Abstract
Objective: The objective of this study was to clarify whether or not pulse volume recoding (PVR) parameters have screening capability equivalent to ankle-brachial pressure index after walking (Ex-ABI) for patients with 0.91 or higher ABI.
Patients and Methods: The subjects were 87 patients (147 limbs) with symptoms of lower extremities with 0.91 or higher ABI. In all patients, upstroke time (UT), percentage of mean artery pressure (%MAP) of PVR and Ex-ABI were measured, and computed tomographic angiography (CTA) was concomitantly performed.
Results: Area under the curve (AUC) of receiver operating characteristics (ROC) curves of Ex-ABI, %MAP, and UT were 0.90, 0.70, and 0.81, respectively. A significant difference was noted in AUC between Ex-ABI and %MAP (p <0.001). When the cut-off values were set at %MAP ≥45% and UT ≥180 msec, the accuracies of %MAP and UT were markedly lower than that of Ex-ABI. When the cut-off values were corrected to the values determined from the ROC curves (%MAP ≥41, UT ≥164 msec), the diagnostic accuracy of UT increased markedly.
Conclusion: In patients with 0.91 or higher ABI, screening capability of PVR parameters was markedly lower than that of Ex-ABI, but UT has screening capability close to that of Ex-ABI when the cut-off value is corrected downward.
Keywords: peripheral arterial disease, ankle-brachial pressure index, pulse volume recording
Introduction
The standard value of ankle-brachial pressure index (ABI) to detect peripheral artery disease (PAD) is 0.9 or lower,1) but sticking to this value increases the probability of overlooking PAD showing boundary (0.91–1.0) and normal (1.1 or higher) values.2) To prevent this overlooking, measurement of ABI immediately after walking (Ex-ABI) is recommended for patients with 0.91 or higher ABI.3) However, although Ex-ABI is a superior method to detect the above PAD, it is not readily introduced into busy clinical sites because it requires an expensive exercise device, multiple manpower, and testing time. Solving these problems was tried by concomitant use of pulse volume recoding (PVR) for measurement of ABI in some studies.4–6) Using oscillometric methods, PVR parameters can be measured simultaneously with ABI, being easy to introduce it into clinical sites. However, it is unclear whether or not these PVR parameters have screening capability equivalent to Ex-ABI. The objective of this study was to clarify whether or not these PVR parameters have screening capability equivalent to Ex-ABI for patients with 0.91 or higher ABI.
Patients and Methods
Patients
Of 111 patients (222 limbs) with one of the following symptoms: intermittent claudication, lower limb coldness, and numbness who underwent Ex-ABI and computed tomographic angiography (CTA), 87 patients (174 limbs) after excluding maintenance hemodialysis patients7) were selected for the subjects. From the 174 limbs, 17 limbs treated with lower limb bypass or endovascular catheterization and 10 limbs with 0.9 or lower ABI were excluded, and the remaining 147 limbs were investigated (Table 1). This study was performed after review by the Ethics Committee of our hospital.
Table 1.
Clinical characteristics of patients
| Number of limbs | 147 |
| Number of patients | 87 |
| Gender (men/women) | 74/13 |
| Age (years) | 70.4 ± 8.9 |
| Hypertension | 56 (64.4) |
| Diabetes melilitus | 26 (29.9) |
| Dyslipidemia | 42 (48.3) |
| Coronary artery disease | 36 (41.4) |
| Cerebrovascular disease | 7 (8.0) |
| Symptoms of lower extremities | |
| Intermittent claudication (limbs) | 65 (44.2) |
| Coldness (limbs) | 69 (46.9) |
| Numbness (limbs) | 42 (28.6) |
| ABI | 1.12 ± 0.10 |
| ABI, 0.91–0.99 (limbs) | 21 (14.3) |
| ABI, 1.0≤ (limbs) | 126 (85.7) |
| UT (ms) | 158.4 ± 28.1 |
| %MAP (%) | 39.3 ± 4.4 |
ABI: ankle-brachial pressure index; UT: upstroke time; %MAP: percentage of mean artery pressure
ABI, percentage of mean artery pressure, and upstroke time
After resting in a supine position for 15 minutes, ABI, upstroke time (UT), and percentage of mean artery pressure (%MAP) were measured using an automatic blood pressure pulse wave measurement device (form BP-203RPE SERIES, OMRON COLIN Co., Ltd., Tokyo, Japan) employing the oscillometric method. ABI was determined by dividing arterial systolic ankle pressure by that of one with a higher brachial pressure. %MAP was determined by dividing the mean waveform area above the baseline on air plethysmography by the waveform amplitude. As UT, the time from the rise to peak on air plethysmography was measured. The cut-off values to detect PAD were set at ABI ≤0.9, %MAP ≥45%, and UT ≥180 msec.8)
Ex-ABI
ABI was measured immediately after walking using the computerized treadmill system (CASE 12, Marquette Electronics, Milwaukee, WI, USA) following a protocol. Each patient’s 12-lead electrocardiogram and heart rate were monitored throughout the walking protocol for 1 minute (2.4 km/h, 12%). However, the walking speed was reduced to 2.0 km/h when the patient was more than 70 years old or was a low left ventricular function case.9)
Walking was discontinued when one of the following conditions appeared: 1) horizontal or downsloping ST depression of 0.1 mV compared with the resting ST at 80 msec after the J point, 2) abnormal blood pressure responses (systolic pressure ≥250 mmHg and diastolic pressure ≥120 mmHg), 3) dangerous or potentially dangerous arrhythmias, 4) the maximum heart rate obtained using the equation “220 – age,” 5) difficulty in walking due to lower limb pain, and 6) symptoms of the cardiopulmonary or neurological system.9)
The definition of PAD required a decrease more than 15% of the Ex-ABI.3)
CTA
CTA was performed in all subjects within 1 week before or after the Ex-ABI measurements to confirm the stenotic lesion. CTA was performed with a 320-detector-row CT scanner (Aquilion one; Toshiba Medical, Tokyo, Japan). The scanning range was planned with a scout view and included the entire vascular tree from the abdominal aorta to ankles. A total of 60–90 mL of contrast media (Proscope 300, Tokyo, Japan) was administered with an automated injector (KCA00226, Nemoto, Tokyo, Japan) at a flow rate of 2–3 mL/sec through a 20-G needle that was placed in a superficial vein. PAD is defined as the presence of stenosis of more than 75% in the case of lesion from an iliac artery to knee on CTA.
Statistical analysis
The receiver operating characteristics (ROC) curves were drawn from the sensitivity and false-positive rate of the tests of EX-ABI, %MAP, and UT. To judge the superiority or inferiority of the tests, area under the ROC curves (AUC) was determined. For analysis of differences in AUC, the c2 test was used. The optimal cut-off point of each test was set to the test value at a point on the ROC curve nearest to the left upper corner of the ROC graph. The p value of <0.05 was considered to be statistically significant.
Results
Significant stenosis lesions were observed in 27 limbs (18.4%) on CTA. ABI became positive after the exercise in 57 limbs (38.8%). UT and %MAP were abnormal in 34 (23.1%) and 16 (10.9%) limbs, respectively.
AUC of %MAP was significantly lower than that of Ex-ABI (p <0.001) (Fig. 1). On the other hand, AUC of UT was not significantly different from that of Ex-ABI (p = 0.067) (Fig. 2).
Fig. 1.
Receiver operating characteristics analysis of exercise-ABI and %MAP in patients with ABI above 0.91. Point A is conventional cut-off value. Point B is optimum cut-off value in this study. ABI: ankle brachial pressure index; %MAP: percentage of mean artery pressure; AUC: area under the ROC curves
Fig. 2.
Receiver operating characteristics analysis of exercise-ABI and UT in patients with ABI above 0.91. Point A is conventional cut-off value. Point B is optimum cut-off value in this study. ABI: ankle brachial pressure index; UT: upstroke time; AUC: area under the ROC curves
No significant difference was noted in AUC of the ROC curve of patients with intermittent culaudicatin (IC) between EX-ABI (0.91) and UT (0.80) (p = 0.087), but %MAP (0.64) was significantly lower than EX-ABI (p = 0.001). On the other hand, in patients with non-IC symptoms showing only numbness and cold sense of the lower limbs, AUC of the ROC curve of EX-ABI (0.88) was not significantly different from that of UT (0.77, p = 0.313) or %MAP (0.70, p = 0.067).
When the cut-off values to detect PAD were set at the conventional values: %MAP ≥45% (Fig. 1, point A) and UT ≥180 msec (Fig. 2, point A), the sensitivities of %MAP and UT were markedly lower than that of Ex-ABI. This tendency was also noted when %MAP and UT were used in combination (Table 2).
Table 2.
Diagnostic accuracy of exercise-ABI and PVR parameters using conventional cut-off value in patients with ABI above 0.91
| Sensitivity (%) | Specifcity (%) | Positive predictive value (%) | Negative predictive value (%) | |
|---|---|---|---|---|
| Ex-ABI | 96.2 | 73.5 | 43.9 | 98.9 |
| UT | 55.6 | 84.2 | 44.1 | 89.4 |
| %MAP | 33.3 | 94.2 | 56.3 | 86.3 |
| Combination of UT and %MAP | 66.7 | 81.7 | 45.0 | 91.6 |
ABI: ankle-brachial pressure index; Ex: exercise; UT: upstroke time; %MAP: % mean artery pressure; PVR: pulse volume recoding
The cut-off value determined from the ROC curve was decreased by 16.8% in Ex-ABI, 41% in %MAP (Fig. 1, point B), and 164 ms in UT (Fig. 2, point B). When the diagnostic accuracy of the cut-off value of Ex-ABI was compared between that determined from the ROC curve and conventional cut-off value, these were the same (Tables 2 and 3). However, when %MAP and UT as an index, the sensitivity of the cut-off value determined from the ROC curve was far superior to the conventional cut-off value (Tables 2 and 3).
Table 3.
Diagnostic accuracy of exercise-ABI and PVR parameters using optimum cut-off value in patients with ABI above 0.91
| Sensitivity (%) | Specifcity (%) | Positive predictive value (%) | Negative predictive value (%) | |
|---|---|---|---|---|
| Ex-ABI | 96.2 | 73.5 | 43.9 | 98.9 |
| UT | 85.2 | 69.2 | 38.3 | 95.4 |
| %MAP | 63.0 | 66.7 | 29.8 | 88.9 |
| Combination of UT and %MAP | 85.1 | 50.8 | 28.0 | 93.8 |
ABI: ankle-brachial pressure index; Ex: exercise; %MAP: % mean artery pressure; UT: upstroke time; PVR: pulse volume recoding
Discussion
Diagnostic accuracy of PVR parameters
When the patients with symptoms suspected as PAD with 0.91 or higher ABI were diagnosed, the PVR parameters played an interesting role. AUC of Ex-ABI was high (0.900), and AUC of UT was not significantly different from that of Ex-ABI, but AUC of %MAP was significantly lower than that of Ex-ABI. These findings suggested that the diagnostic ability of %MAP was low but that of UT was mostly equivalent to Ex-ABI in patients with 0.91 or higher ABI.
Hashimoto et al.6) evaluated the diagnostic abilities of %MAP and UT using CTA as a reference in a group including many patients with 0.9 or lower ABI (range: 0.63–1.11), in which AUC of %MAP was 0.916. In contrast, in our study limiting patients to those with 0.91 or higher ABI, the diagnostic ability of %MAP decreased (AUC = 0.700). %MAP indicates the presence of stenosis lesions in the region with 0.9 or lower ABI, but this ability may not be expected for regions with 0.91 or higher ABI. In the study reported by Hashimoto et al.,6) AUC of UT was 0.798, being almost consistent with that (0.810) in our study. These studies suggest that the diagnostic ability of UT is constantly high regardless of ABI being 0.9 or lower or 0.91 or higher.
Cut-off value
Since no significant difference was noted in AUC between UT and Ex-ABI in patients with 0.91 or higher ABI, UT was considered to have diagnostic ability almost equivalent to that of Ex-ABI, but the cut-off value has to be investigated to support this.
The conventional cut-off values to detect PAD were %MAP ≥45% and UT ≥180 msec. When these cut-off values were used in patients with 0.91 or higher ABI, the sensitivities of %MAP and UT were markedly lower than that of Ex-ABI, that is, application of the conventional cut-off values of %MAP and UT for patients with 0.91 or higher ABI increases overlooking lesions.
The optimum cut-off values were determined from the ROC curves in patients with 0.91 or higher ABI, and the values were %MAP ≥41% and UT ≥164 ms. Both cut-off values were lower than the conventional values, but the sensitivity of UT got close to that of Ex-ABI using these values and UT becomes a suitable test index for screening.
When ABI is 0.91 or higher despite symptoms suggesting PAD, Ex-ABI is the superior method to detect PAD. However, Ex-ABI is difficult to introduce into busy clinical sites. For facilities unable to perform Ex-ABI, downward corrected cut-off value of UT can be used as a substitute for Ex-ABI.
Study limitations
With CTA, it is difficult to visualize lumens with blood flow within severe calcification lesions. In addition, the possibility that patients with multi-vessel occlusive lesions below the knee affected the diagnostic accuracy of each measurement cannot be excluded.
Conclusion
For patients with 0.91 or higher ABI, PAD screening capability of PVR parameters is far inferior to Ex-ABI. However, when the cut-off value of UT is corrected downward, it acquires screening capability almost comparable to that of Ex-ABI.
Acknowledgment
The authors thank Mrs Tomomi Sekino for technical support.
Disclosure Statement
All authors have no conflict of interest.
Author Contributions
Study conception: KT
Data collection: ST, YO, TM, KK, IY, KY, and ST
Analysis: KK and MA
Writing: KT
Funding acquisition: None
Critical review and revision: all authors
Final approval of the article: all authors
Accountability for all aspects of the work: all authors
References
- 1).Orchard TJ, Strandness DE., Jr Assessment of peripheral vascular disease in diabetes. Report and recommendations of an international workshop sponsored by the American Diabetes Association and the American Heart Association September 18-20, 1992 New Orleans, Louisiana. Circulation 1993; 88: 819-28. [DOI] [PubMed] [Google Scholar]
- 2).Natsuaki C, Inoguchi T, Maeda Y, et al. Association of borderline ankle-brachial index with mortality and the incidence of peripheral artery disease in diabetic patients. Atherosclerosis 2014; 234: 360-5. [DOI] [PubMed] [Google Scholar]
- 3).Norgren L, Hiatt WR, Dormandy JA, et al. Inter-Society Consensus for the Management of Peripheral Arterial Disease (TASC II). Eur J Vasc Endovasc Surg 2007; 45: S5-67. [DOI] [PubMed] [Google Scholar]
- 4).Nakajima R, Inoue Y, Sugano N, et al. Upstroke time and percentage of mean arterial pressure with ABI-form. J Jpn Coll Angiol 2005; 45: 7-10. [Google Scholar]
- 5).Goto M, Kimura Y, Ogawa M, et al. Assessment of below-knee artery lesion using an automatic device for blood pressure and pulse wave measurement. J Jpn Coll Angiol 2011; 51: 223-7. [Google Scholar]
- 6).Hashimoto T, Ichihashi S, Iwakoshi S, et al. Combination of pulse volume recording (PVR) parameters and ankle-brachial index (ABI) improves diagnostic accuracy for peripheral arterial disease compared with ABI alone. Hypertens Res 2016; 39: 430-4. [DOI] [PubMed] [Google Scholar]
- 7).Tsuyuki K, Kohno K, Ebine K, et al. Usefulness of exercise-ankle brachial pressure index test and its diagnostic criteria in patients on maintenance hemodialysis. J Jpn Coll Angiol 2015; 55: 21-5. [Google Scholar]
- 8).Yamashina A. Guideline for non-invasive vascular function test (JCS 2013). Guideline for diagnosis and treatment of cardiovascular disease 2013; 3-112.
- 9).Tsuyuki K, Kohno K, Ebine K, et al. Exercise-ankle brachial pressure index with one-minute treadmill walking in patients on maintenance hemodialysis. Ann Vasc Dis 2013; 6: 52-6. [DOI] [PMC free article] [PubMed] [Google Scholar]