...........................................................................................................................................................................................................................................................................................................................

Prediction of post-operative glomerular ®ltration rate after

nephrectomy for renal malignancy
Mats Johansson and Michaela Moonen
Department of Clinical Physiology, GoÈteborg University, Sahlgrenska University Hospital, GoÈteborg, Sweden
Received 18 January 2001; accepted 16 May 2001
Correspondence: Mats Johansson, Department of Clinical Physiology, Sahlgrenska University Hospital, SE 413 45 GoÈteborg, Sweden

...........................................................................................................................................................................................................................................................................................................................

of the tumour affected kidney is preserved, indicating

Summary
an adaptive GFR increase in these cases.
The importance of a correct estimation of contra-
lateral renal function in cases of renal malignancy is Keywords: 51Cr-EDTA clearance, 99mTc-DTPA,
obvious, necessitating a conservative approach to nephrectomy, prediction of GFR, renal malignancy,
tumour resection when function of the contralateral renography.
kidney is markedly reduced. The aim of the present
study was to determine the accuracy of preoperative
Introduction
gamma camera renography and 51Cr-EDTA clear-
ance to predict the glomerular ®ltration rate (GFR) Renal cell carcinomas represent 3% of all human
early and up to 6 months after nephrectomy for renal malignancies. The most effective treatment is radical
malignancy. Patients (n ˆ 40) underwent both gam- nephrectomy, which involves node dissection and
ma camera renography (99mTC-DTPA) and complete removal of the affected kidney and Gerota's
51
Cr-EDTA clearance preoperatively, whereas fascia. Given the risk of renal failure after radical
51
Cr-EDTA clearance was measured within 1 week nephrectomy when the demand for glomerular ®ltra-
and up to 6 months after nephrectomy. The single tion rate (GFR) of the contralateral kidney is
kidney GFR values of the contralateral kidneys were increased, it is important to correctly estimate split
estimated preoperatively and then compared with the renal function prior to surgery. While there are
post-operative 51Cr-EDTA clearance values. The several studies evaluating GFR after nephrectomy in
predicted GFR values were lower compared with healthy kidney donors, there are few such studies in
the measured post-operative 51Cr-EDTA clearance patients operated for renal malignancy (Aurell &
values (45 ‹ 2 vs. 54 ‹ 3 ml min±1 1 week after Ewald, 1981; Argiles et al., 1987; Pfaller et al., 1998).
nephrectomy and 53 ‹ 3 ml min±1 6 months later, Thus, the aim of the present study was to determine
P<0á01, respectively). The difference between the whether gamma camera renography and 51Cr-EDTA
measured and predicted GFR was larger in patients clearance are accurate methods for prediction of the
below the median age of 60 years (P<0á05) and post-operative GFR in patients undergoing nephrec-
con®ned to patients with a relative uptake of >30% tomy for renal cancer. Furthermore, we intended to
by the tumour affected kidney. Prediction of post- study the GFR change of the remaining kidney after
operative GFR by non-invasive renal function tests nephrectomy, and hence, we measured GFR during
performed prior to surgery for renal malignancy the ®rst post-operative week and up to 6 months after
underestimate post-operative GFR when the function surgery.

688 Ó 2001 Blackwell Science Ltd · Clinical Physiology 21, 6, 688±692

 M. Johansson & M. Moonen · Prediction of GFR after nephrectomy for malignancy
............................................................................................................................................................................................................................................................................................................................

taacetic acid (DTPA). Time-activity curves for the

Subjects and methods
regions of interest over the kidneys were created.
The study material comprised patients with renal Renograms were corrected for the extra-renal back-
adenocarcinoma, urothelial cancer or an adrenal ground after normalization for kidney area. Split
tumour who were treated by unilateral nephrectomy kidney function was calculated according to the
(Table 1). Before surgery patients underwent a integral method (Larsson et al., 1975; Moonen et al.,
clinical investigation including computerized tomo- 1994). For calculation of the predicted residual GFR
graphy of the kidneys and/or renal angiography after nephrectomy the percentage uptake by the
and chest X-ray. Only one patient had metastases contralateral kidney, calculated according to the
according to the preoperative examinations. integral method, was multiplied with the preoperative
51
The indication for performing the renal function Cr-EDTA clearance values. For comparison
tests was to estimate split kidney function before between the predicted and measured post-operative
nephrectomy. Patients underwent both gamma GFR measurements the 51Cr-EDTA clearance values
camera renography and 51Cr-EDTA clearance pre- were not normalized to body surface area.
operatively, whereas the 51Cr-EDTA clearance was All patients underwent radical unilateral nephrec-
measured within 1 week and up to 6 months after tomy. Four patients needed a re-operation during the
nephrectomy. The single kidney GFR values of the ®rst 24 h to control bleeding. One of these patients
contralateral kidneys were estimated preoperatively was excluded from analysis because of transient renal
by gamma camera renography and compared with the failure secondary to surgery. During the 6 months
post-operative 51Cr-EDTA clearances. The average follow up, two patients died, skeletal metastases were
time between the preoperative 51Cr-EDTA clearance diagnosed in one patient and in another patient
and renography examinations was 16 ‹ 10 days. metastases to lung and liver were found. These
51
Cr-EDTA clearance was calculated according to patients were included in the analysis of the early
the single injection technique by BroÈchner- post-operative GFR, whereas the 6 months GFR
Mortensen, assuming a 1-compartment model measurements were only available in the patient with
(BroÈchner-Mortensen, 1972). Since the assumption metastases to liver and lung. The other patients were
of a 1-compartment model may overestimate the true followed up to 6 months without signs of recurrent
plasma clearance of 51Cr-EDTA, an equation for malignancy. Hence, 51Cr-EDTA clearance measure-
correction presented by BroÈchner-Mortensen was ments were performed 6 ‹ 2 days after unilateral
applied. Renographic examinations were performed nephrectomy for renal malignancy in 32 patients
in the supine position with the back of the patient and then repeated 179 ‹ 13 days after nephrectomy
against a large ®eld gamma camera (APEX 415, in 27 patients.
Elscint, Israel) in order to visualize the kidneys and The local ethical and isotope committees at
the heart. Ninety-six frames (64 ´ 64 pixels) of 10 s Sahlgrenska University Hospital approved the studies
each were recorded after an intravenous bolus and all subjects gave their consent to participate in the
injection of 100 MBq 99mTc-diethylenetriaminepen- study.

Statistical methods
Table 1 The study population (n = 40).
Results are expressed as mean ‹ SEM values if not
Age (years) 66 ‹ 2
otherwise indicated. Student's t-tests for paired and
Gender (females/males) 17/23
Body mass index (kg m)2) 25 ‹ 1 unpaired observations were used. The values for the
Renal adenocarcinoma 33 relative uptake of 99mTc-DTPA by the kidneys were
Urothelial cancer 6 not normally distributed and hence, Wilcoxon signed
Adrenal tumour 1
rank test for paired comparisons were used. Calcula-
Relative uptake ± involved kidney (%) 40 ‹ 2*
ting the correlation coef®cient according to Pearson
Values represent numbers in each group or mean ‹ SEM. *P<0á01 vs. assessed the relation between two continuous varia-
the contralateral kidney. bles. Statistical signi®cance was de®ned as P<0á05.

Clinical Physiology 21, 6, 688±692 · Ó 2001 Blackwell Science Ltd 689

Prediction of GFR after nephrectomy for malignancy · M. Johansson & M. Moonen
...........................................................................................................................................................................................................................................................................................................................

Results Table 2 Pre-operative estimation of GFR after

nephrectomy, GFR 1 week and 6 months after surgery,
Kidneys affected by tumour showed a reduced uptake respectively.
of 99mTc-DTPA compared with the contralateral
Pre-op. 1 week 6 months
kidneys (P<0á01, Table 1). The preoperative GFR
measurements were 78 ‹ 3 ml min±1*1á73 m2. After GFR (ml min)1) 45 ‹ 2 54 ‹ 3* 53 ‹ 3*
nephrectomy, GFR declined by 30% (P<0á01) and (n = 40) (n = 32) (n = 27)
then remained stable during the 6 months follow up
(Table 2). There was a close relationship between the Values represent numbers in each group or mean ‹ SEM. *P<0á01 vs.
51 estimated GFR pre-operatively.
Cr-EDTA clearance measurements 1 week and
6 months after nephrectomy (Fig. 1). The predicted
GFR values, using renography combined with Table 3 Mean difference (ml min)1) between measured
51
Cr-EDTA clearance, were lower compared with GFR after nephrectomy and residual GFR predicted from
the non-invasive renal function tests prior to surgery.
the post-operative 51Cr-EDTA clearance value within
1 week and 6 months after surgery (Table 2). 1 week 6 months
Underestimation of post-operative GFR was con®ned
Relative uptake >30% by 10 ‹ 2* 11 ‹ 2*
to patients with a relative uptake of >30% by the
the removed kidney (n = 21) (n = 19)
affected kidney, whereas no difference between Relative uptake  30% by 3‹4 2‹2
predicted and measured GFR was found among the the removed kidney (n = 11) (n = 8)
other patients (Table 3 and Fig. 2). The difference
between the measured and predicted GFR was larger Values represent numbers in each group or mean ‹ SEM. *P<0á01 vs.
in younger patients (11 ‹ 2 ml min±1 versus zero.

4 ‹ 2 ml min±1 for patients below and above the

tumour resection and possible requirement of post-
median age of 60 years, respectively, P<0á05).
operative dialysis treatment when function of the
contralateral kidney is reduced. Although gamma
Discussion camera renography in combination with plasma
clearance of 51Cr-EDTA or 99mTC-DTPA are most
The importance of a correct estimation of contralat-
commonly used for estimation of split kidney func-
eral renal function in cases of renal malignancy is
tion prior to surgery, there are no studies that have
obvious, necessitating a conservative approach to
evaluated these methods in patients undergoing
nephrectomy for renal cancer.
The present study shows that gamma camera
renography and 51Cr-EDTA clearance performed
prior to nephrectomy for renal malignancy underesti-
mate the post-operative GFR of the remaining kidney.
This underestimation was con®ned to patients with a
relative uptake of more than 30% by the removed
kidney, indicating a compensatory increase in GFR of
the remaining kidney after nephrectomy in these cases.
Two studies, using probe renography and 131IHippu-
ran as the renal tracer combined with 51Cr-EDTA
clearance, reported higher GFR after nephrectomy
compared with the values predicted by the preopera-
tive examinations (Larsson et al., 1975; Tonnesen
et al., 1976). Tonnesen et al. reported a 42% higher
Figure 1 Graph showing the correlation between the
51
Cr-EDTA clearance values 1 week and 6 months after
GFR compared with the predicted values in patients
nephrectomy. There was a close relationship between those undergoing nephrectomy for renal cancer, obstruction
measurements. or renovascular hypertension in cases where the

690 Ó 2001 Blackwell Science Ltd · Clinical Physiology 21, 6, 688±692

 M. Johansson & M. Moonen · Prediction of GFR after nephrectomy for malignancy
............................................................................................................................................................................................................................................................................................................................

indicating an adaptive increase in GFR of the remain-

ing kidney only when the removed one has a signi®cant
uptake (» 30% relative uptake). This adaptive GFR
increase of the remaining kidney after nephrectomy
could explain the underestimation of post-operative
GFR by the non-invasive renal function tests prior to
surgery. On the other hand, when the removed kidney
is non-functioning, the stimulus for an adaptive
increase of the remaining kidney is absent.
Another explanation to the higher post-operative
GFR values compared with what was predicted prior
to surgery could have been overestimation of the
relative uptake by the tumour affected kidney. Using
gamma camera renography and 99mTc-DTPA as the
radiotracer, we have previously shown that tracer
accumulation within a highly vascularized renal
tumour may affect the calculation of split renal
function (Moonen et al., 1994). Using the integral
method for calculating the split renal function, the
intrarenal background activity is incorporated in the
recorded activity. Hence, an abnormal intrarenal
blood volume such as a hyper-vascularized tumour
could falsely have accentuated the relative uptake
share of the affected kidney and underestimated the
contribution of the contralateral kidney to overall
GFR. Nevertheless, other available calculation meth-
ods of split renal function have limitations that may
increase the error in the estimation of single kidney
GFR (Moonen et al., 1994).
GFR remained stable between 1 week and
6 months after radical unilateral nephrectomy.
Figure 2 Graphs showing the difference between
51 Hence, the major increase in GFR after nephrectomy
Cr-EDTA clearance values 1 week (upper panel) and
6 months (lower panel) after nephrectomy and the GFR for renal malignancy takes place during the ®rst post-
values predicted by gamma camera renography and 51Cr- operative week. This is in concert with previous
EDTA clearance prior to surgery. Patients with a relative ®ndings in both animal experiments and human
uptake of more than 30% by the affected kidney are shown studies of healthy kidney donors (Aurell & Ewald,
in open circles and patients with a relative uptake of 30% or
1981; Argiles et al., 1987; Pfaller et al., 1998). In
less by the affected kidneys are shown in solid circles. A
negative value implies that the preoperative measurements children who underwent nephrectomy for Wilms
overestimated the residual GFR after nephrectomy, and tumour, a similar compensatory increase in GFR of
hence, a positive value implies an underestimation of the the remaining kidney was observed among patients
residual GFR value. and controls who underwent nephrectomy for renal
trauma (Chevallier et al., 1997). An inverse correla-
tion between the compensatory increase in GFR of
affected kidney had some uptake (Tonnesen et al., the contralateral kidney after nephrectomy and age
1976). In contrast, there was no difference between has been reported (Hayslett, 1983; Talseth et al.,
predicted and measured post-operative GFR when the 1986). This is in agreement with the present data,
removed kidney had no uptake of the tracer. Collec- showing a more pronounced underestimation of the
tively, these data are in line with the present study residual GFR in younger patients.

Clinical Physiology 21, 6, 688±692 · Ó 2001 Blackwell Science Ltd 691

Prediction of GFR after nephrectomy for malignancy · M. Johansson & M. Moonen
...........................................................................................................................................................................................................................................................................................................................

The observed variability in the difference between References

predicted and measured post-operative GFR was
ARGILES A., MOURAD G., BASSET N. et al. (1987) Acute
substantial. This difference was less pronounced adaptative changes to unilateral nephrectomy in humans.
6 months after nephrectomy compared with the Kidney Int, 32, 714±720.
values obtained 1 week after nephrectomy, indicating AURELL M. & EWALD J. (1981) The use of living donors.
an effect of surgery on post-operative GFR in some Glomerular ®ltration rate during the ®rst year after donor
patients. We have previously reported a close nephrectomy. Scand J Urol Nephrol Suppl, 64, 137±142.
relationship between predicted and measured post- BROÈCHNER-MORTENSEN J. (1972) A simple method for the
determination of glomerular ®ltration rate. Scand J Clin
operative GFR in seven patients who underwent
Lab Invest, 30, 271±274.
nephrectomy for renal malignancy (Moonen et al., CHEVALLIER C., HADJ-AISSA A., BRUNAT-MENTIGNY M.,
1994). The present data, differ from the previously FARAJ G., SCHELL M. & COCHAT P. (1997) Renal function
reported by showing larger variability for the differ- after nephrectomy for Wilms' tumor. Arch Pediatr, 4,
ence between predicted and observed GFR of the 639±644.
remaining kidney. First, the previously reported data FLOOD H. D., HAMILTON S. & MCDERMOTT T. E. (1991)
were obtained in a small population of patients and Spontaneous recovery of renal function after resection of
contralateral hypernephroma (see comments). J Urol,
secondly, it is important to point out that kidney
146, 155±156.
depth was measured in the previous, but not in the GRANERUS G. & MOONEN M. (1991) Effects of extra-renal
present study. Although the difference in kidney background subtraction and kidney depth correction in
depth is usually less pronounced in the supine the measurement of GFR by gamma camera renography.
compared with the sitting position, a difference in Nuclear Med Commun, 12, 519±527.
kidney depth may have contributed to the variability HAYSLETT J. P. (1983) Effect of age on compensatory renal
between predicted and measured GFR in the present growth. Kidney Int, 23, 599±602.
LARSSON I., LINDSTEDT E., OHLIN P., STRAND S. E. &
study (Granerus & Moonen, 1991). There are other
WHITE T. (1975) A scintillation camera technique for
possible explanations for differences between predic-
quantitative estimation of separate kidney function and its
ted and measured GFR after nephrectomy for renal use before nephrectomy. Scand J Clin Lab Invest, 35,
malignancy, namely case reports of markedly 517±524.
improved GFR after resection of a renal hyperneph- MOONEN M., JACOBSSON L., GRANERUS G., FRIBERG P. &
roma (Flood et al., 1991). Although the mechanisms VOLKMANN R. (1994) Determination of split renal func-
are unclear, a paraneoplastic phenomena including tion from gamma camera renography: a study of three
hyperreninemia or a steal phenomena of the renal methods. Nuclear Med Commun, 15, 704±711.
PFALLER W., SEPPI T., OHNO A., GIEBISCH G. & BECK F. X.
blood supply by a highly vascularized tumour have
(1998) Quantitative morphology of renal cortical struc-
been proposed as explanations to improved GFR after tures during compensatory hypertrophy. Exp Nephrol, 6,
tumour resection (Flood et al., 1991). 308±319.
In conclusion, prediction of the residual GFR after TALSETH T., FAUCHALD P., SKREDE S. et al. (1986) Long-
nephrectomy for renal malignancy by noninvasive term blood pressure and renal function in kidney donors.
renal function tests will underestimate the post- Kidney Int, 29, 1072±1076.
operative GFR as a result of an adaptive GFR TONNESEN K. H., MOGENSEN P., WOLF H., HALD T. &
MUNCK O. (1976) Residual kidney function after unilat-
increase of the remaining kidney. This compensatory
eral nephrectomy. Pre- and post-operative estimation by
GFR increase appears only in patients with preserved renography and clearance measurements. Scand J Urol
function of the tumour affected kidney largely during Nephrol, 10, 130±133.
the ®rst post-operative week.

Acknowledgements

This study was supported by grants from the

Sahlgrenska University Hospital.

692 Ó 2001 Blackwell Science Ltd · Clinical Physiology 21, 6, 688±692