Pre-therapy extrahepatic 68 Ga-DOTATATE avid tumor burden is associated with short-term clinical outcomes of 177 Lu-DOTATATE in advanced metastatic gastroenteropancreatic neuroendocrine tumors

: Lutetium-177 ( 177 Lu)-DOTATATE is an effective systemic therapy for metastatic somatostatin receptor positive neuroendocrine tumors (NETs). Here we report our experience with the use of pre-therapy 68 Ga-DOTATATE PET as prognostic marker for short-term clinical outcomes of 177 Lu-DOTATATE therapy in patients with advanced NETs. Materials and methods : We retrospectively reviewed patients who received at least one dose of 177 Lu-DOTATATE between Dec. 2016 and July 2019 at our institution. 50 patients (63.6 ± 10.0 years) with advanced gastroenteropancreatic neuroendocrine tumors (GEP-NETs) who had pre-therapy 68 Ga-DOTATATE PET were included in the analysis. 68 Ga-DOTATATE avid tumor volumes were determined automatically using an SUV thresholding approach. Total and extrahepatic 68 Ga-DOTATATE avid tumor volumes were measured and dichotomized into large and small tumor volume groups. Association with progression free survival (PFS) and overall survival (OS) were determined at median follow up of 32 months by Kaplan-Meier survival analysis with Log-Rank test. Results: During follow 38 patients (76%) had disease progression and 15 patients (30%) died . Kaplan-Meier analysis of PFS in GEP-NETs showed that smaller extrahepatic 68 Ga-DOTATATE avid tumor volume (<140 mL) is associated with significantly longer PFS (Median PFS 29.0 ± 6.7 months vs 9.0 ± 1.7 months, P = 0.0001). This trend in PFS is less prominent when total 68 Ga-DOTATATE avid tumor volume is analyzed. Similarly, Kaplan-Meier analysis of OS that GEP-NETs patients with smaller extrahepatic 68 Ga-DOTATATE avid tumor volume (<150 mL) is associated with significantly longer OS (Median OS not reached vs 44.0 ± 12.3 months, P = 0.002). This association with OS is not statistically significant when total 68 Ga-DOTATATE avid tumor volume is analyzed. When 68 Ga-DOTATATE avid hepatic tumor volume is grouped into low (<500 mL), medium (500-1000mL) of extrahepatic 68 Ga-DOTATATE avid tumor volume as prognostic marker for PFS and OS at 32 months are moderate at 58% and 72%. Conclusions: Smaller extrahepatic 68 Ga-DOTATATE avid tumor volumes are associated with longer PFS and OS following 177 Lu-DOTATATE treatment in patients with advanced GEP-NETs. The accuracy of extrahepatic 68 Ga-DOTATATE avid tumor volume as prognostic marker for PFS and OS at 32 months are moderate, which may limit its clinical application.


Introduction
Neuroendocrine tumors (NETs) are a group of heterogeneous neoplasms arising from the neuroendocrine system which most commonly includes endocrine islet cells of the pancreas and endocrine cells of the gastrointestinal and respiratory tracts [1][2][3]. The most common sites of primary NETs are the gastrointestinal tract and about 12-22% of the patients had metastatic disease at presentation [4]. In the past decade, innovations in imaging and the development of novel therapeutics such as somatostatin analogues [5,6], Everolimus [7,8], Sunitinib in pancreatic NET [9], Temodar/Capecitabine in pNET [10] and 177 Lu-DOTATATE [11] have significantly improved PFS in patients with metastatic disease [12][13][14].
Lutetium-177 ( 177 Lu) DOTATATE (Lutathera ® ) was approved by the US FDA in January 2018 for the treatment of somatostatin-receptor positive gastroenteropancreatic neuroendocrine tumors (GEP-NETs) after a phase III clinical trial (NETTER-1) showed significantly longer progression-free survival in patients with advanced midgut NETs compared to patients receiving high dose octreotide long-acting release (LAR) [11]. Although it is still debatable at what stage in the course of the disease patients with advanced NETs will benefit the most from 177 111 In-pentetreotide scintigraphy (Octreoscan) and better reflects SSTR2 expression in small lesions [15,16], for which reason 68 Ga-DOTATATE PET has replaced scintigraphy in clinical practice. Given that the NETTER-1 trial used pre-therapy 111 Inpentetreotide scintigraphy for patient selection, 68 Ga-DOTATATE PET imaging is likely to increase the eligible patient population that could potentially receive 177 Lu-DOTATATE therapy. 68 Ga-DOTATATE PET is now widely used in the diagnosis and clinical follow up of patients with well differentiated NETs [17][18][19]. Furthermore, 68 Ga-DOTATATE PET has been studied to predict prognosis of NETs patients [20,21]. Tirosh et al. has demonstrated that higher total 68 Ga-DOTATATE avid tumor volume is associated with worse PFS and increased risk of disease specific mortality [22]. Similarly, our group [23] has shown that total somatostatin receptor expressing tumor volume and other derived metrics have prognostic value for progression free survival (PFS) in patients with well-differentiated NETs. However, only a small fraction of the patients in these studies received PRRT. Here we evaluated the use of pretherapy 68 Ga-DOTATATE as a prognostic marker for short-term clinical outcomes following 177 Lu-DOTATATE therapy.

Patient population
We retrospectively reviewed the images and medical records of 63 patients who received at least one dose of 177 Lu-DOTATATE treatment from December 2016 to July 2019 at our institution. 50 patients (37-80 years old, mean ± SD: 63.6 ± 10.0; 21 male and 29 female) with advanced GEP-NETs who had pre-therapy 68 Ga-DOTATATE PET were included in the study.
Among them, 11 patients were in the NETTER-1 trial expanded access protocol. Institutional Review Board approved this retrospective study and waived the requirement for obtaining written informed consent.
Patients were screened and had clinical laboratory workup prior to each PRRT administration, which was withheld according to the following key exclusion criteria: 1. Serum creatinine > 1.7 mg/dL or GFR <50 mL/min; 2. Hgb < 8.0 g/dL, WBC < 2000/mm 3 , platelets < 75 x 10 3 /mm 3 ; 3. Total bilirubin > 3 x upper limit of normal; 4. Serum albumin <3.0 g/dL unless prothrombin time is within the normal range; 5. Pregnancy or lactation. 6. Any surgery, radioembolization, chemoembolization, chemotherapy and radiofrequency ablation within 12 weeks prior to treatment. 7. Interferons, everolimus or other systemic therapies within 4 weeks prior to treatment; 8. Known brain metastases, unless these metastases have been treated and stabilized. 9. Uncontrolled congestive heart failure or diabetes mellitus.  reported [22,23]. Briefly, using the whole-body automatic contour tool, a SUVbw (SUV body weight) based threshold approach was applied to delineate every 68 Ga-DOTATATE avid tumors.
The PET avid tumors were then compared to the anatomic images to optimize the SUVbw threshold that maximizes the overlap between PET avid and anatomic tumor volumes by visual inspection. For PET/CTs, a SUVbw threshold of 4.0-4.5 were applied to delineate tumor volumes that are most consistent with anatomic tumor volumes. Background non-specific and physiologic uptake, most commonly in the spleen, kidneys, pancreas, stomach, pituitary and adrenal glands were visually inspected and manually removed using the 3D brush tool. The total 68 Ga-DOTATATE avid tumor volume and maximal SUV (SUVmax) measurements were performed automatically. The hepatic uptake was then removed manually, and the remaining extrahepatic tumor volumes were measured automatically. The non-osseous uptake was then removed manually, and the osseous metastases tumor volume were measured automatically.
Overall survival and progression free survival    Table 1.
Most patients tolerated 177 Lu-DOTATATE treatment well. 36 out of 50 patients (72%) completed all four therapy cycles of 177 Lu-DOTATATE (Table 2), which is comparable to 77% observed in the NETTER-1 trial [11]. Eight patients developed adverse events: thrombocytopenia (n=5, 10%), neutropenia and thrombocytopenia (n=1, 2%), low albumin (n=1, 2%), elevated total bilirubin (n=1, 2%). These led to delays or treatment termination (Table 2), which is comparable to the toxicity profile observed in the NETTER-1 trial as well. Two patients recovered from mild thrombocytopenia and completed all four infusions. A few patients had mild nausea and vomiting that were controlled with medication and occurred at the time when a standard amino acid solution was used before it was replaced with two amino acid (arginine and lysine) only compound solution.

PFS and OS analysis
The median follow-up after first dose of 177 Lu-DOTATATE treatment is 32 months (range 21 to 52 months). Kaplan-Meier analysis of PFS found a range of cut-off extrahepatic 68 Ga-DOTATATE avid tumor volumes (range from 130 mL to 180 mL), when used to dichotomize GEP-NETS patients into two groups with large and small extrahepatic tumor volumes, that are associated with significantly longer PFS in the smaller tumor volume group (Table S1). For example, when cut-off tumor volume 140 mL is used, smaller extrahepatic 68 Ga-DOTATATE avid tumor volume (<140 mL, 29 out of 50 patients) is associated with significantly longer PFS (Median PFS 29.0 ± 6.7 months vs 9.0 ± 1.7 months, P = 0.0001) (Fig. 3a). This trend in PFS is less prominent when total 68 Ga-DOTATATE avid tumor volume is analyzed (Table S1), for example, for cut-off total tumor volume 1000 mL, patients with smaller total tumor volume had median PFS 24.0 ± 5.4 months vs 10.0 ± 3.4 months in group with larger total tumor volume, P = 0.04 (Fig. 3b).
Similarly, Kaplan-Meier analysis of OS found a range of cut-off extrahepatic 68 Ga-DOTATATE avid tumor volumes (130 mL to 190 mL), when used to dichotomize patients into large and small tumor volume groups, are associated with significantly longer PFS in the smaller tumor volume group (Table S1). Kaplan-Meier curves were shown for cut-off value 150 mL in  (Table S3, Fig. 4a). Apparent decrease in P value at large cut-off values of 125 mL is considered artifactual. We then analyzed hepatic metastatic volume and its association with PFS. Several cut-off 68 Ga-DOTATATE avid hepatic tumor volumes were found to dichotomize patients into large and small tumor volume groups where an association with statistically different PFS was present, however, such trend is less prominent compared to extrahepatic tumor volume (Table S3, Fig.   4b). However, when 68 Ga-DOTATATE avid hepatic tumor volume is grouped into low (<500 mL), medium (500-1000mL) and large (> 1000 mL) tumor volumes, no statistically significant difference in PFS is observed, P = 0.19 (Fig. 4C). Our finding that larger 68 Ga-DOTATATE-avid hepatic tumor volume is not associated with worse PFS after 177 Lu-DOTATATE therapy is consistent with recent analysis of liver tumor burden among patients enrolled in the 177 Lu-Dotatate arm of NETTER-1 trial [25]. Similarly, no statistically significant difference is observed between groups with higher and lower SUVmax values (Fig. 4D). In addition, when extrahepatic tumor volume was visually inspected and dichotomized into small and large tumor volume groups (Supplement Fig. 1), Kaplan-Meier survival analysis showed similar association where smaller extrahepatic tumor volume is associated with longer PFS, P = 0.005 (Fig. 5).

Discussion
In this single center retrospective study, we evaluated pre-therapy 68 Ga-DOTATATE PET tumor volume as a prognostic marker for short-term clinical outcomes following 177 Lu-DOTATATE therapy. Smaller extrahepatic 68 Ga-DOTATATE-avid tumor volume is associated with longer PFS and OS. As comparison, the association with OS is not statistically significant when total 68 Ga-DOTATATE avid tumor volume is analyzed. To our knowledge, this is the first study that evaluated extrahepatic tumor burden in GEP-NETs patients undergoing 177 Lu-DOTATATE therapy.
Extrahepatic metastatic tumor burden had been recognized as an important prognostic factor for GEP-NETs [26][27][28]. It is not entirely known whether progression to extrahepatic metastases is spread of the low-grade indolent tumor or an indication of transition to higher grade more aggressive tumors. Systemic therapies such as PRRT are currently reserved for these patients. However, our findings suggest the PRRT become less effective in patients with extensive extrahepatic metastases, which is not unexpected. It is not completely clear why hepatic tumor volume alone does not predict PFS as well as extrahepatic metastatic tumor volume. It is likely that liver has large functional reserve and significant tumor involvement is needed to impair its function and leads to mortality. 68 Ga-DOTATATE PET/CT plays a central role in the care of patients with NETs and alters diagnosis and management in one third of the cases [29]. Several groups have assessed somatostatin receptor PET as prognostic factor in long term PFS of NETs patients [21][22][23][30][31][32]. Tirosh et al. demonstrated in a prospective study that 68 Ga-DOTATATE tumor volume correlates with PFS and disease specific survival in NETs patients [22]. Our group has shown that somatostatin receptor expressing tumor volume measured on 68 Ga-DOTATATE PET/CT may have prognostic value of PFS in NETs patients [23]. Ambrosini et al. found that 68 Ga-DOTANOC PET/CT SUVmax is independent prognostic factor in patients with G1 and G2 pancreatic NETs [30]. However, only a small percentage of patients in these studies were treated with 177 Lu-DOTATATE. Our study directly assessed pre-therapy 68 Ga-DOTATATE PET/CT is relation to short term clinical outcomes in patients with GEP-NETs treated with 177 Lu-DOTATATE.
An early study of 68 Ga-DOTANOC PET/CT showed that SUVmax is an independent prognostic factor in patients with low grade NETs [30,33]. A recent study with 64 Cu-DOTATATE PET/CT found association between tumor SSTR density and PFS but not OS [31]. We did not observe association PFS or OS with SUVmax on 68 Ga-DOTATATE PET in this patient cohort. It is uncertain what is the cause of this difference and it is likely related to the use of different PET isotopes and somatostatin analogs, for example, DOTANOC is known to have higher affinity to SSTR-3 in addition to SSTR-2 and 5 [34]. Clinically, unlike SUVmax in 18