Molecular tumor volume on PSMA PET/CT is an independent imaging biomarker associated with progression-free survival in patients with oligorecurrent prostate cancer

Abstract

The purpose of this study is to evaluate the PSMA PET imaging parameters in association with outcomes among patients with oligorecurrent prostate cancer. This retrospective single-center study included 101 patients (median age 71; interquartile range 65-75) with biochemically recurrent prostate cancer who underwent PSMA PET between May 2021 and May 2022, revealing 5 or fewer sites of metastases (oligometastatic disease). Multiple variables including maximum standardized uptake value (SUV_max), mean standardized uptake value (SUV_mean), and molecular tumor volume (MTV) were measured and analyzed on a per-patient basis, along with total MTV and molecular tumor burden (MTB). Multivariable Cox proportional-hazards regression models were used to identify factors associated with progression-free survival (PFS). PSMA PET revealed a total of 216 lesions across all patients, of which 134 (62.0%) involved the lymph nodes and 56 (25.9%) involved the bone. A total of 61 (60.4%) patients received combined metastasis-directed and hormone therapy, and 40 (39.6%) received hormone therapy only. The median subsequent follow-up from PSMA PET detection of oligorecurrent disease was 18.2 months (IQR 10.3-25.0). MTV on PSMA PET was associated with worse PFS (hazard ratio: 1.05, 95% CI 1.00-1.11; P = 0.04). Molecular tumor volume on PSMA PET is associated with worse clinical outcomes in patients with oligorecurrent prostate cancer.

Introduction

Prostate cancer is the second-leading cause of cancer-related death in men, with a rising incidence of the disease by 3% each year and approximately 30,000 new cases diagnosed in the United States in 2024 [1]. Prostate-specific membrane antigen (PSMA) positron emission tomography/computed tomography (PET/CT) has become an indispensable tool for early and accurate detection of prostate cancer both in the primary and recurrent setting. In patients who experience biochemical recurrence with low prostate-specific antigen (PSA), PSMA PET has demonstrated superior disease detection of lesions that are occult on conventional CT or MR imaging, with sensitivity rates of 58% for PSA below 1.0 ng/mL and greater than 76% for PSA above 1.0 ng/mL [2]. Earlier disease detection and the ability for quantitation using PSMA PET present opportunities to deliver targeted therapies and identify prognostic biomarkers, to improve disease control and survival outcomes [3].

Oligometastatic disease in prostate cancer is postulated as a distinctive intermediate state between localized and widespread disease, conferring the unique opportunity to deliver localized metastasis-directed therapy (MDT) to improve patient outcomes with the benefit of lower morbidity compared to systemic therapy [4,5]. MDT is delivered in the form of radiation therapy and surgical resection and has shown promise in preclinical and clinical trials for decreasing disease progression [6,7]. Advancements in radiation therapy aided by improved imaging technologies, treatment planning and delivery, and targeting technologies have ushered the development of MDT techniques, including IMRT and SBRT. These techniques have dose-escalation capabilities that can conform to the prostate and oligometastatic sites with minimal additional toxicity and reduced risk of biochemical recurrence [8]. In clinical practice, MDT is often added to hormone therapy to synergistically improve progression-free survival (PFS) in patients with oligometastatic disease [9-12]. Factors that influence why some patients with oligometastatic prostate cancer fare better than others is an evolving area of investigation.

The ability to prognosticate using imaging and clinicopathologic markers is essential for guiding the clinical management of patients with oligorecurrent disease while minimizing morbidity. The purpose of this study is to determine the PSMA PET imaging markers associated with progression-free survival in patients with oligorecurrent prostate cancer.

Method

Patient cohort

This retrospective single-center study included 123 patients with biochemically recurrent prostate cancer who underwent PSMA-targeted ⁶⁸Ga-gozetotide PET/CT scans between May 2021 and May 2022. The study protocol was approved by the local IRB. All patients had confirmed oligometastatic biochemical recurrence of prostate cancer after primary local therapies, including radical prostatectomy or radiation therapy with adjuvant therapy. Oligometastatic disease was identified on PSMA PET/CT performed at the time of biochemical recurrence. Exclusion criteria included polymetastatic disease (n = 19) or concurrent second active malignancy receiving systemic therapy (n = 3). The final study cohort was 101 patients, of whom 61 underwent MDT combined with hormone therapy and 40 received standard hormone therapy only (Figure 1).

Figure 1.

Flowchart of patient inclusion. PSA: prostate-specific antigen; PSMA: prostate-specific membrane antigen.

Oligometastatic prostate cancer was defined as five or fewer sites of uptake suspicious for metastasis. To those oligometastatic sites, MDT was delivered as radiation therapy (e.g., SBRT or IMRT) or surgical resection. The primary endpoint was progression-free survival, defined as the interval until the occurrence of one of the following events: progression based on clinical or biochemical parameters, radiological evidence of disease progression, or death. Clinical and biochemical parameters followed PSA trends for biochemical recurrence as well as symptomatic changes or decline in functional status which may result in change of therapy. Biochemical recurrence was defined as PSA level of ≥ 0.2 ng/mL on two consecutive tests after radical prostatectomy, or PSA rise of ≥ 2 ng/mL above the nadir after radiation therapy according to the Phoenix definition [13-15]. Imaging evidence of disease progression followed RECIST 1.1 and PERCIST criteria. If no progression or death was observed, PFS was censored at the time of the last documented clinical follow-up. Demographic data and clinicopathological factors, including age, ethnicity, Gleason category, treatment modality, and follow-up, were extracted from the electronic health record.

PET/CT acquisition protocol

PSMA PET/CT examinations were performed on the GE Medical Systems Discovery MI Gen 2 (n = 64) and Siemens Biograph CT 64 (n = 37) PET/CT scanners. Whole body PET/CT images were obtained from scalp to mid-thigh approximately 60 minutes after radiopharmaceutical injection. Noncontrast low-dose CT was used for attenuation correction and anatomic localization.

Image analysis

The number, size, and location of PSMA-avid metastases were recorded. Standardized uptake values (SUVs) for each lesion, including SUV_max and SUV_mean were measured using the attenuation-corrected PET images by Siemens Healthineers syngo.via software. SUV_mean and molecular tumor volume (MTV) were measured using 40% isocontour, in keeping with previous literature [16]. The analysis utilized the maximum lesion values on a per-patient basis for MTV, SUV_mean, and SUV_max. The total MTV represents the sum of MTV values across lesions identified in each patient. Molecular tumor burden (MTB) for each patient was calculated by multiplying MTV with SUV_mean for each lesion and summing up the total number of lesions in the whole body.

Statistical analysis

Demographic summaries were computed for the entire cohort; continuous variables were summarized using the median (interquartile range, IQR), while categorical variables were summarized using the frequency (proportion). Kaplan-Meier curve was constructed to visualize the relationship between clinical and imaging variables and progression-free survival; the log-rank test was applied to evaluate survival differences between dichotomization of each variable (at the mean). Univariate and multivariable Cox proportional-hazards regression models were used to identify clinical and imaging variables independently associated with progression-free survival. The primary analysis modeled each continuous/ordinal variable as linear, but a secondary model was constructed in which the two ordinal variables, Gleason grade and number of metastases, were dichotomized. Gleason grade was dichotomized into low to intermediate-risk Gleason grades 1-3 and high-risk Gleason grades 4-5. The number of metastases was dichotomized into groups containing 1-2 and 3-5. These groups are depicted in Table 1. Variables included in the multivariable model were selected based on the univariable analysis using an inclusion threshold of P < 0.2. Variance inflation factors (VIF) were calculated to assess multicollinearity among predictors; VIF > 5 was considered indicative of potential collinearity based on previous literature [17]. For variables identified as collinear, sensitivity analyses were performed by running separate multivariable analyses in which the collinear variable was excluded while retaining other selected covariates. Each coefficient was exponentiated to reflect fold changes in the hazard ratio (HR) associated with a 1-unit increase in the covariate of interest; 95% confidence intervals and p-values were also calculated. A p-value less than 0.05 was regarded as statistically significant. Statistical analysis was performed using R software (version 4.3.1, R Foundation for Statistical Computing).

Table 1.

Baseline patient and tumor characteristics

Characteristics		Unit
Age*	71 (65-75)	Years (IQR)
Race & Ethnicity
Black	2 (2.0)	No. of patients (%)
Hispanic	3 (3.0)
White	92 (91.1)
Other†	4 (3.9)
PSA Level at Recurrence*	1.0 (0.36-2.70)	ng/mL (IQR)
≤ 0.2	10 (9.9)	No. of patients (%)
> 0.2 to < 2.0	60 (59.4)
≥ 2.0	31 (30.7)
Gleason Grade
1	10 (9.9)	No. of patients (%)
2	13 (12.9)
3	23 (22.8)
4	24 (23.7)
5	31 (30.7)
Metastatic Lesions, No.*	2 (1-3)	Mean (IQR)
1	36 (35.6)	No. of patients (%)
2	29 (28.7)
3	25 (24.8)
4-5	11 (10.9)
Lesion Location (n = 216)
Lymph Node	134 (62.0)	No. of patients (%)
Bone	56 (25.9)
Prostate Bed	18 (8.3)
Lung	5 (2.4)
Other‡	3 (1.4)
Lesion Size*	7 (5-12)	mm (IQR)
≤ 4	13 (12.9)	No. of patients (%)
5-7	32 (31.7)
8-9	8 (7.9)
≥ 10	34 (33.7)
Not measurable	14 (13.8)
Treatment
Metastasis-directed therapy with hormone therapy	61 (60.4)	No. of patients (%)
Hormone therapy alone	40 (39.6)

^*Presented as median and interquartile range; others are presented as number of patients and percentages.

^†Includes Asian and not otherwise specified.

PSA: prostate-specific antigen.

^‡Other lesion locations include the liver, bladder, and adrenal gland.

Results

Patient characteristics

Table 1 describes the patient and tumor characteristics. There were 101 patients with a median age of 71 years (IQR 65-75) at the time of biochemical recurrence. The median PSA level was 1.01 ng/mL (IQR 0.36-2.7) at the time PSMA PET was performed for suspected biochemically recurrent disease. Just over half of patients (54.5%, 55/101) presented with high-grade Gleason group 4-5 disease at the time of initial diagnosis. The majority (64.3%, 65/101) had one or two sites of metastases on PSMA PET. There were 216 metastases across all patients, of which 134 (62.0%) involved the lymph nodes and 56 (25.9%) involved the bones.

PSMA PET parameters and progression-free survival

The median follow-up from PSMA PET detection of oligorecurrent disease was 18.2 months (IQR 10.3-25.0). A total of 61 patients received combined therapy with MDT and hormone therapy, and 40 received standard hormone therapy only. Forty patients experienced progressive disease. No treatment-related difference in PFS was detected in an exploratory unadjusted analysis using the log-rank test. Among all the variables considered, molecular tumor volume - measured as the maximal MTV on a per-patient basis - was the sole parameter that was significantly associated with worse progression-free survival on both univariate (HR: 1.06, 95% CI: 1.01-1.11; P = 0.02) and multivariable analysis (HR = 1.05, 95% CI: 1.00-1.11; P = 0.04) (Table 2). This outcome is depicted on the Kaplan-Meier curve presented in Figure 2 and representative patient case in Figure 3. Total MTV demonstrated significance on univariate analysis (HR = 1.05, 95% CI: 1.00-1.10, P = 0.04) but not multivariable analysis. VIF were 10.48 for MTV, 10.83 for total MTV, 1.00 for Gleason grade, and 1.14 for PSA. Due to collinearity between the variables MTV and total MTV, each was assessed independently in the multivariable analyses in which one of the collinear variables was excluded while retaining the other selected covariates. The results of these analyses are summarized in Table 2. The other variables, PSA and Gleason grade, showed no evidence of collinearity. There was insufficient evidence to conclude that other PET-derived markers were associated with PFS.

Table 2.

Cox proportional hazards analyses assessing the association between PSMA PET parameters and progression free survival

Measured Parameter	Univariable HR 95% (CI)	p	Multivariable HR (95% CI)	p	Multivariable HR (95% CI)	p
SUVmax	1.01 (0.99-1.02)	0.52
SUVmean	1.01 (0.98-1.03)	0.64
MTV	1.06 (1.01-1.11)	0.02	1.05 (1.00-1.11)	0.04
Total MTV	1.05 (1.00-1.10)	0.04			1.04 (0.99-1.10)	0.10
MTB	1.00 (0.99-1.00)	0.11
PSA	1.01 (0.99-1.02)	0.09	1.01 (0.99-1.02)	0.36	1.01 (0.99-1.02)	0.38
Gleason Grade	1.20 (0.94-1.54)	0.15	1.30 (0.79-2.13)	0.31	1.29 (0.78-2.12)	0.33
Number of Metastases	1.11 (0.84-1.46)	0.46
Treatment	1.03 (0.55-1.94)	0.93

Data in parentheses are 95% confidence intervals. SUV_max: maximum standardized uptake value; SUV_mean: mean standardized uptake value; MTV: metabolic tumor volume; MTB: molecular tumor burden; HR: hazard ratio.

Figure 2.

Kaplan-Meier survival curve stratified to MTV values above and below the mean, 2.3 mL. Greater MTV values were associated with worse progression-free survival (P = 0.04, using the log-rank test).

Figure 3.

61-year-old male with rising PSA of 1.67 ng/mL at the time of PSMA PET/CT on March 2022 demonstrating intense uptake to subcentimeter retroperitoneal lymph nodes and right common iliac lymphadenopathy. Maximum intensity projection image (A) shows extent of disease, with representative left para-aortic lymph node MTV of 1.42 cm³ (B) and greatest MTV measuring 3.14 cm³ in the right common iliac lymph node (C). Patient had been previously treated with radical prostatectomy and adjuvant radiation therapy to the prostatic fossa and pelvic lymph nodes in 2019. After oligorecurrence, patient was treated with bicalutamide and radiation therapy to the retroperitoneal and common iliac nodes, with progression of disease detected by rising PSA 21.5 months later.

Discussion

Identifying prognostic biomarkers is essential for guiding clinical management throughout the course of disease for patients with prostate cancer. In the present study, we evaluated the imaging biomarkers on PSMA PET and their associations with progression-free survival in patients with oligorecurrent prostate cancer. MTV was found to be a prognostic factor, with higher MTV associated with worse outcomes.

Previous studies have corroborated the potential value of MTV, a semi-quantitative PET parameter, in tumors such as lymphoma, lung cancer, and endometrial cancer [18-21]. Kubilay et al. (2022) suggested that ⁶⁸Ga-PSMA PET volumetric parameters like MTV have predictive value in the primary staging of prostate cancer, in which an increase in total lesion uptake values was associated with higher pathologic T stage and surgical margin positivity [22]. PSMA PET-derived volumetric measurements for the detection of progressive disease provide further insights into tumor biology and have garnered increasing support for consideration in the PSMA PET progression criteria [23,24]. Seifert et al. found that total molecular tumor volume derived semi-automatically from ⁶⁸Ga-PSMA PET was a repeatable, reader-independent metric that may be used to assess therapeutic response in the context of metastatic disease [25].

Of note, neither SUV_mean nor SUV_max were associated with progression-free survival in this study, within the constraints of this limited-size study cohort. An earlier study by Murad et al. suggested that adjusted SUV measures - SUV_max and segmented SUV_mean - were associated with a lower risk of biochemical progression in patients with oligometastatic prostate cancer who underwent MDT [26]. However, their study sample was smaller, and recurrent disease consisted predominantly of 86% nodal metastases. Our current study included a broader spread of disease sites, including 62% nodal metastases and 38% skeletal or visceral metastases, as well as different therapeutic regimens, which may confer different results. Furthermore, the reliability of SUV measurements is affected by external factors such as patient body composition, radiotracer choice, and scan acquisition parameters.

Our study is a retrospective single-institution study with limitations inherent to such a study design and with a relatively small study sample. At the time of this study, metastasis-directed therapy was delivered in combination with standard hormone therapy for patients with oligometastatic prostate cancer, or standard hormone therapy was given alone. This approach has been supported in previous studies; for example, the OLIGOPELVIS (NCT02274779) is an open-label phase II trial in 2021 that demonstrated better disease control in patients who received combined hormone therapy and pelvic radiation therapy in oligorecurrent pelvic nodal disease while reducing toxicity [27]. The oligometastatic disease state also offers potential therapeutic opportunities to delay not only further metastatic spread but also the initiation of systemic therapies. Although treatment modality was not associated with differences in PFS in this study, these findings should be interpreted cautiously given the size of the study cohort and limitations of a retrospective study. Future prospective studies may be valuable to further investigate the treatment-related associations which are outside the scope of the current study. Molecular profiling and identification of high-risk mutational signatures may also provide another avenue of future study in the ability to complement imaging biomarkers, facilitating risk stratification and improving patient selection for those most likely to respond to localized curative intent therapies and attain sustained disease control [10,28,29].

In conclusion, molecular tumor volume on PSMA PET/CT may serve as an imaging biomarker associated with progression-free survival in patients with oligorecurrent prostate cancer. This finding presents an opportunity to utilize this imaging parameter to facilitate patient selection and therapeutic optimization to improve patient outcomes.

Disclosure of conflict of interest

None.