Argentina

Renal Doppler has also been suggested as a useful tool in evaluating intra-renal perfusion in various settings [ 2 – 8 ]. Hence, intra-renal Doppler-based renal resistive index (RI) has been tested to assess renal allograft status [ 9 , 10 ] and changes in renal perfusion in critically ill patients [ 11 – 13 ] and for predicting the reversibility of an acute kidney injury (AKI) [ 14 , 15 ].

10.1.2.1 Methods

Although 2- to 5-MHz transducers are optimal to measure RI [ 16 , 17 ], various transducers may be successfully used for this purpose, including small phased array transducer. The fi rst step is a B-mode US with a postero-lateral approach allowing location of the kidneys and detection of signs of chronic renal damage. Subsequently, colour Doppler or power Doppler US allows vessels’ localization (Fig. 10.1a ) [ 16 ] and may allow a semi-quantitative evaluation of renal perfusion (Table 10.2 ) [ 18 ].

Either the arcuate arteries or the interlobar arteries are then insonated with pulsed wave Doppler using a Doppler gate as low as possible between 2- and 5-mm [ 16 , 17 ]. In order to obtain repeatable measures, the waveforms should be optimized for the measurements using the lowest pulse repetition frequency (usually 1.2–1.4 kHz) without aliasing (to maximize waveform size), the highest gain without obscuring background noise, and the lowest wall fi lter [ 16 , 17 ]. A spectrum is considered optimal when three to fi ve consecutive similar-appearing waveforms are noted [ 16 , 17 ]. To characterize the intra-renal Doppler waveform, most investigators have used the resistive index (RI) so-called Pourcelot Index (Fig. 10.1b ).

Three to fi ve reproducible waveforms are obtained, and RIs from these wave-forms are averaged to compute the mean RI for each kidney. This easily calculated parameter is defi ned as:

Renal pulsatility index may also be calculated:

RI = [peak systolic shift – minimum diastolic shift]/peak systolic shift

PI = [peak systolic velocity – minimum diastolic velocity]/mean velocity

a

b

Fig. 10.1 Results of a renal colour Doppler ultrasonography showing renal vascularization ( a ). RI measurement using pulsed wave Doppler ( b )

Table 10.2 Colour Doppler for a semi-quantitative evaluation of intra-renal vascularisation [ 18 ] Stage Quality of renal perfusion by colour Doppler

0 Unidentifi able vessels

1 Few vessels in the vicinity of the hilum

2 Hilar and interlobar vessels in most of the renal parenchyma

3 Renal vessels identifi able until the arcuate arteries in the entire fi eld of view

RI might however be more adapted to the study of high-resistance vascular territories. In addition, RI and pulsatility index are closely correlated ( r = 0.92;

P < 0.001) [ 11 ]. Last, pulsatility index has been shown to be subject to wider variations than RI (reproducibility 9–22 % vs. 4–7 %) [ 19 ].

10.1.2.2 Normal Values, Feasibility and Reproducibility

RI can theoretically range from 0 to 1. RI is normally lower than 0.70. In several studies, mean RI (±SD) in healthy subjects ranged from 0.58 (±0.05) to 0.64 (±0.04) [ 20 , 21 ]. The normal RI range is, however, age dependent. Thus, RI values greater than 0.70 have been described in healthy children younger than 4 years [ 22 ] and in individuals older than 60 years and considered healthy [ 23 ]. When the RI is measured for both kidneys, the side-to-side difference is usually less than 5 % [ 24 ].

Renal RI is a simple and non-invasive tool easy to use at the patient bedside.

Feasibility of the measure has been showed to be good, even in the settings of critically ill patients. A recent study suggested a half-day course to be suffi cient to allow inexperienced operators in successfully measuring RI [ 25 ]. Inter-observer reproducibility of RI measurement by senior radiologist or senior intensivist is considered excellent [ 14 , 26 ]. In critically ill patients, the inter-observer reproduc-ibility between senior and inexperienced operator is good and measures seem accurate (absence of systematic bias) although associated with a lack of precision (wide 95 % confi dence interval of ±0.1) [ 25 ].

10.1.2.3 Significance and Usual Confounders

Both physiological and clinical signifi cance of the RI remains debated. Initially considered an indicator of renal vascular resistance and blood fl ow [ 7 ], both experi-mental and clinical studies have demonstrated correlation of RI with vascular resis-tance and blood fl ow to be weak [ 27 , 28 ]. Thus, observed RI changes in response to supra-physiological pharmacologically induced changes in renal vascular resistance are modest (RI changes of 0.047 IU (±0.008) per logarithmic increase in renal resis-tances) [ 29 ]. Both in vitro and ex-vivo studies however demonstrated a strong rela-tionship between vascular compliance (vascular distensibility) and RI [ 27 – 29 ]. This strong relationship between vascular compliance and RI has been confi rmed in a recent large cohort of renal allograft [ 10 ]. In this line, age-related arterial stiffening may explain the progressive increase in RI with age [ 30 ]. Similarly, elevated RI observed in several pathological states such as diabetes mellitus and hypertension may also be related to the infl uence of these diseases on arterial stiffness and to sub- clinical vascular changes related to the underlying disease [ 31 , 32 ].

Macrovascular hemodynamic changes also infl uence RI. Hence, pulse pressure index [(systolic pressure – diastolic pressure/systolic pressure)] had direct and dramatic effects on RI values [ 29 ]. Additionally, since RI depends in part on the minimum diastolic shift, it may be infl uenced by the heart rate [ 33 ]. According to observations performed by Mostbeck and colleagues regarding RI changes as consequences of heart rate variations, a formula has been developed to correct the RI value for heart rate: [Corrected RI = observed RI −0.0026 × (80-heart rate)] [ 33 ].

This formula has, however, never been validated in clinical studies.

In addition to these factors, both oxygen and carbon dioxide levels can affect RI. Several studies have demonstrated that RI varies according to P a O 2 and P a CO 2

levels [ 34 – 36 ]. These studies performed in healthy subjects, patients with chronic obstructive respiratory disease, renal transplant recipients or patients with acute respiratory distress syndrome suggest that hypoxemia and hypercapnia may increase RI [ 34 – 36 ].

Besides vascular and hemodynamic factors, kidney interstitial pressure has been shown to be associated with RI in ex vivo studies [ 28 ]. An increase in interstitial pressure reduces the transmural pressure of renal arterioles, thereby diminishing arterial distensibility and, consequently, decreasing overall fl ow and vascular com-pliance. Similarly, intra-abdominal pressure may affect RI via the same mechanisms.

Thus, incremental changes in intra-abdominal pressure correlated linearly with RI in a porcine model [ 37 ], and reduction in intra-abdominal pressure with paracentesis was followed by a decrease in RI in cirrhotic patients with tense ascites [ 38 ]. Finally, ureteral pressure, likely acting via interstitial pressure, also affects RI [ 6 ].

These numerous confounders suggest RI to be an integrative parameter rather than reliable tool to assess renal perfusion or a substitute for renal biopsy.

10.1.2.4 Clinical Relevancy in ICU

Doppler-based RI has been suggested to monitor renal perfusion in critically ill patients, detect early renal dysfunction in severe sepsis patients or in assessing prog-nosis of AKI.

Renal Doppler has also been proposed to monitor renal perfusion in critically ill patients [ 12 ]. In recent studies, RI was used to assess the impact on renal perfusion of low-dose dopamine infusion and gradual changes in mean arterial pressure in response to norepinephrine infusion in critically ill patients [ 11 , 13 ]. Despite signifi -cant results, the observed RI variations were modest and their real impact on renal perfusion and moreover on renal function remains unclear. Assuming that RI may refl ect renal perfusion, it was recently proposed for the early detection of occult hemorrhagic shock in a small study conducted in normotensive trauma patients [ 39 ]. If patients with occult hemorrhagic shock had higher RI, they also had higher lactate levels and lower base excess. Although these fi ndings are promising, the exact signifi cance remains uncertain. Hence, as mentioned above, RI is infl uenced not only by vascular resistance but also by many other parameters such as age, heart rate, mean arterial pressure, changes in renal perfusion, vascular compliance, and renal interstitial oedema and interstitial pressure [ 27 – 29 ]. A study is currently ongo-ing in way to more clearly underline potential interest of Doppler-based RI in assessing renal perfusion (DORESEP; NCT01473498).

Additionally, several studies assessed interest of Doppler-based RI in detecting early renal dysfunction or in predicting short-term reversibility of AKI [ 14 , 15 , 25 , 40 , 41 ]. In a study conducted in septic critically ill patients, RI measured at admis-sion was higher in patients who developed subsequently AKI [ 14 ]. This fi nding was recently confi rmed in the post-operative setting of cardiopulmonary bypass [ 42 ].

Additionally, several cohort studies suggest Doppler-based RI to be differentiating transient from persistent AKI in selected critically ill patients [ 15 , 41 , 43 ].

Interestingly, semi-quantitative renal perfusion assessment seems to be correlated with Doppler-based RI and associated with reversibility of renal dysfunction [ 25 ].

Despite these promising results, most of these studies were performed in limited patient samples which may have overestimated diagnostic performance [ 15 , 43 – 45 ].

Additionally, a recent study has identifi ed discrepant results regarding RI diagnostic performance in this setting [ 44 ]. Therefore despite the promising preliminary reports, we still lack adequately powered study validating performance of RI in both early detection of renal insult or AKI prognostic assessment.