A novel use of hill function and utility of 99mTc-MIBI scintigraphy to detect earlier lower extremity microvascular perfusion in patients with type 2 diabetes

Abstract We use the Hill function to analyze the dynamics of Tc-99m 2 methoxy-isobutyl-isonitrile (99mTc-MIBI) scintigraphy data and to examine the earlier lower extremity microvascular perfusion of diabetic patients without typical clinical symptoms and with the preserved normal ankle-brachial index (ABI). Eighty-eight participants (30 healthy control, 34 diabetic patients, and 24 diabetic patients with peripheral arterial disease [PAD]) were recruited and applied 99mTc-MIBI scintigraphy. Fourteen diabetic patients with PAD also underwent computed tomography angiography (CTA) examination and were performed endovascular interventions. Diabetic patients with normal ABI already have significantly impaired maximum 99mTc-MIBI muscle perfusion counts (P < .001) and the peak times of the lower extremity muscle perfusion (P < .05). 99mTc-MIBI scintigraphy showed great consistent with ABI and CTA in detecting PAD. 99mTc-MIBI scintigraphy was also found to be effective in evaluating lower extremity circulation after endovascular interventions (P < .05). Hill function-based analysis of 99mTc-MIBI scintigraphy might be effective method to evaluate earlier lower extremity perfusion changes in diabetic patients.


Introduction
Peripheral arterial disease (PAD) is an atherosclerotic disease frequently accompanied with diabetes which severely worsens the lower extremity function. [1] The early diagnosis and evaluation of lower extremity microvascular perfusion in diabetic patients are very important for prompt treatment. Impairment of microcirculation represented by lower extremity muscle perfusion has already occurred in patients without typical clinical symptoms. [2] However, current clinic indices for the evaluation and detection of PAD are usually focused on large vessels. Nuclear imaging approaches have superior in the assessment of muscle perfusion and have been used in quantitatively assessment of lower extremity perfusion in patients. [3][4][5][6] However, its clinical application is limited by the lacking of standard analysis method. Hill function is a concentration-response function, which can be candidate for simulating tissue and cellular uptake of 99m Tc-MIBI dynamics. This study took a novel Hill function-based analysis of 99m Tc-MIBI scintigraphy data, to evaluate earlier lower extremity perfusion changes in diabetic patients.

Study design, imaging acquisition, and analysis
Eighty-eight participants were recruited and assigned to the control group (Cont, had no history of diabetes or PAD, n = 30), diabetes group (DM, with normal ankle-brachial index (ABI), n = 34), and diabetes with the PAD group (PAD_DM, n = 24). ABI is obtained by using the blood pressure differential between the brachial systolic blood pressure and systolic blood pressure from the posterior tibial artery, with a hand-held Doppler (Huntleigh, Cardiff, CF24 5HN, UK). For the lower extremity muscle perfusion assessment, the intravenous injection of a 370-640 MBq (10-20 mCi) bolus of 99m Tc-MIBI (College of Chemistry, Beijing Normal University, China) were performed, and serial dynamic posterior planar images were recorded for 5 minutes in the regions of both calves. Using an image analysis device (Xeleris 3, GE Medical System), serial summed images and time-activity curves were created from the data. Fourteen patients from the PAD_DM group who underwent CTA and performed endovascular interventions were further divided into 3 groups according to the TASC grading method, [7] namely a mild group (TASC A), a moderate group (TASC B), and a severe group (TASC C and D). A flow chart representing the above participants' selection and breakdown of the group intervention was provided ( Fig. 1). Three independent investigators evaluated the images while blinded to the participants' clinical and ultrasound data. This study was approved by the hospital ethical committee and written informed consent was obtained from all the participants.
Originally, the Hill function was proposed to describe the in vivo concentration response relationship, [8] which fits concentration-response curve data. Similarly, the tissue uptake of 99m Tc-MIBI is proportional to blood flow, and its cellular uptake and retention are dependent on mitochondrial and plasma membrane potentials. [9] Hill function parameters were fitted using the greedy algorithm method implemented in Matlab software (Mathworks). Since the matching of target distribution is associated with simulating steps, we used 100,000 steps to ensure the accuracy of our parameter distributions.
The Hill function can be written as V max : the maximum 99m Tc-MIBI perfusion counts in the tissue. t: time-dependent observed 99m Tc-MIBI perfusion counts in the tissue.
K: under the condition of u = v, half-saturation constant 99m Tc-MIBI perfusion counts.
u, v: constants of effect response, and their ratio represents the steepness of the effect-concentration curve, which is a measure of how strongly the 99m Tc-MIBI perfusion counts changes in response to changes in time. Let v , which is the measurement of muscle perfusion ability.

Statistics analysis
All results were presented as mean ± SD. Differences among groups were compared by means of the chi-squared test for categorical variables, and 1-way analysis of variance (ANOVA) with Bonferroni's multiple comparison post-hoc test and Students' t-test for continuous variables of 3 or 2 groups comparison respectively. A 2-sided P value of less than 0.05 was considered to indicate statistical significance. All statistical analyses were conducted with the use of SPSS software, version 13.0 (SPSS Inc.).

Baseline characteristics of the participants
The clinical baseline characteristics of the participants in each group are presented in Table 1. No differences were found in age, gender, serum total cholesterol (TC), triglyceride (TG), and HDLc among groups. Although being apparently distinct from the Cont group, no differences were found in HbA1c, fasting blood glucose, hs-CRP, and LDL-c between DM and PAD_DM groups. The diabetic patients with PAD exhibited profoundly decreased ABI values (P < .001) compared with the other groups.

Earlier changes in lower extremity muscle perfusion in the diabetic patients
The estimated Hill function parameters using the 99m Tc-MIBI muscle perfusion data are also provided in Table 1. However, with normal ABI values, diabetic patients already had significantly decreased maximum 99m Tc-MIBI muscle perfusion counts compared with healthy participants (565.6 ± 126.5 vs 756.6 ± 86.8, P < .001), and decreased lower extremity muscle perfusion became more profoundly in the DM_PAD group. No differences were found in half saturated 99m Tc-MIBI muscle perfusion counts among 3 groups, which may represent the smaller differences of the basic characteristics among groups. The peak times of the lower extremity muscle perfusion of DM and DM_PAD groups were significantly prolonged compared with the Cont group (87.7 ± 24.8 vs 78.0 ± 14.9, P < .05; 92.7 ± 23.8 vs 78.0 ± 14.9, P < .01), whereas the results showed no significant relation with PAD.

Comparison of 99m Tc-MIBI scintigraphy with CTA and ABI in the evaluation of lower extremity perfusion
To examine the sensitivity of Hill function parameters in detecting PAD occurrence, strong consistencies of ABI and the Hill function parameters regarding lower limb microvascular perfusion were observed (Fig. 2). Based on the CTA results and TASC vascular grades, 14 patients who underwent CTA examinations were further divided into 3 groups (i.e., mild, moderate, and severe, 10 limbs in 8 cases, 9 limbs in 7 cases, and 9 limbs in 8 cases, respectively). The maximum 99m Tc-MIBI muscle perfusion counts decreased dramatically with the severity of the vascular lesions represented by the TASC vascular grade increased (Fig. 3). Although no differences of hs-CRP were observed among groups, the maximum 99m Tc-MIBI muscle perfusion counts showed significantly inverse correlation with the hs-CRP level (Fig. 3).

Assessment of the efficiency of endovascular interventions in the lower extremities using 99m Tc-MIBI scintigraphy
Potential application of 99m Tc-MIBI scintigraphy in assessing the efficiency of improvements in lower extremity microcirculation in the diabetic patients who underwent endovascular interventions was also examined. Apparent improvements in muscle perfusion parameters were observed in calves (P < .05), although the peak time of the muscle perfusion failed to reach statistical significance (Fig. 4).

Discussion
This novel study is the first to describe the use of Hill function in analyzing 99m Tc-MIBI scintigraphy data to evaluate earlier lower extremity microvascular perfusion in diabetic patients. We found that asymptomatic diabetic patients with normal ABI already have significantly impaired lower extremity muscle perfusion. 99m Tc-MIBI scintigraphy showed great consistent with ABI and CTA in detecting PAD, whereas exhibited advantages in evaluating lower extremity microcirculation. 99m Tc-MIBI scintigraphy was also found to be effective in assessing lower extremity microcirculation after endovascular interventions.
Hemodynamic changes in the lower extremities frequently occur before the onset of PAD and even before the diagnosis of diabetes, measurements of muscular microcirculation could be of considerable importance in the diagnosis of earlier lower extremity microvascular perfusion. Techniques for restenosis detection include ABI measurements, duplex ultrasound, MR  imaging, and CT angiography. These methods have some shortcomings in the evaluation of the degree and extent of tissue blood perfusion, especially in the early stage of diabetes. [10]99m Tc-MIBI scintigraphy has been used to assess lower extremity perfusion in PAD and diabetic patients, which showed high sensitivity and specificity. [2][3][4][5][6] However, several major limitations restrict its clinic application. The rest and stress perfusion were usually used as the indication of earlier impaired microcirculation. However, there is no unified operational protocol and methods of stress test are diverse. Second, the total counts of the selected regions were used in previous studies, which ignoring the hemodynamic changes of muscle perfusion. Finally, previous statistic analysis needed to select reference 99m Tc-MIBI uptake as control, and variate standards used in different studies lead to poor comparability. We used the Hill function to simulate time-activity data of 99m Tc-MIBI scintigraphy, which largely minimized artificial and bias of statistical methods. In accordance with the biological meanings of the Hill function parameters, we found significant This study has several limitations. First, this is an observational study, and the generality of its results to other countries is still unclear. Second, our study did not include nondiabetic patients with PAD, and the difference between them and diabetic patients with PAD needs further study. Third, due to the lack of well-documented statistical methods on analyzing 99m Tc-MIBI perfusion data, we cannot draw receiver operating characteristic (ROC) curve to evaluate our method's diagnostic accuracy.
This study provided an effective and noninvasive method for the earlier evaluation of lower limb microvascular perfusion in diabetic patients via the use of Hill function based 99m Tc-MIBI scintigraphy analysis. Further studies of the