|Home | About | Journals | Submit | Contact Us | Français|
We sought to determine predictive factors (patient and prostate-specific antigen [PSA] characteristics) for 18F-fluoromethylcholine positron emission tomography-computed tomography (18F-FCH PET/CT) positivity in the context of biochemical recurrence after local treatment of prostate cancer (PCa) with curative intent.
This is a retrospective study including 60 18F-FCH PET/CT scans of patients with biochemical recurrence after initial radical prostatectomy (RP), external beam radiation therapy (EBRT), or focal high-intensity focused ultrasound (HIFU) with curative intent. The results were compared to findings on magnetic resonance imaging (MRI), computed tomography (CT), bone scan (BS), and histological analysis when available. Univariate analysis was performed to correlate results with patient characteristics.
Thirty-eight (63.3%) scans were positive, 17 (28.3%) negative, and 5 (8.3%) equivocal. Of the positive scans, 16 demonstrated local recurrence, 12 regional/distant lymph nodes, five bone metastasis, and five local and distant recurrences. Among the 22 PET/CTs showing metastasis, conventional imaging was performed in 16 patients (72.7%). Of these, it demonstrated the lesion(s) found on PET/CT in eight patients (50.0%), was negative in seven (43.8%), and equivocal in one (6.3%). The trigger PSA (p=0.04), prostate-specific antigen velocity (PSAV) (p=0.03), and prostate-specific antigen doubling time (PSADT) (p=0.046) were significantly different when comparing positive and negative scans. Patients with positive scans were more likely to have received EBRT initially (odds ratio [OR] 11.0, 95% confidence interval [CI] 2.2–55.3). A trigger PSA of 2.6 ng/mL had a sensitivity of 84% and specificity of 65% for a positive scan. PET/CT changed the clinical management plan in 17 patients (28.3%).
18F-FCH PET/CT demonstrates a high detection rate for local and distant recurrences after localized PCa treatment. A trigger PSA above 2.6 ng/mL seems optimal for appropriate patient selection.
Prostate adenocarcinoma (PCa) is the most common cancer among Canadian men. The lifetime risk of developing prostate cancer is one in eight and the lifetime risk of dying from it is one in 27.1 Treatment options for localized PCa include active surveillance in low-risk cancer, and radical prostatectomy (RP), external beam radiation therapy (EBRT), and in some instances, focal therapy —such as high-intensity focused ultrasound (HIFU) — in intermediate-risk disease.2–4 High-risk disease is treated by surgery or a combination of radiation therapy and androgen-deprivation therapy (ADT). Recurrence after treatment is not uncommon, occurring in approximately 20–30% of patients with RP and 20–50% of patients after EBRT.5–8
Differentiating between local or metastatic recurrence is of primordial importance in selecting appropriate treatment. Local salvage treatment with curative intent can be offered in case of local recurrence. Systemic palliative treatment with hormonal therapy and/or chemotherapy remains the current standard for patients with metastatic disease. Therapy directed to oligometastases remains controversial, with a small number of studies showing some benefits.9–11
Computed tomography (CT) and bone scintigraphy (BS) are widely used as part of initial workup of biochemical recurrence. However, these are limited by poor ability to detect disease at low prostate-specific antigen (PSA) levels.12–15 Novel functional imaging techniques have been developed and increasingly studied over the last decade, including 11C-choline and 18F-fluoromethylcholine (18FFCH).16 Choline is an essential component of phospholipid membrane and demonstrates increased metabolism in PCa.17 The limited short half-life of 11C-choline (20.3 minutes) prompted research into synthesis of 18F-FCH with a longer half-life (109.7 minutes). Although conflicting results have been reported in the course of initial staging of PCa, 18FFCH positron emission tomography-computed tomography (PET/CT) has shown promising results for detection of metastasis in the context of biochemical recurrence.16,18–22 Recent reports suggest that this functional imaging technique might directly impact the management of these patients.23–25
The primary objective of this study was to determine predictive factors (patient and PSA characteristics) for 18F-FCH PET/CT positivity in patients with biochemical failure after local treatment for PCa with curative intent. Secondary objectives were to evaluate the impact of PET/CT on patient management and to compare its value to standard imaging techniques.
This study was approved by the hospital research ethics committee (study #13-047) and written informed consent was obtained from all patients. We conducted a single-centre, retrospective analysis including 60 18F-FCH PET/CT scans (59 patients) performed from March 2014–January 2016. Inclusion criteria consisted of: histologically proven PCa, initial treatment with curative intent (RP, EBRT, or HIFU), and biochemical recurrence as defined by the guidelines of the European Association of Urology.26 Patients with distant nodal or bone metastases or with non-diagnostic scans were excluded. Patient characteristics are presented in Tables 1 and and22.
PET/CT preparation consisted of four-hour fasting. Approximately 4 MBq/kg IV of 18F-fluoromethylcholine (18F-FCH) supplied by the Montreal Neurological Institute (Montreal, QC, Canada) were administered. Studies were performed using a hybrid PET/CT scanner (Discovery ST, General Electric Medical Systems, Waukesha, WI, U.S.). Images were acquired 10–30 minutes post-injection from skull base to thighs (6–7 beds, 3–4 minutes/bed). A nuclear medicine specialist then classified the results as positive, negative, or equivocal; a positive examination was defined as 18F-FCH uptake higher than background activity and not explained by physiological process. Examples of positive 18F-FCH PET/CT findings are provided in Fig. 1.
PET/CT findings were compared to magnetic resonance imaging (MRI), CT, BS, and histological analysis when available. Validation criteria included: 1) positive histological analysis; 2) response to treatment, defined by decreased PSA after treatment and/or response on followup imaging; 3) progression on followup PET/CT; and 4) positive conventional imaging. All negative PET/CTs were considered false negatives, as they failed to detect the cause of biochemical recurrence.
A summary of clinical information was provided to two urooncologists, who were not the main treating physicians and were blinded to identifying information, PET/CT results, and information available afterwards. They were then asked to independently determine the theoretical management had the PET/CT not been performed. Finally, the theoretical and actual therapeutic strategies were compared.
All statistical analyses were performed using SAS (v.9.4) (SAS Institute Inc., Cary, NC, U.S.). For descriptive statistics, we computed t-test for continuous variables or performed one-way ANOVA. Results with p<0.05 were considered statistically significant. The Chi-square test for proportions and Fisher exact test were calculated when necessary. Finally, we modeled the probability of having a positive scan using PROC LOGISTIC.
Sixty examinations were performed at a median PSA of 3.9 ng/mL. Of these, 38 (63.3%) PET/CT scans were positive, 17 (28.3%) negative, and five (8.3%) equivocal/indeterminate. Of the 38 positive scans, 16 demonstrated local recurrence (14 in patients with initial EBRT, one RP, and one HIFU), 12 regional and/or distant lymph nodes (10 RP, two EBRT), five bone metastasis, and five local and distant recurrences.
Fifteen (71.4%) of the positive scans for local recurrence (n=21) were validated according to histological analysis in 10 patients (66.7%), response to treatment as defined by a decrease in PSA in two patients (13.3%), and correlation with conventional imaging in three patients (20.0%). One scan was a biopsy-proven false positive.
Eighteen (81.8%) of the positive scans for metastasis (n=22) were validated according to a response to treatment on following imaging studies in eight patients (44.4%), a response to treatment (decreased PSA) in five patients (27.8%), correlation with conventional imaging in four patients (22.2%), and progression on followup PET/CT in one patient (5.6%). One scan was considered false positive after comparison with conventional imaging.
One local, two nodal, and one bone recurrence were missed by PET/CT.
Among the 21 PET/CTs showing local recurrence, prostate MRI was performed in 11 patients (52.4%). Of these, 10 were positive (90.9%) and 1 was equivocal (9.1%).
Among the 22 PET/CT scans showing metastatic disease, conventional imaging was performed in 16 patients (72.7%). Of these, conventional imaging demonstrated the lesion(s) found on PET/CT in eight patients (50.0%), was negative in seven (43.8%), and equivocal in one (6.3%).
More patients with a positive scan were initially treated with EBRT (p=0.002). Moreover, there was a significant difference between trigger PSA, prostate-specific antigen velocity (PSAV), and prostate-specific antigen doubling time (PSADT) when comparing positive and negative PET/CTs (Tables 3, ,4,4, and and55).
Thirty-four patients (89.5%) with a positive PET/CT (n=38) had a trigger PSA ≥2 ng/mL, three (7.9%) 1–2 ng/mL, and none (0%) <1 ng/mL. PET/CT was positive in 77%, 60%, and 0% for patients with a trigger PSA ≥2 ng/mL, 1–2 ng/mL, and <1 ng/mL, respectively.
There was no statistically significant relationship between maximum standardized uptake values (SUVmax) and Gleason scores (p=0.27) or trigger PSA (p=0.14).
A trigger PSA of 2.6 ng/mL and PSADT of 4.4 months had a sensitivity of 84% and 76% and specificity of 65% and 50%, respectively for a positive scan (Fig. 2).
The theoretical and actual management plans differed in 17 cases (28.3%), 11 of which with positive PET/CT. The treatment plan changed from watchful waiting to treatment in seven cases and from treatment to watchful waiting in four cases. In six cases, the therapeutic strategy was modified; two of these involved converting palliative to a curative intent salvage treatment.
In this study, we report our initial local experience with 18FFCH PET/CT in the context of PCa biochemical recurrence after treatment with curative intent.
Reported detection rates for 18F-FCH PET/CT vary from 38–98%. This is probably related to heterogeneous study groups. Our study shows a detection rate of 63.3% for recurrence after localized treatment of PCa, similar to previous studies.27,28 A large number of the positive PET/CTs were validated using pre-established criteria.
PSA parameters were significantly different when comparing positive and negative PET/CT results, which confirms the known relationship with PSA kinetics.27,29–31 We found the best threshold values to predict PET/CT positivity to be a trigger PSA of 2.6 ng/mL, PSAV of 2.0 ng/mL/year, and PSADT of 4.4 months. In comparison, Rodado-Marina et al reported a trigger PSA of 3.5 ng/mL (sensitivity 64%, specificity 76%) and PSADT of six months (sensitivity 58%, specificity 58%) while Cimitan et al reported a sensitivity of 79.5% and specificity of 67% for a trigger PSA of 2 ng/mL.27,30
Initial Gleason score was not significantly different between patients with positive and negative PET/CT (p=0.069). This could possibly be related to our small sample size. In a larger study, Cimitan et al demonstrated that a Gleason score ≥7 was an independent predictive factor for positive PET/CT, even with a low trigger PSA.30
We did not find ADT administration at the time of the scan to negatively affect the detection rate, which is consistent with Chondrogiannis et al.31
In our study, PET/CT did not add significant value to the MRI diagnosis in context of local recurrence. In fact, pelvic MRI remains the modality of choice, with a sensitivity as high as 91% in the context of biochemical recurrence after RP.32,33 Interestingly, 18F-FCH PET/CT detected extraprostatic disease not identified by conventional imaging in seven patients (43.8%).
More specific markers have been developed to increase detection rates of biochemical recurrence. In a study including 125 patients, Bluemel et al found that 68Ga-PSMA PET/CT detected location of recurrences in 43.8% of choline-negative patients.34 In a study including 319 patients, Afshar-Oromieh et al determined that 68Ga-PSMA PET/CT was positive in 92.3%, 71.8%, and 52.9% of cases when PSA >2 ng/mL, 1–2 ng/mL, and ≤1 ng/mL, respectively.35 Therefore, when compared to our results, 68Ga-PSMA PET/CT has a higher detection rate and is positive at lower PSA values than 18F-FCH PET/CT.
We found that PET/CT had an impact on clinical management in 17 patients (28.3%), in the context of biochemical recurrence. This is lower than in other studies, which showed a change of treatment plan in 48–55% of patients.23,24 These studies used retrospective questionnaires completed by treating physicians and were not blinded, in contrast to our study. This could possibly explain the difference between results.
Our study remains limited by its retrospective design and relatively small sample size. Nevertheless, to the best of our knowledge, our study cohort represents the largest Canadian PCa patient group studied so far.
In conclusion, 18F-FCH PET/CT demonstrates a high detection rate for local, as well as distant recurrences after localized PCa treatment with curative intent. A trigger PSA above 2.6 ng/mL seems optimal for appropriate patient selection. Further studies are needed to better understand the role of 18F-FCH PET/CT in the management of PCa patients.
Competing interests: Dr. Probst has been an advisor for Bayer and has participated in a clinical trial for Progenics. The remaining authors report no competing personal or financial interests.
This paper has been peer-reviewed.