Somatostatin receptor expression in small cell lung cancer as a prognostic marker and a target for peptide receptor radionuclide therapy

Despite initial responsiveness to both chemotherapy and radiotherapy, small cell lung cancer (SCLC) commonly relapses within months. Although neuroendocrine characteristics may be difficult to demonstrate in individual cases, a relevant expression of somatostatin receptors (SSTR) on the cell surface has been described. We aimed to evaluate the prognostic value of SSTR-expression in advanced SCLC. We further examined pre-requisites for successful peptide receptor radionuclide therapy (PRRT). 21 patients with extensive stage SCLC were enrolled. All patients underwent positron emission tomography/computed tomography (PET/CT) with 68Ga-DOTATATE to select patients for SSTR-directed therapy. PET scans were visually and semi-quantitatively assessed and compared to SSTR2a and SSTR5 expression in biopsy samples. Peak standardized uptake values (SUVpeak) of tumors as well as tumor-to-liver ratios were correlated to progression-free (PFS) and overall survival (OS). In 4/21 patients all SCLC lesions were PET-positive. 6/21 subjects were rated “intermediate” with the majority of lesions positive, the remaining 11/21 patients were PET-negative. PET-positivity correlated well with histologic SSTR2a, but not with SSTR5 expression. Neither PET-positivity nor SUVpeak were predictors of PFS or OS. In 4 patients with intensive SSTR2a-receptor expression, PRRT was performed with one partial response and one stable disease, respectively. SSTR-expression as detected by 68Ga-DOTATATE-PET and/or histology is not predictive of PFS or OS in patients with advanced SCLC. However, in patients exhibiting sufficient tracer uptake, PRRT might be a treatment option given its low toxicity and the absence of effective alternatives.


INTRODUCTION
Small cell lung cancer (SCLC) represents 15% of all lung cancers and occurs predominantly in smokers [1] Due to its rapid doubling time, high growth fraction, and the early development of metastases, only one-third of patients are diagnosed with localized disease [2]. Although SCLC is highly responsive to both chemotherapy and radiotherapy, it commonly relapses within months despite treatment [3,4]. SCLC represents the most aggressive member of a family of tumors with neuroendocrine features such as typical and atypical carcinoids or large cell neuroendocrine carcinomas (LCNEC). Although neuroendocrine characteristics may be difficult to demonstrate in individual cases, a relevant expression of somatostatin receptors (SSTR) on the cell surface has been described [5,6].
The 68 Ga/ 177 Lu/ 90 Y-labelled compound DOTATATE/-TOC has been shown to reliably and selectively bind to SSTR2a (and SSTR5). It is widely used in imaging ( 68 Ga) and therapy ( 177 Lu, 90 Y) of neuroendocrine tumors [7,8]. The aim of this study was to evaluate the prognostic value of SSTR-expression in patients with advanced SCLC. We further examined pre-requisites for successful peptide receptor radionuclide therapy (PRRT) in this patient group.

SSTR-PET/CT-Imaging
On visual analysis, 4/21 patients were rated "positive" with all SCLC lesions displaying higher tracer uptake than liver parenchyma, 6/21 were rated "intermediate", and 11/21 presented with 68 Ga-DOTATATE-negative disease ( Figure 1). The SUV peak values and the T/L ratios of the maximum tumor to mean liver SUV are listed in Table 1. Both SUV peak values (p<0.001, Kruskal-Wallis test) and T/L ratios (p<0.001, Kruskal-Wallis test) were different in the 3 groups. The highest values (SUV >20, tumor to liver ratio >2,7) were observed in the 4 patients rated positive with increased uptake in all lesions. SUV peak (9.8±2.5; p=0.01, T-test) and T/L ratios (1.4±0.3; p=0.02, T-test) were lower in the intermediate group and even lower (4.6±1.5 and 0.6±0.2, respectively; p<0.001 each, T-test) in the negative group.
SUV peak as well as the T/L ratios were significantly correlated to the IRS score of SSTR2a expression (Spearman's Rho: ρ=0.49 with p=0.03 and ρ=0.61 with p<0.01, respectively) but not to SSTR5 expression (Spearman's Rho: ρ=0.31 with p=0.19 and ρ=0.27 with p=0.27, respectively) ( Figure 2). In the Kruskal-Wallis test for differences in the groups of PET positivity, no significant differences were found for the SSTR5 IRS score (p=0.2). A difference for the SSTR2a IRS score (p<0.01) was attributable to a significantly higher score in the 4 patients rated positive with increased uptake in all lesions; no difference was observed between the "intermediate" and "negative" groups (p=0.96, Mann-Whitney test).
Interestingly, one patient (patient #17) had no detectable SSTR-expression in his biopsy sample taken from the primary tumor in the right bronchus consistent with SSTR-PET imaging, but showed high tracer binding in all metastases.

PRRT
In total, 5/21 patients (4/4 of the positive group and 1/6 of the intermediate group with the vast majority of lesions SSTR-positive) were considered eligible for PRRT. One patient opted for further chemotherapy, leading to a total of 4 SCLC patients undergoing 1, 1, 5 and 6 cycles of PRRT, respectively. Per cycle, 7.6±0.3 GBq of 177 Lu-DOTATATE (patients #2, #9 and #10) / -DOTATOC (patient #17) were administered, resulting in cumulative activities of 39.6, 7.4, 7.6, and 44.2 GBq of 177 Lu, respectively. Treatment was well tolerated by all subjects without any changes in vital signs during therapy or significant therapy-related toxicity. Especially, no kidney failure or myelodysplastic syndrome was observed.
During follow-up, the patient undergoing 5 cycles of PRRT (patient #2) achieved partial response ( Figure 3). PFS was 766 days and OS 907 days in patient #2. The patient with 6 cycles (patient #17) remained stable for 396    days with a survival of 575 days. One of the two subjects with only 1 cycle of 177 Lu-DOTATATE reported on an improvement of his overall performance status as well as of his pain (patient #9). This patient died from pneumonia prior to his second treatment cycle. The remaining patient (patient #10) did not benefit from radiopeptide therapy and deceased 53 days after PET imaging.

DISCUSSION
Management of SCLC remains a therapeutic challenge. Whereas cisplatin-based chemotherapeutic regimens are the mainstay of first-line therapy, a number of agents are offered for second-line treatment (usually employing the topoisomerase inhibitor topotecan). A standard third-line therapy has not been defined yet. PRRT has been investigated in this setting as an alternative for patients with high SSTR expression [9][10][11]. In our study, we could record a benefit from PRRT in 3 of 4 patients treated: One patient achieved partial response and another patient remained stable for more than a year. Given the median survival of <12 months in relapsed/refractory SCLC [12], both patients had a very encouraging outcome with an OS of 29 and 18 months, respectively. A third patient reported on improved performance status and pain relief after the first treatment cycle. Unfortunately, he died from pneumonia prior to the scheduled second PRRT.
Important prerequisites to SSTR-targeted therapy are a robust expression of as well as a specific binding to the target receptor. We could demonstrate a good correlation between imaging results and histopathological SSTR2a expression. However, only few patients qualified for PRRT by demonstrating higher receptor expression in tumors than in normal liver parenchyma. In our cohort, only 4 subjects presented with all SCLC lesions SSTR-positive, 6 presented with a mixture of mostly positive but also negative lesions. This finding highlights the phenomenon of tumor heterogeneity with variable features of both primary and metastatic cancer. Though a robust expression of SSTR has been described for SCLC biopsy specimens [5], receptor expression of the metastases can vary. Additionally, the dynamics of SSTR regulation under therapeutic pressure are unknown. Of note, the primary tumor biopsy sample of the patient undergoing the most cycles of PRRT was negative for SSTR expression. One might speculate that SCLC lesions with retained SSTR expression might still harbor more features of better differentiated neuroendocrine tumors and therefore be associated with a better prognosis. However, we did not observe any influence of receptor expression on progression-free or overall survival. SSTRdirected PET/CT cannot serve as a non-invasive prognostic marker. Nevertheless, it can be used as a theranostic means for PRRT assessment.
Achieved tumor doses play an important role in treatment efficacy. In our study, all results were obtained after administration of standard activities of 177 Lu. Recent in-house dosimetry studies have suggested the feasibility of individualized PRRT with safe administration of higher activities (data not published), possibly resulting in improved outcomes. This study has various limitations. It is retrospective and comprises a very limited number of observations with only 4 patients undergoing radiopeptide therapy. Biopsies were not always obtained in close proximity to PET imaging. However, it is the first report on an anti-tumor effect of SSTR-directed PRRT. Given its excellent tolerability, it can be considered in an otherwise very limited therapeutic scenario. Future research should try to optimize both PRRT procedures (individual dosimetry prior to therapy, shorter intervals between treatment cycles) as well as focus on new potential targets. One of those targets might be represented by chemokine receptor CXCR4, which has been described to be overexpressed in a vast number of malignancies, including SCLC [13,14]. Proof-of-concept for visualization of CXCR4 expression by a radiolabelled PET ligand [15,16] has recently been demonstrated in SCLC patients [17]. As a compound allowing for labelling with various therapeutic α-and β-emitters is available [18], further evaluation of this theranostic approach is warranted.
In conclusion, SSTR-expression as detected by 68 Ga-DOTATATE and/or histology is not predictive of PFS or OS in patients with advanced SCLC. However, in patients exhibiting sufficient tracer uptake, PRRT might be considered an option given its low toxicity and the absence of effective alternative treatment options. Further research on novel therapeutic targets is warranted.

MATERIALS AND METHODS
Due to the retrospective nature of this study, our institutional review board waived the requirement for informed consent. Still, all patients gave written informed consent to receive SSTR-PET/CT imaging on a compassionate use basis for the purpose of PRRT assessment.
After PET/CT, patients were monitored until death with a median follow-up of 3 months (range 0-29 months). Progression-free (PFS) and overall survival (OS) were assessed by serial CT scans (every 3 months) according to Response Evaluation Criteria in Solid Tumors (RECIST) 1.1 [19] and correlated to PET-derived parameters (SUV peak , tumor to liver-ratio) as well as to biopsy-derived SSTR2a and SSTR5 expression. www.impactjournals.com/oncotarget SSTR-PET/CT imaging PET/CT was performed on an integrated scanner (Siemens Biograph mCT 64, Siemens, Knoxville, USA) consisting of a Lutetium oxyorthosilicate full-ring PET and a 64-slice spiral CT. 68 Ga-DOTATATE (114 ± 34 MBq) was injected intravenously. After a period of 40-60 min, transmission and PET emission data were acquired as previously described [20].
All imaging tests were reviewed visually by two experienced nuclear medicine physicians (C.L., R.A.W.) who were blinded to clinical data. Patients were categorized according to SSTR-positivity as "positive" (all tumor lesions PET-positive with 68 Ga-DOTATATE uptake > liver), "intermediate" (the majority of lesions positive) or "negative" (the majority of / all tumor lesions PET-negative).
For semi-quantitative analysis, peak standardized uptake values (SUV peak ) were calculated by assigning spherical volumes of interest of 1.5 cm diameter to the area of highest tracer uptake in tumororus tissue. In addition, a 5 cm spherical volume of interest was drawn to healthy tissue of the right lobe of the liver to determine the liver SUV mean , serving as reference for background activity. Tumor-to-liver ratios (T/L) were derived.
The 3 patient groups were tested by Kruskal-Wallis analysis for differences regarding the parameters age, SUV peak , T/L ratio, PFS, OS, and the IRS scores for the histologically assessed SSTR2a/5 expression.

PRRT
Peptide receptor radionuclide therapy (including renal protection) was performed according to the joint IAEA, EANM and SNMMI practical guidance [22]. In brief, all patients were hospitalized for a total of 3 days. PRRT with a median of 7.6 GBq (range, 7.2-8.5) of 177 Lu-DOTATATE/-TOC was intravenously administered over 30 minutes. Vital signs were documented during the infusion and within 7 days after administration. All patients were followed-up after PRRT including serial blood tests, kidney scintigraphy and SSTR-PET/CT. Treatment was performed with an interval of 8-12 weeks between each cycle.

Statistical analysis
Statistical analyses were performed using IBM SPSS (version 23.0). Data were analyzed with nonparametric tests unless the Shapiro-Wilk test indicated compatibility with normal distribution. Quantitative values were expressed as mean ± standard deviation or median and range as appropriate. The tests used are reported together with the results. All statistical tests were performed two-sided and a p-value < 0.05 was considered to indicate statistical significance.