Clinical and Prognostic Value of PET/CT Imaging with Combination of 68Ga-DOTATATE and 18F-FDG in Gastroenteropancreatic Neuroendocrine Neoplasms

Background To evaluate the clinical and prognostic value of PET/CT with combination of 68Ga-DOTATATE and 18F-FDG in gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs). Method 83 patients of GEP-NENs who underwent 68Ga-DOTATATE and 18F-FDG PET/CT were enrolled between June 2013 and December 2016. Well-differentiated (WD) NETs are divided into group A (Ki-67 < 10%) and group B (Ki-67 ≥ 10%), and poorly differentiated (PD) NECs are defined as group C. The relationship between PET/CT results and clinicopathological characteristics was retrospectively investigated. Result For groups A/B/C, the sensitivities of 68Ga-DOTATATE and 18F-FDG were 78.8%/83.3%/37.5% and 52.0%/72.2%/100.0%. A negative correlation between Ki-67 and SUVmax of 68Ga-DOTATATE (R = −0.415; P ≤ 0.001) was observed, while a positive correlation was noted between Ki-67 and SUVmax of 18F-FDG (R = 0.683; P ≤ 0.001). 62.5% (5/8) of patients showed significantly more lesions in the bone if 68Ga-DOTATATE was used, and 22.7% (5/22) of patients showed more lymph node metastases if 18F-FDG was used. Conclusions The sensitivity of dual tracers was correlated with cell differentiation, and a correlation between Ki-67 and both SUVmax of PET-CTs could be observed. 68Ga-DOTATATE is suggested for WD-NET and 18F-FDG is probably suitable for patients with Ki-67 ≥ 10%.


Introduction
Gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) are a heterogeneous group of neoplasms that arise from cells of the endocrine system [1]. GEP-NENs are rare and present many clinical challenges. Because of their unpredictable biologic behaviors, the diagnosis usually takes place only after the condition has become advanced. Treatment regimens rely mainly on histological grading via biopsy; however, tumor heterogeneity cannot be fully assessed by tumor biopsy [2,3]. Although Ki-67 staining has been shown to have prognostic significance in GEP-NENs, pitfalls may lead clinicians to misjudge the tumor grades. Specifically, first, the current gold-standard method, in which 2000 cells are counted, is heavily dependent on the skill and expertise of the reporting pathologist [4]. Second, the limited tissue in some cases may impede accurate assessment of Ki-67 given the potential for heterogeneity of Ki-67 expression within tumors. Finally, the Ki-67 index may vary over time in the same patient, with changes possible both in response to treatment [5] and over the progression of the disease [6]. We believe that combined dual-tracer PET/CT imaging offers distinct advantages to overcome the above pitfalls.
DOTA-peptides specifically bind to somatostatin receptors 2, 3, and 5 and are usually overexpressed on the surfaces neuroendocrine cells [7,8]. 68 Ga-DOTATATE has been shown to be useful for staging, restaging, surveillance, determining SSTR-based therapy, and monitoring responses to treatments in NENs [9,10]. 18 F-FDG is a glucose analogue, and PET/CT imaging with this tracer has been shown to be correlated with NENs aggressiveness. The presence of increased glucose in NENs highlights an increased propensity for invasion and metastasis, and PET imaging with 18 F-FDG accordingly has higher sensitivity, especially in aggressive and high-grade tumors [11,12]. Therefore, we believe that PET/CT imaging with combination of 68 Ga-DOTATATE and 18 F-FDG PET/CT is a highly efficient whole-body imaging method, and it could be complementary to conventional imaging methods.
A large number of previous studies have evaluated the diagnostic accuracy of both tracers in the presence of a relative shortage of information, regarding the correlation to pathological findings and prognostic value. Only a few studies have compared the clinical impact of both 68 Ga-DOTATATE and 18 F-FDG PET tracers on NENs [13,14]. The present study aimed to determine the clinical value of the complementary PET/CT imaging method in a large histologically proven NENs population.

Materials and Methods
We analyzed the data from 83 (50 males and 33 females) consecutive patients with pathologically proven NENs who underwent contemporaneous PET/CT imaging with 68 Ga-DOTATATE and 18 F-FDG between June 2013 and December 2016. 68 Ga-DOTATATE and 18 F-FDG PET/CT scans were performed within an interval of no more than 2 weeks. No patients were treated during this interval. All NENs were classified according to the histopathological reports, which are based on recent consensus statements of the European Neuroendocrine Tumor Society. According to the grade of differentiation, proliferation index (Ki-67), and mitotic count, the well-differentiated (WD) neoplasms are herein defined as NET and graded G1 (Ki-67 ≤ 2%) or G2 (Ki-67 3-20%) and G3a (Ki-67 > 20%); the poorly differentiated (PD) neoplasms are defined as NEC and G3b. G2 patients were further divided into 2 groups as G2a (3-9%) and G2b (10-20%).

PET/CT Acquisition.
Patients fasted for at least 6 h before PET/CT scan. Images were acquired 1 h after injection of 3.7 MBq/kg 18 F-FDG or 1 h after the injection of 100-200 MBq 68 Ga-DOTATATE. Whole-body scan (brain to mid-thigh) was performed with the patient in the supine position. CT exposure factors for all scans were 120 kV and 100 mA. PET/CT images were reported in consensus by two experienced nuclear medicine physicians who were blinded to the findings of the structural imaging. Any nonphysiological focus of 68 Ga-DOTATATE or 18 F-FDG uptake greater than the normal liver background was considered positive. At the same time, CT imaging was used to differentiate between lesions and physiological uptake. The maximum standardized uptake value (SUV max ) of primary and metastatic lesions was calculated. SUV max generated from each patient was used in the final analysis. The SUV ratio of the tumor relative to the maximal liver uptake was calculated by dividing SUV max of the tumor by SUV max of the liver (SUV T/L ). The ratio between   Table 2 illustrates the sensitivities and SUV max in correlation with primary sites and grades. The uptake values of 68 Ga-DOTATATE in PanNENs were significantly higher than those in GI-NEN (29.87 ± 4.78 versus 16.76 ± 2.62, = 0.011). The values of 18 F-FDG PET/CT in PanNEN had a trend toward a lower SUV than that in GI-NEN (6.51 ± 0.77 versus 7.57 ± 0.86, = 0.067) (Figure 1).

Discussion
68 Ga-DOTATATE and 18 F-FDG PET/CT play a crucial role in the diagnosis and clinical management of NENs with morphologic and functional information. 68 Ga-DOTATATE was found to be superior to 18 F-FDG in WD-NET, whereas 18 F-FDG was more sensitive in PD-NEC [15]. Considering the costs of molecular imaging, choosing the selected patients for the specific PET/CT imaging is of vital importance. Analyzing the results produced using both tracers for different grades and primary sites could be a balanced approach. The aim of the present study was to compare 68 Ga-DOTATATE and 18 F-FDG PET/CT in GEP-NENs and to investigate the relationship between the complementary PET/CT results and histopathologic findings in clinical and prognostic values in a large, histologically confirmed NEN population. 68 Ga-DOTATATE imaging and 18 F-FDG PET/CT imaging have been compared in several studies which have been shown to have variable sensitivities in detecting NENs with a relatively small number of patients. Naswa et al. [16] reported that the sensitivity of 68 Ga-DOTANOC and 18 F-FDG was 91.4% and 42.5%, respectively. Koukouraki et al. [17] demonstrated that the sensitivity of 68 Ga-DOTATOC and 18 F-FDG was 90% and 68%, respectively, and 68 Ga-DOTANOC was more sensitive in the detection of primary sites or metastasis than 18 F-FDG [18]. Notably the patients included in this study were mainly WD-NETs with lower glucose metabolism. In the present study, the sensitivity of 68 Ga-DOTATATE and 18 F-FDG PET/CT was 63.85% and 74.70%, respectively. Subgroup analysis showed that the sensitivity of 68 Ga-DOTATATE was mainly correlated with the degree of differentiation, instead of correlation with Ki-67 index. The sensitivity of 18 F-FDG showed a positive correlation with the Ki-67 index and differentiation. From 18 F-FDG imaging, the sensitivity was under 53% for patients with Ki-67 < 10% and 100% for PD-NEC. SUV max of patients with NEC was low even for those who had positive results under 68 Ga-DOTATATE. We also observed that patients with Ki-67 < 10% showed low uptake in 18 F-FDG PET/CT. On consideration of the weak significance of 68 Ga-DOTATATE for PD-NEC and 18 F-FDG PET/CT for lower-grade NET, our study demonstrated that 18 F-FDG is more suitable for patients with Ki-67 ≥ 10%, and 68 Ga-DOTATATE is less advantageous in PD-NEC and should be tailored to the individual patients.
Our study demonstrated that PET/CT uptake was statistically significantly different between subgroups of GEP-NENs according to grading. The cohort was separated into 3 groups: A (G1 + G2a), B (G2b + G3a), and C (G3b). Group B showed higher sensitivity for 18 F-FDG than group A, and the median SUV max increased significantly, indicating relatively quick proliferation rate associated with Ki-67 index. Strosberg [19] proposed chemotherapy as a treatment option for tumors with Ki-67 ≥ 10%, especially those with higher 18 F-FDG activity. Such tumors showed high proliferative capacity and aggressive behavior and are recommended for chemotherapy after PRRT or somatostatin therapy. In this way, 18 F-FDG PET/CT may be suitable for identifying patients with aggressive conditions in group B who could benefit from chemotherapy. Thus, PET/CT imaging may establish the missing link between histopathologic findings and the treatment regimen. Based on our results, dual tracers' assessment is recommended for WD-NET with Ki-67 ≥ 10% (G2b and G3a). 68 Ga-DOTATATE and 18 F-FDG were complementary in detecting lesions and dual-trace PET/CT showed an advantage in the assessment of SSA, PRRT, and chemotherapy. We proposed that WD-NET patients with Ki-67 ≥ 10% should be examined using dual tracers upon diagnosis. We also suggest that 68 Ga-DOTATATE should be performed solely in WD-NET patients with Ki-67 < 10% and 18 F-FDG is sufficient for PD NEC. Moreover, repeated PET/CT is warranted when disease progresses rapidly, considering the heterogeneous expression and complementary findings to histopathology (Figure 4).
We investigated the correlations between dual tracers and Ki-67 index. SUV max or SUV T/L was positively correlated with Ki-67 index with respect to 18 F-FDG PET/CT and negatively correlated with Ki-67 index with respect to 68 Ga-DOTATATE. SUV max of PET/CT may be a suitable biomarker for evaluation of the biological behavior of NENs. The relationship between SUV T/L and Ki-67 index was weak, but it showed a consistency with SUV max . However, SUV max may be influenced by technical elements, while SUV T/L could reduce the differences attributable to equipment and operation to some extent.
GEP-NENs are a heterogeneous group of neoplasms that display great variability in biological behaviors and clinical outcomes [20]. This requires accurate diagnostic techniques for precise staging and choice of therapy. The standard grading is mainly based on the immunohistochemistry of the proliferation marker Ki-67. However, there are many lesions with variable tracer uptake at different parts of the tumor. Especially within the same organ, these lesions may prevent the biopsy from being a comprehensive reflection of tumor heterogeneity in vivo, leading to inaccurate Ki-67 index results. Therefore, dual-trace PET/CT, which is a wholebody noninvasive alternative, is warranted to overcome the shortcomings of histopathologic grading. The findings of this study included the ability of 68 Ga-DOTATATE to detect heterogeneity in tumors and variable expression in primary sites. SUV max of PanNEN was higher than that of GI-NENs, which was consistent with prior studies [21,22], where the investigators found higher levels of messenger RNA expression of SSTR2 and SSTR5 in pancreatic than in gastrointestinal NENs. In this way, 68 Ga-DOTATATE is more sensitive in PanNEN, but it may miss some lesions in GI-NEN. However, SUV of PanNEN and GI-NEN showed no difference for 18 F-FDG, which is complementary to 68 Ga DOTATATE. Another finding of the present study was that 68 Ga DOTATATE has a superb ability to detect heterogeneity in metastatic lesions, which is beneficial for us to select the optimal protocol. We believe that complementary PET/CT can evaluate the tumors' heterogeneity and influence treatment options.
The morphological findings and Ki-67 index are considered important prognostic markers in NENs. One major limitation of these histopathological parameters as prognostic markers is the requirement for tissue sampling, which is not always feasible. A few studies have demonstrated the prognostic value of 18 F-FDG and 68 Ga-DOTATATE PET/CT in patients with NENs. 18 F-FDG was an independent predictor of PFS [23,24]. 18 F-FDG SUV max > 3 was demonstrated to be independent predictor of disease progression, superior to conventional prognostic factors such as Ki-67 index and serum CgA [25]. SSTR expression was found to be a positive prognostic factor, and it was therefore expected that SSTRbase PET/CT would also have prognostic value in NENs [26]. Sharma et al. [27] demonstrated the prognostic value of SUV max measured with 68 Ga DOTATATE in 37 patients with NET. SUV max ≥ 14.5 was found to distinguish patients with progressive and those with nonprogressive disease. A major limitation of the prior studies was that most of the patients included were WD-NET, which may involve selection bias. Our study enrolled 57 unresectable patients (31 NET and 26 NEC). Unresectable patients positive with 18 F-FDG alone showed the worst prognosis, while those positive with 68 Ga-DOTATATE alone had the best prognosis. NET patients with predominant 18 F-FDG uptake and a negative 68 Ga-DOTATATE scans had worse prognosis. There is a strong association between high 18 F-FDG uptake and worse outcome even in patients with WD-NETs. However, PD-NEC with negative 68 Ga-DOTATATE uptake may lead to worse prognosis. The present study has several limitations. First, our study is retrospective, and the second limitation is that the follow-up time was not long enough to assess treatment response, considering the relatively inert biological behavior of NENs. Noninvasive dual-tracer imaging with 68 Ga-DOTATATE and 18 F-FDG PET/CT seems promising as an alternative to tissue sampling, due to its capacity to reflect two different aspects of tumor biology, SSTR expression and glucose metabolism, respectively. Accordingly, imaging with dual tracer is recommended for WD-NET patients with Ki-67 ≥ 10%, providing information for selection of SSA, PRRT, and chemotherapy. Taking the advantages of dual-tracer imaging in detecting lesions is useful for accurate clinical management. Dual-tracer imaging also shows a clearly linear correlation between SUV max and Ki-67 index. On consideration of heterogeneous expression and complementary findings to histopathology, our results suggested that repeated dualtracer imaging is warranted to evaluate dynamic biological behavior and prognosis.