derived cytology specimens, cell blocks (CB) and post‐ThinPrep PreservCyt (PTPC) supernatant, are reliable sources for molecular testing but differ in DNA concentration and turnaround time (TAT). This study evaluates mutational profiles, cytology characteristics, and clinical correlates to assess the utility of effusion‐based molecular testing in breast cancer.

Methods

The authors screened 2400 malignant effusion cases, identifying 22 cases of breast adenocarcinoma effusions that underwent next‐generation sequencing (NGS). Both CB and PTPC supernatant were evaluated for tumor cellularity, fluid volume, DNA concentration, TAT, and mutation detection. Associations with clinicopathological features and outcomes were explored.

Results

NGS was successful in all 22 cases (mean volume, 426.5 mL; mean tumor cellularity, 50%). Successful sequencing was achieved in low volume samples with high tumor cellularity. Higher tumor cellularity (>50%) was associated with bloody effusions (p = .016). PTPC supernatant yielded 5‐fold higher DNA (mean 177.4 ng/µL vs. 33.8 ng/µL) and shorter TAT by 9 days (15 vs. 24 days). TP53 mutations were more common in estrogen receptor (ER)‐negative and triple‐negative breast effusions (p = .025). Mutations in PIK3CA, TP53, CCND1, and AKT were reliably detected. Time from diagnosis to effusion development was significantly shorter in ER negative disease (p = .001)

Conclusions

Effusion cytology using PTPC supernatant, support successful NGS even in low‐volume samples. ER‐negative and triple‐negative effusions demonstrate more aggressive molecular features including enrichment of TP53 mutations and earlier effusion development. Effusion‐based molecular testing yields clinically relevant genomic information in advanced breast cancer.

Short abstract

Malignant pleural effusions in breast cancer provide a minimally invasive substrate for molecular profiling. This study evaluated 22 breast adenocarcinoma effusions from 2,400 screened cases using next‐generation sequencing on both cell blocks and post‐ThinPrep PreservCyt supernatant. Sequencing was successful in all cases despite variable specimen volumes, with higher tumor cellularity correlating with bloody effusions. PTPC supernatant yielded substantially higher DNA concentrations and reduced turnaround time compared with cell blocks. Clinically relevant mutations including TP53, PIK3CA, CCND1, and AKT were consistently detected, with TP53 enrichment in ER‐negative and triple‐negative disease. Overall, effusion‐based testing enables reliable genomic profiling and reveals aggressive molecular features in advanced breast cancer.

INTRODUCTION

Breast cancer is biologically heterogeneous, therefore, molecular characterization is central to prognostication and therapeutic decision‐making. ¹ Although tissue biopsies remain the gold standard for molecular testing, access to adequate tissues can be limited in patients with advanced or metastatic disease, particularly when repeat sampling is required. Malignant serous effusions, most commonly pleural effusions, frequently occur in advanced breast cancer and represent an underused, accessible source of tumor material for ancillary studies. ² , ³

Effusion cytology specimens are routinely processed using liquid based cytology preparations and cell blocks, both of which can be leveraged for molecular testing. Additionally, tumor cells within effusion specimens often preserve their native microenvironment, enhancing the accuracy and reliability of downstream ancillary testing. ⁴ Cell blocks have traditionally been favored due to their architectural resemblance to tissue biopsies and established validation for immunohistochemistry and next‐generation sequencing (NGS). However, variable tumor cellularity, fixation‐related DNA degradation, and prolonged processing times may limit their molecular utility. Post‐ThinPrep PreservCyt (PTPC) supernatant, typically discarded during cytology processing, has emerged as a promising alternative source of high‐quality, cell‐free, and cellular DNA with shorter turnaround times and potentially higher DNA concentration. ⁵ , ⁶

Despite increasing recognition of effusion‐based molecular testing, data correlating cytomorphologic characteristics, specimen type, molecular profiles, and clinical outcomes in breast cancer effusions remain limited, with most of the literature focusing on lung adenocarcinoma. In particular, the impact of tumor cellularity, fluid characteristics, and cytology specimen preparation on NGS performance and clinical relevance has not been well defined. This study aims to address these gaps by evaluating the cytopathologic features of malignant breast cancer effusions, comparing molecular performance between cell blocks and PTPC supernatant, and correlating mutational profiles with hormone receptor status and survival outcomes.

MATERIALS AND METHODS

This study was approved by the institutional review board of University Hospitals Cleveland Medical Center (UHCMC). We screened 2400 malignant effusion samples from 2017 to 2025 using natural language search in the EPIC software. Twenty‐two (22) effusion samples from breast adenocarcinoma that underwent tumor profiling by NGS using either PTPC fluid or from deparaffinized cell block slides were identified. Cases included in the study represent those in which NGS was clinically requested, introducing potential selection bias related to real‐world ordering practices. All effusion samples were pleural fluid. ThinPrep and cell block slides were prepared for all cytology samples and were reviewed by experienced cytopathologists to confirm the diagnosis and summarize the morphologic features. Immunohistochemistry (IHC) was performed by the UHCMC immunohistochemistry laboratory following manufacturers’ recommendations and confirmed metastatic adenocarcinoma of breast primary in all 22 cases. IHC staining for estrogen receptor (ER), progesterone receptor (PR), and Her2Neu was performed using validated antibodies in our laboratory (Table 1). For ER, ranges for interpretation were as follows: invasive carcinoma cells exhibiting greater than or equal to 10% nuclear staining were considered POSITIVE; invasive carcinoma cells exhibiting less than 10%, but greater than or equal to 1% were considered LOW POSITIVE; and invasive carcinoma cells exhibiting less than 1% staining were considered NEGATIVE. For PR, ranges for interpretation were as follows: invasive carcinoma cells exhibiting greater than or equal to 1% nuclear staining were considered POSITIVE; and invasive carcinoma cells exhibiting less than 1% staining were considered NEGATIVE. Each assay was performed using appropriate positive and negative internal controls. ⁷

TABLE 1.

Hormone receptor immunohistochemistry for primary tumor and cytology effusions.

Antibody	Clone	Dilution	Source
ER	SPI	Predilute	VENTANA
PR	IE2	Predilute	VENTANA
HER2	4B5	Predilute	VENTANA

Abbreviations: ER, estrogen receptor; PR, progesterone receptor.

For HER2, ranges for interpretation was as follows: POSITIVE (3+): greater than or equal to 10% tumor cells with intense and uniform staining; EQUIVOCAL (2+): weak to moderate complete immunoreactivity in >10% of tumor cells or circumferential intense staining in less than or equal to 10% of cells; NEGATIVE (1+): faint weak immunoreactivity in >10% of tumor cells, but only a portion of the membrane is positive; ULTRALOW (0+): Faint/barely perceptible, incomplete membrane staining in <10% of tumor cells; and NEGATIVE (0): No immunoreactivity or immunoreactivity in less than or equal to 10% of tumor cells. All tests were performed using a Ventana Pathway HER‐2/neu (4B5) rabbit monoclonal primary antibody. Only invasive carcinoma was evaluated using the ASCO/CAP scoring system. ⁸ External cell culture and tissue controls stain appropriately.

All IHC stains were evaluated by board‐certified pathologists with expertise in breast pathology, in accordance with established reporting guidelines. Each case was interpreted independently, and the number of reviewing pathologists was consistent with standard institutional practice.

In cases where both cytology effusion samples and corresponding primary tumor specimens were available, the IHC from the effusion specimens was assessed independently and without reference to the primary tumor results to avoid interpretive bias. Clinicopathologic parameters, including patient age, tumor size, total Nottingham score, TNM stage, and follow‐up data such as overall survival (OS) and progression‐free survival (PFS), were documented. Effusion fluid characteristics including volume, gross appearance, and microscopic tumor cellularity were assessed through a review of clinical charts and pathology reports and were reviewed by expert cytopathologists. Molecular analysis by targeted NGS was performed with attention to pre‐analytical tumor cellularity, tumor‐to‐nontumor (T:NT) cell ratio, and the resulting mutational landscape, including variant allele frequencies. NGS testing from PTPC has been validated at our institution. ⁹ PTPC and formalin‐fixed, paraffin‐embedded (FFPE) cell block specimens were processed using the standard institutional workflow. After cytocentrifugation, effusion specimens were centrifuged at 1200g for 5 minutes. The supernatant was removed, and the remaining pellet was used to prepare a cell block. Depending on pellet size, equal volumes of plasma and thromboplastin were added to form a clot, which was then formalin‐fixed and processed into an FFPE cell block. The post‐cytocentrifugation supernatant (approximately 30 mL) was stored at 4°C without additives for 7–29 days (median, 14 days) before extraction. Tumor cellularity was estimated by the pathologist using the hematoxylin–eosin‐stained slide from the corresponding FFPE cell block. A minimum of 200 malignant cells after diagnostic IHC was considered adequate for molecular testing. For extraction, PTPC samples were centrifuged at 4500g for 5 minutes, and ∼1 mL of concentrated material was transferred to a microcentrifuge tube and spun again at 10,000g for 5 minutes. The final pellet was resuspended in 50–100 µL of fluid and brought to 200 µL with phosphate‐buffered saline. DNA and RNA were extracted using Maxwell RSC kits (Promega) following cytology‐specific protocols and quantified using the Quantus Fluorometer. All samples yielded measurable nucleic acid, with DNA ranging from 1.06 to 692 ng. For all cases, we used 15 ng of genomic DNA for library preparation with the expanded Solid Tumor Focus Assay, which interrogates 69 solid‐tumor–related genes. We have previously published the components and validation of the panel. ¹⁰ Template preparation was performed on the Ion Chef using HiQ View kits, and sequencing was performed on the Ion GeneStudio S5 System (15–18 samples per 530 chip). Sequencing data were aligned to the hg19 reference genome, and variant calling was performed using Torrent Variant Caller and Ion Reporter (v5.4). All clinically significant variants were reviewed and interpreted by a board‐certified molecular pathologist. Manual review of variants or genomic regions was performed when needed to confirm variant quality or resolve borderline calls.

Mutations classified as variants of unknown significance were also recorded. The panel is designed to detect targeted clinically relevant single nucleotide variants, insertion and deletions (<30 bp), and whole gene copy number amplifications in a select group of genes. OS was calculated from the date of primary diagnosis and the date of effusion onset to the last follow‐up or death. To compare mortality outcomes Pearson’s χ² test and Fisher exact test were used for categorical variables and Welch two‐sample t‐test was used for continuous variables. Time from breast cancer diagnosis to effusion development was compared across tumor cellularity groups, effusion laterality, loculation, and presence of associated pericardial effusion. Fisher exact test was used to determine any association between tumor cellularity and detected molecular alterations.

RESULTS

Patient characteristics and primary diagnosis

All 22 female patients in our cohort, diagnosed with breast carcinoma and pleural effusions underwent molecular profiling of their effusion samples. The median age was 65 years (range, 39–92 years). The cohort included Caucasian (16 of 22, 73%) and African American (6 of 22, 27%) women. Histological analysis revealed that 21 of 22 patients (95%) had invasive ductal carcinoma, whereas one patient was diagnosed with invasive lobular carcinoma. Metastases occurred in multiple sites: lung (10 patients), bone (13), liver (three), and brain or skin (one patient each). The mean greatest tumor dimension was 2.3 ± 2.4 cm.

Effusion and cytological characteristics

Of 22 patients, 20 developed malignant pleural effusions as a secondary manifestation following the initial diagnosis of breast cancer, whereas two patients presented as pleural effusions as initial manifestation. The mean interval between primary diagnosis and effusion development was 98 months (range, 1–313 months). Effusion characteristics are summarized in Table 2. A mean volume of 426.5 mL (range, 8–1200 mL) of pleural fluid was received as a part of routine diagnostic testing and further molecular testing was ordered after appropriate diagnosis. Tumor cellularity was assessed on cell block preparations and was defined as the proportion of tumor cells relative to all other nucleated cells, including reactive mesothelial and inflammatory cells. The mean tumor cellularity (T:NT ratio) was 50% (range, 10%–90%). Based on tumor cellularity, patients were stratified into two groups: ≤50% (n = 12) and those with >50% cellularity (n = 10) (Figure 1). In general, the submitted pleural fluid volume was adequate (>20 mL). One patient had a limited sample volume of 8 mL; however, the high tumor cellularity (90%) permitted successful molecular testing. In contrast, patients with lower tumor cellularity (10%–20%) had larger submitted sample volumes (>50 mL). Patients with higher tumor cellularity (>50%) were significantly more likely to have hemorrhagic effusions compared with those with lower tumor cellularity (66.7% vs. 33.3%, p = .016). Additionally, effusions with >50% tumor cellularity more frequently occurred in patients with triple‐negative breast cancer (TNBC) (83.3% vs. 16.7%, p = .056).

TABLE 2.

Effusion characteristics in this cohort.

Characteristic	N = 22
Laterality, No. (%)
Unilateral	15 (68)
Bilateral	7 (32)
Concurrent pericardial effusion, No. (%)
Yes	6 (27)
No	16 (73)
Concurrent effusions, No. (%)
Yes	2 (9)
No	20 (91)
Bloody effusions, No. (%)
Yes	7 (32)
No	15 (68)
Mean volume of sample	426.5 mL (range, 8–1200 mL)
Mean Tumor cellularity	50% (range, 10%–90%)
ER positivity on effusions, No. (%)	17/22 (77)
PR positivity on effusions, No. (%)	13/22 (59)
Her2Neu status on effusions, No. (%)	0/22 (0)

Abbreviations: ER, estrogen receptor; PR, progesterone receptor.

FIGURE 1.

Tumor cellularity in effusion cytology preparations. (A and B) Representative photomicrographs of ThinPrep and cell block preparations from an effusion sample with tumor cellularity <50%. Numerous background inflammatory cells with scattered tumor cell clusters are present, resulting in a reduced T:NT ratio (original magnification ×200). (C and D) Representative photomicrographs of ThinPrep and cell block preparations from an effusion sample with tumor cellularity >50%. An almost exclusive population of tumor cells is observed, with minimal background inflammatory cells, resulting in an increased T:NT ratio (original magnification ×200). T:NT indicates tumor‐to‐nontumor.

Molecular testing on effusions (PTPC vs, cell block)

At our institution, molecular testing on cytology specimens is performed using two types of sample sources. The first source consists of FFPE cell block–derived unstained slides, which are macro‐dissected to achieve a minimum tumor cellularity threshold of 10%–20%. These slides are then deparaffinized for DNA/RNA extraction and subsequent NGS analysis. The second source is the PTPC fluid, which refers to the post‐centrifugation supernatant remaining after the ThinPrep slide has been prepared from the PreservCyt vial. Molecular profiling was performed on nine FFPE cell blocks and 13 PTPC supernatant samples. DNA concentration from effusion specimens was 127.2 ± 204.0 ng/µL. DNA concentration from the PTPC sample was 5‐fold higher than that from the cell block samples (mean ± SD 177.4 ± 234.0 ng/µL vs. 33.8 ± 81.2 ng/µL), although this difference did not reach statistical significance (p = .062). Additionally, the turnaround time (TAT) from collection of sample to reporting of molecular results was shorter for the supernatant‐based testing (14.5 ± 8.4 days; range, 7–39) compared with cell block–based testing (24 ± 11 days; range, 13–41), although this difference was not statistically significant. The choice between PTPC and cell block material for NGS was not based on tumor cellularity, but rather on institutional practice patterns, specimen availability, and the timing of assay validation. Earlier in the study period, cell block material was routinely used because PTPC had not yet been validated for NGS at our institution. After validation, PTPC became the preferred specimen type due to its practical and technical advantages.

Hormone receptor status in primary tumor versus effusions

Receptor status was evaluated using IHC on both primary surgical specimens and corresponding cell blocks made from effusion samples. In cases with equivocal HER2/neu IHC results, in situ hybridization (ISH) was performed for confirmation. Receptor status was available in 17 primary tumors and all 22 cytology samples. ER positivity was observed in 15 of 17 (88%) primary tumor samples and in 17 of 22 (77%) effusion specimens. PR was positive in 13 of 17 (76%) primary samples and in 13 of 22 (59%) effusion samples. HER2/neu expression was negative in both primary tumors and effusion samples. Five patients were TNBC based on effusion cytology, and two were triple‐negative on primary tumor. Among the five patients classified as triple‐negative on effusion cytology, two were also triple‐negative in their primary tumors. Of the remaining three cases, two lacked primary‐tumor hormone‐receptor data because the effusion was the patient’s initial presentation and receptor testing had not been performed on the breast biopsy. In the final case, the primary tumor was ER‐ and PR‐positive, whereas the effusion was negative for both markers, consistent with possible receptor conversion.

Quantifiable comparison of hormone receptor status between primary and effusion samples was available for nine cases. A general decrease in both the percentage of positive cells and staining intensity was observed in the effusion samples compared with their corresponding primary tumors. These findings, along with the corresponding mutational profiles for the nine cases, are summarized in Table 3.

TABLE 3.

Hormone receptor status of primary versus effusion specimens (available in nine cases).

S/No.	ER in primary	ER in effusions	PR in primary	PR in effusions	Sample source	Mutations
1	>95% strong	90% moderate	>95% strong	30% moderate	Supernatant	PIK3CA p.H1047R, ESR1 p.Y537N
2	52% strong	<1% weak	82% strong	<1% weak	Supernatant	TP53 p.N288Ifs*63, PIK3CA p.E545K
3	>95% strong	40% moderate	90% strong	10% strong	Cell block	PIK3CA p.H1047R
4	>90% moderate to strong	90% moderate	>90% moderate to strong	90% strong	Supernatant	No mutation detected
5	>95% strong	95% strong	40% moderate	1%–5% moderate	Supernatant	PIK3CA p.Q546H CCND1 amplification
6	>95% strong	Focal and weak	40% moderate to strong	NA	Cell block	No mutation detected
7	>95% strong	10% weak	40% strong	10% weak	Supernatant	ESR1 p.Y537S
8	80% strong	90% strong	40% strong	5% weak	Cell block	PIK3CA p.H1047R, ESR1 p.Y537S and STK11 loss
9	>95% strong	NA	Negative	N/A	Supernatant	PIK3CA p.E545K, ESR1 p.D538G

Abbreviations: ER, estrogen receptor; NA, not available; PR, progesterone receptor.

Mutation spectrum of breast cancer effusions

Disease associated genetic mutations were identified in 18 of 22 effusion samples. The most frequent mutation identified in pleural effusion samples from breast cancer patients was PIK3CA present in 10 of 22 cases (45%), followed by TP53 mutations in six cases (27%) and ESR1 mutation in four (18%) cases. Additionally, CCND1 amplification was detected in three cases (14%), AKT mutations in two cases (9%), and single cases exhibited FGFR4 and MYC amplification. One case each had ALK p.K1525delK and ARAF p.H191Y as variant of unknown significance. No cases demonstrated BRAFV600E mutations or PTEN loss. Most patients (10 of 22, 45%) had a single mutation; six of 22 (27%) had two mutations, and one patient had three mutations. Four patients (18%) had no detectable alterations in the genes assessed by the NGS panel. All cases were microsatellite stable. Food and Drug Administration (FDA)‐approved actionable mutations, including PIK3CA, AKT1, and ESR1 were found in 13 patients (59%).

PIK3CA mutations were more frequently observed in effusions with higher tumor cellularity (p < .05), with six of seven evaluable PIK3CA‐positive cases (86%) exhibiting >50% tumor cellularity. TP53 mutations and CCND1 amplifications were also more commonly observed in samples with >50% cellularity, although these associations were not statistically significant. ESR1 and FGFR4 mutations were detected in samples with >50% tumor cellularity, whereas AKT mutations were observed across both intermediate (11%–50%) and high (>50%) cellularity groups. ER‐negative and triple‐negative breast effusions were significantly associated with TP53 mutation (67% vs. 33% p = .025). No other statistically significant associations were identified between mutation profile and hormone receptor status in our cohort.

Survival outcome

We conducted an exploratory analysis of survival outcomes within our cohort. The mean follow‐up of our patients was 110 months (range, 2–314 months). Of 22 patients, 11 (50%) were deceased at the time of analysis. There was no significant age difference between survivors and nonsurvivors (mean ± SD: 63.1 ± 15.5 vs. 66.6 ± 13.3 years). Although deceased patients more frequently exhibited bilateral effusions, and associated pericardial effusions, these associations did not reach statistical significance. No significant correlation was observed between survival and effusion tumor cellularity. Notably, patients who died had a larger tumor size at presentation (2.1 ± 0.6 cm) compared with those who survived (0.7 ± 0.5 cm; p = .011).

CCND1 amplification was observed only in patients who died (p = .05). PIK3CA mutations were more frequent among deceased patients. In contrast, FGFR4 and AKT mutations were observed in patients who survived. TP53 mutation status in effusions was not associated with mortality.

Time‐to‐development of pleural effusion was defined as the interval between primary cancer diagnosis and the first presentation with pleural effusion. The mean time to effusion was 106.3 ± 88.4 months. This interval was significantly shorter in ER‐negative patients compared with ER‐positive patients (19.5 ± 13.4 vs. 117.9 ± 87.8 months; p < .001), regardless of whether hormone receptor status was assessed in the primary tumor or effusion.

At follow‐up, 11 patients (50%) were alive, and five of these patients (45%) received targeted PIK3CA/AKT pathway therapy guided by actionable PIK3CA or AKT1 mutations identified in effusion samples. Among the 11 patients who were deceased at the time of analysis, two with actionable alterations identified on effusion NGS and were offered targeted therapy. One patient was recommended a PIK3CA inhibitor, and the other was offered a CDK4/6 inhibitor (ribociclib). Of these, one patient declined targeted therapy, whereas the other (CDK4/6 inhibitor) received the recommended treatment but subsequently succumbed to disease progression.

DISCUSSION

From our review of 2400 malignant effusion cases, we identified 22 patients with breast cancer who underwent molecular testing on their effusion samples during their disease course. The occurrence of pleural effusion in cancer patients is generally considered a poor prognostic indicator. ¹¹ However, molecular analysis of effusion samples offers valuable opportunities not only for refined prognostication but also for identifying actionable molecular alterations. ¹² Detection of such biomarkers may inform the use of targeted therapies and, potentially influence the clinical management for these patients. ¹³ Molecular testing on effusion samples offers several advantages: it is minimally invasive and allows for repeated sampling, often represents the only available specimen in advanced disease, preserves cellular morphology and architecture in a near‐native state, and enables a wide range of ancillary studies even from limited material. ¹⁴ In this study, effusion‐derived cytology specimens, including both cell blocks and PTPC supernatant, supported successful molecular testing and yielded clinically actionable genomic data.

The Papanicolaou Society of Cytopathology has established guidelines for lung cytology and recommended methods to optimize cytologic specimens for ancillary studies; however, comparable literature addressing breast cytology remains limited. ¹⁵ Considering recent advances in targeted therapy for breast cancer, multigene testing to identify actionable targets has become increasingly important. Consequently, molecular profiling using NGS has gained widespread adoption, emphasizing the need to optimize and preserve limited cytology specimens for ancillary testing. ¹³ , ¹⁶ Cell block specimens are routinely used in cytopathology for both diagnostic and ancillary studies. More recently, collaborations between cytopathology and molecular pathology laboratories have focused on developing innovative strategies to maximize ancillary testing from limited cytology samples. Among these approaches, previously discarded post‐centrifugation cytology supernatant samples have emerged as a viable alternative. ⁹ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ In our study, DNA concentration from our PTPC supernatant samples was approximately 5‐fold higher than that obtained from corresponding cell block specimens; however, this difference did not reach statistical significance, likely due to the limited sample size. Prior studies largely focused on respiratory cytology have reported shorter TAT for molecular testing performed on centrifuged cytology samples compared with cell block specimens. ⁵ , ⁶ , ²⁰ , ²¹ This improvement is attributed to the elimination of deparaffinization and macrodissection steps required for nucleic acid extraction from cell block slides. Our data similarly demonstrated a shorter TAT for PTPC samples compared with CB specimens in breast cancer effusions, although this difference did not reach statistical significance due to limited sample size. The turnaround time reported in our study was calculated from the time of specimen collection (for cytologic evaluation) to the final molecular report, rather than reflecting the analytic time of NGS alone. As such, the reported TAT incorporates several pre‐analytical and post‐analytical factors that contribute to overall duration. These include the time required for complete cytopathologic evaluation, which may involve additional immunohistochemical workup particularly in diagnostically challenging cases such as triple‐negative profiles or when the effusion represents the initial presentation without a known history of breast carcinoma. Additionally, initiation of molecular testing is contingent upon a formal clinical request, which can introduce variability and delay depending on the clinical context. Consequently, these factors collectively contribute to a longer overall TAT. Importantly, despite these variables, supernatant‐based (PTPC) testing consistently demonstrated a shorter TAT compared to cell block‐based analysis. Given the importance of timely clinical decision‐making, supernatant fluid represents a promising alternative substrate for molecular testing

Regardless of the specimen source used for molecular testing, tumor cellularity and adequacy for successful molecular analysis were assessed on cell block sections. Although specimen volume is often considered an important determinant for any downstream ancillary testing, limited fluid volume and admixture of reactive mesothelial and inflammatory cells with tumor cells may compromise analytical performance. ²⁰ , ²¹ Consequently, tumor cellularity may be a more critical factor than volume alone, particularly in small volume specimens. In this study, tumor cellularity was evaluated on cell block preparations and defined as the proportion of tumor cells relative to nonneoplastic, viable nucleated cells, expressed as the T:NT ratio. Successful NGS analysis was achieved in a specimen with a volume as low as 8 mL, attributable to a high tumor cellularity (90%). Low effusion volume should not necessarily preclude molecular testing when tumor cellularity is sufficient, particularly when high. In cases of low tumor cellularity, processing a larger volume of effusion fluid may increase the likelihood of obtaining adequate nucleic acid yield for molecular analysis. Accurate assessment of tumor cellularity is essential, as specimens that appear highly cellular may predominantly consist of reactive mesothelial or inflammatory cells, with relatively few malignant cells.

In addition to assessing tumor adequacy on cell block preparations, evaluation of hormone receptor status represents a critical component of breast cancer characterization and is routinely performed using standardized IHC. Hormone receptor status is a fundamental biomarker in breast cancer, providing essential prognostic and predictive information that directly guides therapeutic decision‐making. Expression of ER and PR reflects underlying tumor biology and identifies patients who are most likely to benefit from endocrine therapy. ²² , ²³ Reassessment of hormone receptor status at metastatic sites, including pleural effusions, plays an increasingly important role in the management of advanced breast cancer, as receptor discordance between primary and metastatic tumors is well recognized and may significantly impact treatment selection. ²⁴ , ²⁵ In our cohort, paired quantitative comparison of hormone receptor expression between primary tumors and corresponding effusion specimens was available in nine cases. We observed a reduction in both the percentage of ER‐ and PR‐positive tumor cells as well as decreased staining intensity in effusion samples compared with the corresponding primary tumors. These findings are consistent with prior reports, including the study by Chen et al., ²⁶ who demonstrated significantly reduced ER (25%) and PR (57%) expression in metastatic lesions in a large breast cancer cohort. Regardless of specimen source, hormone receptor negative breast cancers particularly TNBC are associated with high histologic grade, increased proliferative activity, limited targeted therapeutic options, and enrichment for TP53 mutations. ²⁷ Consistent with these observations, our cohort demonstrated a significant association between TP53 mutations and TNBC identified in effusion specimens.

The most frequently observed mutations in our study was PIK3CA (45%), followed by TP53 (27%) and ESR1 (18%). Additionally, CCND1 amplification was detected in 14% of cases, AKT mutations in 9%, and a single case exhibited FGFR4 and MYC amplification. The detection of actionable mutations in malignant effusions: PIK3CA (45%) and ESR1 (18%) underscores the value of effusion‐based NGS in precision oncology. These biomarkers are clinically meaningful, as they can predict response to PI3K‐targeted therapy such as alpelisib and oral selective estrogen receptor degraders like elacestrant. ²⁸ , ²⁹ , ³⁰ Reinhardt et al. ³¹ reported a somatic PIK3CA mutation frequency of 26.7% with mutations significantly more common in ER‐ and PR‐positive, Her‐2‐negative cases. The frequency of PIK3CA mutations observed in effusions in our cohort was comparable to rates reported in the literature for primary breast tumors. PIK3CA mutations are more frequently identified in hormone receptor–positive breast cancers and are relatively uncommon in TNBCs. ²⁹ Consistent with prior studies, most PIK3CA‐mutated cases in our cohort were hormone receptor–positive. However, PIK3CA mutations were also identified in two of five triple‐negative cases; one of these cases was hormone receptor–positive in the primary tumor but negative in the effusion, whereas the receptor status for the other primary tumor was unavailable. Notably, four of five TNBCs, including the one with PIK3CA cases, also harbored concurrent TP53 mutations, making it unclear whether the presence of PIK3CA mutations is biologically significant in this context or potentially influenced by the coexisting TP53 alterations. The clinical implications of PIK3CA mutations remain controversial; some studies associate these mutations with favorable prognosis, whereas others report poorer outcomes. These discrepancies likely reflect differences in patient populations, disease stages, and methodological approaches across studies. ³² , ³³ , ³⁴ Interestingly, we observed a significant association between PIK3CA mutations and a higher tumor cellularity (>50%), compared to samples with lower tumor cellularity. This may suggest a link between PIK3CA‐driven signaling and increased cellular proliferation or metastatic potential. However, validation in larger, homogeneous cohorts of breast cancer patients with malignant effusions is necessary to confirm these preliminary observations. The overall incidence of TP53 mutations in breast cancer ranges from 20% to 35%. Unlike PIK3CA mutations, TP53 alterations are more frequently observed in nonluminal subtypes and are strongly associated with poor prognosis, therapy resistance, and genomic instability. ³⁵ , ³⁶ , ³⁷ , ³⁸ In our cohort, the incidence of TP53 mutations in effusions was 27%, consistent with previously published data on primary breast tumors. TP53 mutations were significantly associated with ER‐negative and triple‐negative breast cancers in effusion specimens, consistent with prior report in primary tumors. ³⁸

Molecular profiling of breast cancer‐associated effusions remains poorly characterized in the literature. Our review identified only a limited number of studies exploring the molecular landscape of effusions in this context. Notably, Beg et al. ³⁹ analyzed 72 serous fluid samples using NGS, including eight derived from breast cancer patients; however, no tumor‐type‐specific subgroup analysis was performed, and KRAS was the only driver mutation identified in the breast cancer cases. Another group reported enrichment of TP53, PIK3CA, TTN, MUC16, and USH2A in pericardial effusion samples from patients with breast cancer. ⁴⁰ In contrast, our cohort did not demonstrate KRAS, TTN, MUC16, or USH2A mutations.

Survival outcomes following the development of malignant pleural effusion were heterogeneous, with half of the patients alive at the time of analysis. Age and effusion tumor cellularity were not associated with survival. Although bilateral pleural effusions and concomitant pericardial effusions were more common among nonsurvivors, these associations did not reach statistical significance, likely reflecting limited statistical power. Primary tumor size emerged as a significant prognostic factor, with deceased patients presenting with larger tumors at diagnosis, consistent with previously published literature. ⁴¹

Time‐to‐pleural effusion was significantly shorter in ER‐negative patients compared with ER‐ positive patients, regardless of whether hormone receptor status was assessed in the primary tumor or effusion cytology. This finding reinforces the aggressive biology of hormone receptor‐ negative disease and underscores the clinical relevance of receptor reassessment at metastatic sites.

Nearly half of the surviving patients in our cohort received targeted therapy directed against the PI3K/AKT pathway based on actionable alterations identified through molecular analysis of pleural effusion specimens. This observation underscores the clinical utility of effusion‐based molecular testing, particularly in the metastatic setting where tissue biopsies may be limited, infeasible, or pose procedural risk. The ability to derive therapeutically relevant genomic information from cytology specimens highlights the evolving role of pleural effusions as both diagnostic and predictive specimens.

Although constrained by the small cohort size, our results support the integration of comprehensive cytologic evaluation with ancillary molecular testing in malignant effusions to inform prognosis and enable personalized treatment selection in advanced breast cancer. These findings reinforce the expanding role of cytopathology in precision oncology, emphasizing the importance of optimizing specimen handling and molecular workflows to maximize the clinical value of cytology‐derived material.

The limitations of this study include its retrospective design and single institution setting, which may restrict the generalizability of the findings. The NGS assay employed detects targeted, clinically relevant alterations such as single nucleotide variants, small insertions and deletions (<30 bp), whole‐gene copy number amplifications, and gene fusions within a predefined set of genes, but it cannot identify mutations below the assay’s limit of detection or alterations in genes not covered by the panel.

In conclusion, effusion cytology extends beyond its traditional diagnostic role, serving as a minimally invasive avenue for understanding tumor evolution and enabling therapy personalization in metastatic breast cancer. Our findings demonstrate that molecular profiling is feasible on effusion specimens, although adequate tumor cellularity remains crucial for successful sequencing. Supernatant samples yielded higher quantities of genomic material and shorter molecular TAT compared with deparaffinized cell blocks, supporting their growing utility for molecular testing. PIK3CA and TP53 were the most frequently mutated genes identified. Importantly, the detection of actionable FDA alterations enabled several patients to receive targeted therapies, underscoring the potential of effusion based molecular testing to inform precision in advanced disease settings.

AUTHOR CONTRIBUTIONS

Harpreet Virk: Data curation; formal analysis; investigation; writing—original draft; and writing—review and editing. Claire Michael: Writing—review and editing. Keri Ann Pfeil: Methodology. Navid Sadri: Methodology; data curation; validation; writing—review and editing; and software. Philip Bomeisl: Writing—review and editing. Aparna Harbhajanka: Conceptualization; methodology; validation; writing—original draft; writing—review and editing; and supervision.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.