Radiopharmaceuticals for evaluation of the kidney s are classified into three general categories: cortical agents, tubular agents, and glomerular agents.

In recent years, much has been written regarding radiation dose limitation, specifically the concept of “ALARA – as low as reasonably achievable.” This has led authors to favor imaging modalities that do not utilize ionizing radiation whenever possible, such as ultrasonography or magnetic resonance imaging. In situations where this is not feasible or when US or MRI does not perform as well as nuclear or radiographic studies, one needs to choose the study that offers the greatest diagnostic information and with the least radiation exposure.

The majority of nuclear studies, with the exception of cystography, carry a significant radiation burden (Table 6.2). In most cases, nuclear medicine studies used in pediatric urology offer similar or less radiation than other comparable imaging modalities. An exception would be scrotal scintigraphy, whose nonnuclear imaging counterpart is Doppler.^99mTc-DMSA renal cortical imaging, because of its selective tubular uptake and delayed excretion, provides one of the higher effective radiation doses. This is similar to that of IVP but offers greater sensitivity and specificity [18]. Radionuclide cystography, due to the relatively short period of time and distribution limited mostly to the bladder, provides a much lower effective dose [19] than contrast voiding cystourethrography (VCUG). However, strides have been made in reducing the radiation exposure during VCUG by limiting fluoroscopy time and technical refinements of the instrumentation. The use of grid-controlled pulsed fluoroscopy decreases radiation doses across all age groups by 10–50 % compared to continuous fluoroscopy with an effective dose comparable to that of DRC [19, 20].

Traditionally, GFR has been determined by clearance of a substance that is (1) completely filtered by the glomerulus; (2) not synthesized, destroyed, reabsorbed, or secreted by the renal tubule; (3) physiologically inert; and (4) not bound to plasma proteins [21]. Agents such as inulin, urea, and creatinine have been used to measure GFR. While inulin clearance is generally acknowledged as the gold standard, it has limited clinical utility as it must be continuously infused and requires multiple timed blood samples. ^99m Tc-DTPA, however, despite having minimal protein binding, meets the above criteria and requires only one injection [22].

GFR can be estimated via DTPA scintigraphy with serial timed plasma samples or via a camera-only method, which is less cumbersome, requires less time, and does not involve blood sampling. Briefly, DTPA is drawn up into a syringe, a 1-min preinjection measurement of the amount of radioactivity in the syringe is made, it is rapidly injected intravenously, a 1-min postinjection measurement of the empty syringe is made, and serial images of both kidneys are acquired at 1-min intervals between 1 and 3 min postinjection. The 2–3-min interval is the best time for GFR measurement. Both total and fractional GFR can be estimated. The entire duration of the study is less than 10 min.

The camera-only technique was described by Gates who compared DTPA measurements of GFR to 24-h creatinine clearance values obtained for each patient. There was an excellent linear correlation (r ² > 0.9) between the two measurements across a wide range (1–113 mL/min) of renal function. The DTPA study was repeated 24–48 h later, and a remarkable correlation coefficient (99 %) with the first study confirmed the technique’s reproducibility [23]. Some studies suggest that^99mTc-DTPA is not reliable for estimating GFR in young children [24, 25]. Chandhoke et al. reported significant inaccuracy in estimation of GFR when compared to clearance of continuously infused iothalamate in children. They postulated that this may be due to the fact that the adult renal depth correction used in the Gates method may not be applicable to children [26].

The very high uptake of^99mTc-DMSA by the pars recta of the proximal tubules and its minimal glomerular filtration and loss in the urine allows for detailed images of the renal cortex (Fig. 6.1). Traditionally,^99mTc-DMSA cortical imaging has been used to delineate areas of cortical hypoperfusion that accompany acute pyelonephritis and also to demonstrate renal scarring from prior infections. Lesions associated with acute pyelonephritis appear as areas of photopenia, with preservation of the normal renal contour. In contrast, renal cortical scars appear as wedge-shaped areas of photopenia which distort the renal contour (Fig. 6.2). This information can aid in surgical planning and has been proposed to identify patients who should be screened for vesicoureteral reflux, the so-called top-down approach [27, 28].

^99mTc-DMSA cortical scintigraphy also can provide information about differential renal function (Fig. 6.3). However, because it is filtered to a small degree, and blood clearance occurs over a period of hours, variations in renal DMSA uptake are not always equivalent to true variations in function. Thus,^99mTc-DMSA should not be used to estimate absolute renal function or GFR [29].

In an effort to standardize reporting and methodology for renal cortical scintigraphy, several consensus statements and procedure guidelines have been issued, but there is still some variation in the timing of image acquisition (1.5–4 h postinjection), the need for sedation, type of views (planar and pinhole magnification vs. SPECT), and reporting of results [30, 31]. For clarity, the described technique is that of the Society of Nuclear Medicine Guideline publication in 1997, with some recent modifications by Zeissman et al. [31]. Single-photon emission computed tomography (SPECT) imaging will be described separately as there is considerable disagreement as to its utility versus standard planar and pinhole magnification images and also serious difficulties in multicenter standardization [31–34].

A large or standard field of view gamma camera equipped with a parallel-hole high-resolution collimator and/or a pinhole collimator with a 4 mm aperture is recommended. The patient is placed supine for parallel-hole imaging and prone for pinhole imaging. Imaging can begin at 1.5–2 h postinjection, although superior image quality is obtained by waiting at least 4 h [31, 35]. Classically, images are acquired in the posterior and the left and right posterior oblique planes. Differential function is calculated from the posterior planar view using the parallel-hole collimator. Background correction can be undertaken by selecting a region of interest (ROI) around both kidneys in the planar projection and then drawing circumferential background regions approximately 2 pixels in width and 2 pixels away from the kidney. Acquisition time for the images varies by technique, with pinhole magnification requiring fewer counts (150,000 vs. 300,000) but longer acquisition time than the parallel-hole collimator (10 min vs. 5 min). For pinhole images, the kidney should fill approximately 2/3 of the field of view. In cases of significant hydronephrosis, furosemide can be given and/or delayed images can be performed up to 24 h after injection of radiotracer.

Various techniques have been described to adequately image the cortex; most centers rely on pinhole magnification to properly delineate renal scarring and acute pyelonephritis. These images should be correlated with other imaging modalities such as ultrasonography and voiding cystography, when available. A sample grading protocol used in the RIVUR trial is described in Table 6.3 in which the renal cortex is divided into numbered segments and severity graded by the quantity of segments involved [31].

Single-photon emission computed tomography, or SPECT, provides a panoramic multi-image view of both kidneys versus the standard three views of planar and pinhole magnification. Some suggest that SPECT is more sensitive and easier to interpret than planar and pinhole images. Other disagree contending that SPECT is less specific than planar and pinhole imaging. In an international multicenter survey of nuclear medicine experts, SPECT was performed routinely by only 22 % of respondents and seldom performed by 45 % of respondents and never used by 33 % of the respondents [30]. Further complicating matter is the absence of an accepted standard technique, with several authors reporting a range of projections (60–120) and varying increments of scanning time [35–37].

^99mTc-DMSA planar imaging has been shown to have excellent interobserver agreement in several reports using various measures. Consensus opinion regarding “normal” and “abnormal” among expert observers was 93.5 % and 90.5 %, respectively [38, 39]. While the reliability and reproducibility of planar and pinhole imaging is well established, little has been published on SPECT interobserver agreement, and the absence of a standard acquisition protocol hinders accurate comparisons among studies. Craig et al. suggest decreased reproducibility and interobserver agreement in interpretation of SPECT images compared to planar images [40]. Cost also can be a factor; by one estimate, SPECT costs up to 36 % more to perform than planar imaging. This is presumably due to longer image acquisition times and a possible greater need for sedation [41].

In comparison to standard planar and pinhole imaging, SPECT is comparable, possibly at the expense of decreased specificity and higher false-negatives. The performance of standard planar and pinhole magnification DMSA cortical imaging in the diagnosis of acute pyelonephritis has been widely investigated, and most authors agree on a sensitivity around 90 % and specificity between 90 and 100 % [42, 43].

Craig et al. report the results of a meta-analysis of published animal trials evaluating the overall ability of^99mTc-DMSA cortical scintigraphy to diagnose acute pyelonephritis. He initially found an average sensitivity of 84 % and an average specificity of 88 %; however, after correcting statistical errors and excluding some unusable data, there were no significant difference between SPECT and planar imaging [41].

In one elegantly designed study correlating histopathology and cortical scintigraphy in piglets, Majd et al. directly compared SPECT and pinhole imaging for detection of pyelonephritis 1–10 days after infection. After evaluating all zones of pyelonephritis over the entire time period, SPECT carried a sensitivity of 91 % vs. 86 % for pinhole imaging but was less specific, 82 % vs. 95 % compared to planar imaging. Overall accuracy was similar for both techniques, 88 % for SPECT and 90 % for pinhole imaging (Table 6.4) [42].

Rossleigh et al. compared SPECT to standard planar imaging in the diagnosis of renal cortical scarring in a refluxing piglet model with post-study histopathologic confirmation. Each animal after VCUG-confirmed VUR and^99mTc-DMSA-confirmed pyelonephritis underwent another planar^99mTc-DMSA and SPECT scan 3 months later. They reported similar sensitivity, specificity, and overall accuracy between planar images, pinhole magnification, and SPECT (Table 6.5) [34].

There are several human studies demonstrating the ability of SPECT to detect more renal lesions in acute pyelonephritis and renal cortical scars than planar and pinhole imaging. Tarkington et al. compared the ability of SPECT and pinhole imaging to detect renal cortical defects of various etiologies in 33 children, about 60 % of who did not have UTI, VUR, or prior pyelonephritis. In this heterogeneous population, 63 % of the pinhole image studies read as “normal” were found to have cortical abnormalities on SPECT. They also found that SPECT “clarified or enhanced” the pinhole imaging in 71 % of the kidneys studied. Overall SPECT found 55 % more cortical defects than pinhole imaging [44]. Yen et al. compared SPECT to planar imaging in 27 infants and 17 children upon diagnosis of acute pyelonephritis and at 3 months posttreatment and found SPECT identified significantly more scarring than planar imaging (33 % vs. 9.5 %, p < 0.05) at 3 months, particularly in the presence of high-grade VUR than those found using planar imaging [36]. The prospective data of Applegate et al. seem to confirm prior human studies that SPECT detects more “definite” lesions than pinhole and planar images. Cortical defects were classified as “possible” or “definite” and “single” or “multiple” as defined by a consensus of three blinded experts. However, they found that no technique was superior for detecting multiple as opposed to solitary defects [32].

Data from studies in human subjects seem to corroborate previous animal studies that demonstrate increased sensitivity with SPECT, but without histopathologic analysis, false-positives cannot be excluded with certainty. As such, the pooled data from the animal studies is slightly less impressive but probably more accurate than that reported in human subjects. Regardless of technique, the performance of^99mTc-DMSA renal cortical scintigraphy in the diagnosis of acute pyelonephritis and the detection of renal cortical scarring is excellent. At best, SPECT is slightly more sensitive than, and is comparable in specificity to standard pinhole imaging. It seems that the combination of SPECT and pinhole imaging may improve interobserver agreement and may be more useful than either test alone.

In 1992 the Society for Fetal Urology and the Pediatric Nuclear Medicine Council released a consensus statement regarding performance of the “well-tempered diuretic renogram” in response to a lack of standardization of technique and interpretation of dynamic renal imaging. However, this protocol has not been adopted universally, and debate continues regarding such technical aspects as method of pretest hydration, need for urethral catheterization, timing of diuretic administration, and even the method of calculation of urinary drainage [45, 46].