Evaluation of SUV normalized by lean body mass (SUL) in ⁶⁸Ga-PSMA11 PET/CT: a bi-centric analysis

In the last decade, positron emission tomography/computer tomography (PET/CT) has gradually emerged as the standard-of-care imaging modality in the diagnosis and treatment response monitoring of different oncological malignancies [1, 2]. Moreover, standardized uptake value (SUV), which is the commonly used quantitative parameter in PET/CT showed a high predictive value for treatment outcome [3, 4]. However, quantitative SUV is still hampered by a number of physiological, technical and physical factors widely discussed in the literature [5].

In ¹⁸F-FDG PET/CT, SUV normalized by patient’s weight is known to be highly dependent on body weight [6]. Since ¹⁸F-FDG does not significantly accumulate in adipose tissue in the fasting state, the use of SUV can falsely lead to high values in patients with high body mass. Subsequently, SUL (lean body mass (LBM)–corrected SUV) has been proposed as a more appropriate quantitative method, with Janmahasatian formulation for LBM showing most accurate results [7].

The liver is typically used as reference organ in PET imaging, with a 3-cm spherical volume of interest (VOI) computed to measure the liver background activity [8]. In ¹⁸F-FDG PET/CT, liver SUL showed only a fair repeatability between two time points in the same patient [9].

In the past 5 years, prostate-specific membrane antigen (PSMA), a transmembrane protein highly expressed in prostate cancer, has become a promising target for PET in prostate cancer imaging [10]. ⁶⁸Ga-PSMA11 PET/CT has shown enhanced accuracy compared to conventional imaging modalities in lesion detection [11], with SUV being largely used as a quantitative PET-derived parameter [12, 13]. However, the effect of body weight on SUV in ⁶⁸Ga-PSMA11 PET/CT has not been yet investigated. Since ⁶⁸Ga-PSMA11 does not typically accumulate in adipose tissue [14], we hypothesized that liver SUV is dependent from body weight.

The aim of the present study was to investigate whether SUL is a more appropriate quantitative method compared to the commonly used SUV normalized by body weight in ⁶⁸Ga-PSMA11 PET/CT imaging.

To the best of our knowledge, this is the first report evaluating the potential of using SUL as compared to the commonly used SUV as quantitative parameter in ⁶⁸Ga-PSMA11 PET/CT. Our data indicate that a weak but significant positive correlation is present between liver SUV and body weight. Contrarily, SUL as alternative parameter seems to be unaffected from body weight. Since ⁶⁸Ga-PSMA11 does not typically accumulate in adipose tissue the use of SUL can be recommended to avoid any influence from patient habitus.

PET-derived parameters, such as SUV_mean or SUV_max, are increasingly used for therapy response monitoring or patient outcome prediction. Therefore, highly accurate computed parameters should be addressed given their potential decisive role for the clinical image–based decisions. The present study attempted to reproduce a clinical setting where quantitative ⁶⁸Ga-PSMA11 PET/CT scans are used in the framework of ¹⁷⁷Lu-PSMA radioligand therapy.

We investigated the relation of body weight with SUV and SUL in a patient cohort including subcohorts of both European and North-American patients. Consistently among the subgroups, our findings indicate that liver SUL calculated based on Janmahasatian formula annuls the body weight dependence of liver SUV. Noteworthy, even though a significant correlation was found between body weight and SUV, its strength is rather weak (r = 0.35, p = 0.02). However, since SUL (r = 0.23, p = 0.07) annuls and lowers the strength of the positive correlations to BW, the use of SUL should be preferred over SUV. In FDG PET/CT, liver SUV showed a moderate correlation for women (r = 0.58, p < 0.001) and men (r = 0.54, p < 0.001) with body weight, which was annulled and reduced by the Janmahasatian formula, respectively [7].

In addition, we have to stress that the high interpatient repeatability of both liver SUV and SUL between two time points (ICC = 0.821 and 0.818) demonstrates an acceptable mean difference of 1.69% and 0.80%. Despite potential changes of liver ⁶⁸Ga-PSMA11 uptake during ¹⁷⁷Lu-PSMA therapy upon shifts of biodistribution depending on tumor sink effect, the use of liver as the reference organ to threshold malignancy comparing subsequent timepoints is feasible [19]. Notably, it has been shown that only high differences in tumor burden have significant implications on liver ⁶⁸Ga-PSMA11-uptake, with low vs. high tumor load exhibiting a liver SUV_mean of 4.34 vs. 3.27, p < 0.001 [13].

Androgen deprivation therapy (ADT) has shown to increase the PSMA-ligand uptake in the first weeks after treatment initiation in metastatic sensitive prostate-cancer [20]. Moreover, continuous long-term ADT significantly decreased lesion uptake in ⁶⁸Ga-PSMA11 PET imaging [21]. However, these findings might not be valid for patients with metastatic castration-resistant prostate cancer, since most of their tumor lesions are not responding properly to first-line ADT. No significant differences were noted between liver SUV in patients receiving ADT versus not receiving [22]. Same analysis further evaluated the ⁶⁸Ga-PSMA11 uptake of other tissues such blood pool (SUV_mean 1.08) or muscle (SUV_mean 0.50). Nevertheless, the liver showed the most feasible values (SUV_mean 4.73) to be used for PSMA PET quantification. The mean liver SUV obtained in the present analysis (4.36) is in concordance with those obtained by Jansen et al (4.73) and Gaertner et al. (4.25) [13].

Interestingly, in an analysis including 64 patients who received ¹⁸F-DCFPyL PET/CT no correlations were found between both liver SUV and SUL with body weight [23]. Comparing the results, for ¹⁸F-DCFPyL were obtained higher liver SUV and SUL values: 5.1 ± 0.7 vs. 4.4 ± 1.5 and 3.8 ± 0.5 vs. 3.3 ± 1.1 respectively. Similar to FDG, PSMA-ligands do not typically accumulate in adipose tissue, therefore a positive correlation between SUV and BW annulated by SUL was expected to be found for both radiopharmaceuticals.

For PSMA-targeted radioligand therapies, ⁶⁸GaPSMA11 PET imaging is typically used at baseline for patient selection, as well as during treatment for radiographic response assessment. However, the clinical utility of ⁶⁸GaPSMA11 PET in metastatic castration-resistant prostate cancer goes beyond the radioligand therapy, being increasingly used for evaluation of treatment response in patients receiving, e.g., taxanes [24]. Since the traditional SUV has shown a significant correlation with body weight, its clinical value in ⁶⁸Ga-PSMA11 PET quantification remains questionable. Thus, our findings may have clinical implications especially in treatment response assessment. However, further studies comparing the prognostic value of both SUV and SUL for imaging response evaluation are warranted.

Notable limitations of the current analysis are the retrospective nature of the study and the inclusion of a selected patient cohort scheduled for ¹⁷⁷Lu-PSMA radioligand therapy. However, as our patients have shown both high and low tumor load the potential influence of different tumor burden is already acknowledged. Finally, our analysis is focused on the liver as most important normal organ in PSMA-ligand PET imaging severing as reference tissue. We have not investigated the influence of body weight on other normal tissues.

Our results indicate that the Janmahasatian formula annuls the positive correlations between absolute SUV and body weight, suggesting that SUL is preferable to SUV for quantitative analyses in ⁶⁸Ga-PSMA11 PET. Future studies, are necessary to determine the clinical significance of the differences between SUV and SUL for different clinical applications, such as thresholds for delineation of tumor volumes or monitoring tumor response to therapy.