Visual Abstract
Abstract
Our previous study found that the prostate-specific membrane antigen (PSMA) PET/CT response of primary prostate cancer (PCa) to neoadjuvant therapy can predict the pathologic response. This study was designed to investigate the association between [68Ga]PSMA PET/CT changes and biochemical progression-free survival (bPFS) in high-risk patients who underwent neoadjuvant therapy before radical prostatectomy (RP). Methods: Seventy-five patients with high-risk PCa in 2 phase II clinical trials who received neoadjuvant therapy before RP were included. The patients received androgen deprivation therapy plus docetaxel (n = 33) or androgen deprivation therapy plus abiraterone (n = 42) as neoadjuvant treatment. All patients had serial [68Ga]PSMA PET/CT scans before and after neoadjuvant therapy. Age, initial prostate-specific antigen level, nadir prostate-specific antigen level before RP, tumor grade at biopsy, treatment regimen, clinical T stage, PET imaging features, pathologic N stage, and pathologic response on final pathology were included for univariate and multivariate Cox regression analyses to identify independent predictors of bPFS. Results: With a median follow-up of 30 mo, 18 patients (24%) experienced biochemical progression. Multivariate Cox regression analyses revealed that only SUVmax derived from posttreatment [68Ga]PSMA PET/CT and pathologic response on final pathology were independent factors for the prediction of bPFS, with hazard ratios of 1.02 (95% CI, 1.00–1.04; P = 0.02) and 0.12 (95% CI, 0.02–0.98; P = 0.048), respectively. Kaplan–Meier analysis revealed that patients with a favorable [68Ga]PSMA PET/CT response (posttreatment SUVmax < 8.5) or a favorable pathologic response (pathologic complete response or minimal residual disease) had a significantly lower rate of 3-y biochemical progression. Conclusion: Our results indicated that [68Ga]PSMA PET/CT response was an independent risk factor for the prediction of bPFS in patients with high-risk PCa receiving neoadjuvant therapy and RP, suggesting [68Ga]PSMA PET/CT to be an ideal tool to monitor response to neoadjuvant therapy.
Patients with high-risk prostate cancer (PCa) have a significant risk of biochemical recurrence and distant metastases after radical prostatectomy (RP) (1), despite the standard therapies of external-beam radiation therapy in combination with long-term androgen deprivation therapy (ADT) and RP plus extended lymph node dissection (2). Though neoadjuvant therapy before RP for patients with high-risk PCa still remains investigational, results from phase II trials have indicated a favorable pathologic response to neoadjuvant ADT combined with new-generation androgen receptor pathway inhibitors (3–8) or docetaxel chemotherapy (9,10).
A favorable pathologic response, defined as a pathologic complete response (pCR) or minimal residual disease (residual tumor ≤ 0.5 cm), has been widely applied as the primary endpoint to evaluate the efficacy of neoadjuvant therapy (3,4,6,8). A significant correlation between pCR and improved long-term oncologic outcomes has been verified in breast (11) and bladder cancer (12). However, whether a favorable pathologic response after neoadjuvant therapy can translate to better long-term oncologic outcomes, such as progression-free and metastasis-free survival, remains unclear in patients with high-risk PCa.
Prostate-specific membrane antigen (PSMA)–based PET is a promising technique for both initial staging (13) and restaging of biochemical recurrence (14,15). Recently, accumulative evidence also indicated that [68Ga]PSMA PET/CT could be applied to monitor response in patients receiving systematic therapies (16–19). Our previous study demonstrated that [68Ga]PSMA PET/CT performed better than prostate-specific antigen (PSA) in diagnosing a pathologic response to neoadjuvant ADT plus abiraterone, with SUVmax being an independent predictive factor for a favorable pathologic response (19). However, the relationship between [68Ga]PSMA PET/CT changes and oncologic outcomes in neoadjuvant settings remains unknown.
To investigate the potential relationship between [68Ga]PSMA PET/CT response and biochemical progression, this study analyzed patients with high-risk localized or locally advanced PCa treated with neoadjuvant therapy and RP who, in 2 clinical trials, had serial [68Ga]PSMA PET/CT scans before and after neoadjuvant therapy.
MATERIALS AND METHODS
Study Design and Participants
Our center conducted 2 phase II clinical trials (NCT04356430 and NCT04869371) that were designed to investigate the efficacy of neoadjuvant ADT plus docetaxel or abiraterone for patients with high-risk localized or advanced PCa. Patients who met the following criteria were included: ADT plus docetaxel or abiraterone as neoadjuvant treatment; serial [68Ga]PSMA PET/CT scans before and after neoadjuvant therapy; and at least 12 mo of follow-up since randomization, unless biochemical progression had occurred. The exclusion criterion was any adjuvant treatment (ADT or radiotherapy) after RP or persistence of PSA after RP (PSA > 0.2 ng/mL 8 wk after RP). Finally, 75 patients were included for analysis; the study flowchart is shown in Figure 1. The study was approved by the Ethics Committee of the Drum Tower Hospital (2019-214 and 2020-314), and all patients provided written informed consent. Clinical covariates including initial age, PSA level during each visit, preoperative clinical T stage, and International Society of Urological Pathology grade at biopsy were documented.
Study flowchart with excluded patients and reason for exclusion.
Treatment Intervention
Patients received 2 different therapies: ADT plus docetaxel or ADT plus abiraterone according to previously published protocols (3,9). Briefly, ADT was applied by a luteinizing hormone–releasing hormone analog every 12 wk. The ADT-plus-docetaxel group was additionally administered docetaxel, 75 mg/m2 of body surface area, every 3 wk for 6 cycles plus 5 mg of prednisone acetate twice a day. The ADT-plus-abiraterone group took an additional 1,000 mg of abiraterone acetate and 5 mg of prednisone acetate orally once a day. After 6 mo of neoadjuvant therapy, the participants underwent robot-assisted RP and extended lymph node dissection. The median interval between the first PET/CT scan and the initiation of neoadjuvant therapy was 8 d (interquartile range [IQR], 6–12 d), the median interval between the initiation of neoadjuvant therapy and the second PET/CT scan was 165 d (IQR, 157–179 d), and the median interval from the second PET/CT scan to surgery was 9 d (IQR, 4–11 d).
Follow-up and Outcomes
PSA and testosterone levels were assessed every 4 wk during neoadjuvant treatment, 2 d before RP, and every 4 wk after RP. Biochemical progression was defined as a postoperative serum PSA level greater than 0.2 ng/mL on 2 separate occasions at minimally 2-wk intervals (20). Biochemical progression-free survival (bPFS) was defined as the time from randomization to biochemical progression or death.
PET/CT Imaging Acquisition and Evaluation
[68Ga]PSMA-11 PET/CT scanning was performed 1 h after intravenous injection of [68Ga]PSMA-11. With a uMI 780 PET/CT scanner (United Imaging Healthcare), a CT scan (130 keV, 80 mAs) and a static emission scan were performed from the vertex to the proximal legs, corrected for dead time, scatter, and decay (19). [68Ga]PSMA-ligand PET/CT images were reviewed by 2 nuclear medicine physicians with over 10 y of reading experience in the interpretation of PSMA-targeted PET. Lesions were delineated by higher uptake than background by a RadiAnt DICOM viewer (version 2022.1.1; Medixant). The PSMA intensity of the lesions was measured as the SUVmax in the delineated area. For patients with multiple lesions, the one with the highest SUVmax was recognized as the index tumor and recorded. For patients with no obvious PSMA uptake after neoadjuvant therapy, SUVmax was determined at the location of the same tumor as found on the first scan, by comparing the anatomic position through other tissues such as bladder or bone and excluding respective normal organs that demonstrate high uptake as part of normal biodistribution, including the bladder. Twelve patients did not have any obvious uptake on follow-up scans, with a median SUVmax of 3.06 (IQR, 2.23–3.37). The median time frame between 2 subsequent scans was 179 d (IQR, 169–188 d). The change in SUVmax between the 2 scans was defined as the SUVmax decline percentage, which was calculated by .
Whole-Mount Histologic Imaging and Pathologic Response
After robot-assisted RP, a whole-mount histologic sample was fixed and stained as previously described (19,21). To obtain the final pathologic result, all whole-mount histology slides were subsequently digitalized by a scanning system (NanoZoomer Digital Pathology) and interpreted by 2 dedicated genitourinary pathologists masked to clinical information. Residual tumors in the posttreatment surgical resection specimen were determined from the bidimensional diameters of the primary tumor bed as previously described (22). A favorable pathologic response was defined as pCR or as minimal residual disease whose largest cross-sectional dimension was less than 5 mm (22). Pathologic T stage, lymph node metastasis, and a positive margin were also recorded.
Statistical Analysis
Continuous nonnormally distributed variables were reported by median and IQRs. Univariable and multivariable Cox regression was applied to identify factors associated with clinical outcomes. The cutoff for the posttreatment SUVmax of the index tumor for prediction of bPFS was determined by X-tile plotting (23). Kaplan–Meier analysis was used to test the ability of selected variables to determine the survival probability, and the log-rank test was used to compare differences among groups. A significance level of 5% was applied. All analyses were conducted by SPSS software (version 22.0; IBM Corp.)
RESULTS
Patient Characteristics
The clinical and pathologic variables of the 75 patients are shown in Table 1. The median age was 70 y (IQR, 65–73 y). The initial PSA level before biopsy was 40.10 ng/mL (IQR, 19.24–80.02 ng/mL), followed by a nadir PSA of 0.04 ng/mL (IQR, 0.01–0.12 ng/mL) before RP. Thirty-three patients (44%) received ADT plus docetaxel, and 42 (56%) received ADT plus abiraterone. According to final pathology, 22 patients (29.3%) showed lymph node metastases and 15 patients (20.0%) had a positive surgical margin. Notably, 25 patients (33.3%) achieved a favorable pathologic response (pCR or minimal residual disease) on the final pathology. Most index tumor lesions underwent a significant decline in [68Ga]PSMA-11 intensity, from a median pretreatment SUVmax of 18.9 (IQR, 12.45–27.6) to a median posttreatment SUVmax of 5.61 (IQR, 4.51–7.91). The median follow-up for all participants was 30 mo (IQR, 20.0–41.5 mo). Eighteen patients (24%) experienced biochemical progression at a median follow-up of 30 mo since randomization.
Pre- and Postoperative Characteristics of 75 High-Risk PCa Patients with [68Ga]PSMA PET/CT Scanning Before and After Neoadjuvant Treatment
Univariate and Multivariate Cox Regression Analyses of Clinical and PET Imaging Parameters for Prediction of bPFS
Among all incorporated variables, clinical staging of T3b, pretreatment SUVmax, posttreatment SUVmax, SUVmax decline percentage, and a favorable pathologic response on final pathology were significantly associated with bPFS according to Cox proportional-hazards regression (Table 2), with hazard ratios of 4.68 (95% CI, 1.04–21.02; P = 0.04), 1.02 (95% CI, 1.00–1.05; P = 0.02), 1.04 (95% CI, 1.02–1.06; P = 0.00), 1.00 (95% CI, 0.99–1.00; P = 0.05), and 0.10 (95% CI, 0.01–0.65; P = 0.02), respectively.
Univariate Cox Regression Analyses for Risk of Biochemical Progression
To avoid the possible dependence of different PET-based variables, we made the multivariate model of pathologic response with each SUV-based variable separately (Table 3). We found that only posttreatment SUVmax and a favorable pathologic response on final pathology were independent variables for the prediction of bPFS, with hazard ratios of 1.02 (95% CI, 1.00–1.04; P = 0.02) and 0.12 (95% CI, 0.02–0.98; P = 0.048), respectively (model 2). However, when posttreatment SUVmax was not included in the model, only a favorable pathologic response on final pathology was an independent variable for the prediction of bPFS, with hazard ratios of 0.11 (95% CI, 0.01–0.89; P = 0.04) in model 1 and 0.10 (95% CI, 0.01–0.80; P = 0.03) in model 3.
Multivariate Cox Regression Analyses for Risk of Biochemical Progression with SUV-Based Variable
Predictive Value of PET Imaging Parameters and Pathologic Response for bPFS
With a cutoff of 8.5, Kaplan–Meier analysis revealed a significant difference in bPFS between patients with a posttreatment SUVmax of more than 8.5 and of less than 8.5, with a 36-mo biochemical progression-free rate of 29.4% (IQR, 7.6%–51.2%) and 97.6% (IQR, 92.6%–100%), respectively (log-rank P < 0.001) (Fig. 2A). Patients with and without a favorable pathologic response also had a significant difference in bPFS (P = 0.002), with a 36-mo biochemical recurrence-free rate of 100% (IQR, 100%–100%) and 55.2% (IQR, 35.0%–75.4%), respectively (Fig. 2B). Two representative cases, with and without biochemical progression, are shown in Supplemental Figures 1 and 2, respectively (supplemental materials are available at http://jnm.snmjournals.org). The patient who experienced biochemical progression had a higher posttreatment SUVmax and an unfavorable pathologic response.
bPFS in patients with different pathologic responses on final pathology (A) and after neoadjuvant therapy [68Ga]PSMA PET/CT (B). Favorable pathologic response was defined as pCR or minimal residual disease < 0.5 cm (pCR or minimal residual disease), whereas favorable [68Ga]PSMA PET/CT response was defined as posttreatment SUVmax < 8.5 on [68Ga]PSMA PET/CT.
DISCUSSION
This study was designed to investigate the relationship between response on [68Ga]PSMA PET/CT and bPFS in patients with high-risk localized or locally advanced PCa who received neoadjuvant therapy and RP. Our results indicated that [68Ga]PSMA PET/CT–derived SUVmax after neoadjuvant therapy was an independent risk factor for the prediction of bPFS. Patients with favorable responses on [68Ga]PSMA PET/CT after neoadjuvant therapy (SUVmax < 8.5) had better bPFS than those with unfavorable responses. This is the first study, to our knowledge, to suggest that response on [68Ga]PSMA PET/CT could be applied as an ideal tool to predict the oncologic outcomes of PCa patients receiving neoadjuvant therapy.
The pathologic response was set as the primary endpoint in several phase II clinical trials designed to investigate the efficacy and safety of neoadjuvant ADT in combination with androgen receptor pathway inhibitors for high-risk localized PCa. In addition, pCR was set as the coprimary endpoint in the ongoing phase III clinical trial, which was designed to determine whether treatment with apalutamide plus ADT before and after RP in patients with high-risk localized or locally advanced PCa (NCT03767244, PROTEUS trial) can bring benefit to those patients. In breast cancer and bladder cancer, the pathologic response has been well indicated to correlate significantly with improved long-term oncologic outcomes (11,12). Moreover, residual breast cancer burden after neoadjuvant therapy has been shown capable of predicting oncologic outcomes after neoadjuvant chemotherapy (22). Therefore, the pathologic response was set as the primary endpoint to evaluate the efficacy of neoadjuvant therapies in these cancers (24–27). However, the positive association between a favorable pathologic response and better long-term oncologic outcomes, such as bPFS and metastasis-free survival, remains unclear (28). In a pooled analysis, a favorable pathologic response after neoadjuvant therapy was demonstrated to be significantly associated with a better 3-y biochemical recurrence-free survival (4,5). In our study, a favorable pathologic response, defined as pCR or minimal residual disease, was found to be significantly associated with a lower rate of biochemical progression in a median follow-up of 30 mo, a finding that was consistent with previously published data (4).
Significant heterogeneity was found in pathologic response after neoadjuvant therapy, with a favorable pathologic response rate of 15.7%–62% in the previously published studies (3–6,19). Though some preliminary results suggested pathologic response as a surrogate endpoint to evaluate the efficacy of neoadjuvant therapy, efficacy could be revealed only after RP. A noninvasive biomarker to monitor response during or after neoadjuvant therapy is urgently needed to adopt novel treatment approaches and identify candidates for the subsequent RP.
PSMA PET/CT is currently recommended by guidelines for initial staging and restaging because of its high sensitivity and specificity (13–15). Patterns of change in PSMA PET/CT have been well indicated to be significantly associated with response in patients with metastatic hormone-sensitive or castration-resistant PCa to docetaxel chemotherapy or new-generation androgen receptor pathway inhibitors (29–33). Unlike pathologic response, which could be revealed only after RP, [68Ga]PSMA PET/CT, as a noninvasive and repeatable imaging tool, could provide predictive information during or after neoadjuvant therapy, suggesting [68Ga]PSMA PET/CT to be an ideal biomarker to monitor treatment response. In fact, we previously reported the utility of [68Ga]PSMA PET/CT in the prediction of pathologic response in patients with high-risk localized or locally advanced PCa receiving neoadjuvant ADT plus abiraterone for 6 mo (19). With a median follow-up of 30 mo, the present study revealed that PSMA uptake on PET/CT after neoadjuvant treatment was an independent risk factor to predict bPFS. In addition, patients with a better response on [68Ga]PSMA PET/CT after neoadjuvant therapy (SUVmax < 8.5) had a significantly lower rate of biochemical progression than those with poor responses. Our results further verified the positive association between the response of [68Ga]PSMA PET/CT and response in high-risk patients receiving neoadjuvant therapies. Of interest, SUV decline was not an independent risk factor for the prediction of bPFS in the multivariate analysis, though it was associated with biochemical progression in univariate analysis. Apparently, posttreatment SUVmax could better reflect residual tumor burden, which has been demonstrated to be significantly associated with longer oncologic outcomes (19).
The inherent limitation of this study is the limited sample size because of the relatively strict inclusion criteria. However, with patients pooled from 2 prospective cohorts, basic characteristics and treatment procedures were well balanced and standardized despite the retrospective design. Another limitation is the relatively short follow-up time, allowing us to apply only bPFS as the clinical outcome and not longer oncologic outcomes such as metastasis-free survival or castration-resistant PCa–free survival. A larger prospective study with longer follow-up is needed for further validation. In addition, we included only [68Ga]PSMA PET/CT–derived SUV in the Cox regression analysis. The role of [68Ga]PSMA PET/CT–derived radiomics in predicting bPFS needs to be further investigated. However, to our knowledge, our study is the first to reveal the role of [68Ga]PSMA PET/CT response in the prediction of oncologic outcomes in high-risk patients receiving neoadjuvant therapy.
CONCLUSION
Our study indicated the predictive role of PSMA PET for patients with high-risk localized or locally advanced PCa receiving neoadjuvant therapies. Patients with better responses on [68Ga]PSMA PET/CT after neoadjuvant therapies had significantly longer bPFS than did those with poor responses. Combined with our previous results indicating the association between [68Ga]PSMA PET/CT changes and pathologic response, our studies suggest that [68Ga]PSMA PET/CT is an ideal tool to monitor the response of primary PCa to neoadjuvant therapies and that patients with a higher posttreatment SUVmax (>8.5) could get limited benefits from neoadjuvant therapy after RP. Radiotherapy might be a better option for these patients. In addition, posttreatment SUVmax could be considered an idea biomarker for adjustment of neoadjuvant therapy regimens.
DISCLOSURE
This study is supported by grants from the National Natural Science Foundation of China (82172639, 81972388), the Mobility Programme of the National Natural Science Foundation of China (M-0670), and the Natural Science Foundation of Jiangsu Province (BK 20210023). No other potential conflict of interest relevant to this article was reported.
KEY POINTS
QUESTION: Could response on [68Ga]PSMA PET/CT be a surrogate endpoint for patients with high-risk localized PCa receiving neoadjuvant therapy?
PERTINENT FINDINGS: In this pooled cohort of 75 patients from 2 clinical trials evaluating the efficacy of neoadjuvant treatment in high-risk PCa, we found that SUVmax derived from posttreatment [68Ga]PSMA PET/CT and pathologic response on final pathology were independent factors for the prediction of bPFS.
IMPLICATIONS FOR PATIENT CARE: [68Ga]PSMA PET/CT is an ideal tool to monitor response to neoadjuvant therapy.
Footnotes
Published online Jul. 20, 2023.
- © 2023 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication January 2, 2023.
- Revision received May 10, 2023.