DE LA UNIVERSIDAD TECNOLÓGICA EQUINOCCIAL

within histological subtypes of NSCLC

using dynamic 18_{F-FDG PET}

Tineke W.H. Meijer Lioe-Fee de Geus-Oei Eric P. Visser Wim J.G. Oyen Monika G. Looijen-Salamon Dimitris Visvikis Ad F.T.M. Verhagen Johan Bussink Dennis Vriens

Thanks to Nicolle Peters for help with patient inclusion and Peter Kok’s team of PET-technologists for support during acquisition

Radiology 2017;283:547-559

532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer Processed on: 20-8-2019 Processed on: 20-8-2019 Processed on: 20-8-2019

Processed on: 20-8-2019 PDF page: 84PDF page: 84PDF page: 84PDF page: 84

84

Abstract

Purpose To assess whether dynamic 18_{F-FDG PET has added value over static} 18_{F-FDG PET}

for tumor delineation in non-small cell lung cancer (NSCLC) radiotherapy planning using pathology volumes as Reference Standard, and to compare pharmacokinetic rate constants of 18_{F-FDG metabolism, including their regional variation, between NSCLC}

histological subtypes.

Materials and methods In this prospective observational study, one-hour dynamic

18_{F-FDG PET/CTs were acquired in 35 patients (36 resectable NSCLCs) between 2009-2014.}

Static and parametric images of glucose metabolic rate (MR_glc) were obtained to determine lesion volumes using three delineation strategies. Pathology volume was calculated from three orthogonal dimensions (n=32). Whole tumor and regional rate constants (K₁-k₃) and fractional blood volume (V_B) were computed using compartment modeling.

Results Pathology volumes were larger than PET volumes (median difference 8.7-25.2 cm3_{, Wilcoxon-signed-rank-test, p<0.001). Static FLAB volumes corresponded best with}

pathology volume (ICC=0.72, p<0.001). Bland-Altman analyses showed highest precision and accuracy for static FLAB volumes. MR_glc and 18_{F-FDG phosphorylation rate (k}

3) were

higher in squamous cell carcinomas than adenocarcinomas, whereas V_B was lower (Mann- Whitney-U-test or t-test, p=0.003, p=0.036, p=0.019). MR_glc, k₃ and V_B were less heterogeneous in adenocarcinomas than squamous cell carcinomas (Friedman’s ANOVA).

Conclusion Parametric images are not superior to static images for NSCLC delineation. FLAB-based segmentation on static 18_{F-FDG PET is in best agreement with pathology}

volume and could be useful for NSCLC auto-contouring. Differences in glycolytic rate and V_B between squamous cell carcinomas and adenocarcinomas are relevant for research in targeting tumor glucose metabolism.

532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer Processed on: 20-8-2019 Processed on: 20-8-2019 Processed on: 20-8-2019

Processed on: 20-8-2019 PDF page: 85PDF page: 85PDF page: 85PDF page: 85

85

5

Introduction

Combined radiochemotherapy is standard treatment for irresectable stage III non-small cell lung cancer (NSCLC). Concurrent chemoradiation decreases locoregional progression rate from 34% to 28% at 3 years and improves overall survival, compared to sequential chemoradiation. This reflects the importance of optimizing locoregional therapy [1,2].

Tumor delineation for radiotherapy planning using 18_{F-fluorodeoxyglucose positron}

emission tomography (18_{F-FDG PET) is classically based on static images. Using dynamic} 18_{F-FDG PET, distinction can be made between unmetabolized} 18_{F-FDG and bound} 18_{F-FDG-6-phosphate [3,4]. Dynamic PET allows calculation of glucose metabolic rate}

(MR_glc; nanomol.mL-1_.min-1_{). MR}

glc images give an increased signal to background ratio

relative to static images due to absence of unmetabolized 18_{F-FDG in the background [5].}

Therefore, we hypothesized that tumor delineation using dynamic 18_{F-FDG PET is more}

accurate than static 18_{F-FDG PET, which might translate into improved locoregional}

treatment.

Additionally, biological factors like reprogrammed tumor energy metabolism influence response to radiotherapy and might be a target to improve locoregional treatment [6]. Several metabolites, such as glucose, lactate and glutathione, differ between NSCLC and normal lung tissue [7-9]. Emerging evidence supports differences in glucose metabolism between adeno- and squamous cell carcinomas, based on the expression of metabolic transporters and enzymes, static 18_{F-FDG PET and level of several metabolites [10-12].}

Compared to adenocarcinomas, squamous cell carcinomas show higher glucose transporter 1 (GLUT1) expression and standardized uptake values (SUV) on 18_{F-FDG PET}

[12,13]. Also, adenocarcinomas might be better vascularized/perfused than squamous cell carcinomas (higher vascular density on immunohistochemistry and higher Ktrans _on

dynamic contrast-enhanced computed tomography (CT) and magnetic resonance imaging (MRI)) [12-15]. However, blood flow and volume do not differ on these imaging modalities [14-16]. Insight into the metabolic rate of glucose metabolism and metabol- ic-vascular heterogeneity is required to further phenotype NSCLC. This biological information is likely to be relevant for research in targeting agents and radiotherapy dose escalation. Using dynamic 18_{F-FDG PET, rate constants of} 18_{F-FDG metabolism and blood}

volume can be calculated (Figure 1) [4].

The aim of this prospective cohort study was dual: to assess whether dynamic

18_{F-FDG PET has added value over static} 18_{F-FDG PET for tumor delineation in NSCLC}

radiotherapy planning using pathology volumes as Reference Standard, and to compare pharmacokinetic rate constants of 18_{F-FDG metabolism, including their regional variation,}

532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer Processed on: 20-8-2019 Processed on: 20-8-2019 Processed on: 20-8-2019

Processed on: 20-8-2019 PDF page: 86PDF page: 86PDF page: 86PDF page: 86

86

Materials and methods

Study was approved by the Commission Medical Research Involving Human Subjects Region Arnhem–Nijmegen, the Netherlands. All patients gave written informed consent.

Patients

Therapy-naïve, non-diabetic patients with newly diagnosed or suspected NSCLC, stage IB-limited stage IIIA (TNM 7th _{edition), who underwent a primary resection were}

consecutively included in this prospective observational cross-sectional study between 2009-2014. All subjects were routinely staged using contrast-enhanced CT of chest/upper abdomen and 18_{F-FDG PET/CT with additional histological staging of the mediastinum or}

other suspicious sites when necessary. Patients with histology proven non-NSCLC or high suspicion for metastasis were ineligible. Tumors had to be at least 30 mm to minimize partial volume effect [17]. Dynamic 18_{F-FDG PET was performed within 7 days of surgery.}

A total of 38 patients were included. Three patients were excluded because of no malignancy after lobectomy (infarction with inflammation), absence of 18_{F-FDG uptake on}

dynamic PET or severe mismatch between dynamic PET and CT which could not be correctly realigned before reconstruction. Thirty-six lesions in 35 patients remained (mean age of males 66 years (range 45-82); mean age of females 64 years (range 48-77); age not

Figure 1 The irreversible, two-tissue compartment model for 18_{F-FDG metabolism}

The measured PET signal (thin dashed line) is a combination of the intracellular activity concentration of free 18_{F-FDG (}18_{F-FDG in tissue), the intracellular activity concentration of} 18_{F-FDG-6-phosphate}

(18_{F-FDG-6-PO4 in tissue) and a fraction of the activity concentration of} 18_{F-FDG in blood plasma}

(V_B).

In dynamic PET, pharmacokinetic rate constants K₁ (rate constant of transport of 18_{F-FDG into tumor}

cells; mL.g-1_.min-1_{), k}

2 (rate constant of export of 18F-FDG out of tumor cells; min-1) and k3 (rate

constant of cytoplasmic phosphorylation of 18_{F-FDG; min}-1_{), and VB (blood volume fraction; mL.mL}-1₎

can be calculated using non-linear least squares regression.

Abbreviations: 18_F-FDG: 18_{F-fluorodeoxyglucose;} 18_F-FDG-6-PO4: 18_{F-fluorodeoxyglucose-6-phosphate.}

532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer 532953-L-sub01-bw-Meijer Processed on: 20-8-2019 Processed on: 20-8-2019 Processed on: 20-8-2019

Processed on: 20-8-2019 PDF page: 87PDF page: 87PDF page: 87PDF page: 87

87

5

significantly different between males and females (Independent-Samples t test); one patient had two synchronous primary NSCLCs (an adenocarcinoma and a squamous cell carcinoma). Measurement of all pathology dimensions was not available in 3 cases and no resection took place in one patient due to detection of stage IV peroperatively (Figure 2). Clinicopathological characteristics, shown in Table 1, were similar for adenocarcinomas and squamous cell carcinomas (Table 2).