Locally advanced prostate cancer treated with neoadjuvant therapy combined with surgery: a multicenter retrospective cohort analysis

Abstract

Recent data from clinical trials have shown that neoadjuvant therapies significantly improve the pathological outcomes of prostate cancer patients. This study aimed to assess the specific pathological and prognostic effects of these therapies in a real-world, multicenter cohort. Additionally, we explored how factors such as the duration of neoadjuvant therapy and pretreatment imaging modality impact overall treatment outcomes within this therapeutic framework. Data were collected from 407 patients with locally advanced prostate cancer (LAPC) who underwent radical prostatectomy following neoadjuvant therapy. Kaplan-Meier estimates were used to evaluate the four primary clinical endpoints. The log-rank test was used to assess whether significant differences existed between patients grouped according to neoadjuvant therapy duration and pretreatment imaging modality. After a median follow-up period of 36 months, the median progression-free survival (PFS) for the entire cohort was 19 months. An analysis of different durations of neoadjuvant therapy revealed that compared with a 3-month regimen, a 6-month regimen was significantly associated with a greater extent of pathological downstaging and more favorable values for drug response indicators (Pearson test, P = 0.018). Additionally, the 6-month regimen significantly improved the clinical endpoints of PFS (log-rank test, P = 0.0075) and metastasis-free survival (MFS; log-rank test, P = 0.0069). Kaplan-Meier analysis of patients grouped according to preoperative imaging modality revealed that the use of ⁶⁸Ga-labeled prostate-specific membrane antigen-directed positron emission tomography/computed tomography (⁶⁸Ga-PSMA PET/CT) before treatment, as opposed to traditional imaging, led to significant improvements in the clinical endpoints of PFS (log-rank test, P = 0.0059) and radiographic progression-free survival (rPFS; log-rank test, P = 0.016).

INTRODUCTION

Prostate cancer is the second most common cancer in men worldwide.1 In recent years, the incidence of advanced-stage prostate cancer has gradually increased, with locally advanced prostate cancer (LAPC) accounting for approximately 20% of newly diagnosed cases.2 Patients with prostate cancer have high risks of recurrence and metastasis, with biochemical recurrence rates as high as 50% and a 15-year disease-specific mortality rate ranging from 22% to 37%.3,4 LAPC is typically associated with a greater likelihood of positive surgical margins (PSMs), metastasis, and recurrence, leading to a generally poor prognosis.5,6 Consequently, there is an urgent need to explore more effective treatment options for patients with LAPC. According to the latest guidelines from the National Comprehensive Cancer Network (NCCN), a multimodal treatment approach involving radiotherapy, radical prostatectomy (RP), and androgen deprivation therapy (ADT) is recommended for patients with LAPC.7 Given the higher clinical stage of LAPC, the proportion of patients with a PSM, metastasis, and postoperative complications is higher among LAPC patients than among those with earlier stage prostate cancer; thus, LAPC has suboptimal prognostic outcomes.8,9 The addition of neoadjuvant therapy to RP can potentially mitigate the limitations of RP alone in patients with LAPC and increase treatment efficacy.10,11 In terms of efficacy, neoadjuvant therapy has been shown to provide therapeutic benefits, including improved survival rates, across various solid tumors.12 However, the survival benefits of neoadjuvant therapy have yet to be confirmed by large-scale studies, largely owing to the limited scope of relevant studies, the low-risk nature of the study populations, and the late appearance of prognostic markers in prostate cancer patients.13,14 Nonetheless, neoadjuvant therapy has been shown to reduce PSM incidence and lower postoperative pathological staging in LAPC patients.15 With advancements in diagnostic technologies and the development of novel therapeutic agents, exploratory clinical trials have progressively yielded robust clinical insights. Compared with neoadjuvant ADT alone or no neoadjuvant therapy, ADT combined with chemotherapy or androgen receptor signaling inhibitors (ARSIs) significantly improved pathological outcomes and clinical endpoints such as biochemical progression-free survival (bPFS).13,16,17 However, large-scale, evidence-based studies to investigate the overall prognostic outcomes of such multimodal therapies in the real-world settings are lacking. Furthermore, there is an urgent need for standardized frameworks to evaluate the impact of specific diagnostic and therapeutic approaches, particularly regarding patient selection criteria and treatment duration.

Therefore, we performed this retrospective, multicenter cohort study in a real-world setting to evaluate the overall prognostic outcomes with this treatment approach. We also developed relevant prognostic models and explored the associated influencing factors. On the basis of these findings, we further discuss the impact of the duration of neoadjuvant therapy and the role of auxiliary imaging modality during the diagnosis and treatment decision-making on the overall treatment and prognosis of LAPC patients.

PATIENTS AND METHODS

Patients

This study was conducted in accordance with the Declaration of Helsinki of 1975, as revised in 2013, and approved by the Human Research Ethics Committee of Affiliated Drum Tower Hospital, Medical School of Nanjing University (Nanjing, China; Approval No. 2023-047-01). All participants provided written informed consent on admission, authorizing the use of their data from clinical practice and subsequent publication. Data were collected from male patients who received neoadjuvant hormonal therapy or neoadjuvant chemotherapy at four treatment centers (Affiliated Drum Tower Hospital, Medical School of Nanjing University; The First Affiliated Hospital, Fujian Medical University, Fuzhou, China; The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, China; and The First Affiliated Hospital of Guangzhou Medical University, Guangzhou, China) between January 2017 and December 2023. All included patients had complete clinical data and follow-up records that extended over 1 year. Patients who did not subsequently undergo surgery or whose postoperative pathological results indicated small cell carcinoma, neuroendocrine features, or sarcoma were excluded. The final study cohort comprised 407 patients. Clinical tumor-node-metastasis (TNM) staging was performed on the basis of physical examination and pretreatment imaging evaluation (including magnetic resonance imaging [MRI], bone scintigraphy, or ⁶⁸Ga-labeled prostate-specific membrane antigen-directed positron emission tomography/computed tomography [⁶⁸Ga-PSMA PET/CT]), with stage determined according to the American Joint Committee on Cancer (AJCC) 8^th edition staging guidelines.18

Treatment

Patients were diagnosed with LAPC through pretreatment imaging studies (enhanced MRI, bone scintigraphy, or ⁶⁸Ga-PSMA PET/CT). Patients subsequently underwent neoadjuvant chemotherapy (ADT combined with docetaxel) or neoadjuvant hormonal therapy (ADT combined with ARSIs). The duration of neoadjuvant treatment varied among the patients. The frequency distributions of treatment duration in the cohort were as follows: 3 months (128 patients, 31.4%), 4 months (36 patients, 8.8%), 5 months (23 patients, 5.7%), 6 months (213 patients, 52.3%), and more than 6 months (7 patients, 1.7%). After completing their respective cycles of neoadjuvant therapy, patients underwent preoperative imaging to confirm the absence of distant metastases, followed by RP with standard pelvic lymph node dissection (PLND) or extended pelvic lymph node dissection (ePLND).

Survival endpoints and definitions

We used four representative survival parameters to reflect the overall outcomes, including metastasis-free survival (MFS), radiographic progression-free survival (rPFS), castration-resistant prostate cancer-free survival (CRPC-FS), and progression-free survival (PFS). MFS was defined as the time from the start of neoadjuvant therapy until any of the following events: (1) death from any cause; or (2) distant metastasis, including bone, visceral, and nonpelvic (e.g., inguinal and supraclavicular) lymph node metastasis. rPFS was defined as the time from the start of neoadjuvant therapy until any of the following events: (1) death from any cause or (2) distant metastases and local radiographic recurrence. CRPC-FS is defined as the time from the start of neoadjuvant therapy until any of the following events: (1) death from any cause or (2) castration-resistant prostate cancer (CRPC). CRPC is defined as that, despite achieving castrate levels of serum testosterone (usually ≤50 ng dl⁻¹) after ADT, shows a 25% or greater increase in prostate-specific antigen (PSA) levels from the nadir, with an absolute increase of at least 1 ng ml⁻¹, or new radiographic progression. PFS was defined as the time from the start of neoadjuvant therapy until any of the following events occurred: (1) death from any cause; (2) distant metastasis; (3) initiation of salvage or adjuvant treatment, such as radiation, ADT, or anti-androgen therapy; or (4) PSA progression. PSA progression was defined as two consecutive PSA measurements (at least 2 weeks apart) exceeding 0.2 ng ml⁻¹.

A pathologic complete response (pCR) was defined as the absence of residual cancer cells in the prostatectomy specimen. Minimal residual disease (MRD) is defined as a residual tumor with a maximum size of <5 mm in the prostatectomy specimen.

Statistical analyses

Relevant data, including documented clinical notes and imaging reports, were determined on the basis of existing medical records and postoperative follow-up data. An experienced surgeon and a radiologist independently interpreted the imaging reports. The baseline demographic and clinical characteristics of the patients, as well as the pathological outcomes after treatment, are expressed as mean ± standard deviation (s.d.). Categorical variables are presented as counts and percentages. Univariate and multivariate Cox regression analyses were conducted using age, initial PSA level, biopsy Gleason score, initial T stage, initial N stage, pCR or MRD status, pathological margins, pathological T stage, and lymph node metastasis as variables to explore the independent factors influencing the four study endpoints. In the Kaplan-Meier subgroup analysis for the four primary clinical endpoints, pCR/MRD was categorized as either “pCR/MRD” or “no pCR/MRD”. The duration of neoadjuvant therapy was classified as either “3 months” or “more than 3 months”. Imaging methods were divided into “⁶⁸Ga-PSMA PET/CT” or “traditional imaging”, with traditional imaging including MRI and bone scans. Kaplan-Meier survival curves (including those for subgroup analyses) were generated, and the log-rank test was used to compare the survival outcomes between the subgroups. The threshold for statistical significance was set at P < 0.05, and all P values were two-sided. A two-tailed Pearson’s correlation test was used to verify the corresponding effects and influences of different parameters during treatment. All the data analyses were performed using the SPSS software version 22.0 (IBM, Armonk, NY, USA), and R version 4.2.3 (R Foundation for Statistical Computing, Vienna, Austria) was used for data visualization.

RESULTS

Our cohort consisted of 407 patients with a median follow-up of 36 (interquartile range [IQR]: 25–49) months. Baseline characteristics are presented in Table 1, and posttreatment pathological characteristics are detailed in Supplementary Table 1.

Table 1.

Pretreatment patient characteristics in 407 patients

Characteristic	Value
Age (year), median (IQR)	70.0 (65.0–73.0)
>65 years, n (%)	290 (71.3)
≤65 years, n (%)	117 (28.7)
Initial PSA (ng ml−1), median (IQR)	49.0 (19.5–91.0)
Biopsy Gleason score, n (%)
3+3=6	11 (2.7)
3+4=7	30 (7.4)
4+3=7	60 (14.7)
4+4=8	157 (38.6)
4+5=9	87 (21.4)
5+4=9	45 (11.1)
5+5=10	17 (4.2)
TNM stage, n (%)
T3aN0	122 (30.0)
T3bN0	110 (27.0)
T4N0	29 (7.1)
TxN1	146 (35.9)
Pretreatment imaging modality, n (%)
Traditional imaginga	210 (51.6)
68Ga-PSMA PET/CT	197 (48.4)

^aTraditional imaging techniques include enhanced MRI and bone scintigraphy. IQR: interquartile range; PSA: prostate-specific antigen; TNM: tumor-node-metastasis; ⁶⁸Ga-PSMA PET/CT: ⁶⁸Ga-labeled prostate-specific membrane antigen-directed positron emission tomography/computed tomography; MRI: magnetic resonance imaging

Supplementary Table 1.

Pathologic characteristics

Characteristics	Patients (n=407)
Posttreatment outcome
PSA, ng/mL
≤0.1	186 (45.7)
>0.1	221 (54.3)
Pathological outcome
Pathological treatment translation
0	22 (5.4)
1	139 (34.2)
2	219 (53.8)
3	27 (6.6)
pCR	27 (6.6)
MRD	42 (10.3)
pCR or MRD	69 (17.0)
Positive margin	140 (34.4)
Clinical TNM stage
TxN0	296 (72.7)
T0	17 (4.2)
T2	128 (31.4)
T3a	94 (23.1)
T3b	56 (13.8)
T4	1 (0.2)
Positive lymph node metastasis	111 (27.3)
1–3 lymph nodes with metastasis	86 (21.1)
4 or more lymph nodes with metastasis	25 (6.1)

Data are expressed as, n (%). pCR: pathological complete response; MRD: minimal residual disease; Clinical TNM stage: clinical tumor, node, metastasis stage

All patients received standard neoadjuvant therapy, including ADT combined with either docetaxel or first- and second-generation androgen receptor inhibitors. Three patients experienced all-cause mortality, with no cases of cancer-specific death. A total of 161 patients exhibited PSA progression, and these patients had a median PFS of 19 months. Twenty-eight patients developed CRPC, 44 presented clinical recurrence, and 19 had distant metastases. The Kaplan-Meier survival curves for PFS, CRPC-FS, rPFS, and MFS are shown in Figure 1.

Figure 1.

Kaplan–Meier curves of four primary clinical endpoints in the overall cohort. Median PFS was 19 months. (a) PFS, (b) CRPC-FS, (c) rPFS, and (d) MFS. PFS: progression-free survival; CRPC-FS: castration-resistant prostate cancer-free survival; rPFS: radiographic progression-free survival; MFS: metastasis-free survival.

Univariate and multivariate Cox regression analyses were conducted to explore the factors influencing the four study endpoints. The multivariate Cox regression analyses for PFS and rPFS are presented in Supplementary Table 2 and 3. PFS is affected primarily by PSA progression; the most reliable indicators (independent risk factors) for PSA progression-related PFS events were treatment efficacy, pCR/MRD, a PSM, and positive lymph node metastasis. Multivariate Cox regression analysis revealed significant correlations between rPFS and patient age, Gleason score, pretreatment N stage, and postoperative pathological findings of positive lymph node metastasis. Multivariate Cox regression analyses results for CRPC-FS and MFS are provided in Supplementary Table 4 and 5. For CRPC-FS, the initial PSA level at diagnosis and Gleason score were identified as the independent risk factors for CRPC events. In the analysis of the factors potentially influencing MFS, positive lymph node metastasis was identified as an independent risk factor for MFS. However, owing to the low number of positive CRPC and distant metastasis events, there is a possibility of model overfitting and instability of the results.

Supplementary Table 2.

Univariate and multivariate cox regression model for progression-free survival

Factors	Univariate		Multivariate
	HR (95% CI)	P	HR (95% CI)	P
Age	0.994 (0.976–1.013)	0.549	-	-
Initial PSA	1.001 (1.000–1.002)	0.038*	1.001 (0.999–1.002)	0.370
Biopsy Gleasona	1.116 (1.006–1.238)	0.038*	1.066 (0.921–1.234)	0.340
Initial T stagea	1.271 (1.077–1.501)	0.005*	0.968 (0.805–1.164)	0.119
Initial N stage	1.478 (1.147–1.907)	0.003**	1.062 (0.796–1.416)	0.684
pCR or MRD	2.075 (1.388–3.101)	<0.001***	1.649 (1.086–2.504)	0.019*
Pathological margin	3.056 (2.349–3.976)	<0.001***	3.497 (2.327–5.256)	<0.00001***
Pathological T stageb	1.382 (1.211–1.578)	<0.001***	1.172 (1.008–1.362)	0.039*
Positive lymph node metastasisc	1.734 (1.436–2.094)	<0.001***	1.306 (1.034–1.649)	0.025*

*P<0.05; **P<0.01; ***P<0.001. ^aBiopsy Gleason score and T stage (T0, T2, T3a, T3b, and T4) were assigned values of 0, 1, 2, 3, and 4, respectively. ^bPathological T stage (T0, T1, T2, T3a, T3b, and T4) were assigned values of 0, 1, 2, 3, 4 and 5, respectively. ^cPositive lymph node metastasis was categorized as 0 for no metastasis, 1 for 1–3 nodes, and 2 for 4 or more nodes, and these were included in the analysis as continuous variables. HR: hazard ratio; CI: confidence interval; pCR: pathological complete response; MRD: minimal residual disease; PSA: progression-free survival

Supplementary Table 3.

Univariate and multivariate Cox regression model for radiographic progression-free survival

Factors	Univariate		Multivariate
	HR (95% CI)	P	HR (95% CI)	P
Age	0.943 (0.905–0.982)	0.005*	0.937 (0.897–0.978)	0.003**
Initial PSA	1.001 (0.998–1.004)	0.633	1.000 (0.999–1.002)	0.455
Biopsy Gleasona	1.458 (1.139–1.865)	0.003**	1.406 (1.085–1.822)	0.010*
Initial T stagea	1.481 (0.982–2.234)	0.061	1.102 (0.885–1.372)	0.386
Initial N stage	3.599 (1.893–6.843)	<0.001***	2.422 (1.229–4.775)	0.011*
pCR or MRD	8.348 (1.374–2.463)	0.036*	3.333 (0.722–15.389)	0.123
Pathological margin	1.917 (1.035–3.552)	0.039*	1.321 (0.695–2.512)	0.395
Pathological T stageb	1.408 (1.011–1.961)	0.043*	0.900 (0.609–1.332)	0.599
Positive lymph node metastasisc	2.178 (1.483–3.201)	<0.001***	1.705 (1.091–2.664)	0.019*

*P<0.05; **P<0.01; ***P<0.001. ^aBiopsy Gleason score and T stage (T0, T2, T3a, T3b, and T4) were assigned values of 0, 1, 2, 3, and 4, respectively. ^bPathological T stage (T0, T1, T2, T3a, T3b, and T4) were assigned values of 0, 1, 2, 3, 4 and 5, respectively. ^cPositive lymph node metastasis was categorized as 0 for no metastasis, 1 for 1–3 nodes, and 2 for 4 or more nodes, and these were included in the analysis as continuous variables. HR: hazard ratio; CI: confidence interval; pCR: pathological complete response; MRD: minimal residual disease; PSA: progression-free survival

Supplementary Table 4.

Univariate and multivariate Cox regression model for castration-resistant prostate cancer-free survival

Factors	Univariate		Multivariate
	HR (95% CI)	P	HR (95% CI)	P
Age	0.987 (0.933–1.044)	0.650	-	-
Initial PSA	1.004 (1.002–1.006)	<0.001***	1.004 (1.001–1.006)	0.004**
Biopsy Gleasona	1.850 (1.369–2.501)	<0.001***	1.885 (1.337–2.657)	<0.001***
Initial T stagea	1.911 (1.135–3.217)	0.015*	1.184 (0.655–2.142)	0.576
Initial N stage	2.537 (1.186–5.428)	0.016*	1.683 (0.687–4.120)	0.255
pCR or MRD	27.020 (0.383–1904.399)	0.129	-	-
Pathological margin	2.413 (1.141–5.101)	0.021	-	-
Pathological T stageb	1.248 (0.847–1.839)	0.263	-	-
Positive lymph node metastasisc	2.171 (1.371–3.440)	<0.001***	1.516 (0.887–2.590)	0.128

*P<0.05; **P<0.01; ***P<0.001. ^aBiopsy Gleason score and T stage (T0, T2, T3a, T3b, and T4) were assigned values of 0, 1, 2, 3, and 4, respectively. ^bPathological T stage (T0, T1, T2, T3a, T3b, and T4) were assigned values of 0, 1, 2, 3, 4 and 5, respectively. ^cPositive lymph node metastasis was categorized as 0 for no metastasis, 1 for 1–3 nodes, and 2 for 4 or more nodes, and these were included in the analysis as continuous variables. HR: hazard ratio; CI: confidence interval; pCR: pathological complete response; MRD: minimal residual disease; PSA: progression-free survival

Supplementary Table 5.

Univariate and multivariate Cox regression model for metastasis-free survival

Factors	Univariate		Multivariate
	HR (95% CI)	P	HR (95% CI)	P
Age	0.970 (0.907–1.037)	0.371	-	-
Initial PSA	1.001 (0.996–1.005)	0.757	-	-
Biopsy Gleasona	1.521 (1.056–2.189)	0.025*	1.342 (0.872–2.069)	0.181
Initial T stagea	1.549 (0.836–2.869)	0.164	-	-
Initial N stage	6.093 (2.019–18.388)	0.001**	2.876 (0.854–9.682)	0.088
pCR or MRD	26.738 (0.143–4986.172)	0.218	-	-
Pathological margin	3.107 (1.253–7.704)	0.014*	1.562 (0.599–4.076)	0.362
Pathological T stageb	1338 (0.819–2.185)	0.245	-	-
Positive lymph node metastasisc	3.735 (2.171–6.424)	<0.001***	2.544 (1.358–4.766)	0.004**

*P<0.05; **P<0.01; ***P<0.001. ^aBiopsy Gleason score and T stage (T0, T2, T3a, T3b, and T4) were assigned values of 0, 1, 2, 3, and 4, respectively. ^bPathological T stage (T0, T1, T2, T3a, T3b, and T4) were assigned values of 0, 1, 2, 3, 4 and 5, respectively. ^cPositive lymph node metastasis was categorized as 0 for no metastasis, 1 for 1–3 nodes, and 2 for 4 or more nodes, and these were included in the analysis as continuous variables. HR: hazard ratio; CI: confidence interval; pCR: pathological complete response; MRD: minimal residual disease; PSA: progression-free survival

Duration of neoadjuvant therapy

To further explore the impact of the duration of neoadjuvant therapy on the overall treatment outcomes for patients with LAPC, we categorized the different neoadjuvant treatment durations into “3 months” and “more than 3 months”. First, we visualized the changes in pathological stage before and after treatment using bar charts (Figure 2). Compared with shorter treatment duration, a longer neoadjuvant therapy duration had a more significant effect on pathological downstaging, including better downstaging effects on N1 disease and higher organ-confined tumor (ypT2) rates. This may have positive implications for treatment outcomes.

Figure 2.

Changes in tumor stage before and after surgery for patients grouped according to neoadjuvant therapy duration. TNM stage: tumor-node-metastasis stage.

Next, we directly compared the four primary survival endpoints between the two treatment duration-related subgroups; the Kaplan–Meier method was used to fit the survival curves (Figure 3). Notably, there were significant differences in PFS and MFS between patients grouped according to the duration of neoadjuvant therapy (log-rank test, P = 0.0075 and P = 0.0069, respectively).

Figure 3.

Kaplan–Meier curves for the four primary clinical endpoints of patients grouped by the duration of neoadjuvant therapy. Red curves: more than 3 months subgroup; blue curves: 3 months subgroup. The longer-duration subgroup demonstrated a significant advantage in PFS and MFS, with median survival times of 21 months and 15 months for PFS in the more than 3 months subgroup and 3 months subgroup, respectively. (a) PFS, (b) CRPC-FS, (c) rPFS, and (d) MFS. PFS: progression-free survival; CRPC-FS: castration-resistant prostate cancer-free survival; rPFS: radiographic progression-free survival; MFS: metastasis-free survival.

In the context of tumor drug therapy, pCR/MRD serves as an objective indicator of the effectiveness of neoadjuvant drug treatment. In our study cohort (n = 407), 27 patients achieved a pCR, and 42 patients presented MRD. To investigate the impact of neoadjuvant treatment duration on treatment outcomes, we analyzed the pCR/MRD rates between treatment duration-related subgroups. The frequency of pCR/MRD in the 3 months group was 9.7% (11/113), whereas in the more than 3 months group, it was 19.7% (58/294). We also conducted a two-tailed Pearson’s correlation test to assess the relationship between different durations of neoadjuvant treatment and pCR/MRD. The results revealed a significant correlation between the duration of neoadjuvant therapy and the occurrence of the pCR/MRD (Pearson test, P = 0.018). The frequency of a PSM was 41.6% (47/113) in the 3 months group and 31.6% (93/294) in the more than 3 months group, and there was a significant difference between the two subgroups (Pearson test, P = 0.032). In addition, we analyzed the correlation between the occurrence of pCR/MRD and positive lymph node metastasis in the cohort. The detection rate of positive lymph node metastasis in the entire study cohort was 27.0% (110/407) and that in the pCR/MRD group was 13.0% (9/69). A significant correlation was observed between pCR/MRD and positive lymph node metastasis (Pearson test, P = 0.024).

pCR/MRD and clinical endpoints

To provide an overall assessment of the relative prognosis of patients with pCR/MRD, we classified the patient cohort into two groups on the basis of posttreatment response: “pCR/MRD” and “no pCR/MRD”. The log-rank test was used to compare the survival curves between the different subgroups (Figure 4). The results revealed significant survival differences between the pCR/MRD subgroups in terms of PFS, CRPC-FS, and rPFS (log-rank test, P < 0.001, P = 0.015, and P = 0.029, respectively).

Figure 4.

Kaplan–Meier curves for the four primary clinical endpoints of patients grouped by neoadjuvant therapy response. Red curves: no pCR/MRD subgroup; blue curves: pCR/MRD subgroup. The occurrence of pCR/MRD demonstrated significant associations with three primary clinical endpoints. (a) PFS, (b) CRPC-FS, (c) rPFS, and (d) MFS. PFS: progression-free survival; CRPC-FS: castration-resistant prostate cancer-free survival; rPFS: radiographic progression-free survival; MFS: metastasis-free survival; pCR/MRD: pathologic complete response/minimal residual disease.

⁶⁸Ga-PSMA PET/CT in neoadjuvant therapy

⁶⁸Ga-PSMA PET/CT is a prostate cancer-specific radiological imaging method. To investigate the impact of ⁶⁸Ga-PSMA PET/CT on overall treatment outcomes, we divided the patients in the cohort into two subgroups on the basis of the imaging method used: the ⁶⁸Ga-PSMA PET/CT subgroup and the traditional imaging subgroup. We then used the Kaplan-Meier method to fit survival curves for the four primary endpoints and conducted a log-rank test to analyze the differences between subgroups within the survival curves (Figure 5). Compared with preoperative use of traditional imaging, the preoperative use of ⁶⁸Ga-PSMA PET/CT significantly affected PFS and rPFS (log-rank test, P = 0.0059 and P = 0.016, respectively).

Figure 5.

Kaplan–Meier curves for the four primary clinical endpoints of patients grouped by pretreatment imaging modality. Red curves: ⁶⁸Ga-PSMA PET/CT subgroup; blue curves: traditional imaging subgroup. The subgroup undergoing ⁶⁸Ga-PSMA PET/CT imaging demonstrated significant advantages in PFS and rPFS, with median PFS times of 27 months and 14 months in the ⁶⁸Ga-PSMA PET/CT subgroup and traditional imaging subgroup, respectively. (a) PFS, (b) CRPC-FS, (c) rPFS, and (d) MFS. PFS: progression-free survival; CRPC-FS: castration-resistant prostate cancer-free survival; rPFS: radiographic progression-free survival; MFS: metastasis-free survival; ⁶⁸Ga-PSMA PET/CT: ⁶⁸Ga-labeled prostate-specific membrane antigen-directed positron emission tomography/computed tomography.

DISCUSSION

In our large multicenter retrospective cohort study, our systematic analysis of this exploratory treatment approach provided further evidence and discussion of this combined therapy. In the Kaplan-Meier survival curves for the four primary clinical endpoints, the median survival time was only reached for PFS, determined according to PSA progression; the median PFS was 19 months, similar to the 17 months reported by Qian et al.16 Positive lymph node metastasis and PSMs are among the key factors associated with postoperative adverse pathological outcomes. These indicators, along with pCR/MRD status, prompt clinicians to develop more aggressive treatment strategies for improved disease management. The most direct indicator of drug treatment efficacy, pCR/MRD, is increasingly being used as a surrogate short-term oncological endpoint for biochemical recurrence (BCR).19 Further research into whether pCR/MRD can be linked to long-term oncological outcomes, such as MFS, is ongoing, and our cohort data (Figure 4) support this surrogate role to some extent.

In neoadjuvant therapy for prostate cancer, a classic double-edged sword phenomenon is observed in the context of treatment duration. Neoadjuvant therapy often leads to clear pathological responses, improvements in pathological staging and grading, and tumor shrinkage. However, as treatment progresses, the biological characteristics of cancer cells may react adversely, and excessively prolonged therapy can lead to pathological recurrence.20 Several controlled trials have explored the optimal duration of neoadjuvant hormonal therapy (NCH), with mixed conclusions. Selli et al.21 reported that a 6-month NCH regimen significantly improved tumor organ confinement compared with a 3-month regimen. In contrast, Gleave et al.22 reported that the incidence of PSMs increased significantly with longer treatment durations within the 3- to 8-month range. Given that NCH alone does not show significant prognostic benefits, subsequent research in this area remains limited. Current clinical evidence demonstrates that compared with RP alone or monotherapy with either ADT or ARSIs, dual-agent neoadjuvant therapy (ADT combined with chemotherapy or ARSIs) achieves superior outcomes in terms of both pathological phenotypes (e.g., tumor downstaging and margin negativity) and key clinical endpoints (e.g., BFS and MFS).13,16,17

Building upon these foundational studies, we extended this line of inquiry to determine the optimal treatment duration for maximizing therapeutic efficacy in patients with LAPC. In our cohort of patients from four treatment centers, we classified the cases according to treatment durations of “3 months” and “more than 3 months”. The proportion of patients with the most direct indicator of drug efficacy, pCR/MRD, was 9.7% (11/113) in the 3 months group and 19.7% (58/294) in the more than 3 months group. This is comparable to the proportion of patients with pCR/MRD (9.6%) reported in a clinical trial by Qian et al.16 involving 142 patients with LAPC treated with 4 months of neoadjuvant ADT and chemotherapy, supporting the reliability of our results. Additionally, a phase II trial of intensified neoadjuvant hormonal therapy revealed no cases of MRD or positive lymph node metastasis,23 suggesting that sustained pathological responses may be linked to the eradication of micrometastatic foci. We performed a correlation analysis of MRD and positive lymph node metastasis in our cohort and found a significant correlation between pCR/MRD and positive lymph node metastasis (Pearson test, P = 0.024). This finding indicates high consistency in treatment responses between primary tumors and micrometastases, providing valuable information for posttreatment diagnostics and subsequent treatment decision making. This study also highlights the high effectiveness of prolonged neoadjuvant therapy in clearing micrometastases, potentially leading to positive outcomes in LAPC treatment.

From the analysis of the impact of treatment duration (3 months versus more than 3 months) on the four main endpoints (Figure 3), Kaplan-Meier survival curves and log-rank tests revealed significant differences in PFS and MFS between the two treatment duration-related subgroups. For PFS, the median survival time was 15 months for the 3 months group and 21 months for the more than 3 months groups. Notably, a 17-month median survival period with a 4-month regimen was reported by Qian et al.,16 and a linear relationship between treatment duration and short-term oncological outcomes, particularly PFS, was evident within the 3- to 6-month window.

On the basis of the subgroup analyses of PFS and MFS related to treatment duration, we conclude that a longer neoadjuvant treatment period within the 3- to 6-month range is correlated with better survival outcomes. This is likely due to improved pathological responses in primary tumors and more effective eradication of micrometastases.23 In our study, a consistent and positive response to treatment within a specific time frame was demonstrated in terms of various pathological responses, including overall posttreatment pathological downstaging, improvements in PSM rates, increased organ confinement rates, improvements in terms of direct indicators of drug efficacy (pCR/MRD), and lymph node micrometastasis eradication. This contrasts with the negative pathological responses often observed with NCH regimens over time,20 possibly because of the positive effects of combined chemotherapy or intensified hormonal treatment.

Notably, none of the patients in our study experienced distant metastasis during the neoadjuvant treatment period, reinforcing our confidence in the use of longer treatment durations. The majority of the more than 3 months group received a 6-month regimen (213/294); according to the evidence and external data, we conclude that a 6-month neoadjuvant treatment duration offers clear advantages over a 3-month regimen in terms of primary tumor pathological response, micrometastasis eradication, and short-term oncological outcomes such as PFS and MFS. Additionally, we believe that longer neoadjuvant treatment durations may offer potential benefits for overall patient survival.

⁶⁸Ga-PSMA PET/CT is a specialized imaging technique tailored to the biological characteristics of prostate cancer. Its primary advantage over traditional imaging methods is its significantly increased sensitivity and specificity for the detection of small metastases. Clinical studies have shown that ⁶⁸Ga-PSMA PET/CT has a notable advantage over traditional imaging methods (such as enhanced MRI and bone scintigraphy) in detecting small metastatic lesions, with a sensitivity of up to 85%.24,25 ⁶⁸Ga-PSMA PET/CT before surgery have been found to yield more accurate pathological staging results than traditional imaging and thus may benefit patients. However, a clinical consensus has not yet been reached in terms of the necessity of ⁶⁸Ga-PSMA PET/CT in high-risk prostate cancer management, and its potential prognostic impact has not been verified in cohorts or explored through the corresponding mechanisms.

In our study, the use of ⁶⁸Ga-PSMA PET/CT had a significant positive effect on oncological outcomes such as PFS and rPFS. Although distant metastasis events are infrequent, which might affect the model’s fit, ⁶⁸Ga-PSMA PET/CT has the potential to contribute to long-term oncological outcomes, such as MFS. In our clinical cohort, ⁶⁸Ga PSMA PET/CT was more sensitive than traditional imaging in identifying small lymph node metastases. On the basis of the postoperative lymph node dissection results combined with preoperative imaging reports and assuming that no new lymph node metastases developed during treatment, ⁶⁸Ga-PSMA PET/CT achieved a detection rate of 90.0% (45/50) for lymph node-positive patients, whereas traditional imaging achieved a detection rate of 50.8% (31/61). The increased sensitivity in detecting micrometastases provides clinicians with more accurate tools for diagnosis and treatment decision-making. This may lead to more targeted therapeutic approaches, such as omission of surgery, extended neoadjuvant therapy, more aggressive ePLND, and postoperative salvage radiotherapy, thereby improving patient survival outcomes.26

While current knowledge primarily links lymph node micrometastasis with short-term oncological outcomes, such as BCR,27 long-term outcomes remain unclear owing to the need for extended observation and potential treatment biases. Previous clinical trials have shown no significant difference in the long-term oncological outcomes between PLND and ePLND.28 However, these trials did not focus on patients with LAPC who inherently have a higher risk of lymph node involvement and do not undergo neoadjuvant therapy. Thus, the potential synergistic effects of neoadjuvant treatment combined with surgical intervention on lymph node micrometastasis and prognosis are not well understood.

On the basis of these findings, we conclude that a more precise diagnosis of micrometastasis via ⁶⁸Ga-PSMA PET/CT before treatment positively impacts the overall outcomes of patients with LAPC. However, this hypothesis is based solely on clinical data and remains theoretical. More targeted prospective studies with less bias and greater depth are needed to obtain a comprehensive understanding of the necessity of ⁶⁸Ga-PSMA PET/CT in patients with LAPC and the prognostic impact of prostate cancer micrometastases.29

CONCLUSION

Given the suboptimal treatment outcomes of LAPC, alternative approaches that integrate multidisciplinary strategies to improve the overall outcomes of these patients are needed. Neoadjuvant therapy followed by RP is a promising approach for the treatment of LAPC. In this exploratory treatment paradigm, the duration of the neoadjuvant phase and the choice of imaging modality may influence overall therapeutic outcomes. In our cohort study, we found that a 6-month neoadjuvant regimen not only yielded better overall treatment outcomes than a 3-month regimen but also prevented disease progression during treatment. These findings support the efficacy and safety of a longer neoadjuvant course in patients with LAPC. Among different pretreatment imaging modalities, ⁶⁸Ga-PSMA PET/CT showed greater specificity and sensitivity in detecting micrometastases in lymph nodes, which potentially contributed to improved outcomes in terms of PFS, rPFS, and MFS. However, more long-term evidence-based data are needed to clarify comprehensive oncological outcomes with this pioneering treatment approach. Further prospective studies are also needed to elucidate the impact of various treatment-related factors during diagnosis and treatment decision-making, to facilitate refinement of treatment protocols, and to improve therapeutic efficacy.

AUTHOR CONTRIBUTIONS

JLZ proposed the concept of the research, and JHW established the methodology. ZGW, YWW, and WC provided critical resources for data acquisition. SJL, JHW, and YJ collaboratively performed data curation. SHC conducted investigations. YJ carried out formal analysis, followed by validation by SHC, LFX, and XYZ. XFQ and DG were responsible for visualization of the results. SZ and HQG jointly oversaw supervision of the research process. NX and HQG were responsible for project administration. Funding for the study was granted to HQG and JLZ. SJL drafted the initial manuscript, and JLZ led the manuscript review and editing. All authors read and approved the final manuscript.

COMPETING INTERESTS

All authors declare no competing interests.

SUPPLEMENTARY INFORMATION

Statistical considerations

This study focused on exploring the overall impact of neoadjuvant therapy regimens in patients with LAPC. Therefore, cases using different neoadjuvant treatment drugs have been discussed collectively. To minimize the bias that could arise from the use of different neoadjuvant drugs, a two-tailed Pearson correlation test was conducted to compare the pCR/MRD rates, an indicator directly reflecting the pharmacological treatment effect, among cases using different drugs within the same treatment duration. Pairwise comparisons of different drugs and their corresponding treatment responses over the same period revealed no significant differences (Pearson's test, P > 0.2).