The utility of next‐generation sequencing in metastatic prostate cancer FNA biopsies

ABSTRACT

Background

Current American Society of Clinical Oncology guidelines state that patients with metastatic prostate cancer (MPC) should undergo germline and somatic DNA sequencing. The authors examined the utility of next‐generation sequencing (NGS) on fine‐needle aspiration (FNA) biopsies in which NGS was performed on cell block (CB) and/or smears.

Methods

A retrospective review was performed of cytology cases with diagnosis of MPC either before and/or after NGS on FNA material. Clinical and NGS data were obtained from the medical record. Androgen receptor, NKX3.1, INSM1, synaptophysin, chromogranin, Rb, PTEN, and Ki67 immunohistochemical stains were performed on CB if not originally done.

Results

Slides and NGS data were available for 46 MPC FNA biopsies from 45 patients from 2015 to 2024. Metastatic sites included 20 lymph node, 12 liver, five lung, four soft tissue, two pleura, two bone, and one adrenal gland. Ten patients (22%) had change or potential change in therapy based on NGS results. For one patient with poorly differentiated carcinoma previously thought to be urothelial, a TMPRSS2:ERG fusion confirmed prostatic origin. NGS confirmed lung origin for one patient diagnosed initially as metastatic prostatic adenocarcinoma. For one patient, NGS demonstrated TP53 and RB1 mutations, supporting transformation to high‐grade neuroendocrine carcinoma. Change or potential change in therapy was planned for two patients with CDK12 mutations, one with IDH1 mutation, three with BRCA2 mutations, and one with PTEN and TP53 mutations.

Conclusions

NGS on cytology material showed diagnostic and therapeutic utility in a subset of patients, with 10 of 46 patients (22%) having a change or potential in therapy based on NGS results.

Short abstract

The authors examined the use of next‐generation sequencing (NGS) in 46 metastatic prostate cancer fine‐needle aspiration biopsies and found that NGS showed diagnostic and therapeutic utility in a subset, with 10 of 46 patients (22%) having a change or potential change in therapy based on NGS results.

INTRODUCTION

For patients with metastatic prostate cancer (MPC), androgen deprivation therapy (ADT) is a mainstay of treatment. ¹ Despite this therapy, many patients on ADT develop castration‐resistant prostate cancer (CRPC). ¹ Although androgen receptor (AR) signaling drives the majority of these CRPC cases, ² one significant risk is treatment resistance resulting in transformation to high‐grade neuroendocrine carcinoma (HGNEC). Detection of transformation to HGNEC is critical, as HGNEC is treated with platinum chemotherapy. ¹ , ³ Another risk that patients with CRPC face is transformation to so‐called “double negative prostate cancer,” an aggressive carcinoma that lacks both neuroendocrine markers and AR markers. ³

Various morphologic patterns are observed in biopsies from patients with MPC. In addition to conventional acinar morphology, HGNEC, and double negative prostate cancer, a range of intermediary forms with neuroendocrine marker expression and/or high Ki67 proliferation index also exist. ¹ , ² However, there is no established consensus regarding the classification or management of these intermediary cases. ² In an effort to develop a reproducible classification system of the various morphologies that can be observed in the metastatic setting, a working group of genitourinary pathologists, oncologists, and urologists recently recommended the use of an immunohistochemical (IHC) staining panel that includes NKX3.1, AR, synaptophysin, INSM1, and Ki67. ²

In the past, patients with MPC were often not biopsied late in disease. ² However, this practice has shifted in recent years. The current American Society of Clinical Oncology (ASCO) guidelines state that MPC patients should undergo germline and somatic DNA sequencing. ⁴ The ASCO guidelines also state that a metastatic biopsy can be considered when a patient has a significant change in clinical status. ⁴ Next‐generation sequencing (NGS) can be helpful both diagnostically and in detecting therapeutic biomarkers. ² With HGNEC, for example, RB1 and TP53 mutations on NGS support the diagnosis. ³ NGS can also have diagnostic utility in confirming prostate origin for poorly differentiated carcinoma cases, as the detection of a TMPRSS2:ERG fusion is highly specific for prostate. ⁵ , ⁶

Several targetable alterations can be detected using NGS. For example, patients with homologous recombination repair gene mutations, such as BRCA1 and BRCA2, can be treated with the poly(adenosine diphosphate ribose) polymerase (PARP) inhibitor olaparib. ⁷ The PARP inhibitor olaparib is also Food and Drug Administration (FDA)‐approved for metastatic CRPC patients with cyclin‐dependent kinase 12 (CDK12) mutations. ⁸ Microsatellite instability is an FDA‐approved tumor agnostic biomarker for immunotherapy treatment. ⁹ Several clinical trials are currently underway for prostate cancer patients with isocitrate dehydrogenase 1 (IDH1) mutations. ¹⁰ Recent evidence suggests that PTEN (phosphatase and tensin homolog on chromosome 10) status can help predict which patients will respond to immunotherapy. ¹¹

Cytology cell blocks (CB) and smears have been previously shown to be excellent material for NGS testing in many organs. ¹² , ¹³ , ¹⁴ Although several studies have examined the use of IHC in MPC fine‐needle aspiration biopsies, ¹ , ¹⁵ , ¹⁶ , ¹⁷ the use of NGS in MPC FNA biopsies has not been well reported. The goal of this study was to examine the diagnostic and therapeutic utility of NGS in MPC FNA biopsies. If tissue was available, an extensive IHC panel was also performed to further characterize the FNA biopsies as conventional, intermediary, HGNEC, or double negative prostate cancer.

MATERIALS AND METHODS

Under institutional review board approval, patients with FNA biopsies with a diagnosis of MPC either before and/or after our institution’s NGS panel was performed were identified. Patient age, prostate‐specific antigen (PSA) at biopsy, outcome (dead, alive, or lost to follow‐up), and biopsy site data were obtained from the medical record. Therapy history information, including use of ADT, taxane, platinum, androgen receptor signaling inhibitor (ARSI), and radiation therapy information was obtained from the medical record. Clinical notes were reviewed to determine how the NGS results were used.

If not performed originally as part of the clinical work up, the following IHC stains were performed on cell block material, if sufficient tissue remained: NKX3.1, AR, insulinoma‐associated protein 1 (INSM1), synaptophysin, chromogranin, Rb, PTEN, and Ki67. Cell blocks had been prepared using the collodion bag method. ¹⁸

NGS data was obtained from the medical record. The NGS assay at our institution (University of California, San Francisco) targets coding regions of 529 cancer‐related genes, select introns from 73 genes (for detection of structural variants including gene fusions), and the TERT promoter, with a total sequencing footprint of approximately 2.8 mega‐base pairs. ¹⁹ Somatic single‐nucleotide variants and indels were visualized and verified using the Integrative Genomics Viewer. Genome‐wide copy number analysis based on on‐target and off‐target reads was performed by CNVkit and Nexus Copy Number (Biodiscovery, Hawthorne, California). Microsatellite instability analysis was performed with MSIsensor2. ²⁰ Tumor mutational burden was calculated from the number of somatic mutations in the coding regions of the genes on the panel, counting single nucleotide variants and small indels, with exclusion of germline variants. Only pathogenic or likely pathogenic molecular alterations are reported and included in the analysis. For each sample, tumor content, DNA yield (ng/µL resuspended in 100 µL), and average number of reads were recorded. Although a quantity of 250 ng DNA is considered optimal for library preparation and sequencing, a minimum threshold of 50 ng was used.

To assess the impact of high proliferation on overall survival, we performed a univariable Cox proportional hazards model for time to death as a function of Ki67 percentage stratified by the median value, with findings shown graphically by Kaplan‐Meier plot. This analysis was done in R using the survival and survminer packages.

RESULTS

Clinical features

Fifty FNA biopsies from 2015 to 2024 with MPC on which UCSF500 NGS had been performed or attempted were identified. For three cases, aspirate slides and CB were not available, and so the three cases were not included in the study. For one case, NGS was attempted but failed due to insufficient amount of tumor. A total of 46 FNA biopsies were included (Table 1). The FNA biopsies were from 45 patients, with an average age of 72.4 years (SD, 6.5 years). PSA at time of biopsy ranged from 0.02 to 1753 ng/mL, with average 131 (SD, 335) and median 4.6. Of the 45 patients, 21 were alive as of 3–45 months after biopsy, 21 patients were dead of disease, ranging from 0.3 to 36 months after biopsy, and three were lost to follow‐up, ranging from 4 to 12 months after biopsy (Table 1). Nine of the 46 biopsies had a concurrent core needle biopsy; for all nine cases, NGS had been performed on the cytology material, not core material.

TABLE 1.

Status after FNA biopsy and FNA biopsy sites.

FNA biopsy	No.
Status, time range, months
Alive	21	3–45
Lost to follow‐up	3	4–12
Dead of disease	22	0.3–36
Site
Lymph node	20	6 mediastinal, 5 supraclavicular, 3 neck, 3 inguinal/pelvic, 2 axilla, 1 peri‐esophageal
Liver	12
Lung	5
Soft tissue	4
Pleura	2
Bone	2
Adrenal gland	1

Abbreviation: FNA, fine‐needle aspiration.

The biopsy sites included 20 lymph nodes, 12 liver, five lung, four soft tissue, two pleura, two bone, and one adrenal gland. Of the 20 lymph node biopsies, there were six mediastinal, five supraclavicular, three neck, three inguinal/pelvic, two axilla, and one peri‐esophageal (Table 1). A majority of patients had received either ADT (91%), ARSI therapy (67%), or radiation therapy (76%). Prior treatment history data is shown in Table 2.

TABLE 2.

Treatment history.

Treatment history	% cases
ADT	42/46 (91%)
Taxane	16/46 (35%)
Platinum	5/46 (11%)
ARSI	31/46 (67%)
Radiation	35/46 (76%)
No therapy	4/46 (9%)

Abbreviations: ADT, androgen deprivation therapy; ARSI, androgen receptor signaling inhibitor.

Morphological and immunohistochemical results

IHC stains were performed if not done as part of the original workup for the case and if sufficient tissue remained in the CB after NGS. The vast majority (40 of 45, 88.9%) of cases were positive with NKX3.1 (Table 3). Approximately one‐fifth (8 of 42, 19%) showed staining with a neuroendocrine marker (synaptophysin, chromogranin, INSM1). A total of 17 of 40 (42.5%) cases showed PTEN loss, and nine of 40 (22.5%) showed Rb loss (Table 3; Figure 1). The median Ki67 proliferation index was 20%. A Ki67 proliferation index greater than 20% was associated with a 4.1‐fold increased risk of death (95% confidence interval [CI], 1.6–10.6, p = .004) (Figure 2).

TABLE 3.

IHC staining results.

IHC stain(s)	Result	% cases
NKX3.1	Positive	40/45 (88.9%)
INSM1/synaptophysin/chromogranin	Any staining	8/42 (19%)
PTEN	Loss	17/40 (42.5%)
Rb	Loss	9/40 (22.5%)
Ki67	Median of 20%

Note: IHC was performed either as part of the original clinical work up or as part of this study. There were several cases in which insufficient tissue remained after NGS for staining to be performed.

Abbreviations: IHC, immunohistochemical; NGS, next‐generation sequencing.

FIGURE 1.

Cell block of metastatic prostatic adenocarcinoma fine‐needle aspiration biopsy, with H&E stain in (A). The tumor shows both PTEN loss (B) and Rb loss (C).

FIGURE 2.

Ki67 proliferation index on FNA cell block and survival. The median Ki67 proliferation index was 20%. A Ki67 proliferation index >20% was associated with a 4.1‐fold increased risk of death (95% confidence interval, 1.6–10.6, p = .004).

There were four cases in which insufficient tissue remained after NGS for neuroendocrine stains and Ki67 stains to be performed. Of the remaining 42 cases, 11 cases showed conventional acinar morphology with low proliferation index (Ki67 <20%) and no neuroendocrine staining. Of these 11 patients, seven were alive at 3–45 months and four had died of disease at 8–36 months (Table 4).

TABLE 4.

Categorization of FNA cases by immunohistochemical staining results and outcomes.

Category	Definition	No.	Outcome (follow‐up time, months)
Conventional	+NKX3.1, +AR, –SPH, –CHG, –INSM1, Ki67 <20%	11	7 alive (3–45) 4 dead of disease (8–36)
Intermediary	Ki67 ≥20%	20	7 alive (3–42) 2 lost to follow‐up (4–12) 11 dead of disease (1.5–26)
Intermediary	Any SPH, CHG, or INSM1 staining present	7	4 alive (6–21) 1 lost to follow‐up (6) 2 dead of disease (0.3–7)
High‐grade neuroendocrine carcinoma	–NKX3.1, –AR, +SPH, +CHG, +INSM1, elevated Ki67	1	1 alive at 5 months
Double‐negative prostate cancer	–NKX3.1, –AR, –SPH, –CHG, –INSM1	3	3 dead of disease at 2, 6, and 6 months

Abbreviations: CHG, chromogranin; FNA, fine‐needle aspiration; SPH, synaptophysin.

There was one case with transformation to HGNEC, with negative NKX3.1 and AR stains, positive neuroendocrine staining, and Ki67 proliferation index of 80% (Figure 3). This patient was alive at 5 months (Table 4). Three patients showed transformation to so‐called “double negative prostate cancer” with lack of staining with NKX3.1, AR, and neuroendocrine markers (Figure 4). All three of these patients were dead of disease at 2, 6, and 6 months (Table 4).

FIGURE 3.

One patient showed transformation to high‐grade neuroendocrine carcinoma (H&E stain of cell block in A). Tumor cells were negative for NKX3.1 (B), positive for synaptophysin (C), and had a Ki67 proliferation index of approximately 80% (D).

FIGURE 4.

Three cases were so‐called “double negative prostate cancer.” Papanicolaou stain on Thin Prep is shown in (A) and H&E stain of cell block is shown in (B). Tumor cells show loss of staining with NKX3.1 (C), androgen receptor (D), INSM1 (E), and synaptophysin (F).

A subset of patients showed intermediary features, with either Ki67 of 20% or greater (n = 20) or neuroendocrine staining (n = 7). One such example is shown in Figure 5; this patient’s biopsy appeared to show the beginning of a transformation to HGNEC, with a subset of tumor with acinar morphology and a subset with HGNEC morphology. The subset of the tumor with HGNEC morphology was positive with synaptophysin but showed a low Ki67 proliferation index. Survival data for these intermediary patients with Ki67 of 20% or greater or with any neuroendocrine staining is shown in Table 4.

FIGURE 5.

For one patient, the biopsy appeared to show the beginning of a transformation to HGNEC, with a subset of tumor with acinar morphology and a subset with HGNEC morphology (H&E of cell block in A). The subset of tumor with HGNEC morphology was positive with synaptophysin (B) but showed a low Ki67 proliferation index (C). HGNEC indicates high‐grade neuroendocrine carcinoma.

NGS results

NGS was performed on the cell block for 41 cases, smears for four cases, and pooled material from cell block and smear for one case. The number of pathogenic/likely pathogenic mutations ranged from 1 to 24, with a median of 4, and average of 4.7 (SD, 3.5).

NGS results are shown in Table 5. For the purpose of this study, we focused on the clinically relevant alterations seen in prostate carcinoma: TMPRSS2:ERG fusions, androgen axis, cell cycle (RB1, TP53, PTEN, CDK12, and NF2), homologous recombination and DNA damage repair, ubiquitination (SPOP), transcription factors (including FOXA1, FOXO1, FOXP1, MYC, CTNNB1, and SMAD4), GTPase (KRAS), WNT signaling (APC), and oxidative decarboxylation (IDH1).

TABLE 5.

NGS results.

Molecular alteration type	Gene	n
TMPRSS2:ERG	TMPRSS2:ERG	Fusion: n = 11 Rearrangement/structural change: n = 8 Deletion of TMPRSS2: n = 1
Androgen axis	AR	Amplification: n = 11 Point mutations: n = 4
Cell cycle	RB1	n = 7
	TP53	n = 27
	PTEN	n = 21
	CDK12	n = 2
	NF2	n = 1
Homologous recombination/DNA damage repair	BRCA2 MLH1	BRCA2: n = 5 MLH1: n = 1
Ubiquitination	SPOP	n = 2
Transcription factors	FOXA1/FOXO1/FOXP1	n = 10
	MYC	n = 5
	CTNNB1	n = 2
	SMAD4	n = 1
GTPase	KRAS	n = 2
WNT signaling	APC	n = 8
Oxidative decarboxylation	IDH1	n = 1

Abbreviation: NGS, next‐generation sequencing.

There were 20 patients with TMPRSS2:ERG alterations, 11 of whom had fusions, eight with rearrangement/structural change with fusion partner not identified by DNA sequencing, and one with deletion of TMPRSS2 with breakpoint in intron 1 that would be predicted to result in TMPRSS2:ERG fusion by deletion. Fifteen patients had androgen receptor mutations, with amplifications in 11 and point mutations in four. For cell‐cycle genes, there were seven with RB1 mutations, 27 with TP53 mutations, 21 with PTEN mutations, two with CKD12 mutations, and one with NF2 mutation.

Regarding homologous recombination/DNA damage repair genes, five patients had BRCA2 mutations and one had MLH1 mutation. The MLH1 mutation was accompanied by a high MSIsensor2 score of 54.12% and tumor mutation burden of 54.3 mutations/Mb, consistent with a microsatellite unstable (high) tumor. Two patients had SPOP mutations, a gene involved in ubiquitination. Among transcription factors, 10 had FOXA1/FOXO1/FOXP1 mutations, five had MYC mutations, two had CTNNB1 mutations, and one had SMAD4 mutation. Other mutations included two KRAS mutations, eight APC mutations, and one IDH1 mutation (Table 5). A summary of NGS results with cases categorized as either conventional, intermediary, high‐grade neuroendocrine carcinoma, or double negative prostate cancer is shown in Table S1.

Outcome and therapy results

Based on review of clinician notes, there were 10 of 46 (22%) patients with either a change or potential change in therapy based on NGS results (Table 6). There were two patients for whom a change in diagnosis was made due to NGS findings. One patient had been previously diagnosed with poorly differentiated carcinoma presumed to be of urothelial origin. NGS of his biopsy showed a TMPRSS2:ERG fusion, confirming prostatic origin, and he was switched to prostate cancer therapy. For a second patient with a history of lung adenocarcinoma and prostatic adenocarcinoma, his lung FNA biopsy was originally diagnosed as metastatic prostatic adenocarcinoma. However, NGS of his lung biopsy showed a similar profile to his previously sequenced lung adenocarcinoma, with CDKN2A, TP53, KEAP1, and NAV3 mutations. The FNA diagnosis was amended, and the patient was then treated with cyberknife therapy to his lung.

TABLE 6.

Patients with change in diagnosis or change/potential change in therapy per clinician notes based on NGS results.

NGS results	Previous diagnosis	New diagnosis
Two patients with change in diagnosis based on NGS results
TMPRSS2:ERG fusion	Poorly differentiated carcinoma, favor urothelial origin	Metastatic prostatic adenocarcinoma
Lung origin: similar profile to previously sequenced lung adenocarcinoma, with CDKN2A, TP53, KEAP1, and NAV3 mutations	Metastatic prostatic adenocarcinoma	Metastatic lung adenocarcinoma

NGS results	No. of patients	Change/potential change in therapy per clinician notes
Eight patients with change or potential therapy change based on NGS results
TP53, RB1	1	Docetaxel and carboplatin
CDK12	2	1 started olaparib 1 started pembrolizumab
IDH1	1	Plan for use of IDH1 inhibitor ivosidenib
BRCA2	3	1 started PARP inhibitor 2 had future use of olaparib
PTEN/TP53	1	Potential future option of immuno‐oncology therapy AMG757

Abbreviation: NGS, next‐generation sequencing.

For one patient with transformation to HGNEC (Figure 3), NGS supported the diagnosis, with TP53 and RB1 mutations. The patient was switched to docetaxel and carboplatin therapy (Table 6). For two patients with CDK12 mutations, one was started on olaparib and one on pembrolizumab. For the one patient with an IDH1 mutation, the IDH1 inhibitor ivosidenib was planned. For three patients with BRCA2 mutations, one started a PARP inhibitor and olaparib was planned for two. For one patient with PTEN and TP53 mutations, the immuno‐oncology therapy AMG757 was listed as a potential future option.

DISCUSSION

Although historically patients with MPC were not biopsied late in the course of disease, ² this practice has shifted in recent years. The current ASCO guidelines recommend germline and somatic DNA sequencing for MPC patients. ⁴ In addition, patients are often biopsied if conversion to HGNEC is suspected. ² Although the use of IHC in MPC FNA biopsies is well‐described in the cytology literature, ¹ , ¹⁵ , ¹⁶ , ¹⁷ the use of NGS in MPC FNA biopsies is less so. In this study, we examined the utility of NGS in a cohort of 46 MPC FNA biopsies. Based on clinician notes, NGS resulted in either a change in diagnosis or therapy/potential therapy in 22% of patients, with a change in diagnosis for two patients and a change or potential change in therapy for eight patients. However, this percentage is likely an underestimation, as it only includes cases in which there was explicit mention of a specific NGS finding in the clinician’s note. For example, there were five patients with BRCA2 mutations, yet the clinician notes only explicitly mentioned a change or potential change in therapy for three of those patients. In addition, MLH1 mutation and microsatellite instability in one patient would potentially confer consideration for immunotherapy.

Accurate diagnosis in the setting of MPC can be difficult, as numerous morphologic and IHC staining patterns are encountered. In addition to conventional acinar morphology, HGNEC, and double negative prostate cancer, a range of intermediary forms exist. ¹ , ² In this study, if enough tissue remained in the cell block after NGS testing, the IHC staining panel recently described by Haffner et al. ² was performed. In addition to the panel, PTEN and Rb stains were also performed. PTEN and RB1 are tumor suppressor genes, mutations that have been shown to be associated with treatment resistance and decreased survival. ²¹ We successfully showed that the FNA cell blocks can be an excellent source for the IHC panel recommended by Haffner et al. ² Of the 42 cases with sufficient tissue remaining for immunohistochemical staining, 11 were classified as conventional, 27 as intermediary, one as HGNEC, and three as double negative prostate cancer. In addition, PTEN loss was observed in 42.5% and Rb loss was observed in 22.5%. A median Ki67 of 20% was observed, which is the same as has been reported for CRPC biopsies. ² Strikingly, a Ki67 proliferation index higher than 20% was associated with a 4.1‐fold increased risk of death.

In addition to identifying targetable molecular alterations, there are also scenarios in which NGS has diagnostic value by confirming prostatic origin. One such example is high‐grade prostatic adenocarcinoma that morphologically mimics urothelial carcinoma, which can occur in post‐treatment settings. ²² , ²³ NGS can be helpful in such a scenario, as TMPRSS2 fusions are considered pathognomonic of prostatic origin. ²⁴ TERT mutations, on the other hand, are present in approximately 70% of urothelial carcinomas but are extremely rare in prostatic adenocarcinoma. ²³ NGS also has utility in patients with metastatic disease of unknown primary. For example, in one large study ²⁴ that examined the frequency of TMPRSS2:ERG fusions in various tumor types, six tumors that were initially diagnosed as squamous cell carcinoma were found to have TMPRSS2:ERG fusions. The index case of the study was a patient with a large retrovesical mass originally diagnosed as squamous cell carcinoma. Sequencing of the patient’s retrovesical mass, primary prostatic adenocarcinoma, and a metachronous prostatic adenocarcinoma showed a TMPRSS2:ERG fusion in all three, and the patient was started on androgen deprivation therapy. In summary, NGS confirmation for diagnosis should be considered in cases where IHC findings are equivocal, in the post‐treatment setting, and in metastatic carcinomas of unknown primary.

For patients with MPC who are undergoing biopsy, we advocate for concurrent FNA and core biopsies if feasible. Rapid onsite evaluation (ROSE) of FNA biopsies is beneficial both for confirming that the site has been accurately targeted and for appropriately triaging diagnostic material. ¹⁴ An additional advantage of performing both types of biopsy is that if tissue is exhausted from IHC stains on a core biopsy, the FNA biopsy can serve as a backup for ancillary testing such as NGS, ¹⁴ , ²⁵ or vice‐versa. If the CB is inadequate for NGS testing, FNA smears can be used. FNA material can also be used in clinical trials. ²⁵ An additional advantage of FNA biopsies is that they can be performed in sites where cores are typically not performed, such as mediastinal lymph nodes.

Limitations of our study include that it is a retrospective study with a relatively small number of patients, with maximum follow‐up time of 45 months. Additionally, clinical utility was assessed based on explicit documentation in clinician notes, which could have led to underestimation of the true impact of NGS results on therapeutic decision‐making. Last, the study did not compare NGS results from cytology specimens with matched core biopsy specimens.

To our knowledge, however, this is the first study of the utility of NGS in MPC FNA biopsies. Based on clinician notes, NGS resulted in a change or potential change in therapy for 22% of patients. In addition, the CB material was significantly robust that for 42 of 46 cases an extensive IHC staining panel was performed, and cases could be classified as either conventional, intermediary, HGNEC, or double negative prostate cancer. Current ASCO guidelines state that MPC patients should undergo germline and somatic DNA testing. ⁴ This study shows that cytology samples are an adequate source of NGS testing, which has both diagnostic and therapeutic utility. Future directions could include validating these findings in larger, multi‐institutional cohorts to strengthen generalizability. Prospective studies could further evaluate the real‐world impact of FNA‐based NGS testing on treatment outcomes and survival. In addition, as molecular classifications and therapeutic targets evolve, integration of cytology‐based NGS into clinical trials could offer more personalized treatment options for patients with MPC.

AUTHOR CONTRIBUTIONS

Deepika Sirohi: Conceptualization; investigation; methodology; visualization; writing—original draft; writing—review and editing; formal analysis; data curation. Chien‐Kuang Cornelia Ding: Writing—review and editing. Bradley A. Stohr: Writing—review and editing. Ronald Balassanian: Writing—review and editing. Poonam Vohra: Writing–review and editing. Rahul Aggarwal: Writing—review and editing. Emily Chan: Writing—review and editing. Nancy Y. Greenland: Conceptualization; investigation; methodology; visualization; writing—original draft, writing—review and editing; project administration; formal analysis; data curation.

CONFLICT OF INTEREST STATEMENT

Ronald Balassanian reports fees for professional activities from Cerus Corporation. Chien‐Kuang Cornelia Ding reports grant and/or contract funding from Bristol‐Myers Squibb and Intuitive Surgical Operations, Inc. Bradley A. Stohr reports consulting fees from Alessa. The other authors declare no conflicts of interest.

Supporting information

Table S1

Sirohi D, Cornelia Ding C‐K, Stohr BA, et al. The utility of next‐generation sequencing in metastatic prostate cancer FNA biopsies. Cancer Cytopathol. 2025;e70038. doi: 10.1002/cncy.70038

DATA AVAILABILITY STATEMENT

Data that support the findings of this study are available from the corresponding author on reasonable request.