2.2.1. Impact of Delayed Imaging on Lesion Positivity Rate P-P

A total of 551 lesions in 114 patients were detected on early scans. Twenty-nine of these lesions were local recurrent findings, and 326 were LNM (262 pelvic LNM and 64 extra-pelvic LNM). In delayed images, 328 LNM (262 pelvic LNM and 66 extra-pelvic LNM) were found. A total of 106 patients showed no additional malignant lesions in late images. Three lesions were only found in the late imaging (two extra-pelvic LNM and one bone metastasis). Comparison of tracer accumulation in pathologic lesions between baseline and delayed scans was statistically significant (*p* < 0.001 pelvic LNM, bone metastases), but this increase in maximum standardized uptake value (SUVmax) did not correspond to a significant influence of late images on the lesion positivity rates (LPR) (Table 3, Figure 1).

**Figure 2.** Comparison of baseline and delayed [68Ga]Ga-PSMA-11 positron-emission tomography/computed tomography (PET/CT) regarding tracer uptake of prostate-specific membrane antigen (PSMA)-positive lesions (P-P).



Abbreviations: LNM, lymph node metastases; SUVmax, maximum standardized uptake value; *p* < 0.05 is considered significant.

#### 2.2.2. Impact of Delayed Imaging on Lesion Positivity Rate P-I

In 33 patients, 186 lesions were found in baseline PET/CT (Figure 1). By comparison, the 33 patients showed 187 findings in late imaging. No additional PCa lesions were shown in 32 patients. In total, a single lesion in one patient was noted 3 h p.i. (one local recurrent PCa lesion in the prostate bed). The comparison of SUVmax in pathologic lesions between early and late images was statistically

significant (*p* = 0.008 pelvic LNM). However, there was no significant impact of delayed imaging on LPR.

2.2.3. Impact of Delayed Imaging on Lesion Positivity Rate P-T

All 13 [68Ga]Ga-PSMA-11 PET/CT-positive patients showed 47 lesions (Figure 1). No additional PCa lesions were identified by late imaging.

2.2.4. Total Comparison of Biphasic Lesion Detection

Eleven patients with discordant results showed 51 discordant lesions. All unclear lesions (33 lesions moderately suspicious of malignancy) detected by [68Ga]Ga-PSMA-11 PET/CT on standard imaging (1 h p.i.) could be clarified by additional late images (3 h p.i.). The decision to classify the lesions as malignant was made on the basis of various criteria such as a higher tracer uptake with increased SUVmax in the late images compared to the early images (Table 3), and an improved contrast as well as presentation of the lesions with a more focal character. The assessment was carried out by nuclear medicine and radiological specialists with several years of diagnostic experience with regard to the analysis of oncologic PET/CT imaging of PCa foci.

• LPR on PET, but not on CT:

By comparison of PET and CT imaging separately, nine LNM with high PSMA avidity were detected on PET, but these were not suspect of malignancy on CT alone.

• Lesions only detected on early imaging:

Seven PSMA-avid lesions (five LNM and two bone metastases) were only shown on early imaging (two P-P and zero P-I). These findings could not be confirmed as PCa lesions on delayed images.

• Lesions only detected on delayed imaging:

In this study, 11 lesions suspicious of malignancy were detected exclusively by delayed imaging (eight P-P and one P-I).

• Additional impact of delayed imaging:

A total of 33 PSMA-avid lesions (of which 15 were LNM, eight were bone metastases, and five were lesions in the prostate bed) suspected of being malignant were confirmed as malignant by increased tracer uptake in the delayed scans.

• No additional impact of delayed imaging/concordant lesions:

In 222 of 233 evaluable patients (95%), the baseline PET/CT and the delayed PET/CT were concordant.

• Time dependency of LPR:

PSMA avidity in pelvic LNM was related more often to scan time than in other metastases (e.g., extra-pelvic LNM, bone metastases, visceral metastases) (*p* < 0.001). Comparing early and late PET/CT imaging, the mean SUVmax of pelvic LNM was 25% higher (*p* < 0.001) and the mean SUVmax of extra-pelvic LNM was 14% higher (*p* = 0.003), respectively.
