Treatment Response Assessment of Skeletal Metastases in Prostate Cancer with ¹⁸F-NaF PET/CT

Abstract

Purpose

To determine the utility of ¹⁸F-sodium fluoride positron emission tomography-computed tomography (¹⁸F-NaF PET/CT) in the imaging assessment of therapy response in men with osseous-only metastatic prostate cancer.

Methods

In this Institutional Review Board–approved single institution retrospective investigation, we evaluated 21 ¹⁸F-NaF PET/CT scans performed in 14 patients with osseous metastatic disease from prostate cancer and no evidence of locally recurrent or soft-tissue metastatic disease who received chemohormonal therapy. Imaging-based qualitative and semi-quantitative parameters were defined and compared with changes in serum PSA level.

Results

Qualitative and semi-quantitative image-based assessments demonstrated > 80% concordance with good correlation (SUVmax κ = 0.71, SUVavg κ = 0.62, SUVsum κ = 0.62). Moderate correlation (κ = 0.43) was found between SUVmax and PSA-based treatment response assessments. There was no statistically significant correlation between PSA-based disease progression and semi-quantitative parameters. Qualitative imaging assessment was moderately correlated (κ = 0.52) with PSA in distinguishing responders and non-responders.

Conclusion

¹⁸F-NaF PET/CT is complementary to biochemical monitoring in patients with bone-only metastases from prostate cancer which can be helpful in subsequent treatment management decisions.

Introduction

Prostate cancer is the most common cancer in men and the second leading cause of cancer death, with a high incidence of osseous metastases [1]. Traditionally, conventional computed tomography (CT) and bone scintigraphy with ^99mTc-based radiotracers, such as ^99mTc methylene diphosphonate (MDP), have been used for the detection and monitoring of bony metastases [2]. However, while ^99mTc-based radiotracers are sensitive, radiotracers are taken up in a variety of other disease processes, including infection, trauma, non-infectious inflammation, and metabolic bone diseases limiting its specificity for malignancy and treatment monitoring [3, 4]. Due to these limitations, there has been an increased interest in the role of positron emission tomography (PET) for the evaluation and monitoring of metastatic prostate cancer [5].

¹⁸F-sodium fluoride (¹⁸F-NaF) is a positron-emitting radiopharmaceutical analog of the hydroxyl group found in hydroxyapatite bone crystals, first used for skeletal scintigraphy in the 1970s [6]. With fast bone uptake and rapid blood clearance, ¹⁸F-NaF provides high target to background uptake, leading to high-quality images in less than an hour after intravenous administration [7]. However, due to its shorter half-life, technical limitations, and cost at the time, it was replaced by ^99mTc-phosphonates. Yet, with the expansion of availability and access to PET/CT, there has been a resurgence of interest in the use of ¹⁸F-NaF for bone metastasis imaging [8]. In a cohort of patients with high-risk prostate cancer, Evan-Sapir et al. showed ¹⁸F-NaF PET/CT to have a sensitivity and specificity of 100% for the detection of osseous metastases, compared with 70% and 57% for ^99mTc-MDP [9], with several subsequent studies supporting the utility of ¹⁸F-NaF PET/CT [10–12].

Currently, monitoring the treatment response of bony metastases involves a combination of clinical assessment, biochemical markers with prostate-specific antigen (PSA) and serum alkaline phosphatase (ALP), and imaging with CT and ^99mTc-MDP [13]. However, these methods have been inadequate and often non-specific compared with those used for soft tissue disease [14]. The purpose of this study was to investigate the utility of ¹⁸F-NaF PET bone scans in evaluating response of osseous metastatic prostate cancer to treatment.

Methods

Institutional review board approval was obtained for this retrospective study. Patient selection criteria for this study were as follows: (1) men above the age of 21 years with a prior histological diagnosis of prostate adenocarcinoma; (2) PSA relapse, defined as post-radical prostatectomy PSA level exceeding 0.2 ng/mL [15] or post-radiation therapy PSA rise of 2 ng/mL or more above the nadir after external beam radiation therapy [16]; (3) patients with bone-only metastases who underwent a baseline and follow-up ¹⁸F-NaF PET/CT scan after undergoing medical treatment. Patients undergoing treatment for both metastatic castrate-sensitive prostate cancer (mCSPC) and metastatic castrate-resistant prostate cancer (mCRPC) were included in this study. Exclusion criteria included history of cancer other than prostate cancer, active infection, active inflammatory conditions, recent or complicated nonhealing fracture, and hip or knee arthroplasty. Medical therapy was chosen at the discretion of the treating physicians, who were made aware of the results of the scans.

Records of all men who underwent baseline and follow-up ¹⁸F-NaF PET/CT scans from 2010 to 2012 were obtained and reviewed for specific eligibility criteria as defined above. All ¹⁸F-NaF PET/CT scans (Biograph Duo LSO; Siemens) were performed 60 min after the intravenous administration of 10 mCi of ¹⁸F-NaF. The fused ¹⁸F-NaF PET/CT studies were interpreted by a board-certified fellowship-trained nuclear radiologist with more than 20 years of experience of interpreting PET/CT studies. Suspicious skeletal lesions were identified based on previously published guidelines [17]. The maximum standardized uptake values (SUVs) of suspicious lesions, defined as foci of nonphysiological uptake above regional background bone activity, were obtained using 3D region of interest (ROI) software (Siemens). PET variables included maximum SUV value of the most active lesion (SUVmax), average maximum SUV value of all lesions (SUVavg), and the sum of the maximum SUV values of all lesions (SUVsum). PSA labs were available within 50 days of each scan.

Semi-quantitative imaging treatment response and PSA-based treatment response criteria were defined as complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD) (Table 1). Qualitative imaging assessment of the ¹⁸F-NaF PET/CT was the general impression that was reported by the interpreting nuclear radiologist in the final report combining the overall PET and CT findings. Patients were further categorized if they demonstrated progressive disease, progressors (P=PD) vs. non-progressors (NP=SD+PR+CR), and if they demonstrated complete or partial response, responders (R=CR+PR) vs non-responders (NR=SD+PD).

Table 1.

Definition of treatment response criteria

Semi-quantitative imaging treatment response criteria		PSA-based treatment response criteria
CR	Disappearance of all target lesions	PSA decline to undetectable (< 0.2 ng/mL)
PR	≥ 25% decrease in SUVsum, SUVavg, or SUVmax	≥ 25% PSA decline and ≥ 2 ng/mL decline
SD	< 25% decrease/increase in SUVsum, SUVavg, or SUVmax	< 25% PSA decline/rise or < 2 ng/mL decline/rise
PD	Any new lesions or a ≥ 25% increase in SUVsum, SUVavg, or SUVmax	≥ 25% PSA increase and ≥ 2 ng/mL increase

CR complete response, PR partial response, SD stable disease, PD progressive disease

For patients who underwent multiple ¹⁸F-NaF PET/CT scans, each interval scan was analyzed separately, with the most recent ¹⁸F-NaF PET/CT scan serving as the new baseline. Descriptive statistics were used to summarize the data, and Bowker’s test was performed to assess the concordance of PSA-based treatment response, semi-quantitative treatment response, and qualitative imaging treatment response. Statistical analysis was performed in (STATA, College Station, TX), reporting unequal variance two-sided p values at preset significant level of 0.05.

Results

On review of all patients who underwent ¹⁸F-NaF PET/CT scans, there were 14 patients who met our eligibility criteria, 7 of which underwent multiple scans, allowing for 21 interval comparisons. All patients’ disease was confined to the bones, with no evidence of soft tissue disease by CT. At the time of the baseline ¹⁸F-NaF PET/CT scan, the average PSA was 29.3 ng/mL (range 0.3–130.6 ng/mL). The average time between baseline and follow-up scans was 218 days (range 85–489 days). The baseline ¹⁸F-NaF PET/CT scans with avid lesions demonstrated average values for SUVsum of 200.8 (range 19.7–549.5), SUVavg of 39.8 (range 6.6–95.0), and SUVmax of 53.3 (range 8.8–172). One patient with biochemically recurrent disease had no detectable lesions on baseline scan.

Tumor response assessment by PSA-based treatment response, semi-quantitative imaging treatment response, and qualitative imaging treatment response can be found in Table 2. Qualitative imaging demonstrated a moderate correlation with PSA in assessing responders and non-responders, (κ = 0.52, 95%CI 0.16–0.89), and was concordant in 76% of cases (Fig. 1). However, no correlation was noted between qualitative reports and PSA in the assessment of progressors vs. non-progressors (κ = 0.05, 95%CI − 0.38–0.47, concordant in 52% of cases). Of the 10 patients with no disease progression by PSA-based treatment response criteria, four demonstrated progressive disease on qualitative assessment with ¹⁸F-NaF PET/CT, two of which showed new metabolically active lesions (Fig. 2).

Table 2.

Tumor response assessment results

PSA-based treatment response		Semi-quantitative imaging treatment response			Qualitative imaging treatment response
		SUVsum	SUVavg	SUVmax
PD	11	9	9	9	9
SD	3	6	6	7	8
PR	6	6	6	5	4
CR	1	0	0	0	0

Entries are number of scans compared with a prior scan

CR complete response, PR partial response, SD stable disease, PD progressive disease

Fig. 1.

Concordance among treatment response criteria, treatment changed (P progressor, NP non-progressor, R responder, NR non-responder, N number of lesions)

Fig. 2.

Discordance among response criteria, treatment changed (P progressor, NP non-progressor, R responder, NR non-responder, N number of lesions)

Semi-quantitative imaging treatment response demonstrated a moderate correlation between SUVmax and PSA-based treatment response (κ = 0.43, 95%CI 0.04–0.82) and was concordant in 71% of cases. PSA-based treatment response and the semi-quantitative features SUVsum and SUVavg were both concordant in 67% of cases; however, no statistically significant correlation was demonstrated. PSA-based disease progression and semi-quantitative parameters demonstrated no significant correlation, with 43% concordance between PSA-based disease progression and all measurements. The qualitative reports and semi-quantitative analysis demonstrated good correlation in assessing treatment response: SUVmax (κ = 0.71, 95%CI 0.41–1.0, concordant in 86% of cases), SUVavg (κ = 0.62, 95%CI 0.28–0.95, concordant in 81% of cases), SUVsum (κ = 0.62, 95%CI 0.28–0.95, concordant in 81% of cases), and very good correlation in assessing disease progression (κ = 0.81, 95%CI 0.56–1.0, concordant in 91% of cases for all semi-quantitative measurements).

Discussion

There is a wide array of therapies available for patients with metastatic prostate cancer. However, regardless of initial treatment choice, a significant number of patients will develop resistance sometime during therapy [18]. Thus, accurately monitoring tumor response during therapy is crucial to ensure patients are receiving the most optimal treatment. Yet, current methods of biochemical monitoring and imaging of osseous metastases with CT and ^99mTc-MDP are not sufficiently sensitive and specific [13]. Advances in the understanding of the complex biology of prostate cancer have paved the way for molecular imaging with PET. Several PET-based radiotracers are in use or under active investigation in prostate cancer, including ¹⁸F- or ¹¹C-choline, radiotracers based on prostate-specific membrane antigen (PSMA), 16β-¹⁸F-fluoro-5α-dihydrotestosterone targeted to the androgen receptor, and the synthetic L-leucine analog ¹⁸F-fluciclovine [12]. ¹⁸F-FDG PET/CT has shown both prognostic value in patients with metastatic prostate cancer and correlation with successful treatment response [19–21]. Assessing therapy response with ¹⁸F-choline PET/CT [22, 23] and ⁶⁸Ga-PSMA PET/CT [24] has also shown promising results, although larger studies are needed to confirm their utility.

The recent resurgence of ¹⁸F-NaF PET imaging has offered a potentially better method for the evaluation of osseous metastases. Even-Sapir et al. showed ¹⁸F-NaF PET to have a sensitivity of 100% and specificity of 62% for the detection of bony metastases compared with 70% sensitivity and 57% specificity with ^99mTc-MDP, with a sensitivity and specificity of 100% for combined ¹⁸F-NaF PET/CT scans [10]. Furthermore, ¹⁸F-NaF PET/CT has been shown to be a useful diagnostic tool in otherwise radiologic occult metastases [25]. Shen et al. conducted a meta-analysis on the diagnostic performance of ¹⁸F-NaF PET/CT for the detection of bone metastases and found a sensitivity and specificity of 96% and 91% respectively, with better diagnostic accuracy compared with ^99mTc-MDP bone scan (sensitivity 88%, specificity 80%) and FDG PET/CT (sensitivity 73%, specificity 98%) [26].

Initial studies evaluating the utility of ¹⁸F-NaF PET in the monitoring treatment response have shown encouraging results [27]. Cooke et al., in a pilot study of five patients with metastatic castrate-resistant prostate cancer treated with ²²³RaCl₂, showed concordance between the mean SUVmax of patients bone metastases and PSA-treatment response [28]. The functional burden of metastatic castrate-resistant prostate cancer has also been shown to have a strong correlation with response to chemotherapy and androgen receptor pathway inhibitors, and can be predictive of progression-free survival [29]. In an imaging companion trial of a multicenter metastatic castration-resistant prostate cancer tissue biomarker-guided therapeutic trial from the American College of Radiology Imaging Network (ACRIN 6687), changes of ¹⁸F-NaF PET average SUVmax correlated with bone alkaline phosphatase levels, although there was no correlation with PSA or progression-free survival [30, 31]. Results from the National Oncology PET Registry (NOPR) have demonstrated that not only ¹⁸F-NaF PET is useful in assessing treatment response, but can change treatment plans in up to 40% of patients [32]. Additional NOPR data have also shown that ¹⁸F-NaF PET results are highly associated with patient survival and subsequent hospice claims, aiding patients and their physicians in the decisions on whether to continue treatment or pursue palliative care [33].

Our study demonstrated a strong concordance between PSA-based treatment response and qualitative and semi-quantitative imaging, especially when further stratifying patients into responders vs. non-responders. These findings suggest that ¹⁸F-NaF PET/CT may serve as a highly accurate method for identifying patients who do not respond to treatment, which can potentially lead to changes in treatment management. Of note, the one patient with complete response by PSA-based criteria had residual stable disease by ¹⁸F-NaF PET/CT and subsequently had a rise of PSA on follow-up studies. This is suggestive of ¹⁸F-NaF PET/CT detecting residual smoldering disease. Additionally, in patients with no progressive disease by PSA-based criteria, two of the four patients with discordant ¹⁸F-NaF PET/CT findings had new metabolically active lesions. Therapy-induced flare phenomenon has been reported for ¹⁸F-NaF PET [34]; however, this process is thought to be limited to sites of treated metastases with increased osteoblastic activity as demonstrated on CT. While few cases of discordance may reflect treated small marrow only lesions that were previously radiologically occult, it is more likely that ¹⁸F-NaF PET/CT provided a better evaluation of overall disease burden and disease progression. This suggests ¹⁸F-NaF PET/CT may offer advantages over biochemical monitoring in bone-dominant metastases and at the least should be used in conjunction with PSA-monitoring for determining treatment management.

Several limitations must be considered when interpreting the results of our study. The study had a relatively small sample size and was conducted at a single institution. Differences in clinical referrals, equipment, and imaging protocols may have an influence on the outcomes in an alternative setting. Patients with both mCSPC and mCRPC were analyzed as a group despite the differing tumor biology and treatment regimen. Additional larger studies are needed to assess potential differences in the utility of ¹⁸F-NaF PET/CT between mCSPC and mCRPC. Furthermore, conventional ^99mTc-MDP imaging was not performed for a majority of the patients; thus, the comparison between ^99mTc-MDP treatment response and ¹⁸F-NaF PET/CT response was out of the scope of this study. Additionally, there was a wide range of time intervals between the ¹⁸F-NaF PET/CT studies, which may limit the efficacy of treatment assessment. Lastly, the length of time between PSA collection and ¹⁸F-NaF PET/CT scans may have influenced results; however, all patients had PSA levels drawn within 50 days from the ¹⁸F-NaF PET/CT scans and most prostate cancers are relatively slow growing within this time frame [35].

Conclusion

¹⁸F-NaF PET/CT is an accurate imaging modality in the assessment of treatment response in patients with bone-only metastases from prostate cancer. This modality is complementary to biochemical monitoring and potentially can serve as a useful tool for determining further treatment management.

Abbreviations

CT

computed tomography

PET

positron emission tomography

PSA

prostate-specific antigen

MDP

methylene diphosphonate

NaF

sodium fluoride

P

progressor

NP

non-progressor

R

responder

NR

non-responder

Funding Information

We acknowledge National Institutes of Health grants R01-CA111613 (PI: H. Jadvar) and P30-CA014089 (USC Norris Comprehensive Cancer Center).

Compliance with Ethical Standards

Conflict of Interest

Erik M. Velez, Bhushan Desai, and Hossein Jadvar declare no conflicts of interest. There was no direct funding for this study.

Ethical Approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed Consent

Informed consent was obtained from all individual participants included in the study.