Short Reports      250

Spectrum of severe chronic kidney disease in India: A clinicopathological study
K. S. DHARAN, G. T. JOHN, N. NEELAKANTAN,
A. KORULA, N. BALAKRISHNAN,
M. G. KIRUBAKARAN, C. K. JACOB

ABSTRACT
Background. The healthcare burden due to chronic kidney disease has increased worldwide in the past decade. Elucidating the aetiology of chronic kidney disease may help in identifying strategies for prevention, both in the population and the individual patient. Only a clinicopathological study can
define the exact spectrum of chronic kidney disease since epidemiological studies have not shown a consistent aetiological profile. The histological evidence used to support the diagnosis varies with the degree to which renal biopsy is done. Renal biopsy is the gold standard in making an aetiological diagnosis in renal failure, but as a diagnostic tool in chronic kidney disease it is underutilized.
Methods. This prospective study done at Christian Medical College, Vellore in southern India from 1998 to 2003 aimed to determine the aetiological profile of severe chronic kidney disease by analysing renal biopsies. The value of pre-renal biopsy clinical judgement in predicting the histological diagnosis was also assessed. Patients with diabetic nephropathy were excluded from the study.
Results. Four hundred and fifty-seven patients had evidence of chronic kidney disease as evidenced on biopsy as well as on clinical parameters. Three hundred and twenty-two of these patients (70.5%) had glomerulonephritis as the histological diagnosis. Fifty-five (12%) had interstitial nephritis, 30 (6.6%) had hypertensive arteriosclerosis and 28 (6.1%) had metabolic nephropathies. The positive predictive value of a pre-biopsy clinical diagnosis in predicting interstitial nephritis was very low (33%). A large number of patients clinically diagnosed to have chronic interstitial nephritis had other aetiologies of chronic kidney disease.
Conclusion. Glomerulonephritis was the most common cause of chronic kidney disease, not including diabetic nephropathy, followed by interstitial disease and benign arterionephrosclerosis. In patients with unidentified severe chronic kidney disease, renal biopsy provided an aetiological diagnosis.

Natl Med J India 2006;19:250–2

INTRODUCTION
The aetiological profile of chronic kidney disease (CKD) is disputed, as the relative prevalence of the common causes differ among studies.1–5 Many studies2,3 have found that chronic glomerulonephritis is the most common cause while others rate interstitial nephritis1,4,5 as the commonest aetiology. Most studies are based on clinical presentation and are not supported by renal biopsy, which is the gold standard in making an aetiological diagnosis.
   Renal biopsy is underutilized in CKD as a diagnostic tool. Although CKD was considered a relative contraindication previously, this approach has changed in recent times because of the precision and safety offered by ultrasound-guided biopsy techniques and the use of automated biopsy devices.
   A biopsy-based study is the most appropriate method to profile the spectrum of CKD. This would avoid the bias that may arise from studies based primarily on clinical diagnosis and would give a more accurate picture of the causes of CKD. We attempted to identify the relative incidence of various causes of CKD in India as confirmed by renal biopsy.
METHODS
This prospective study was done by analysing all renal biopsies in patients with CKD stage 4 and 5 during 1998–2003 at Christian Medical College, Vellore in southern India. From 1998, all patients with CKD stage 4 and 5 were offered an ultrasound-guided renal biopsy by an experienced nephrologist if the kidney size (interpolar length >8 cm) and cortical thickness were adequate. Although western studies report the normal size of a kidney to be approximately 11 cm,6 the normal size in Indians is not known. Patients with CKD who gave consent for renal biopsy had a pre-biopsy diagnosis made which was based on clinical and biochemical parameters. Patients with clinical features suggestive of diabetic nephropathy were excluded unless there were atypical features such as absence of retinopathy, presence of dysmorphic red blood cells or erythrocyte casts in the urine, rapid deterioration of renal function of unknown cause, elevated serum creatinine without urinary abnormalities or massive proteinuria. Multiple sections of all biopsies were examined and stained with haematoxylin and eosin, periodic acid-Schiff and Jones methenamine silver. Immunoflourescence microscopy was done with IgA, IgG, IgM and C3.
   For the purpose of classification, the renal histology was divided into 5 broad categories.

1. Chronic glomerular disease included all biopsy diagnoses where the primary disease was glomerular in origin by either light microscopy or immunofluorescence or both. If patients had focal glomerulosclerosis and hyalinosis with no evidence of a primary disease that would have caused hyperfiltration glomerulopathy, a diagnosis of focal glomerulosclerosis was made. IgA nephropathy, membranous nephropathy and membranoproliferative glomerulonephritis was diagnosed based on both glomerular changes and immunofluorescence.
2. Chronic interstitial nephritis included all biopsies where the interstitial changes predominated without evidence of a primary glomerular disease.
3. Benign arterionephrosclerosis was diagnosed when the biopsy showed shrinkage of glomerular tuft and ischaemic obsolescence, hyaline deposits of media and intima of arterioles and small arteries and intimal thickening.
4. Metabolic nephropathies included histologies such as amyloidosis, light chain deposition disease, oxalosis, etc. in which the primary disease was a metabolic abnormality.
5. End-stage parenchymal disease included those biopsies in which a definite primary aetiology could not be determined on the basis of histological appearance.

   A pre-renal biopsy clinical diagnosis was made based on history, clinical features and investigations, and patients were grouped into different categories. Chronic glomerulonephritis was diagnosed in patients with long standing oedema, history of a nephritic or proteinuric illness in the past or nephrotic range proteinuria. Chronic interstitial nephritis was diagnosed in patients with an insidious onset of renal failure, onset of oedema only at late stages of renal failure and proteinuria <1 g with a normal serum albumin. Benign arterionephrosclerosis was diagnosed when there was a long standing history of documented hypertension, subnephrotic proteinuria and hypertensive retinopathy.
   The pre-biopsy diagnosis was compared with the histological diagnosis. The sensitivity, specificity and predictive values of pre-biopsy diagnosis in predicting the exact histological picture were calculated. SPSS 11.0 was used for data analysis.

RESULTS
Of the 3070 renal biopsies done during the 6-year study period (1998–2003), 457 patients had severe CKD. The average annual number of CKD patients seen by the hospital is approximately 800. The mean (SD) age was 38.2 (14.5) years. Twenty-two patients were >65 years of age. Two hundred and eighty-three (62%) patients were men. One hundred and twenty-two patients underwent haemodialysis before biopsy. In 305 patients, the glomerular filtration rate (GFR) as calculated by the formula followed in the MDRD study7 was below 15 ml/min/1.73 m2 (CKD stage 5) while in 152 patients it was 15–30 ml/min/1.73 m2 (CKD stage 4).
   The most common cause of CKD in the series was glomerular disease 322 (70.5%), followed by interstitial nephritis 55 (12%), benign arterionephrosclerosis 30 (6.6%), metabolic nephropathies 28 (6.1%), end-stage parenchymal disease 12 (2.6%) and other 10 (2.2%). There was no significant difference in the incidence of aetiologies among patients with CKD stage 4 and stage 5.
   The role and utility of pre-biopsy diagnosis and its correlation with histological diagnosis was assessed. Before biopsy, 326 patients (71.3%) were diagnosed to be having glomerular disease, 108 (23.6%) with chronic interstitial nephritis and 16 (3.5%) with benign arterionephrosclerosis based on clinical and biochemical parameters. The sensitivity, specificity, positive predictive value and negative predictive value of pre-biopsy diagnosis in each of these categories along with kappa statistic are given in Table I.
   The biopsy sample was inadequate in 10 patients. In 12 patients, it was not possible to clearly identify the primary disease and was denoted as end-stage parenchymal disease.
   Of the 108 patients clinically thought to have chronic interstitial nephritis, only 36 (33%) were shown to be correct on histology. Glomerular disease was found in 44 (40.7%), benign arterio-

nephrosclerosis in 19 (17.6%) and metabolic nephropathies in 9 (8%). Of the 326 patients who had a pre-renal biopsy diagnosis of glomerular disease, 271 (83.1%) had glomerular disease on biopsy, 14 (4.3%) were found to have interstitial nephritis, 11 (3.4%) had benign arterionephrosclerosis and 27 (8.3%) had metabolic nephropathies. Of the 16 patients thought to be having benign arterionephrosclerosis, 7 (43.8%) were found to have the same diagnosis on biopsy; 5 (31.3%) were found to have interstitial nephritis and 4 (25%) had glomerular disease.
   Two of the patients had macroscopic hematuria and required hospitalization beyond 24 hours. No patient required nephrectomy for control of haematuria.

DISCUSSION
The incidence of chronic renal failure is rising.8 A clear understanding of the spectrum of causes leading to severe CKD may help in planning the strategies to counter the increasing incidence. Most epidemiological studies on the aetiology of chronic renal failure have been based mainly on clinical and biochemical parameters. Only a small proportion of patients have had a renal biopsy to prove the aetiology. In spite of new and less invasive tests, renal biopsy is still an irreplaceable tool in establishing the diagnosis. It is well known that in CKD, clinical diagnosis may not always correlate with the histological diagnosis.9 Hence, only a biopsy-based study can accurately profile the spectrum of CKD.
   In a population-based study published earlier from the same region,1 2028 patients with CKD were assigned a tentative diagnosis and were analysed (only 23.4% of patients had a biopsy examination). The results showed that chronic interstitial nephritis (27.9%) was the commonest aetiology followed by diabetic nephropathy (26.8%). On re-analysis after excluding diabetic nephropathy and other conditions where renal biopsy is usually not done such as polycystic kidney disease and obstructive uropathy, interstitial nephritis (46%) remained the predominant cause, while glomerulonephritis (30%) remained second. Ball et al.10 showed a higher proportion of Indians having interstitial nephritis (17/75) as compared with Caucasians (13/223). Another study2 from the same region which analysed 453 consecutive pre-end stage renal disease CKD patients (in which only 15.6% of patients had a biopsy examination) showed chronic glomeru-lonephritis (25.6%) as the predominant cause. In a population-based study in Scotland,3 46% of cases were of unknown cause, interstitial nephritis and diabetic nephropathy were the most frequent specific causes, and only 4% had glomerulonephritis. In another study of 369 patients with CKD who were under the care of nephrologists, a much higher proportion (21%) had glomerulonephritis.4 Barsoum5 in a review from North Africa describes interstitial nephritis as the commonest cause followed by glomerulonephritis.
   Our study describes the spectrum of CKD from southern India. It shows that the predominant cause of CKD is glomerular disease. Interstitial disease was the second common group, but much
less than glomerular disease. This was followed by benign arterionephrosclerosis and metabolic nephropathies, respectively. The positive predictive value of pre-renal biopsy diagnosis was low especially with interstitial nephritis. This offers the possibility that in non-biopsy-based studies, a proportion of patients with glomerulonephritis might have been diagnosed to have interstitial nephritis based on clinical and biochemical data, accounting for the higher incidence of interstitial nephritis.
   Another question to be answered is whether renal biopsy is being done for fewer patients with interstitial nephritis at CKD stages 4 and 5 when compared with other aetiologies especially glomerular disease. There are no data to suggest that kidneys with interstitial disease contract faster than those due to other causes and become unsuitable for biopsy at end stage. The present study shows that there are no differences in the relative incidence of aetiologies between CKD stages 4 and 5 suggesting there is no significant difference in kidney size among aetiologies between stages 4 and 5. Further, another study11 from our institution involving 5258 biopsies irrespective of renal function has shown a low incidence of interstitial disease (6.8%). Studies from the USA11 have hypertensive vascular disease as one of the most important causes of CKD (30%), but in our study its incidence was low.
   In conclusion, glomerulonephritis is the predominant cause of CKD, not including diabetic nephropathy, followed by interstitial disease and benign arterionephrosclerosis. A renal biopsy needs to be attempted in all patients with acceptable kidney size and sonological features as a clinical diagnosis may not be accurate. Awareness of the relative incidence of the aetiologies in a geographical locality may help in planning preventive strategies to combat the rising incidence of CKD in that area.

REFERENCES

1. Mani MK. Chronic renal failure in India. Nephrol Dial Transplant 1993;8:684–9; discussion 683.
2. Sakhuja V, Jha V, Ghosh AK, Ahmed S, Saha TK. Chronic renal failure in India. Nephrol Dial Transplant 1994;9:871–2.
3. Khan IH, Catto GR, Edward N, MacLeod AM. Chronic renal failure: Factors influencing nephrology referral. QJ Med 1994;87:559–64.
4. Landray MJ, Thambyrajah J, McGlynn FJ, Jones HJ, Baigent C, Kendall MJ, et al. Epidemiological evaluation of known and suspected cardiovascular risk factors in chronic renal impairment. Am J Kidney Dis 2001;38:537–46.
5. Barsoum RS. End-stage renal disease in North Africa. Kidney Int Suppl 2003:
   S111–S114.
6. Standring S. Gray’s anatomy: The anatomical basis of clinical practice. 39th ed. London:Churchill Livingstone; 2005:1269.
7. Levey AS, Greene T, Kusek JW, Beck GJ. A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 2000;11:155A.
8. Agarwal SK. Chronic kidney disease and its prevention in India. Kidney Int Suppl 2005;68:S41–S45.
9. Pfister M, Jakob S, Frey FJ, Niederer U, Schmidt M, Marti HP. Judgment analysis in clinical nephrology. Am J Kidney Dis 1999;34:569–75.
10. Ball S, Cook T, Hulme B, Palmer A, Taube D. The diagnosis and racial origin of 394 patients undergoing renal biopsy: An association between Indian race and interstitial nephritis. Nephrol Dial Transplant 1997;12:71–7.
11. Balakrishnan N, John GT, Korula A, Visalakshi J, Talaulikar GS, Thomas PP, et al. Spectrum of biopsy proven renal disease and changing trends at a tropical tertiary care centre, 1990–2001. Indian J Nephrol 2003;13:29–35.
12. Valderrabano F, Gomez-Campdera F, Jones EH. Hypertension as cause of end-stage renal disease: Lessons from international registries. Kidney Int Suppl 1998; 68:S60–S66.