br The acquisition was made by low dose
The acquisition was made by low dose CT imaging (autoMA, 120 KVp) with the patient in supine position with the arms ele-vated, followed by 3D PET with a matrix size of 128 × 128 pixels and an acquisition time of 3–4 min per bed position in the initial study. The time was increased to 6 min for delayed studies. The images were reconstructed with the VUE Point iterative reconstruction algorithm implemented in the tomograph.
The following data were also collected: patient age (years), weight (kg), height (m), Gleason score, PSA value (ng/ml) PSA dou-bling time (PSAdt) (months) and the type of treatment received.
The images were analyzed blinded to the clinical data and taking into account the radiotracer dose administered (MBq), the SUV of the prostate or prostatic bed as well as the presence of lesions in either, adenopathies, bone or other visceral metastases (measuring the corresponding SUVmax , SUmean and lesion size in these cases [mm]).
The OsiriX v.8.0.2 program was used for image review and quan-tification. The criteria of interpretation included the classification of all the deposits observed which could not be justified as physi-ological uptake or by disease not related to the primary tumor and congruent with the natural history of the disease as pathological uptake related to prostate cancer.12,13
SUVmax and SUVmean values (isocontours of 40%) were obtained tracing volumes of interest on the pathological lesions identified. The SUVmean of the L3 vertebra was considered the reference value.
Statistical analysis was performed using the SPSS (IBM; SPS 22.0). The variables are expressed as mean ± standard deviation (median and interquartile range) or as number (n) (%). Variables without a normal distribution underwent logarithmic transfor-mation. The Student’s t and Chi-squared tests were used as appropriate.
In order to anaylze the parameters associated with pathologi-cal findings, we used forward stepwise multiple logistic regression and Wald statistic applying the Hosner and Lemeshow goodness of fit statistic, considering an abnormal PET as a dependent variable, and age, body mass index, having received androgen deprivation, the Gleason scale, the PSAdt and PSA >1 ng/ml as independent vari-ables.
Fig. 1. Lymph node recurrence. Gleason 8(4 + 4) prostate adenocarcinoma. pT3b pN0 cM0, treated with radical prostatectomy and CORM-3 node dissection one year previously.
Radical radiotherapy 7 months before biochemical recurrence. New relapse with PSA 1.32 ng/ml. Negative imaging studies. PET/CT with 68 Ga-PSMA: bilateral iliac involvement.
The study was carried out using an anonymized Excel database and according to the bioethical laws of the Declaration of Helsinki. The study was approved by the Ethics Committee of Investigation of Malaga (Spain). All the patients provided signed informed consent prior to initiation of the study.
Of the 58 studies performed, 5 cases were excluded: 3 for hav-ing been referred for staging and the remaining 2 for evaluation of the indication of treatment with Lutetium-177. Fifty-three studies carried out in 50 patients were included, and of these 36 (68%) were positive (Fig. 1). Table 1 shows the results of the technical and clin-ical variables studied based on whether the study was positive or negative.
Two cases presented positive supradiaphragmatic adenopathies with extensive iliac and retroperitoneal lymph node involvement considered as pathological.
It was of note that we only found significant differences between positive and negative results in the PET study in relation to the Gleason scale, PSA levels, PSAdt and having received androgen deprivation treatment (p < 0.05).
the patients with a PSA < 1 ng/ml and a positive result, 5 showed involvement at the level of the prostate or prostatic bed, 3 had adenopathies (one at a perirectal level and the other 2 at a retroperi-toneal level), and 2 presented bone metastasis. Fig. 3 shows the distribution of the number of negative and positive studies by PSA levels.
Table 2 shows the main PET findings according to the different territories. Only one territory was affected in most of the patients presenting pathological involvement (35 cases; 60.3%), the most frequent being isolated lymph node or prostatic involvement in 15 and 14 patients, respectively.
Patients with a pathological study had a mean PSAdt value of 12.97 months and a median of 6.75 months. Of the patients with a pathological study, the PSAdt was less than 6 months in 46.4% and greater than 6 months in 53.6%.
On multivariate analysis only the Gleason scale was inde-pendently associated with an abnormal PET finding, despite adjustment for other covariables (Table 3).