Determinants of 18F-NaF uptake in femoral arteries in patients with type 2 diabetes mellitus.

BACKGROUND: The goal of this study was to investigate the potential determinants of 18F-NaF uptake in femoral arteries as a marker of arterial calcification in patients with type 2 diabetes and a history of arterial disease. METHODS AND
RESULTS: The study consisted of participants of a randomized controlled trial to investigate the effect of vitamin K2 (NCT02839044). In this prespecified analysis, subjects with type 2 diabetes and known arterial disease underwent full body 18F-NaF PET/CT. Target-to-background ratio (TBR) was calculated by dividing the mean SUVmax from both superficial femoral arteries by the SUVmean in the superior vena cava (SVC) and calcium mass was measured on CT. The association between 18F-NaF TBR and cardiovascular risk factors was investigated using uni- and multivariate linear regression corrected for age and sex. In total, 68 patients (mean age: 69 ± 8 years; male: 52) underwent 18F-NaF PET/CT. Higher CT calcium mass, total cholesterol, and HbA1c were associated with higher 18F-NaF TBR after adjusting.
CONCLUSION: This study shows that several modifiable cardiovascular risk factors (total cholesterol, triglycerides, HbA1c) are associated with femoral 18F-NaF tracer uptake in patients with type 2 diabetes.

Introduction

Arterial calcification on CT is used as a marker for the presence of intimal calcification (atherosclerosis) and medial arterial calcification. Microcalcifications (< 50 µm) are observed in the early stages of the calcification process and are possibly associated with plaque rupture, whereas macrocalcifications contribute to stability of atherosclerotic plaques.1 In addition, microcalcifications in the internal elastic lamina and in the tunica media,2,3 may be related to arterial stiffening.4 CT allows to identify macrocalcifications of over 200 µm in diameter.5 Sodium 18F-fluoride (18F-NaF) PET/CT is a novel imaging technique that can visualize smaller calcifications.6–10 In the femoral arteries, 18F-NaF PET can visualize ongoing (medial and intimal) calcification.11 In the coronaries 18F-NaF PET is associated with high risk plaque features and plaque rupture.6 In patients with diabetes, additional to atherosclerosis, medial arterial calcification (MAC or Mönckeberg’s sclerosis) is highly prevalent and is associated with increased arterial stiffness, hypertension, distal symmetrical neuropathy, chronic kidney disease, cardiovascular and all-cause mortality.12–14 MAC is observed in arteries of the lower extremity across the entire length.3 A recent study showed that increased 18F-NaF PET accumulation in the femoral arteries is associated with diffuse calcium deposition and significantly correlates with cardiovascular risk factors.11 Several studies investigated determinants of 18F-NaF uptake in oncologic patients.11,15 While the use of 18F-NaF is increasing, the available data are limited regarding its uptake values and potential uptake modifiers in high CVD risk groups such as diabetes patients. The goal of this study is to investigate the potential determinants of 18F-NaF uptake as a marker of arterial calcification in patients with type 2 diabetes and a history of arterial disease.

Method

Study Design and Population

The study population consisted of participants of a randomized controlled trial (ClinicalTrials.gov Identifier: NCT02839044).16,17 Patients were recruited at the University Medical Center Utrecht between June 2016 and September 2018. The goal of the study was to investigate the effect of vitamin K2 supplementation on arterial calcification in patients with type 2 diabetes. Inclusion criteria were age > 40 years with type 2 diabetes and documented presence of cardiovascular disease. The presence of arterial diseases was defined as coronary artery disease, stroke, peripheral artery disease, abdominal aortic aneurysm or an ankle brachial index (ABI) of < 0.9. The study protocol was approved by the local medical ethical committee (NL53572.041.15, METC 15/571) and all patients provided written informed consent. In this prespecified analysis, baseline 18F-NaF PET/CT scans and patient data were used to determine the potential contributors of increased 18F-NaF uptake. Patient data were collected from questionnaires and available medical records. All participants completed a questionnaire on medical history, medication, smoking history, and a physical examination including height, weight and blood pressure. Non-fasting blood samples were collected to measure cholesterol levels, HbA1c levels and creatinine.

PET/CT Imaging and Quantification

All patients underwent a full body 18F-NaF PET/CT scan, performed on a Siemens Biograph 40 scanner (Siemens Healthineers, Erlangen, Germany). Imaging was performed 90 min after intravenous injection of 2.0 MBq/kg 18F-NaF, with a maximum dosage of 200 MBq. Circular regions of interest (ROI) were drawn bilaterally around the superficial femoral arteries (see Figure 1A and B) from the bifurcation of the femoral artery to the femur condyles on every sequential axial slice to determine the 18F-NaF uptake. To derive the average maximum standardized uptake value (SUVmax), maximum values measured per ROI were summed and averaged. Target-to-background ratio (TBR) was calculated by dividing the mean SUVmax from both femoral arteries by the SUVmean from three sequential slices in the SVC (see Figure 1C), the circular ROI size was not fixed as the vena cava size differs between individuals. The femoral artery TBR was used as the measure of femoral 18F-NaF PET activity. To evaluate inter-rater reproducibility on radiotracer uptake ten participants were randomly selected and analyzed by thee blinded raters.

Figure 1

Axial CT image A showing circular calcification in the superficial femoral artery and fusion image B with a circular region of interest (ROI) drawn for quantification 18F-NaF SUVmax in the femoral artery. Axial fusion image (C) demonstrating background measurement of 18F-NaF SUVmean in the superior vena cava using a circular ROI

Calcification Quantification

Non-contrast-enhanced CT scans were visually evaluated using in-house built software to identify calcium scores in the femoral arteries. A threshold of ≥ 130 Hounsfield units was used and calcification was quantified using the calcium mass equivalent score which considers calcification volume and density.

Statistical Analyses

Normality of the data was tested using Quantile–Quantile plots. Descriptive data were expressed as mean ± standard deviation (SD) for normally distributed continuous variables, and median (Q1-Q3) for non-normally distributed continuous variables. Categorical variables are displayed as count (percentage). Reliability for TBR measurements was assessed using intraclass correlation coefficient. Univariate linear regression (unstandardized betas) was used to identify associations between all variables and 18F-NaF TBR. Multivariate linear analysis was performed to correct for age and sex. No adjustments were made for multiplicity of testing, and no imputation was used for missing values (only 3 patients had missing data on calcium mass). A p value < 0.05 was considered statistically significant. Statistical analyses were performed using SPSS software (IBM Corp, Version 24.0. Armonk, NY, USA).

Results

Patient Characteristics

The study population consisted of 68 participants (76% male) with a mean age 69 ± 8 years. Glucose lowering drugs were used by 60 (88%) patients, while insulin was used in 32 patients (47%). In total 25 patients used both insulin and glucose lowering drugs and only one patient was neither on insulin nor glucose lowering drugs. Fifty-two participants (77%) were treated with statins. As defined by the inclusion criteria all patients had a history of cardiovascular disease. Femoral arterial calcification on CT was present in 97% of the study population. Arterial 18F-NaF was measured on average on 76 slices per patient (38 per side). Baseline characteristics of the patients are reported in Table 1.

Table 1

Patient characteristics of the study population

Clinical characteristics	N = 68
Age (years)	69 ± 8
Male	52 (76)
Blood pressure
Systolic (mmHg)	137 ± 18
Diastolic (mmHg)	72 ± 11
Body mass index (kg/m2)	31.2 ± 5.3
Smokers
Former	42 (62)
Current	10 (15)
Femoral 18F-NaF PET TBR	2.2 ± 0.7
Femoral calcium mass by CT	118 (14–451)
Cholesterol
Total (mmol/l)	4.4 [3.5–5]
LDL (mmol/l)	2.1 [1.4–2.5]
HDL(mmol/l)	1.1 ± 0.3
Triglycerides (mmol/l)	2.9 [1.6–3.4]
HbA1c (mmol/mol)	58.2 [47.3–63.0]
Serum creatine (μmol/l)	85.2 ± 23.4
History of arterial disease
CAD	41 (60)
CVA, TIA	23 (34)
PAD	16 (24)
ABI < 0.9	18 (27)
AAA	6 (9)
Medication use
Insulin	32 (47)
Glucose lowering drugs	60 (88)
Statins	52 (77)
Lipid lowering drugs	31 (46)
Bisphosphonates	1 (2)
Antihypertensive drugs	60 (88)
Platelet inhibitors	31 (46)
Calcium supplements	9 (13)
Vitamin D supplements	21 (31)

Values are represented as mean ± SD, median [Q1-Q3] or n (%)

TBR, target-to-background ratio; LDL, low-density lipoprotein; HDL, high-density lipoprotein; HbA1c, glycated hemoglobin; CAD, coronary artery disease; CVA, TIA, cerebrovascular disease, transient ischemic attack; PAD, peripheral artery disease; ABI < 0.9, ankle brachial index < 0.9; AAA, abdominal aortic aneurysm

18F-NaF PET Activity

Higher femoral calcium mass on CT, total cholesterol, and HbA1c were associated with higher 18F-NaF TBR (Table 2). Age, blood pressure, creatine and body mass index (BMI) were not associated to 18F-NaF TBR. When applying multivariate linear regression adjusting for age and sex, CT femoral calcium mass, total cholesterol and HbA1c remained significantly associated (Table 2). A trend towards significance was seen for higher triglycerides and higher 18F-NaF PET TBR (p = 0.052). 18F-NaF PET TBR was lower in patients on statins with a trend towards significance (p = 0.058). No significant differences in TBR were observed for smokers, male sex or other medication (Table 2). The intraclass correlation coefficient for inter-rater reliability of radiotracer uptake (measured as TBR) was excellent with a value of 0.98 (95% CI 0.94-0.99).

Table 2

Uni- and multivariate linear regression of TBR

Clinical characteristics	Univariate β (95% CI)	p value	Multivariate β (95% CI)	p value
Age (years)	0.05 (− 0.015 to 0.025)	0.614	NA
Female sex	− 0.025 (− 0.428 to 0.349)	0.838	NA
Blood pressure
Systolic (mmHg)	− 0.002 (− 0.012 to 0.007)	0.644	− 0.003 (− 0.013 to 0.007)	0.516
Diastolic (mmHg)	− 0.010 (− 0.026 to 0.005)	0.189	− 0.010 (− 0.027 to 0.007)	0.229
Body mass index (kg/m2)	0.019 (− 0.012 to 0.050)	0.222	0.026 (− 0.008 to 0.060)	0.127
Femoral calcium mass by CT*	0.349 (0.204–0.494)	< 0.001	0.417 (0.251–0.582)	<0.001
Cholesterol
Total (mmol/l)	0.159 (0.034–0.284)	0.013	0.168 (0.040–0.296)	0.011
LDL (mmol/l)	0.111 (− 0.071 to 0.292)	0.227	0.120 (− 0.069 to 0.308)	0.210
HDL (mmol/l)	− 0.012 (− 0.626 to 0.602)	0.969	− 0.045 (− 0.706 to 0.616)	0.891
Triglycerides (mmol/l)	0.044 (− 0.003 to 0.090)	0.064	0.048 (0.000–0.095)	0.052
HbA1c (mmol/mol)	0.011 (0.000–0.021)	0.042	0.014 (0.003–0.025)	0.015
Serum creatinine (μmol/l)	0.001 (− 0.006 to 0.008)	0.847	0.000 (− 0.008 to 0.008)	0.965
Current smokers	− 0.042 (− 0.507 to 0.424)	0.858	− 0.030 (− 0.504 to 0.444)	0.900
Medication use
Insulin	0.245 (− 0.080 to 0.570)	0.137	0.283 (− 0.056 to 0.622)	0.100
Statins	− 0.349 (− 0.728 to 0.030)	0.070	− 0.237 (− 0.762 to 0.014)	0.058
Glucose lowering drugs	− 0.199 (− 0.708 to 0.311)	0.439	− 0.268 (− 0.812 to 0.276)	0.329
Lipid lowering drugs	0.009 (− 0.322 to 0.340)	0.957	0.014 (− 0.323 to 0.350)	0.934
Antihypertensive	− 0.146 (− 0.656 to 0.365)	0.571	− 0.136 (− 0.656 to 0.385)	0.605
Platelet inhibitors	0.155 (− 0.251 to 0.561)	0.448	0.147 (− 0.277 to 0.571)	0.490
Calcium supplements	0.383 (− 0.094 to 0.860)	0.114	0.399 (− 0.097 to 0.895)	0.113
Vitamin D supplements	0.242 (− 0.110 to 0.594)	0.175	0.234 (− 0.130 to 0.597)	0.203

Multivariate linear regression age and sex adjusted. β is per one unit change

*log (coronary calcium score plus 1)

Discussion

This study provides several observations that contribute to our understanding of the determinants of 18F-NaF PET in the femoral arteries of patients with type 2 diabetes and a history of arterial disease. Firstly, we showed that femoral calcium mass on CT is positively associated with 18F-NaF uptake. Second, several cardiovascular risk factors including total cholesterol, triglycerides and HbA1c were associated with 18F-NaF uptake in patients with type 2 diabetes. Together, these data suggest that 18F-NaF uptake is an important imaging biomarker of arterial disease burden both related to dyslipidemia and diabetes control even in extensively treated patients.

Previous research showed that multiple cardiovascular risk factors are associated with 18F-NaF uptake. Derlin et al15 examined 18F-NaF in the carotid arteries in 269 oncologic patients. 18F-NaF uptake was significantly associated with age, male sex, hypertension and hypercholesterolemia. In a different study among 409 oncologic patients, the association between 18F-NaF accumulation in femoral arteries with cardiovascular risk factors and calcified plaque burden was demonstrated.11 In both studies, hypercholesterolemia was associated with increased 18F-NaF uptake, which is in line with our study. An association between HbA1c and 18F-NaF uptake, had been observed in healthy controls.18 In patients with diabetes, high HbA1c is known to be associated with osteocalcin, which supports the concept of vascular calcification in patients with diabetes.19 Blomberg et al18 investigated 18F-NaF uptake in the coronary arteries of 89 healthy adults. According to their results age, female sex and BMI were independent determinants of increased coronary 18F-NaF uptake. In our study we did not find an association between age, sex or BMI and 18F-NaF uptake. The differences in associations between cardiovascular determinants and 18F-NaF uptake compared to our study could be explained by the differences in study population (our cohort consisted predominantly out of males with a high BMI), investigated vessels beds20 and inclusion criteria.

18F-NaF uptake is significantly higher in microcalcifications compared to macrocalcifications.21 Irkle et al21 demonstrated using preclinical μPET/μCT that microcalcifications have no barriers for penetration, while 18F-NaF cannot penetrate into the deeper layers of macrocalcifications. Also, microcalcifications have a relatively big surface area compared to their volume. In a prior longitudinal study, we observed that areas without calcification on CT, but with increased 18F-NaF uptake, had more arterial calcification at follow-up.17 These could represent microcalcifications not yet detectable on CT.17 In the current study CT detected arterial calcification was related to 18F-NaF uptake, suggesting a significant effect of large calcium depositions on tracer uptake with clinical PET/CT systems.

An important strength of the current study is that images were acquired 90 min after tracer injection. The 90 min time-point is considered to be advantageous, because of the balance between signal-to-background ratio and patient comfort. For comparability with other studies we used TBR as our measure of 18F-NaF uptake for research.15 There is debate on how to quantify arterial calcification with NaF PET.22–24 We calculated the TBR using the average of the SUVmax of each femoral artery slice. Though SUVmax is sensitive to noise, it can also be a good measure of the maximum disease burden in a slice.22,23 By averaging (76 slices per patient) the noise effect is reduced to some extent. Others advocate the SUVmean for an individual slice,24 which is less limited by noise, but also averages non-diseased parts of the artery and the blood pool. Differences in PET SUV quantification will influence the observed result therefore it would be important to achieve uniform quantification methods. Other limitations are the relative small sample size and the lack of prospective outcome data.

In conclusion, we showed that cardiovascular risk factors including total cholesterol, triglycerides and HbA1c are associated with 18F-NaF uptake and calcification in the femoral arteries in patients with type 2 diabetes mellitus and a history of arterial disease. Our findings show that 18F-NaF uptake can be an imaging biomarker of arterial disease burden both related to dyslipidemia and diabetes control. Larger studies are needed to investigate whether these findings are generalizable to non-diabetic patients with cardiovascular diseases and if (further) modification of these risk factors is beneficial for 18F-NaF uptake and outcome.

New Knowledge Gained

18F-NaF uptake may be an important arterial imaging biomarker related to dyslipidemia and diabetes control (HbA1c) even in extensively treated patients with type 2 diabetes.

Below is the link to the electronic supplementary material.

Supplementary material 1 (DOCX 12 kb)

Supplementary material 2 (MP3 3557 kb)

Supplementary material 3 (PPTX 7371 kb)