Fatima N
Karachi Institute of Radiotherapy and Nuclear Medicine (KIRAN), Karachi, Pakistan

Zaman Mu
Karachi Institute of Heart Disease (KIHD), Karachi, Pakistan

Kamal S
Atomic Energy Medical Centre, JPMC, Karachi, Pakistan

Hameed A
Karachi Institute of Radiotherapy and Nuclear Medicine (KIRAN), Karachi, Pakistan

Correspondence:

Maseeh uz Zaman, M.D.
Department: Karachi Institute of Heart Disease (KIHD)
Address: Karachi Institute of Heart Disease (KIHD), Karachi, Pakistan
City: Karachi
Country: Pakistan

Background and Aims:

Cisplatin is an effective agent for the treatment of solid tumors and nephrotoxicity isits serious side effect. This nephrotoxicity is dose dependent and results in debilitating renal failure. The aim of this study was to determine the early nephrotoxicity by serial estimation of glomerular filtration rate (GFR) using plasma two sample clearance method (PSC2) as the reference method and compared with other methods.

Methods:

Thirty three patients (26 males and 7 females; mean age 45 ±14 yrs) with different solid tumors whowere scheduled to receive cisplatin (mean dose of 114 ±23 mg/m2) in each cycle. GFR was determined simultaneously by 6 methods; (1) 24 Hours Creatinine Clearance (24HCrCl); (2) Gamma camera uptake method after 99mTc-diethylene triamine pentaacetic acid (DTPA) injection (Gates method); (3) Predicted creatinine clearance by Modification of Diet and Renal Diseases (MDRD); (4) Cockcroft-Gault's equation (CG's); (5) PSC 2; (6) Plasma one Sample Clearance method after 99mTc-DTPA injection (PSC 1). PSC 2 was chosen as reference method.

Results:

Three out of thirty three patients developed severe acute nephrotoxicity (GFR <50 ml/min). In remaining 30 patients, average fall in GFR after 6th cycle of cisplatin was 43.86 ± 16.10 ml/min/1.73 m2 (p <0.001) estimated by PSC2 which was more significant after 3rd cycle. Correlation coefficient (r) for PSC1, Gates, 24HCrCl, CG's and MDRD methods were 0.95, 0.91, 0.51, 0.59 and 0.67 respectively.

Conclusions:

GFR is the most sensitive indicator of early cisplatin induced nephrotoxicity and PSC1 andGates method are reliable substitute of PSC2 method. CG's, MDRD and 24 HCrCl methods are not reliable for detection of early nephrotoxicity. It is recommended that serial GFR should be estimated before and after every cisplatin dose, if not possible, than at least after the third cycle for precise detection of early nephrotoxicity.

Keywords: Cisplatin, Nephrotoxicity, GFR, PSC2, Gates, PSC1, MDRD, CG's

Cisplatin, a heavy metal complex, remains a major antineoplastic drug for the treatment of solid tumors. Full therapeutic potential of cisplatin is limited by long-lasting and potentially debilitating toxicity, most notable and principal target organ for cisplatin toxicity is kidney, which evolves slowly and predictably after initial and repeated exposure (1, 2). This toxicity is manifested by reduced renal function and leads to serum-electrolyte and pathological changes in the urine analysis. Doses of cisplatin whichproduce changes in renal function may cause no histopathological changes. Higher doses of the drug lead to interstitial nephritis (3-5). This drawback is of importance due to its high incidence and its impact on patient morbidity and mortality.

Thus, the assessment of renal function of in patients being treated with cisplatin is necessary at the early stage to avoid permanent renal damage (1-6). Glomerular filtration rate (GFR) is considered to be a representative parameter for evaluating the functional status of the kidneys. Measurement of GFR is based on the renal clearance of a marker in plasma, expressed as the volume of plasma completely cleared of the marker per unit time. Markers used to measure GFR include endogenous (creatinine, urea) or exogenous (inulin, iothalamate) substances (7-9). Inulin clearance is proved as the gold standard for GFR estimation. However, this method is not performed routinely in clinical practice, because of technical complexity and limited availability. The intrinsic creatinine clearance (CrCl) has been widely performed as only alternative to inulin clearance in routine practice. This method, however, is not as accurate when compared to the inulin clearance.

Therefore, simple and accurate determination of the GFR is still a clinical challenge (8, 10, 11). Numerous formulae have been developed to estimate GFR or creatinine clearance from serum creatinine and other variables. In 99mTc- DTPA (Diethylene Triamine Pentaacetic Acid) renography, GFR is calculated without blood or urine sampling. This method (Gates Method) has been the most common in the routine setting, although the diagnostic accuracy of the gamma camera methods is debatable. Several formulae have been developed to estimate either CrCl or 51Cr-EDTA (Ethylene Diamine  Triacetic Acid) measurements of GFR including the Cockroft-Gault (CG), Jelliffe and Wright formulae. Plasma sample method following a single-injection of radioactive or non-radioactive markers has been proved effective as an alternative to the continuous infusion method with inulin for the determination of GFR in clinical practice (12-14).

The aim of this study was to detect early cisplatin induced nephrotoxicity by serial estimation of GFR and comparison of various methods of GFR estimation.

The study was conducted after informed consent from patients was taken and approval by the Hospital Ethics Committee at the Department of Nuclear Medicine, Karachi Institute of Radiotherapy and Nuclear Medicine (KIRAN), Karachi. Thirty three subjects with various solid tumors (26 males and 7 females, mean age 45 ±14 yrs; median age 45 yrs), were enrolled in this study prospectively after consideration of inclusion and exclusion criteria on referral for a GFR measurement by the outpatient department of the medical oncology unit.

Sample selection of the study was non- probability purposive and inclusion criteria of the study were planned cases for cisplatin therapy in higher dose ≥50 mg/m2 for various solid tumors, non- ypertensive/ non-diabetics and with adequate baseline renal status (GFR ≥70 ml/min/1.73m2) patients. Those who were taking other nephrotoxic drugs or low dose of cisplatin < 50 mg/m2, known cases of renal failure or GFR <70 ml/min/1.73m2, hypertensive, diabetics, or hemodynamically unstable patients or advanced age >70 years were excluded from the study. GFR of each subject was measured simultaneously by plasma two samples clearance method (PSC2), plasma one sample clearance method (PSC1), Gates method (gamma camera based uptake method), predicted creatinine clearance by modified diet in renal diseases (MDRD) method and Cockcroft-Gault (CG) method at baseline (two days prior to start of cisplatin) and within 3 days after the completion of every successive cycle (total 12 estimations).

Plasma Sample Clearance and Gates methods

99mTc-DTPA was prepared by following the manufacturer's instructions; aseptically two 5-mCi (185 MBq) aliquots using a 3-cc syringe with a 22-gauge needle were drawn. One aliquot was set aside as the standard and the remainder were used for patient dose. The standard and the dose were calibrated carefully in a way that the percent difference between standard and dose should exceed 5% not under any circumstances. The patient was injected under the gamma camera (Nucline Spirit DH-V, MEDISO single-Head Digital Gamma Camera) for a dynamic renal study. The empty and full syringes counts were recorded by well counter and also under gamma camera. GFR was automatically calculated by the Gates method in ml/min/1.73 m2.

In two samples method, the blood samples was drawn at 60 and 180 minutes, and in one sample method single sample was drawn at 180 minutes after radiotracer injection from the contralateral arm in a collection bottle containing EDTA, mixed well and samples centrifuged for 10 minutes. 0.1 ml of filtrate was pipette out by using micropipette into labeled test tube. The standard and sample test tubes were counted in a gamma scintillation counter set for 140 KeV 99mTc photopeak with a 20% window, background correction was done and sample was counted for one minute to ensure good counting statistics. PSC2 and PSC1 methods were automatically calculated by Biodex Medical System Atomlab 950 version 3.08 by Russell method (15, 16).

In this method, only the blood sample of the patients was required for serum creatinine, serum albumin and BUN and by putting the age and sex factors in the following formula GFR was calculated in ml/min/1.73m²:

GFR (ml/min/1.73m²)= 170 × (Scr)^-0.999 × (Age) ^-0.175 × (BUN)^-0.130× (Alb) ^0.318 × (0.762 if female) × (1.180 if black)

Where Scr= Serum Creatinine,

BUN= Blood Urea Nitrogen

Alb= Serum Albumin

No specific precaution was required for this test. All patients were asked to collect their urine in a wide mouth bottle for a period of 24 hours and fresh blood sample was drawn for serum creatinine on deposition of urine sample.

Creatinine clearance was calculated by the following formula:

C =  TV * UC / 440 × SCr

Where C = Creatinine clearance

TV =Total Volume of Urine

UC = Urinary creatinine

SCr = Serum creatinine

Cockcroft-Gault's method (CG's)

By putting the serum creatinine, age, height and sex factors in the following formula the GFR in ml/min was calculated:

GFR (ml/min) = ((140- Age) ×Weight / 72*Scr)

GFR (ml/min) values as obtained by the 6 methods were normalized for the body surface area (BSA) of 1.73 m2 in order to interpret the result and compare it with the reference range. Values of BSA were estimated from patient's height and weight using the following Haycock formula:

BSA (m²) = ^0.024265 × Wt ^0.5378 × Ht ^0.3964

Where Wt = Patient's body weight in Kilograms (Kg)

Ht = Patient's Height in centimeters (cm)

Data was entered and analyzed by using Microsoft Excel 2003 and, Statistical Programme for Social Sciences (SPSS) v.11.0 data base programme. Population distribution of sample was also normal. In order to observe statistical significance, patient's data including mean change in GFR from baseline, in response to each cisplatin cycle were compared using paired sample t-test. For the comparison of methods, standard linear least-squares regression analysis was used. A p-value of 0.001 or less was considered significant.

Mean dose of cisplatin administered to the study population was 114 ± 22mg/m2 per cycle. Three patients developed severe renal dysfunction (GFR less than 50 ml/min/1.7 m2), one after the 1st and two after the 4th cycle. PSC2 (reference method) estimated the average fall in GFR of 43.86 ± 16.10 ml/ min/1.73 m2 (p <0.001). While average fall in GFR was 45.88 ± 15.68, 45.69 ± 14.97, 47.82 ± 24.85, 45.29 ± 13.92 and 35.94 ± 15.55 ml/min/1.73m2 as estimated by PSC1, Gates, 24 HCrCl, MDRD and CG's method respectively (Table 1).

Significant decline in GFR was seen in each cycle of cisplatin even after the first dose by all methods but PSC2 method detected these changes with more precision particularly after the 3rd cycle as demonstrated by smaller standard deviation values as compared with other methods (Table 2 and 3).

A phenomenon of recovery (improvement in GFR from post-last cycle to pre-next cycle values) from the nephrotoxic effect of cisplatin was seen after every cycle but maximum recovery was noted after the 1st cycle. A progressive decline in recovery was seen in the remaining cycles (Figure 1).

Table 1. Fall in GFR from the baseline to the end of Cisplatin therapy by all methods

GFR, Glomerular Filtration Rate; SD, Standard Deviation.

*Significantly higher (p-value<0.001)

Table 2. Mean fall in GFR in each Cisplatin cycle estimated by all methods

PSC 2, Plasma two sample method; PSC 1, Plasma one sample method; 24 HCrCl, 24 Hours Creatinine Clearance; MDRD, Modification in Diet and Renal Disease; CG, Cockcroft-Gault's equation.

*Significantly higher (p-value<0.001)

Table 3. Mean, SD and CID values of PSC 2 sample method in each CDDP cycles

SD, Standard Deviation; CID, Confidence interval of Difference.

Figure1. GFR values by PSC 2 method in each pre and post-Cisplatin cycles

The regression equation of the PSC 1 and the Gates against the PSC 2 was Y = 11.562 +0.905X (r = 0.958, p =0.000, SEE = 2.6 ml/min/1.73 m2) and Y =8.766+ 0.935X (r = 0.91, p= 0.000, SEE= 2.8ml/min/1.73 m2), respectively (Figure 2 and 3). No-significant correlation was seen for 24 HCrCl, CG's and MDRD methods (r value 0.511, 0.595 and 0.675 respectively, p value=0.001).

GFR was overestimated by 24 HCrCl by 32% (-27.38 ml/min/1.73m2, p value=0.000), PSC 1 method by 4% (-3.38 ml/min/1.73m2, p value nonsignificant) and Gates method by 4% (-3.44 ml/min/1.73m2, p value non-significant). But GFR was underestimated by CG's method by 14% (11.84 ml/min/1.73m2, p value=0.000) and Predicted Creatinine Clearance by MDRD method by 10% (8.29 ml/min/1.73m2, p value=0.000) of PSC 2 method (Figure 4).

Figure 2. Scatter plot of Plasma one Sample Clearance method (PSC 1) against Plasma two Sample Clearance method (PSC 2) and the solid line indicates the regression line

Figure 3. Scatter plot of Gates method of GFR estimation against Plasma two Sample Clearance method (PSC 2) and the solid line indicates the regression line

Figure 4. The average difference of GFR between Plasma two Sample Clearance method and other methods

In our study, 3/33 patients (9%) developed severe renal dysfunction (GFR < 50 ml/min/1.73 m2) in spite of standard hydration protocol prior to the chemotherapy. This incidence is much less than reported 25% after 50-100 mg/m2 dose which is reversible in majority of cases (17, 18). This may be explained by the biased study population due to strict exclusion criteria, i.e. no co-morbids like diabetes, hypertension or advanced age, etc. It was noted after the first cycle in one patient, most probably represents an idiosyncratic effect while exaggerated cumulative dose dependant toxicity might be the plausible explanation in remaining two patients (19).

Average decline in GFR at the completion of 6 cycles in remaining 30 patients was 43.86 ± 16.10 ml/min/1.73 m2 which represents an expected nephrotoxicity by the average cumulative dose of 114± 22.32 mg/m2 per cycle for six cycles (17). Reported incidence of long-term cisplatin induced persistent decline in GFR is up to 30%, however, we could not follow these patients which is a limitation in this study (20). Significant decline in mean GFR has been noted after every cycle of cisplatin with better precision (small SD values) for all methods after 2nd cycle. Similarly, significant improvement in the GFR has also been noted between post-previous and pre-next cycle GFR (Table 2, 3 and Figure 1). This indicates an attempt of kidneys to recover from acute insult, which starts within 3 hours (tubular dysfunction) and after 72 hours (decline in GFR) of cisplatin dose (18). But the magnitude of recovery shows a progressive decline from the second cycle which is explained by the partial phenomenon of recovery in response to escalating cumulative dose.

GFR estimated by PSC1 and Gates methods show good correlation with PSC2 method. The reason for this good correlation is that in our study population, serum creatinine did rise but did not exceed 4 mg/dl as beyond this limit, the accuracy becomes questionable (21). PSC1 overestimated the GFR by 4% and is due to protein binding of DTPA which increases with time, from 3% at 1 hour to 6% at 3 hour after injection (16). Gates method overestimated GFR by 4% and this may be attributed to the insufficient correction of background counts in the kidneys (21). In this study, non-significant correlation of 24 HCrCl, CG's and MDRD methods with PSC2 signifies that these methods are not true substitute for the reference method of GFR estimation as has been reported by our group (14). The CG's method underestimated GFR by 14% and MDRD method by 10% with reference to PSC2 method. Lin and colleagues described the sources of errors in CG's method when compared with MDRD equations due to inaccuracies of formulae caused by intra-assay and intra-individual SCr and GFR variability, lack of calibration of SCr estimation across laboratories, and measurement of other variables in predicted equations which can affect the precision and accuracy of renal clearance prediction equations (21). 24 HCrCl overestimated the GFR by 32% which may be due to inaccuracies in 24 hr collection of urine by patients (22-24).

However, significant correlation of the renal uptake of 99mTc-DTPA against the 24-hour creatinine clearance has promoted this method for clinical application in routine practice (12). We conclude that GFR is the most sensitive indicator of early cisplatin induced nephrotoxicity. PSC1 and Gates method are reliable substitutes of PSC2 method. CG's, MDRD and 24 HCrCl methods are not reliable for detection of early nephrotoxicity. It is recommended that serial GFR should be estimated before and after every cisplatin dose, if not possible, than at least after the third cycle for precise detection of early nephrotoxicity.

Conflict of Interest

None declared.

1. Dorner GT, Pehamberger H, Kornek G, et al. Cisplatininduced renal effects and thromboxane A2 receptor blockade.Eur J Clin Invest. 1997;27:836-9.

2. Dillman RO, Seagren SL, Propert KJ, et al. A randomized trial of induction chemotherapy plus high-dose radiation versus radiation alone in stage III non-small-cell lung cancer. N Engl J Med. 1990;323:940-5.

3. Arany I, Safirstein RL. Cisplatin nephrotoxicity. Semin Nephrol. 2003;23:460-4.

4. Miyaji T, Kato A, Yasuda H, Fujigaki Y, Hishida A. Role of the increase in p21 in cisplatin-induced acute renal failure in rats. J Am Soc Nephrol. 2001;12:900-8.

5. Schrier RW. Cancer therapy and renal injury. J Clin Invest. 2002;110:743-5.

6. Elliott WC, Newcom SR, Houghton DC, Baines-Hunter J, Bennett WM. cis-diamminedichloroplatinum(II) nephrotoxicity: tubular function after rescue with sodium diethyldithiocarbamate in rats. Cancer Res. 1983;43:3759-62.

7. Levey AS. Measurement of renal function in chronic renal disease. Kidney Int. 1990;38:167-84.

8. Itoh K. Determination of glomerular filtration rate by means of newly developed plastic scintillation counter both with and without dilution procedures. J Nucl Med. 2001;42:1484-8.

9. Poggio ED, Wang X, Greene T, Van Lente F, Hall PM. Performance of the modification of diet in renal disease and Cockcroft-Gault equations in the estimation of GFR in health and in chronic kidney disease. J Am Soc Nephrol. 2005;16:459-66.

10. Smith H. The kidney structure and function in health and disease. New York: Oxford University Press,1951:231-8.

11. Itoh K. Comparison of methods for determination of glomerular filtration rate: Tc-99m-DTPA renography, predicted creatinine clearance method and plasma sample method. Ann Nucl Med. 2003;17:561-5.

12. Laterza OF, Price CP, Scott MG. Cystatin C: an improved estimator of glomerular filtration rate? Clin Chem.2002;48:699-707.

13. Itoh K, Tsushima S, Tsukamoto E, Tamaki N. Accuracy of plasma sample methods for determination of glomerular filtration rate with 99mTc-DTPA. Ann Nucl Med. 2002;16:39-44.

14. Fatima N, Zaman M, Niaz K, et al. Estimation of GFR using Tc-99m DTPA plasma sample 1, Gate's methods, Cockcroft-Gault method, PCC method: A prospective comparative analysis with plasma 2 sample clearance method. JDUHS. 2008;2:65-70.

15. Russell CD, Bischoff PG, Kontzen FN, et al. Measurement of glomerular filtration rate: single injection plasma clearance method without urine collection. J Nucl Med. 1985;26:1243-7.

16. Biggi A, Viglietti A, Farinelli MC, Bonada C, Camuzzini G. Estimation of glomerular filtration rate using chromium- 51 ethylene diamine tetra-acetic acid and technetium-99m diethylene triamine penta-acetic acid. Eur J Nucl Med. 1995;22:532-6.

17. Gonzalez R, Borrego A, Zamora Z, et al. Reversion by ozone treatment of acute nephrotoxicity induced by cisplatin in rats. Mediators Inflamm. 2004;13:307-12.

18. Yao X, Panichpisal K, Kurtzman N, Nugent K. Cisplatin nephrotoxicity: a review. Am J Med Sci. 2007;334:115-24.

19. Begg AC, van der Kolk PJ, Dewit L, Bartelink H. Radiosensitization by cisplatin of RIF1 tumour cells in vitro. Int J Radiat Biol Relat Stud Phys Chem Med. 1986;50:871-84.

20. Brigden ML. The monitoring of potential long-term complications in treated adult cancer patients. Ann Saudi Med. 1997;17:622-8.

21. Mittal BR, Kher V, Gulati S, et al. Measurement of GFR using 99m Tc DTPA-A comparative analysis of gates and plasma clearance technique. Ind J Nephrol. 1996;6:4-8.

22. Lin J, Knight EL, Hogan ML, Singh AK. A comparison of prediction equations for estimating glomerular filtration rate in adults without kidney disease. J Am Soc Nephrol. 2003;14:2573-80.