Introduction
Prostate cancer (PCa) is the second most common cancer in men, with a high prevalence
and incidence worldwide.[1]
Optimal treatment in high-grade tumors with a Gleason score ≥7 includes purportedly
curative prostatectomy and/or radiotherapy. A cancer-specific survival rate of 85%
at 10 years has been demonstrated.[2] However, at least 12% of patients have metastatic involvement and consequently a
shorter survival time.[3]
The well-known low sensitivity at low prostate-specific antigen (PSA) levels of the
previous gold standard radiolabeled choline led to the search for new tracers with
better performance.[4]
Prostate-specific membrane antigen (PSMA) targeted positron emission tomography (PET)/computed
tomography (CT) has proven to be a highly accurate method of detecting PCa lesions,
with better diagnostic accuracy than that of choline. PSMA overexpression on the cell
surface of PCa cells, especially at low PSA levels, is well documented.[5]
The 68Ga-labeled PSMA-targeted PET/CT tracer Glu-urea-Lys(Ahx)-HBED-CC (PSMA-11) has proven
to be highly sensitive in the detection of recurrent PCa,[6] but it has some limitations, such as a short half-life and limited activity per
synthesis.[7]
Several 18F-labeled analogs were introduced into clinical PET imaging in recent years, which
present many advantages over gallium: production on large scale, reasonable cost due
to cyclotron 18F production, lower positron energy with higher image quality, and a more practical
half-life that allows for the acquisition of delayed images. The most studied 18F-labeled tracers are 18F-DCFBC, 18F-DCFPyL, and, most recently, 18F-PSMA-1007.[8]
[9]
A novel agent, 18F-AlF-PSMA-11, was produced at our center with suitable radiochemical purity for clinical
purposes[10] and with the additional advantage of being cheap, since the precursors are synthesized
in situ.
Most data outlining the utility of PSMA-PET techniques came from the setting of biochemical
recurrence after definitive therapy.
The purpose of this study was to prospectively compare the diagnostic performance
of 18F-AlF-PSMA-11 and 68Ga-PSMA-11 in the detection of metastasis in patients with intermediate-/high-risk
PCa at initial staging prior to therapy and their potential impact on patient management.
Materials and Methods
Between September 2017 and March 2019, we prospectively analyzed 66 patients (mean
age: 63 years; range: 52–78; median PSA: 26.5 ng/mL; range: 1.7–152 ng/mL; [Table 1]) with localized intermediate-/high-risk PCa according to the D'Amico classification,[11] prior to radical prostatectomy (Gleason score ≥6). Patients with confirmed metastasis,
concomitant cancers, or benign pathology were excluded. Patients underwent clinical
and laboratory preoperative staging with PSA, biopsy Gleason score grading, bone scintigraphy,
and CT.
Table 1
Patients' characteristics
|
Patient's characteristics
|
|
Age (y)
|
Mean: 63.1 ± 6.5; median: 63; range 52–78
|
|
PSA (ng/mL)
|
Mean: 26.5 ± 31.4; median: 14; range 1.7–152
|
|
Gleasson score of 6
|
8 patients (12.1%)
|
|
Gleasson score of 7
|
38 patients (57.6%)
|
|
Gleasson score of 8
|
11 patients (16.6%)
|
|
Gleasson score of 9
|
8 patients (12.2%)
|
|
Gleasson score of 10
|
1 patient (1.5%)
|
Abbreviation: PSA, prostate-specific antigen.
Radiopharmaceuticals
18F-AlF-PSMA-11
The complex (18F)AlF-PSMA was obtained from an automated Tracerlab FXFN (GE) platform. A radiochemical
purity higher than 90% (95 ± 3%) was achieved. Stability was verified for 4 hours
in the final formulation vial and for up to 1 hour in human plasma.[10]
68Ga-PSMA-11
68Ga-PSMA-11 was produced using PSMA-11 (HBED-CC) from ABX as a precursor and 68Ga eluted from an 68Ge/68Ga generator (ITG, Germany). The precursor (3.2–3.6 nmol) was dissolved in ultrapure
water and mixed with 1.00 mL of 0.25-M sodium acetate and 650- to 1,450-MBq 68GaCl3 in 4 mL of 0.05-M HCl. After 5 minutes of incubation at 100°C, 68Ga-PSMA-11 was purified and sterilized. The radiochemical purity was 99.2 ± 1.7% and
specific activity was 170 ± 76 MBq/nmol.
Imaging
All patients underwent a 64-slice PET/CT scan with both 18F-AlF-PSMA-11 and 68Ga-PSMA-11 within 1 to 2 weeks. The scans were performed 60 minutes after tracer injection,
with a dose of 4.0 MBq/kg for 18F-AlF-PSMA-11 and 2.0 MBq/kg for 68Ga-PSMA-11.
The CT parameters were the following: tube voltage, 120 kVp; autoMA, 80 to 180 mA;
index noise, 30, “GE SmartMa dose modulation”; rotation time, 0.8 seconds; rotation
length/full helical thickness, 3.75 mm; pitch, 1.375:1; and speed, 55 mm/rotation.
PET data were acquired in 3D with a scan duration of 2 minutes (18F-AlF-PSMA-11) or 3 minutes (68Ga-PSMA-11) per bed position and with 11-slice overlap. Images were reconstructed
using an ordered subset expectation maximization algorithm (OSEM) with time-of-flight
correction (matrix size: 128 × 128 pixels) with 2 iterations/24 subsets.
Image Analysis
Images were evaluated by two board-certified specialists in nuclear medicine and by
one radiologist.
Suspicious metastatic lesions (“positive” study) were defined as any focal uptake
at any location, higher than surrounding background or normal tissue, localized by
hybrid images, excluding joint processes and areas of physiological uptake.[12] For image interpretation, we also consider the Promise criteria, the recently published
E-PSMA (European Association of Nuclear Medicine-EANM standardized reporting guidelines
por PSMA-PET imaging in prostate cancer) and the PSMA-RADS (Prostate-specific Membrane
Antigen Reporting and Data System).[13]
[14]
Lesions were evaluated regarding their localization (bone, lymph node [LN], or soft
tissue metastases) and their maximum standardized uptake values (SUVmax) selecting
gluteal musculature as background.
We measured the lesion-to-background ratio (lesion-SUVmax/background-SUVmax; LBR)
in all prostate glands and in all coincident lesions. Afterward, prostatectomy and
pelvic LN dissection were performed in nonmetastatic patients.
Gold Standard
Histopathology, correlative imaging, and/or clinical follow-up were considered the
reference standard to assess the diagnostic performance on a per-patient and per-lesion
basis as follows:
Statistical Analysis
A systematic comparison was performed between the results obtained with both tracers
regarding the number of detected PET-positive lesions, the SUVmax value, the LBR,
the PSA, and Gleason score. The assessment of the normality distribution of variables
was done via the Shapiro–Wilk test. The nonparametric Spearman rho test and the Pearson
correlation coefficient were used to measure the strength of association between SUVmax
and LBR in all concordant lesions. The tumor SUVmax and tumor-to-background ratio
signals from the same lesions in both the 18F-AlF-PSMA-11 and 68Ga-PSMA-11 studies were statistically analyzed using a nonparametric two-sided Wilcoxon
signed-rank test and the Student paired t-test. The two-sided McNemar test was used to analyze whether 68Ga-PSMA-11 PET/CT detects significantly more lesions suggestive of PCa when compared
with AlF-based PET/CT. Sensitivity and specificity were selected to describe the diagnostic
performance of PET/CT in detecting metastasis involvement. A p-value of 0.05 was considered significant. For statistical analyses, we used the Statistical
Package for the Social Sciences (SPSS) version 23.
Results
We enrolled a total of 66 patients. The mean age was 63.1 ± 6.5 years (median: 63
years; range: 52–78 years) and the mean PSA level was 26.5 ± 31.4 ng/mL (median: 14 ng/mL;
range: 1.7–152 ng/mL).
Both PET tracers demonstrated abnormal findings in 56 patients (positivity rate:85%)
and were negative in 10 patients (15%). At least one lesion suspicious for metastasis
was detected in 20/66 (30%) and 21/66 (32%) patients for 68Ga-PSMA-11 and 18F-AlF-PSMA-11 PET/CT, respectively ([Table 2]).
Prostatic focal lesions that showed moderate to high radiotracer uptake above the
SUVmax of 2.5 were seen in 54/66 patients (81.8%) in 68Ga-PSMA-11 PET/CT scans and in 53/66 patients (80.3%) in 18F-AlF-PSMA-11 PET/CT scans. In the remaining patients, the primary tumor showed no
tracer accumulation or only diffuse tracer accumulation.
A total of 145 extra-prostatic lesions were detected by at least one radiopharmaceutical
in bone (n = 56), LNs (n = 88), and lung (n = 1), for a total of 131 for 68Ga-PSMA-11 and 123 for 18F-AlF-PSMA-11. Discriminating by radiotracer in the total population, PET/CT with
gallium and fluorine identified 42 and 53 bone lesions and 88 and 69 LN lesions, respectively
([Figs. 1] and [2]).
Fig. 1 Maximum intensity projection (MIP) and axial slices of fusion positron emission tomography
(PET)/computed tomography (CT) of a patient with early-stage prostate cancer scanned
with (A) 18F-AlF-PSMA-11 and (B) 68Ga-PSMA-11 PET/CT. (C) Prostate lesion, (D) right external iliac lymph node metastasis, and (E) bone lesion in T7 demonstrated with both tracers.
Fig. 2 Maximum intensity projection (MIP) of 18F- AlF-PSMA-11 (A ) and 68Ga-PSMA-11 (B ) and axial slices of fusion positron emission tomography (PET)/computed tomography
(CT) demonstrating coincident metastatic lymph node and bone lesions (C –E ). Note the better contrast and uptake intensity of 18F-AlF-PSMA images in some bone lesions (yellow arrows).
In concordant lesions (n = 177), a significant correlation was found between the SUVmax of both radiopharmaceuticals
(r = 0.90; 95% confidence interval [CI]: 0.87–0.93; p = 0.001). The correlation between the 68Ga-PSMA-11 and 18F-AlF-PSMA-11 SUVmax ratio and background lesions was also significant (0.91; 95%
CI: 0.89–0.94; p = 0.001). The global median SUVmax showed higher values for 68Ga-PSMA-11 (9.1 ± 9.6 vs. 7 ± 9.5; r=0.91); also, the median was higher for 68Ga-PSMA-11 LBR (13.3 ± 23.6 vs. 12.0 ± 16.7; p = 0.001 for both tests).
We found a significantly higher median SUVmax for 68Ga-PSMA-11 compared with that of 18F-AlF-PSMA-11 in LN (n = 69) and prostate foci (n = 69): 10.2 ± 12.5 (3.00–74.12) versus 7.07 ± 12.8 (2.91–72.92) and 7.86 ± 7.2 (2.79–34.93)
versus 7.4 ± 6.3 (2.00–34.93) for each tracer, respectively (p = 0.001). The 18F-AlF-PSMA-11 SUVmax and LBR were higher in bone foci (n = 39) compared with that of 68Ga-PSMA-11 (8.92 ± 12.8 vs. 7.2 ± 12.5 and 13.0 ± 8.6 vs. 10.9 ± 8.9, respectively;
p = 0.02; [Fig. 3]).
Fig. 3
18F-AlF-PSMA-11 and 68Ga-PSMA-11 positron emission tomography (PET)/computed tomography (CT) images show
abnormal tracer uptake in unsuspected metastases in the right ischiopubic ramus (A - B), left acetabulum (C-D), coccyx (E-F), and in the spinous process of T12 (G-H). Note the poor bone uptake with 68Ga-PSMA-11.
Rising SUVmax values were not associated with rising Gleason categories in 68Ga-PSMA-11 (r = 0.01; p = 0.88) or 18F-AlF-PSMA-11 PET/CT (r = 0.02; p = 0.753). No significant associations were observed between PSA and SUVmax gallium
(Spearman's rho = 0.001; p = 0.986) or fluorine (Spearman's rho = 0.089; p = 0.188). No relationship was found between the SUVmax and the location of the detected
abnormal foci.
There was a strong correlation between the number of lesions detected by gallium and
by fluorine (r = 0.96; Spearman's rho = 0.95; p = 0.001). We also found a significant correlation between the Gleason score and the
number of lesions detected by both fluorine and gallium (r = 0.214 and 0.215; Spearman's rho = 0.31 for each tracer; p = 0.01).
The correlations between PSA and the number of lesions detected were weakly positive
(r = 0.355 with gallium and r = 0.25 with fluorine; Spearman's rho = 0.24 and 0.25, respectively; p = 0.06 and 0.04, respectively).
We studied the diagnostic accuracy of both techniques in identifying regional and
distant metastases. Of the total patient sample (n = 66), 28 underwent radical prostatectomy with extended pelvic lymphadenectomy. The
rest of the patients were treated with radiation therapy (n = 38; 57.6%), associated or not associated with hormone therapy.
For the detection of systemic metastasis on a per-patient basis, the sensitivity values,
with their 95% CI, were the same for both techniques: 0.90 (95% CI: 0.68–0.98). We
calculated specificities of 0.96 (95% CI: 0.85–0.99) and 0.94 (95% CI: 0.82–0.98)
for 68Ga-PSMA-11 and 18F-AlF-PSMA-11, respectively.
In a per-lesion analysis, the total number of lesions studied was 225, of which 219
were gold standard positive. There were six false-positive bone results in three patients
detected with both tracers, corresponding to lesions with moderate diffused uptake.
PET/CT changed clinical management in 45/66 patients (68%) who were upstaged due to
the detection of bone lesions or extra-pelvic LNs that were undetectable by other
modalities (bone scan, technetium imaging, or magnetic resonance [MR] imaging). Treatment
was thus shifted from surgery or local radiation to systemic therapy in those patients.
Discussion
Our preliminary data highlight the growing body of evidence supporting the use of
radiolabeled PSMA-PET/CT for primary staging of intermediate-/high-risk PCa patients.
The introduction of PET/CT and PET/MR technology based on molecular information has
led to a revolution in imaging diagnosis in PCa, improving the detection of metastatic
disease compared with CT or MRI.[15]
Radioactive labeled choline showed a high specificity (95%) but only a moderate sensitivity
(33–50%) in the detection of metastatic LN spread,[16] and resulted in research for more sensitive radiotracers.[17]
The glutamate carboxypeptidase type 2 PSMA is overexpressed on the cell surface of
PCa cells specially in advanced-stage and castration-resistant metastatic PCa.[18] The most widely studied agent based on low-molecular-weight urea type structures
is the 68Ga-labeled PSMA inhibitor Glu-NH-CO-NH-Lys(Ahx)-HBED-CC. In addition, after this tracer
binds to PSMA, internalization occurs, and this can be used for a theragnostic approach
using 177lutetium.[19]
Several studies have demonstrated the diagnostic superiority of 68Ga-PSMA versus choline PET/CT in assessing PCa patients.[12]
[17]
[18] This was also reported by our group.[20] However, 68Ga-labeled compounds are produced with generators providing limited activity per synthesis
(1–4 patients per batch) and have some shortcomings, such as a short half-life (68 minutes)
and nonideal energies.[7]
Compared with 68Ga-labeling compounds, 18F-PSMA agents constitute an attractive alternative that offers promising advantages
concerning availability, production amount, and image resolution.[21] Dietlein et al demonstrate the advantage of labeling this type of peptide with 18F, with the additional advantage of transportation from in situ cyclotron to distant
PET centers.[7]
Mease et al described a first generation of 18F-labeled PSMA ligand, 18F-DCFBC, with the potential for the detection of metastatic PCa.[8] Chen et al have introduced a second generation of 18F-labeled PSMA ligand,18F-DCFPyL, suggesting its high potential for PSMA radiolabeled PET imaging.[22] Recently, it has been demonstrated that 18F-PSMA-1007 presents similar behavior to that of 68Ga-PSMA-11 and other 18F-labeled PET tracers, with reduced urinary clearance.[9]
In addition, recent research has emerged using Al18F labeling in patients. Yu et al[23] performed an evaluation of integrins labeled with Al18F in healthy volunteers. In recent years, the possibility of labeling PSMA with 18F has emerged using the AlF + complex.
Based on the work of Allott et al,[24] tagging and purification of 18F-AlF-PSMA was optimized, transferring the manual synthesis to an automatic module
and producing a batch of the radiopharmaceutical with high activity in a safe and
effective way.
Our group was the first to optimize the synthesis of an Al18F radiofluorinated GLU-UREA-LYS(AHX)-HBED-CC PSMA ligand in an automated synthesis
platform with suitable radiochemical purity for clinical purposes expecting that 18F-AlF-PSMA-11 would perform comparably to other 18F-labeled PSMA targeted tracers, with the additional advantage of low cost.[10]
Recently, 18F-AlF-PSMA-11 has been suggested as a novel and attractive alternative to 68Ga-PSMA-11 and other 18F-PSMA tracers with accessibility and commercial advantages as well as similar tumor
uptake among them.[25]
[26]
Interestingly, a recent study made a head-to-head comparison between 18F-AlF-PSMA-11 and 68Ga-PSMA-11 in diagnosing PCa. They demonstrated that 18F-AlF-PSMA-11 can bind to receptors stably and persistently; in the meantime, a clearer
metabolic background was manifested when compared with 68Ga-PSMA-11 PET, indicating that 18F-AlF-PSMA-11 has great potential as an alternative PSMA imaging agent to 68Ga-PSMA-11 since it allows for further characterization of lesions through delayed
imaging.[27]
Finally, while many studies have been published on the role of radiolabeled PSMA-PET/CT
in the recurrent PCa biochemical scenario, only a small number of studies explored
its use in the primary staging of patients with intermediate-/high-risk PCa prior
to therapy.[28] Notably, radiolabeled PSMA-PET imaging in combination with multiparametric MRI could
enable a complete staging with increased accuracy and additional molecular information.[29]
In this prospective, descriptive, and observational study, we compared the diagnostic
value of 68Ga-PSMA-11 PET/CT with that of 18F-AlF-PSMA-11 for the detection of metastatic disease in 66 patients with intermediate-/high-risk
PCa at initial staging prior to radical prostatectomy and their overall impact on
patients' management plans.
Each lesion was delineated as a true- or false-positive/negative result based on follow-up,
imaging, biopsy, and follow-up PSA levels after treatment.[30] Lesions indicative of PCa in PET/CT were detected in 56/66 patients (85%) with both
tracers.
Abnormal prostate gland foci were clearly observed with 68Ga-PSMA-11 (n = 54; 81.8%) or 18F-AlF-PSMA-11 (n = 53; 80.3%), while the rest of the patients showed non/diffuse/irregular uptake.
These data are similar to previous reports in which it was revealed that more than
90% of primary PCa show moderate to high PSMA expression levels by 68Ga-PSMA-PET.[31] It is known that less than 10% of primary cancer tumors may not overexpress PSMA[28] and that false-negative scans may occur if PCa is poorly differentiated or displays
neuroendocrine aberrations.[18] Furthermore, low PSMA expression caused by tumor heterogeneity[32] might be responsible for occasional false-negative PET/CT results.
It has been proven that 68Ga-PSMA-11 PET/CT reveals the highest contrast in LN metastases, followed by bone
metastases, local relapses, and soft tissue metastases.[12] Due to low background signal, 68Ga-PSMA-11 PET/CT allows the detection of bone and organ metastases,[17] which may lead to systemic therapy, but if excluded, may lead to curative therapy.[33]
A total of 131 lesions were detected in 20 patients using 68Ga-PSMA-PET/CT, and 123 lesions were detected in 21 patients using 18F-AlF-PSMA-11 PET/CT. Our results corresponded with those reported in the literature.[31] In concordant lesions, there is a strong correlation between the values of 68Ga-PSMA and 18F-AlF-PSMA-11 SUVmax (r = 0.9; p = 0.001).
The mean SUVmax in the concordant PSMA-positive lesions was 11.9 for 68Ga-PSMA-11 and 10.5 for 18F-AlF-PSMA-11 (p = 0.001; n = 177 lesions). We also found a significantly different median LBR for both tracers:
13.3 (0.95–176) and 12.0 (2.67–132) for 68Ga-PSMA-11 and 18F-AlF-PSMA-11, respectively (p = 0.001).
68Ga-PSMA-11 demonstrated a slightly higher SUVmax, tumor-to-background ratio, and median
and mean values as compared with 18F-ALF-PSMA-11 in coincident lesions. These data are also similar to previous reports
concerning these PSMA ligands.
Dietlein et al[7] did not find significant differences in SUVmax and tumor-to-background ratios between
(18F)DCFPyL and (68Ga)Ga-PSMA-HBED-CC. However, additional skeletal metastases were observed for (18F)DCFPyL as compared with (68Ga)Ga-PSMA-HBED-CC. Despite these differences, good correspondence between both tracers
was observed. In our study, the 18F-AlF-PSMA-11 SUVmax and LBR were higher in bone-concordant foci compared with68Ga-PSMA-11 (7.2 0 vs. 8.92 and 14.4 vs. 13.1, respectively). However, a significant
time-dependent degradation of [18F]AlF-PSMA-11 has been reported and PET acquisition should take place no later than
1 hour postinjection. The limited instability and consequently physiological nonspecific
bone uptake due to the potential release of free fluoride might influence SUVmax and
LBR calculations, or eventually hamper the visualization of small PCa bone metastases.[25]
[26]
A strong correlation was found between the SUVmax values of gallium and fluorine (r = 0.83; p = 0.001) and between the number of lesions detected by both radiopharmaceuticals
(r = 0.96). Although there is a very high correlation, we clarify that we cannot conclude
identical detection capabilities, since subtle differences in lesion characterization
(e.g., intensity, size) could be missed by one tracer over the other.
Hoffman et al found a tendency toward increasing SUVmax with rising PSA.[34] Moreover, Afshar-Oromieh et al described a strong association between PSA level
and positive 68Ga-PSMA-11 PET/CT scans.[12] Koerber et al observed a significantly higher mean SUVmax in tumors with higher
D'Amico risk classification and Gleason score from biopsy (p = 0.001 for grouped analyses).[35] However, in our research, rising SUVmax values were not associated with rising Gleason
categories, PSA levels, or the type of injury detected. These contradictions can be
explained by differences in the selected population or characteristics of the research.
Hoffmann et al studied several patients with suspected (but not confirmed) prostate
carcinoma and he found a “tendency,” but not significant correlation between variables.[34] Afshar-Oromieh et al's results concern the scenario of recurrent PCa, but not initial
staging.[12] Koerber et al[35] obtained statistical significance in high-risk but not intermediate-risk tumors.
There is controversy in the literature concerning distant metastasis and regional
LN involvement. Based on the results of Steuber et al's research, choline-derivative
PET/CT scans did not prove to be useful for LN staging in localized PCa prior to treatment
and should not be applied if clinically occult metastatic disease is suspected, suggesting
the research of new tracers with higher affinity.[36]
As normal lymphatic or retroperitoneal fatty tissue does not exhibit PSMA expression,
a metastatic LN can be detected with a favorable lesion-to-background ratio.[31] Initial studies in recurrent PCa scenarios with 68Ga-PSMA described the first promising results.[34] At initial staging, some authors achieved a sensitivity of 88.1% for LN detection
with 68Ga-PSMA-PET/CT.[12]
Maurer et al suggest that preoperative LN staging with 68Ga-PSMA-PET is superior to standard routine imaging.[31] They were the first to study a large cohort of 130 patients with intermediate-/high-risk
PCa, and in a patient-based and template-based analysis, they observed a sensitivity
of 65.9 and 68.3%, respectively, with a specificity of 98.9 and 99.1%, respectively,
for LN staging.
Many published studies do not recommend routine clinical use of PET/CT technology
to detect occult LN metastasis prior to initial treatment because they suggest that
detection of microscopical tumor metastases might be missed due to the poor sensitivity
of the technique.[36]
Giesel and his team demonstrated an excellent sensitivity of 94.7% with 18F-PSMA-1007 PET/CT in detecting LN metastases in the pelvis, including nodes as small
as 1 mm,[9] although in other series, the sensitivity to these very small nodes was limited.
On the other hand, an analysis evaluating 68Ga-PSMA-PET for LN staging in 30 high-risk PCa patients reported a low sensitivity
of 33% and a specificity of 100%.[37] Concerning the excellent sensitivity of 18F-PSMA-1007 (94.7%),[9] versus the low sensitivity of 68Ga-PSMA-PET (33.0%)[37] for LN staging, the mentioned study suffered from several limitations, including
a long interval between imaging and surgery or the lack of standardization in regard
to imaging protocols and documentation of findings. Therefore, it cannot be concluded
that the higher sensitivity in the first cohort is caused solely by the improved tracer.
Nevertheless, LN metastases with median diameters of 5 mm are close to the technical
resolution limits of PET with 68Ga-PSMA tracers and, therefore, it would be reasonable that 18F-PSMA tracers be perform at a higher level.
Also, concerning the two different sensitivity values for 68Ga-PSMA-PET in LN staging (65.9% for Maurer et al[31] and 33.0% for Budäus et al[37]), the reasons for this finding can only be speculated as has been highlighted in
a recent reply by the authors: a highly dedicated but nonroutine pathological workup
with PSMA immunohistochemistry of all resected LN has been performed. However, and
probably more relevant, a central standardized review of the PSMA-PET scans performed
at external institutions by an experienced nuclear medicine physician was completely
missing.
In our study, for the detection of systemic metastasis on a per-patient basis, the
sensitivity was the same for both techniques: 0.90 (95% CI: 0.68–0.98). We calculated
specificities of 0.96 (95% CI: 0.85–0.99) and 0.94 (95% CI: 0.82–0.98) for 68Ga-PSMA-11 and 18F-AlF-PSMA-11, respectively.
Three LNs in two different patients were histologically positive for metastasis, indicating
a false-negative PET/CT result. This is in line with the findings of Budäus et al,
who concluded that 68Ga-PSMA-PET/CT performance may be susceptible to the micro-size of tumor deposits
(<5 mm).[37]
In our initial selection, five patients were excluded because they had additional
tumors besides prostate tumors. PSMA expression has been reported in the neovasculature
of some solid tumors, such as those related to colon, breast, bladder, and renal cancer,
and this fact can limit the technique's specificity.[38]
Six false-positive lesions found were in the bone area (2/6 detected by 68Ga-PSMA-11 and 6/6 for 18F-AlF-PSMA-11; [Table 2]*). One of them corresponded to a rib pitfall and the others were verified retrospectively
by CT as trauma/fracture. In all cases, patients underwent the intended curative primary
prostate treatment, and the PSA turned indetectable, confirming the false-positive
result. Recent studies have reported occasionally increased radiolabeled PSMA uptake
in rib fractures, as well as in benign bone diseases, such as fibrous dysplasia and
Paget's disease.[39] As such findings must be taken into account when reporting PET/CT results, the presence
of benign disease was an exclusion criterion for our clinical assay.
Although the literature refers to false-positive findings in relation to uptake in
retroperitoneal celiac ganglia,[40] in this analysis, they were easily identified and not taken into account.
Concerning ethical limitations, positive PET/CT extra-pelvic findings were not histopathologically
validated in most patients, as many studies reported in the literature.[35] However, it is not possible to assess all lesions in multimetastatic patients, and
on the other hand, small lesions are difficult to reach anatomically. The high sensitivity
in this research may be related to the relatively low average follow-up time, which
may have prevented distant metastases from becoming clinically evident in our time
window. Obviously, it is important to note that our study focuses on patient-based
diagnostic values. Finally, although our study has more patients than other published
studies, we highlight the need for prospective studies with adequate numbers of patients
to achieve strong statistical significance.
Both PET/CT studies accordingly changed the management in 20/66 patients (30%) from
supposedly curative therapy options (radiotherapy/prostatectomy) to systematic therapy
in cases of distant metastasis, radically changing patients' management.
