GFR Estimation and Correlation for Oncology Patients by Two Methods, Gates Method and Dual Time Point Plasma Sampling Method

Introduction

With the increasing number of oncology cases per year, there has been a parallel surge in the usage of chemotherapeutic drugs too, for its treatment. To detect nephrotoxicity and provide a personalized dosing strategy, routine evaluation of kidney functioning is assessed by the treating physicians. There are in general, two pathways involved for excretion of the drug: glomerular filtration and tubular secretion. Tubular secretion can be performed but is laborious and of unclear clinical significance.[1] Hence, glomerular filtration rate (GFR) assessment is preferred. GFR is defined as a rate of filtration of fluids by the kidneys and a method for evaluating kidney function.[2] The gold standard method for GFR estimation, however, is through constant intravenous infusion of inulin and bladder catheterization.[3] Being an invasive process and difficult to perform, leads to its limited usage in clinical practice.[4] Second, the inulin process is time-consuming, expansive, and poorly available posing a challenge for routine usage. In addition to this method, 24 h endogenous creatinine clearance method can also be adopted but reliability is too dependent on accurate urine collection.[567891011] Of the various tests available, GFR under gamma camera with the help of the Gates’ method has gained importance, as it gives a good index of overall kidney functions.[1213] Second, it is one of the best measures of functioning kidney mass.[1] It was Gary Gates, who first evaluated GFR from radiopharmaceutical uptake study in the kidneys. Since then, GFR measurement with a camera-based method using ^99mTc diethylenetriaminepentaacetate (DTPA) has been quite an economical and easy method for renal function evaluation.[314] Second, it can help determine unilateral renal blood flow, that too immediately. Another alternate and old method for GFR estimation was multisample technique, wherein multiple blood samples were collected (plasma method) after regular intervals postinjection of radiotracer for generating time activity curve to help evaluate GFR using Russel’s formula. With the passage of time and to achieve more functionality and making it less exhaustive, single plasma and double plasma methods were derived from multisample technique.[215] We at our institution carry out both methods (Gates’ method and dual plasma sample method) for better evaluation of GFR in cancer patient management for coming chemotherapy cycle, especially in borderline cases (GFR <65 mL/min).

Methods

This is a retrospective study wherein patients’ databases were evaluated from June 2022 to September 2022 at Mahamana Pandit Madan Mohan Malaviya Cancer Centre, Varanasi. Thirty patients were included in the study as per the following inclusion:

• Age 18 and above

• Serum creatinine report of not more than 2 weeks from the date of study

• Patient not pregnant

• Having both dual-time plasma sampling and Gates’ camera-based method data.

The study review process was approved by the Institutional Review Board of Mahamana Pandit Madan Moan Malaviya Cancer Centre, Varanasi (OIEC/11000694/2023/00001, dated November 2, 2023).

Glomerular filtration rate measurement under gamma camera

The radiopharmaceutical, ^99mTc-DTPA used in routine clinical practice for conducting the study under gamma camera was prepared in-house by using freshly eluted ^99mTc supplied by SDS life sciences ⁹⁹Mo-^99mTc generator and cold kit of DTPA, procured from Board of Radiation and Isotope Technology, Mumbai, India. The acquisition of images was done as per the standard operating procedure guideline. Half an hour before the study, each patient was encouraged to drink at least 300 mL of water and each patient’s height, weight, age, and sex were recorded in the system. First, presyringe/full syringe count of ^99mTc DTPA activity of 185MBq, measured with the help of calibrated dose calibrator-Capintec INC, CRC 25tW, used for the study was taken. It was followed by injecting intravenously the same radiopharmaceutical (which had a labeling yield of over 97%) as a bolus under a gamma camera. A total of 7-min dynamic study was acquired in a 64 × 64 matrix in two phases using GE’s NMCT/670 dual head gamma camera with a low-energy high-resolution collimator (LEHR) and the patient positioned in a supine position with arms elevated over the head. First phase was 2 s per frame for 1-min duration followed by 15 s per frame for the rest of the time. This sequential imaging following the passage of radiotracer through the kidney constitutes the renogram. The syringe with the remaining activity used in the clinical study was counted as postinjection activity after renogram acquisition. The pre- and postsyringe counts under gamma camera were taken for 1 min each on a 128 × 128 matrix, which helped in calculating the net activity injected in the body. For the processing of GFR study acquired under a gamma camera, regions of interest (ROI) were drawn around the kidney and a background at the inferior pole of the kidney avoiding any other tissue to obtain time activity curves. The background area was drawn since the kidney ROI contains the contribution from tissues over and underlying the kidney, which should be removed before the estimation of GFR [Figure 1a]. Based on the 2–3-min period postinjection images, GFR was calculated. All necessary correction: background activity, linear attenuation, and kidney depth correction based on the height and weight of the patient using Tonnesen’s formula was applied as shown in Table 1.

Close

Figure 1 (a) Region of interest image around left and right kidney and its background region, (b) Computed tomography image used for depth evaluation for left and right kidney

Export to PPT

Table 1 Tonnesen’s formula and its parameters

Parameters	Calculation
Right kidney depth (cm)	13.3 (weight/height) +0.7
Left kidney depth (cm)	13.2 (weight/height) +0.7
Corrected counts of the right kidney	Right kidney counts-background/e-µ xR
Corrected counts of the left kidney	Left kidney counts-background/e-µ xL

Height measured in cm, weight measured in kg. µ: Linear attenuation coefficient: 0.153/cm for ^{99 m}Tc, xR: Depth of right kidney, xL: Depth of left kidney

The accuracy of camera-based method is dependent on the estimated kidney depth as is evident in the above-mentioned equation. It was also calculated manually with the help of previous CT images available in the picture archiving and communication system of the institution. Axial views of the previously acquired CT images were chosen and renal depth was defined as the distance from the middle point of anteroposterior diameter to the body surface on the back in each view [Figure 1b].

Plasma sampling method

A standard solution was prepared with the same amount of activity which was used for clinical study and a presyringe count was acquired for 1 min under gamma camera. After this, the activity was emptied in a conical flask, and postsyringe counts were acquired at the same time under the gamma camera. Venous blood samples (4 mL) were collected from the contralateral arm at 120 and 240-min postinjection time. Plasma counting was done under a calibrated well counter (Capintec INC, CRC 25tW) the next day after it was separated from the collected blood by separation through centrifuging at 3000 g for 15 min. About 1 mL of plasma from each sample was taken out with the help of pipette, taking care to avoid disturbing the interface between plasma and the red cells. An equal volume of standard was counted in a calibrated well counter with necessary background corrections. Counting of all the samples was done after selection of proper energy peak and window of ^99mTc. Quality control was performed on all the equipment used in the study (dose calibrator and well counter). Finally, Russell’s formula, as shown in Table 2, was used for GFR estimation.

Table 2 Statistical analysis was done to help evaluate a correlation coefficient between the methods (Gates’ method with and without computed tomography-based manual depth correction and dual-time point plasma sampling)

D: Dose measured under gamma camera (preinjection counts), R: Residue dose after injection (postinjection counts), ADS: Activity of 1 mL dilution of standard in 1000 mL, DS: Dilution of standard (1000 mL), AS: Activity of standard sample measured under gamma camera, D: Administered activity (counts per min), P₁: Activity at T₁ (counts/min/mL), P₂: Activity at T₂ (counts/min/mL), T₁=120 min, T₂=240 min, GFR: Glomerular filtration rate

Statistical analysis was done to help evaluate a correlation coefficient between the methods (Gates’ method with and without CT-based manual depth correction and dual-time point plasma sampling).

Results

A total of 30 patients’ data were analyzed, consisting of 20 (66.67%) males and 10 (33.33%) females. Among these, fifteen had carcinoma of the head-and-neck region, three with carcinoma of the cervix, two with carcinoma of esophageal, one with renal cell carcinoma, one with gallbladder carcinoma, one with carcinoma endometrium, one carcinoma anal canal, one carcinoma penis, one acute myeloid leukemia, one with metastasis of unknown origin, one myeloma, one urothelial carcinoma, and one with urinary bladder carcinoma. The mean age was 53.97 ± 10.79 years (range 33–73 years). The mean height was 159.2 ± 9.72 cm and the mean weight was 54.93 ± 12.44 kg. The mean GFR calculated by Gates’ method and dual-time point plasma sampling method (DPSM) were 81.69 ± 24.76 and 66.15 ± 23.66, respectively. The representation of the demographics is shown in Table 3. GFR calculated by Gates’ method and dual-time point plasma sampling are shown with the help of a scatter plot and regression line [Figure 2]. Based on the above plots, the correlation coefficient of Gates’ method GFR and plasma sampling GFR was 0.66, which can be interpreted as a moderate correlation and was statistically significant with a P < 0.0001. The Bland–Altman plot [Figure 3] showed good agreement between GFRs calculated by Gates’ method and plasma sampling (P < 0.0001) as most of the differences between GFR calculated by Gates’ method and plasma sampling fall within ±1.95 standard deviation. The GFR calculated by dual-time point plasma sampling did not fall between ±10% of GFR calculated by Gates’ method as shown in Figure 4 and the box-and-whiskers plot [Figure 5] showed that the median GFR calculated by Gates’ method was more than that of the GFR calculated by plasma sampling. However, when we compare GFR calculated by Gates’ method and CT-based manual depth correction, it was found to have a high positive relationship [r = 0.85, Figure 6a] but, there was a moderate positive relationship [r = 0.65, Figure 6b] between dual-time point plasma sampling and GFR with CT-based manual depth correction. There is a statistically significant difference between GFR calculated by Gates’ method and GFR with CT-based manually corrected depth method (P < 0.0001) as well as between plasma sampling and GFR with CT-based manually corrected depth method (P < 0.0001).

Close

Figure 2 Correlation between glomerular filtration rates calculated by Gates’ method and plasma sampling (r = 0.66; y = 36.062 + 0.689x). GFR: Glomerular filtration rate

Export to PPT

Close

Figure 3 Bland–Altman plot for glomerular filtration rates calculated by Gates’ method and plasma sampling (P < 0.0001). GFR: Glomerular filtration rate

Export to PPT

Close

Figure 4 Line plot showing the glomerular filtration rate (GFR) calculated by plasma sampling whether lies within ±10% GFR calculated Gates’ method. GFR: Glomerular filtration rate

Export to PPT

Close

Figure 5 Box-and-whiskers plot for comparing the glomerular filtration rates calculated by Gates’ method and plasma sampling. GFR: Glomerular filtration rate

Export to PPT

Close

Figure 6 (a) Correlation between Gates’ method glomerular filtration rate (GFR) and GFR-depth correction (r = 0.85; y = 20.745 + 0.969×), (b) Correlation between plasma sampling and GFR-depth correction (r = 0.65; y = 48.362 + 0.779×). GFR: Glomerular filtration rate

Export to PPT

Discussion

Accurate estimation of GFR helps the oncologist with more precise chemotherapeutic drug dosing. Therefore, it is imperative that the best possible method be chosen to calculate GFR. Indeed, Gates’ method is one of the most convenient and faster processes.[314] It has the added advantage of providing differential function while others provide global function only. However, many factors such as renal depth,[16] extravasation, height, weight, and hydration status can affect the outcome. Second, ^99mTc-DTPA is diffusible and rapidly enters the extravascular space after injection.[17] Hence, background corrections are a must before the final processing of GFR to avoid contribution from tissues over and underlying the kidney.[1819] In addition to this, renal depth is also an important factor. It becomes more significant when working with LEHR collimators as the full-width half maximum varies considerably with the distance of the source to collimator as compared to low-energy all-purpose.[17] It has been clearly observed that the renal depth calculation is being underestimated in the Gates’ method using Tonnessen’s formula and our results are in consensus to the previous study done by other authors.[17202122] Renal depth variation can cause GFR error.[2324] Ma et al. showed in their study that a 1 cm error in kidney depth leads to an 18% difference in adult’s GFR.[25] Hence, in our study, we did the calculation of GFR with manually measured renal depth on the previously acquired CT image, which helps in giving a better and near-accurate GFR with correct values of renal depth application in the Gates’ formula. An alternative to using CT images for depth correction can be acquiring a lateral view after completion of renal dynamic scan if feasible, which can reduce the radiation burden too.[26] On the other hand, plasma sampling stands out to be a better method for evaluation avoiding the above-mentioned limitation. However, this technique also has its own limitations: erroneous preparation of standard, incorrect sampling timing, error in pipetting, and unintentional extravascular injection of tracer.[27] Instead of multisample technique, the one adopted at our center of two-sample technique can be taken into consideration as has been shown by Kumar et al. in their study, wherein the two-sample method a derivation of the multisample method has shown to be of significant correlation. In another study on the Indian population by Hephzibah et al., a poor correlation[27] is seen between Gates’ method and the single/double plasma sampling method. Our study, however, showed a moderate correlation and this could be due to the challenges of overcoming all possible errors in routine practice while performing this method.

Conclusion

In our study, a direct comparison of GFR values was done between the three methods (automated kidney depth Gates’ method, manual depth correction in Gates’ method, two sample plasma sampling). Although inulin clearance is considered a gold standard, in GFR measurement, this methodology is time-consuming, expensive, and not easily available. The endogenous markers, used for assessing GFR are not ideal and occasionally do not perform well. In a busy department like ours, it becomes a challenge collecting samples at an accurate time and it confines one person dedicated to the task for the entire procedure. One can completely agree that quantification of GFR is a challenging activity as there are innumerable variables with Gates and DPSM, leading to less correlation among each. Clinicians need to understand the limitations of each method and see if the renal depth corrections can be done with the help of CT or lateral images of Nuclear Medicine (NM) for near-accurate GFR and in the case of selecting dual plasma sampling, errors to be minimized in pipetting and sample counting. Hence, it will be better to use both methods for coming to a conclusion. Manual depth correction is better when working with LEHR. Accurate estimation of GFR shall help in more precise dosing of chemotherapeutic drugs as well as detection of early nephrotoxicity. The GFR values acceptable at our institute are >65 mL/min/1.73 m².