Objectives: Tc-99m-diethylenetriaminepentaacetic acid renal scintigraphy is a noninvasive method for assessment of graft function. Here, we evaluated the diagnostic accuracy of quantitative scintigraphy parameters in differentiating ischemic complications (acute tubular necrosis) from nonischemic complications (acute rejection, tubulointerstitial nephritis), using tissue biopsy as the reference standard.
Materials and Methods: We retrospectively analyzed 63 kidney transplant recipients who underwent Tc-99m-diethylenetriaminepentaacetic acid renal scintigraphy followed by renal biopsy within 6 months posttransplant. Quantitative parameters derived from time-activity curves included delta peak, peak perfusion activity, plateau perfusion activity, peak perfusion activity-to-plateau perfusion activity ratio, peak function activity, peak perfusion activity-to peak function activity ratio, and the 20/3 ratio. We compared scintigraphy findings with biopsy results categorized as ischemic, nonischemic, or normal. For statistical analysis, we used the Mann-Whitney test and receiver operating characteristic curve analysis.
Results: Of the 63 patients, 59 had abnormal biopsy findings (5 ischemic, 54 nonischemic). Scintigraphy findings were concordant with biopsy results in 61 cases. Delta peak (P = .⁰⁴³) and peak perfusion activity-to peak function activity ratios (P = .⁰¹⁵) differed significantly between ischemic and nonischemic groups. Receiver operating characteristic curve analysis showed that delta peak was a good predictor of nonischemic complications (area under curve = 0.772), whereas peak perfusion activity-to peak function activity ratio was a strong predictor of ischemic complications (area under curve = 0.822). A delta peak cutoff value of 6.5 seconds yielded 63% sensitivity and 80% specificity for identifying nonischemic complications, whereas a peak perfusion activity-to peak function activity ratio cutoff of 1.31 provided 80% sensitivity and 82% specificity for detecting ischemic complications.
Conclusions: Quantitative parameters derived from Tc-99m-diethylenetriaminepentaacetic acid renal scintigraphy are useful for differentiating ischemic from nonischemic renal transplant complications and correlate well with biopsy findings. These parameters may aid in the early diagnosis and management of posttransplant graft dysfunction.

Key words : Acute rejection, Acute tubular necrosis, Kidney transplantation, Tubulointerstitial nephritis

Introduction
Since the first successful human renal transplant performed by Joseph Murray in 1954, renal transplant has become a standard treatment for patients with end-stage renal disease.¹ Although transplant offers superior survival and quality of life compared with hemodialysis, transplant is associated with both surgical and nonsurgical complications.² Various invasive and noninvasive diagnostic tools are available for evaluating renal transplant complications, including serum creatinine measurement, Doppler ultrasonography, renal scintigraphy, computed tomography, magnetic resonance imaging, and tissue biopsy. Renal scintigraphy is a noninvasive imaging modality that was first used for graft evaluation in 1956, only 2 years after the first human renal transplant procedure. This technique uses different radiotracers to assess graft function from the renal arteries to the ureters.³ Tc-99m-diethylenetriaminepentaacetic acid (Tc-99m-DTPA) and Tc-99m-mercaptoacetyltriglycine are among the most commonly used radiotracers for renal graft evaluation.⁴ Renal scintigraphy assesses graft function in 3 phases: an initial perfusion phase immediately after tracer injection, a parenchymal uptake phase, and an excretory phase through the pelvicalyceal system.⁴ Time-activity curves (TACs) generated by computer software visually represent these phases, from which several quantitative parameters can be derived.⁵ However, no consensus exists on the diagnostic value or optimal cutoff points of these indices, and many previous studies have not correlated scintigraphy findings with histopathology as the gold standard. Therefore, in this study, we aimed to evaluate the diagnostic accuracy of quantitative Tc-99m-DTPA renal scintigraphy parameters for differentiating acute ischemic and nonischemic renal transplant complications using tissue biopsy as the reference standard.

Materials and Methods

Patient selection
We retrospectively evaluated 63 patients with end-stage renal disease who underwent Tc-99m-DTPA renal scintigraphy within 6 months after renal transplant (between September 2022 and August 2024) at the Department of Nuclear Medicine, Labbafinejad Hospital, Tehran, Iran. All patients subsequently underwent renal allograft biopsy due to clinical suspicion of graft dysfunction. This study was approved by the Ethics Committee of Shahid Beheshti University of Medical Sciences. We retrieved demographic and clinical data, including age, sex, time interval between transplant and scintigraphy, and time interval between scintigraphy and biopsy, from the hospital database. Patients with suboptimal biopsy samples or poor bolus injection during scintigraphy were excluded.

Image acquisition
Imaging was performed after intravenous bolus injection of 259 to 370 MBq (7-10 mCi) of Tc-99m-DTPA with the patient in the supine position and the gamma camera positioned anteriorly over the pelvic region. Immediately after tracer injection, 10 mL of normal saline was administered intravenously. Images were acquired using a Siemens ECAM single-head gamma camera (Siemens Medical Solutions) equipped with a low-energy all-purpose collimator, a 128 × 128 matrix, a 20% energy window, and a 140-keV photopeak. Dynamic imaging was obtained at 1 frame per second for 1 minute (perfusion phase), followed by 15 seconds per frame for 40 minutes (clearance phase). In cases of suspected urinary obstruction, intravenous furosemide (0.5 mg/kg) was administered 20 minutes into the clearance phase. A single experienced technologist manually drew regions of interest over the graft kidney, background (semilunar region of interest inferolateral to the graft), and ipsilateral iliac artery. Perfusion and clearance TACs were generated, and quantitative parameters were derived.

Scintigraphy interpretation
All scintigraphy studies were independently reviewed by 2 nuclear medicine physicians who were blinded to biopsy results. Normal or near-normal perfusion with impaired function, increasing background activity, and a descending functional TAC were interpreted as ischemic complications (eg, acute tubular necrosis [ATN]). Impaired perfusion and function with a flat functional TAC were interpreted as nonischemic complications (eg, acute rejection or tubulointerstitial nephritis).

Quantitative perfusion parameters
We studied the following perfusion parameters: delta peak (DP), defined as the difference between time to graft peak perfusion and time to iliac artery peak perfusion, peak perfusion activity (P), plateau perfusion activity (PL), and P/PL ratio.

Quantitative function parameters
We studied the following function parameters: peak function activity (U), P/U ratio, and 20/3 ratio (activity at 20 minutes divided by activity at 3 minutes).

Allograft biopsy
Final diagnoses were based on histopathological examination by 2 experienced pathologists using the Banff 2015 classification. Patients were classified into 3 groups: nonischemic complications (acute rejection and tubulointerstitial nephritis), ischemic complications (ATN), and normal.

Statistical analyses
Quantitative parameters were compared between ischemic and nonischemic groups using the Mann-Whitney U test. We used receiver operating characteristic (ROC) curve analysis to determine diagnostic accuracy and optimal cutoff values. We calculated sensitivity, specificity, and likelihood ratios. We used SPSS version 20.0 for statistical analyses.

Results
We evaluated 63 patients (median age of 48 years; range, 15-80 years; male-to-female ratio of 50:13). The median interval between scintigraphy and biopsy was 3 days (range, 0-18 days). Most scintigraphy studies (94%) were performed more than 3 months after transplant. Of the 63 biopsies, 59 (93.65%) showed abnormal findings, with 5 ischemic (7.94%) and 54 nonischemic (85.71%) complications. Scintigraphy identified 56 nonischemic (88.89%), 4 ischemic (6.35%), and 3 normal (4.76%) cases (Table 1).

Ischemic group
The ischemic group included 5 patients (median age of 42 years; range, 15-57 years; male-to-female ratio of 4:1). All scans were performed within 30 days of transplant and within 6 days of biopsy. Scintigraphy indicated ATN in 4 cases and nonischemic complications in 1 case. Median (range) values for DP, P/PL ratio, P/U ratio, and 20/3 were 5 (0-10), 1.57 (1-3), 1.50 (1-3), and 0.73 (0.30-1), respectively (Figure 1).

Nonischemic group
The nonischemic group included 54 patients (median age of 48 years; range, 15-80 years; male-to-female ratio of 42:12). All scans were performed within 6 months of transplant. Median (range) values for DP, P/PL ratio, P/U ratio, and 20/3 were 10 (1-30), 2 (0.25-1.75), 1 (0.70-1.75), and 0.71 (0.29-1), respectively (Figure 2). Median DP and P/U differed significantly between the 2 groups (P = .⁰⁴³ and P = .⁰¹⁵, respectively) (Table 2). The ROC curve analysis demonstrated that DP had good accuracy for identifying nonischemic complications (area under the curve = 0.772), whereas P/U showed strong accuracy for identifying ischemic complications (area under the curve = 0.822) (Table 3 and Figure 3). A DP cutoff of 6.5 seconds yielded 63% sensitivity and 80% specificity for nonischemic complications. A P/U cutoff of 1.31 yielded 80% sensitivity and 82% specificity for ischemic complications (Figure 4).

Discussion
This study evaluated the diagnostic accuracy of quantitative renal scintigraphy parameters using histopathology as the reference standard. Among the 4 quantitative indices assessed (ΔP, P/PL, P/U, and 20/3), 2 parameters (ΔP and P/U) showed significant differences between ischemic and nonischemic complications. Both ΔP and P/PL were higher, whereas P/U was lower, in the nonischemic group compared with the ischemic group. Although the 20/3 index did not differ significantly between the 2 groups, the 20/3 was slightly higher in patients with nonischemic complications. Similar findings were reported by Yazici and colleagues, who demonstrated that the 20/3 ratio and half-time of the peak perfusion had good sensitivity for differentiating ATN and acute rejection from normal grafts; however, these indices were not statistically significant in distinguishing ATN from acute rejection.⁶ In ROC curve analysis, ΔP showed good diagnostic accuracy for identifying nonischemic transplant complications, whereas P/U demonstrated good accuracy for detecting ischemic complications. These findings are consistent with the study by Gupta and colleagues, which reported that the hilum index and ΔP were more accurate than other parameters for evaluating renal transplant complications.⁷ In contrast, Aktas and colleagues and Jackson and colleagues reported higher sensitivity and specificity for the P/PL index in diagnosing acute rejection.^8,9 These discrepancies may be related to differences in patient populations, timing of imaging, and reference standards used across studies. In the present study, a ΔP cutoff value of 6.5 seconds showed good specificity and acceptable sensitivity for diagnosing nonischemic complications, whereas a P/U cutoff value of 1.31 provided good sensitivity and specificity for identifying ischemic complications. Gupta and colleagues proposed a similar ΔP cutoff value of 6.7 seconds for differentiating ischemic from nonischemic complications.¹⁰ Our findings are also consistent with those of Yazici and colleagues, who reported that a ΔP cutoff of 5 seconds predicted elevated serum creatinine at 1 year with a sensitivity of 59.3% and specificity of 74.4%.¹¹ Histopathologic diagnosis of acute rejection is based on cellular infiltration of the renal interstitium by T lymphocytes, involvement of tubular epithelial cells (tubulitis), and, in more advanced stages, penetration of the vascular endothelium (intimal arteritis).¹² In early stages (Banff grades I and IIA), interstitial infiltration and tubulitis predominate, whereas intimal arteritis appears later in the disease course (Banff grades IIB and III). Consequently, early acute rejection is expected to cause functional impairment without marked perfusion abnormalities on Tc-99m-DTPA scintigraphy, whereas both perfusion and function are affected in advanced stages. In the present study, most patients with acute rejection exhibited impairment in both perfusion and function phases, which may reflect referral bias, as only patients with strong clinical suspicion of rejection were evaluated. Among the 63 patients studied, scintigraphy and histopathologic findings were discordant in only 2 cases. One patient had a normal biopsy with scintigraphy findings suggestive of nonischemic complications, and another had biopsy-proven ATN with scintigraphy features of nonischemic pathology. Possible explanations for this discordance include sampling error, heterogeneous pathologic involvement, and inadequate tissue sampling from different graft regions.¹⁰ Acute tubular necrosis is the most common cause of delayed graft function in the early posttransplant period and occurs more frequently in deceased donor grafts than in living donor grafts. Acute tubular necrosis results from tubular epithelial cell necrosis and sloughing into the tubular lumen, leading to tubular obstruction.¹³ The perfusion phase of Tc-99m-DTPA renal scintigraphy has generally been described as relatively preserved in ATN.^13,14 However, in the present study, 1 patient with biopsy-proven ATN demonstrated impaired perfusion and was therefore not categorized as having ischemic complications on scintigraphy. Similar findings were reported by Gupta and colleagues, who observed abnormal perfusion in some patients with histologically confirmed ATN. The investigators attributed this phenomenon to postoperative inflammation and edema, which may cause a compartment effect and compromise graft perfusion.¹⁰ This study had several limitations. First, part of the study period coincided with the COVID-19 pandemic, during which the number of renal transplantations in our center, one of the main transplant centers in the country, was reduced. This contributed to the small sample size, particularly in the ischemic and normal groups. Second, patients were selected based on high clinical suspicion of graft dysfunction, which likely explains the low proportion of normal cases. Third, histopathology was used as the sole reference standard; although biopsy is a strong diagnostic tool, combining histologic findings with clinical follow-up might provide more accurate classification. Another limitation was the variability in the interval between transplantation and scintigraphy, which resulted from the retrospective design and the absence of routine baseline scans in our institutional protocol. The main strength of our study lies in the direct comparison of scintigraphy findings with histopathology obtained within a short time interval. Notably, approximately 80% of biopsies were performed within 6 days of scintigraphy, minimizing the likelihood of temporal changes affecting the correlation between imaging and pathology.

Conclusions
Quantitative Tc-99m-DTPA renal scintigraphy parameters, particularly ΔP and P/U ratio, are useful for differentiating ischemic from nonischemic renal transplant complications. These indices correlate well with tissue diagnosis and may facilitate early and noninvasive assessment of graft dysfunction.