Clinical findings and kidney morphology in chronic kidney disease of unknown cause in India

Chronic kidney disease of unknown cause (CKDu) is an emerging health problem in India and other countries worldwide. However, clinical descriptions, including kidney pathology, are scarce.


Introduction
Chronic kidney disease of unknown cause (CKDu) has emerged as a severe health problem in several rural regions with hot and humid climates around the globe [1]. In Central America, where this disease was first described, it is often called Mesoamerican nephropathy (MeN). However, as similar disease entities have been described in regions outside of Mesoamerica, the term CKDu is now more often used. The disease is characterized by decreased kidney function with minimal urinary findings and affected patients lack traditional risk factors for kidney diseases, such as hypertension, diabetes mellitus, or obesity.
In India, CKDu has also been termed Uddanam Nephropathy, named after the Uddanam region in the state of Andhra Pradesh, where the disease was first reported [2]. A cross-sectional study of 2210 participants in the region reported a 14% prevalence of CKD, defined as eGFR <60 mL/min/1.73 m 2 , and 73% of the CKD cases were of unknown cause [3]. Furthermore, another study found that the prevalence of CKD (eGFR < 60 mL/min/1.73 m 2 ) was as high as 32% in the Andhra Pradesh region, and working as a farmer was identified as a risk factor [4]. A high proportion of CKDu among patients with CKD has also been reported in southeastern India (states of Tamil Nadu and Puducherry), as well as an endemic pattern in which some districts have a higher CKDu prevalence was found [5], similar to what has been described in Central America.
The clinical picture, biochemical findings, and renal biopsy morphology of patients with CKDu in India remain largely unknown. Gowrishankar et al. reported a brief description of biopsy findings from six patients with Uddanam nephropathy and described the main findings of chronic tubulointerstitial nephritis with varying degrees of tubulointerstitial pathology, nonspecific glomerular damage, and negative immunofluorescence (IFL) [6]. Furthermore, in a study from the Chhattisgarh region, kidney biopsies were collected from two participants, and the findings were glomerulosclerosis, periglomerular fibrosis, chronic IFTA changes of a mild-to-moderate degree (20%), and vascular changes [7].
To compare the endemic regions with other CKDu regions worldwide, detailed kidney morphology descriptions of CKDu in India are urgently needed. Notably, a common aetiology is possible if endemics have similar morphologies and biochemical patterns. Morphological findings may also provide clues regarding underlying pathophysiology.
Our group previously described the renal biopsy findings in three different case series of patients with CKDu in Central America and Sri Lanka [8][9][10]. In this study, we strived to provide a detailed description of kidney morphology and biochemical findings in a case series of patients with CKDu from rural areas in the Andhra Pradesh region, using a similar study protocol as in previous studies on CKDu.

Methods
CKD patients from rural areas with an endemic prevalence of CKDu were invited to participate in this study. Inclusion criteria were 20-65 years of age and eGFR 30-80 mL/min/1.73 m 2 . The exclusion criteria were diabetes mellitus (fasting blood glucose >126 mg/dL), uncontrolled hypertension (>140/90 or more than one hypertensive medica-tion), proteinuria >1 g/24 h, or other known or clinically diagnosed kidney diseases.
Participants underwent percutaneous ultrasoundguided kidney biopsy according to standard routine and were observed 24 h after the procedure. Blood and urine samples were collected before the biopsy, and the participants completed a questionnaire regarding their medical history, environmental exposure, and work history.
TEM tissue samples were transported to Dr. Lal PathLabs, Chennai, and stained with uranyl acetate and lead citrate according to standard protocols. Dr. Alok Sharma evaluated and examined the TEM sections at Dr. Lal PathLabs. A report, including a few representative TEM pictures, was sent to the research group for each participant.

Ethical statement
This project was approved by the Institutional Ethics Committee of Sri Ramachandra Medical College and Research Institute (IEC-NI/17/ JUN/60/73) and the Regional Ethical Review Board of Stockholm (Swedish Ethical Review Authority) (Dnr 2012/441-31/3; 2019-03119). All the participants provided informed consent.

Results
This study was conducted in December 2019 in Nellore, India. Fourteen patients (3 females and 11 males) from endemic areas in the Andhra Pradesh region with a mean age of 53 ± 11 (36-65) years and a mean eGFR creatinine of 53 ± 15 (range 29-78) mL/min/1.73 m 2 were included in the study.

Questionnaire
The clinical and questionnaire data are presented in Table 1. Three participants reported having family members with CKD. The mean liquid intake was 3.2 L during the workday and mostly consisted of water (84%-100% of the liquid consumed). The water source was groundwater for 11 participants and borewells for 3 participants. Most participants kept water in steel vessels; a few used clay pots or plastic containers.
Nine participants did not take any medication. Two participants received hypertension treatment (Losartan 50 mg and Amlodipine 5 mg respectively). Two participants were treated with potassium supplements, two with calcium tablets, two with sodium bicarbonate, and three participants with proton pump inhibitors. Intermittent NSAID use was reported by 57% of participants.
All but one participant had blood pressure (BP) ≤140/90 (median 110/70) mmHg on the day of the biopsy. One participant had a BP of 180/100 mmHg initially, but after a single dose of nifedipine (5 mg) the BP was lowered to 100/70 mmHg. Ten participants reported no specific symptoms, whereas four participants reported Note: Results are presented as mean ± standard deviation (range) or count (percentage). a Fertilizer Novagran was used by two participants. One participant reported exposure to organophosphate pesticides. The remaining participants were unaware of the agrochemicals they had been exposed to.

Biochemical results
The biochemical results are summarized in Table 2. Among the participants, 10 individuals tested negative for protein in the urinary dipstick test, whereas 2 showed trace-positive results, one had a 1+ result, and one had a 2+ result. Urinary sediment analysis revealed no erythrocyturia (<5 RBC/HPF), with leukocyte counts below 5 per HPF for all but 2 participants (4-6 and 10-15 WBC per HPF, respectively). Epithelial cells were generally observed at 1-2 per HPF (ranging from 0 to 4 per HPF). Urine cultures were obtained from 11 participants, all of which yielded negative results.  Screening for hepatitis B, hepatitis C, and HIV infections returned negative results. The anti-GBM-Ab (anti-glomerular basement membrane antibody) levels were below 7 U/mL in all cases. IFL ANA (IFL assay antinuclear antibody) testing showed positive results (2+ with a homogenous pattern) in one participant, but subsequent anti-dsDNA (anti-double-stranded DNA) analysis was negative. The complement C3 and C4 levels were within normal ranges. Weak P-ANCA (perinuclear antineutrophil cytoplasmic antibody) positivity was observed in two participants with IFL, whereas C-ANCA (cytoplasmic antineutrophil cytoplasmic antibody) was negative in all cases. Unfortunately, PR3 or MPO (proteinase 3 or myeloperoxidase) antibodies were not analysed; however, the subjects did not exhibit clinical or morphological signs of vasculitis.

Kidney biopsy results
Kidney biopsies were collected from 13 of the 14 participants. Kidney biopsy samples were not collected from one participant because of the risk of bleeding due to aspirin use before the scheduled biopsy. No adverse events were reported after the biopsy. Kidney biopsies contained 3-22 glomeruli per specimen (mean: 10 glomeruli). IFL was performed in all but one participant (patient 4) and showed no signs of immune complex disease.
Polarized light revealed no birefringent crystals in any of the specimens. The main results of the LM evaluation are presented in Table 3, and the detailed results are presented in Tables S1-S4.
Glomerular pathology. All patients had glomerulosclerosis of varying degrees (17%-54% of the included glomeruli) (Fig. 1a). Glomerular hypertrophy was observed in all but one participant (Fig. 1b), and signs of glomerular ischaemia, that is, wrinkling of the glomerular basement membrane and/or periglomerular fibrosis, were observed in 8 of 13 biopsies (Fig. 1c). Segmental scleroses were found in two participants (patients 1 and 14; Fig. 1d). All participants had normal mesangial areas, and no endocapillary cell proliferation was observed.
Tubulointerstitial pathology. In most cases, interstitial fibrosis and tubular atrophy were mild (Fig. 2a), affecting less than 25% of the cortex. Four specimens had mild tubulitis (patients 1, 3, 4, and 12); in patient 1, both lymphocytic and neutrophilic tubulitis were observed, and the other cases had only lymphocytic infiltration (Fig. 2b). Two patients had PMNs in the tubular lumen (patients 1 and 3), one of whom also had PMN casts (patient 1). Thyroidization patterns of the tubules were observed in patients 7 and 8. In most cases, interstitial inflammation was absent or mild; Table 3. however, two participants had severe inflammation, mainly consisting of lymphocytes. One participant (patient 3) displayed focal dysplasia in a few tubular cross sections (Fig. S1). Two participants (patients 6 and 10; Fig. 2c) had intracytoplasmic silver-positive granules, as described in a previous publication on CKDu [15]; however, silver staining was weak in several other biopsies, possibly interfering with the results.
Vascular pathology. In most biopsies, the blood vessels had no or mild pathology, consisting of mild arterial intimal thickening and/or mild arterial smooth muscle hyperplasia (Fig. 1a). The arterioles were normal or showed mild hyalinosis in most specimens (Fig. 1c,d). Two biopsies had moderate arterial changes.
Transmission electron microscopy. In 11 of the 13 samples subjected to TEM, glomeruli were found in the sections, whereas two biopsies only had tubulointerstitial fractions. The mean thickness of the GBM was 355 ± 71 nm. One biopsy showed an increased mean GBM thickness of 517 nm (patient 5), whereas all others were normal. All biopsies showed focal process effacement ranging between 15% and 30%, with a median of 20% (Fig. 3a).
Occasional electron-dense deposits were found in the mesangium of patient 14, but the appearance was not typical for immune complexes, and the patient had a negative IFL, making the finding unspecific. No electron-dense extraglomerular deposits were observed. Endothelial cells did not show tubuloreticular inclusions. In tubular cells, electron-lucent vacuoles were observed in 9 of 11   Table S5.
Kidney morphology diagnoses. Of the 13 available kidney biopsies, 6 (patients 2, 5, 6, 7, 11, and 13) displayed morphology characterized by a combination of glomerulosclerosis, glomerular hypertrophy, and mild tubulointerstitial damage, similar to that described in CKDu cases from Central America [8,9,16]. Two biopsies (patients 1 and 3) showed acute and chronic tubulointerstitial nephritis (TIN), of which one also displayed focal tubular dysplasia in a few cross-sections. One biopsy (patient 4) showed chronic TIN with severe inflammation and tubulointerstitial changes. A biopsy of patient 9 showed nephrosclerosis, and another biopsy (patient 12) showed a combination of TIN and nephrosclerosis. One biopsy (patient 14) showed signs of glomerular hyperfiltration with probable secondary FSGS; primary FSGS was a possible differential diagnosis. The biopsy from participant 10 had ≤4 glomeruli and, therefore, not enough tissue to make a reliable morphological diagnosis.

Comparison of kidney biopsy morphology with other
CKDu endemic regions. A summary of the LM findings in kidney biopsies in the present study from Andhra Pradesh, India, along with results from previously published kidney biopsy studies by our group from El Salvador [8], Nicaragua [9], and Sri Lanka [10], are presented in Table 4. Common traits in all four studies from different endemic regions were glomerulosclerosis, glomerular hypertrophy, mild-to-moderate tubulointerstitial changes, and mostly mild vascular changes. Most participants in all four studies had signs of glomerular ischaemia (wrinkled GBM and/or periglomerular fibrosis).

Discussion
CKDu is a serious health problem in parts of India [3,17], but descriptions of its clinical picture and renal biopsy findings are scarce and incomplete. This study presents the detailed renal morphology and clinical and biochemical data of 14 patients with CKDu (3 females and 11 males, 36-65 years of age) from Andhra Pradesh, a state with a high disease prevalence. We compared kidney biopsy morphology from other cohorts with MeN/CKDu that we examined in previous studies [8][9][10].
The included participants had a mean eGFR creatinine of 53 (range 29-78) mL/min/1.73 m 2 , and the main morphological findings were a combination of glomerulosclerosis, glomerular hypertrophy, signs of glomerular ischaemia, and mild-to-moderate chronic IFTA changes. These findings are similar to those we (Table 4) and other research groups have described for CKDu in Central America and Sri Lanka [8][9][10]16], although some differences were observed. For example, signs of glomerular ischaemia are less frequently observed in biopsies from India and Sri Lanka than in those from Central America. Although some biopsies in this study showed pronounced inflammation and tubulointerstitial damage, potentially representing an active form of the disease, similar features have been described by Fischer et al. in agricultural workers with acute MeN/CKDu in Nicaragua [18]. Nevertheless, our overall findings conclude that MeN in Central America and CKDu in Sri Lanka and India most likely share a common pathological background.
Compared with our prior MeN/CKDu studies in Central America, vascular pathology was slightly more common in this sample, but this was consistent with our study in Sri Lanka. Possible explanations could be age-related factors, such as the higher mean age in this study and a higher smoking frequency.
In 2020, Gowrishankar et al. reported kidney morphology in six patients with Uddanam nephropathy to be chronic TIN; however, they also reported varying degrees of glomerulosclerosis and glomerular changes, including periglomerular fibrosis [6], findings that are compatible with ours.
A small area of focal tubular dysplasia was observed in one participant. The clinical relevance of this finding is unclear, and follow-up is recommended. Dysplasia has not been previously described in patients with CKDu and may be an incidental finding. In another endemic type of CKD, aristolochic acid nephropathy/Balkan nephropathy, upper urothelial malignancies have been described [19]. However, the morphology of aristolochic acid nephropathy is typically widespread interstitial fibrosis [19], which is incompatible with the morphology observed in CKDu.
number of TEM images that could be evaluated. Hence, although there was no obvious increase in the presence of dysmorphic lysosomes, a detailed analysis of this feature could not be performed. However, dysmorphic lysosomes can also be found in healthy kidney biopsies and other kidney diseases and are therefore not pathognomonic for CKDu but could indicate tubular stress [20,21].
Clinically and biochemically, the participants were similar to those in other CKDu regions: no haematuria, absence of, or low-grade proteinuria. Low potassium, sodium, and magnesium levels have been reported in patients both in Central America, Sri Lanka, and India [7,10,22]. In our case series, potassium and sodium levels were mostly normal, but in the lower reference interval. Further, two participants were treated with potassium supplements. Polyuria was found in most participants in this study, indicating tubular injury. Polyuria has also been reported as a common finding among patients with CKDu in Central America [9,22].
The cause of CKDu is not completely understood; however, studies from Central America have suggested that heat stress and episodes of dehydration are important risk factors and potential causes [23][24][25][26][27]. Studies of sugarcane workers in Nicaragua have shown that sugarcane cutters, the category most exposed to heat stress, experience the highest risk of kidney injury during harvest season and that an intervention providing 'water, rest and shade' decreased the risk of kidney injury [25]. Occupational heat stress is currently considered the main driver of the CKDu epidemic in Central America [28]. The extent to which heat exposure and strenuous physical work contribute to CKDu in India and Sri Lanka is unknown, but the type and intensity of heat stress differs and is likely to be more severe in sugarcane workers in Central America than in rice paddy farmers in Sri Lanka and India. Nevertheless, a study from Brazil shows that an increased ambient temperature of 1°C increased the hospitalization rates for renal disease by approximately 1% [29]. In support of the heat stress hypothesis, CKDu has been reported mainly in rural areas with high ambient temperatures and humid climates. Individuals with CKDu in India and other CKDu-endemic regions often have poor socioeconomic status and inadequate housing and water sources, making them more prone to occupational and environmental heat stress.
The development of chronic interstitial nephritis after heat stroke has been reported [30,31], and a recent study in Taiwan reported a substantially higher risk of CKD after an acute heat illness [32]. An interesting case report described a longdistance runner who repeatedly drank inadequate water during exercise and developed acute kidney injury and subsequent CKD. Interestingly, the renal biopsy specimen from this patient showed a combination of tubulointerstitial and ischaemic glomerular changes [33], similar to those found in CKDu patients [8][9][10]. Whether subclinical heat stress with volume depletion can cause CKD is not yet well established. However, animal studies have reported chronic kidney damage after repeated episodes of heat stress and dehydration [34,35].
Signs of glomerular ischaemia without significant chronic vascular changes were common in this case series, as in our previous studies on CKDu in Nicaragua, El Salvador and Sri Lanka [8][9][10]. Glomerular ischaemia is usually observed in hypertensive patients, but hypertension is rarely observed in typical CKDu cases and must therefore be induced by another mechanism. We have previously argued that a possible mechanism for the glomerular ischaemia seen in many patients with CKDu may be chronic sodium depletion (secondary to sweating/heat stress) activating the renin-angiotensin-aldosterone system, whereby angiotensin II causes glomerular ischaemia by constricting the glomerular capillaries [9]. However, this hypothesis warrants further investigation.
In addition to heat, many environmental factors, including contaminated drinking water (heavy metals, arsenic, and water hardness), pesticides/agrochemicals, infections, and air pollution, have been proposed as potential causes of CKDu; however, no conclusive evidence has been presented [36]. Silica exposure from drinking water [37] or inhalation [38] has also been proposed as a cause. Recently, the herbicide paraquat was reported as a potential aetiology of CKDu in Central America [39]. In our study, exposure to indoor smoke from open fires (86%) and agrochemicals (71%) were both common (any patient did not specifically report paraquat); however, the findings may simply reflect the living conditions in rural areas where CKDu is prevalent, and without a control group, no conclusions can be drawn.
Genetic factors have also been proposed because many patients with CKDu report a family history of CKD [40]. However, family clustering of cases may be due to similar environmental and workrelated exposures. In Sri Lanka, genetic studies have shown significant associations between single nucleotide polymorphisms in the SLC13A3 gene [41] and KCNA10 gene [42], both expressed in proximal tubular cells. These associations are interesting and require further investigation.
Low birth weight may be another potential risk factor because it is associated with a lower nephron count, which leads to nephron hyperfiltration, glomerular hypertrophy, and increased susceptibility to kidney injury [43]. The relationship between low birth weight and socioeconomic status has been documented in previous studies [44,45]. Countries affected by CKDu have a higher prevalence of low birth weight, such as Sri Lanka with a prevalence of 17% [46], India with 18% [44], and Central America with 15% [47], compared to high-income countries, where the prevalence is around 7%. While the birth weights in this study are unknown, we acknowledge the potential role of low birth weight as a contributing factor to the development of CKDu, albeit not as the primary cause.
In conclusion, it is striking how many clinical, biochemical, and kidney morphology similarities were shared in our case series of CKDu patients from India with CKDu patients from Central America and Sri Lanka. Although these similarities and the lack of classical risk factors indicate a shared pathophysiology, we cannot determine whether the factors driving CKDu development are identical across geographical locations. Additional crosssectional and epidemiological studies are required to determine the aetiology of CKDu in India and globally. Although firm evidence that heat stress is a driving factor behind CKDu in India has yet to be established, a cautious approach is to recommend ready access to clean drinking water and protection against heat stress in endemic locations. As extreme heat weather conditions become more common, global research efforts to understand and prevent kidney injury caused by heat stress are of utmost importance.

Supporting Information
Additional Supporting Information may be found in the online version of this article: Table S1: Glomerular changes observed by light microscopy in kidney biopsies from 13 participants with CKDu from the Andhra Pradesh region. Table S2: Tubular changes detected by light microscopy in kidney biopsies from 13 participants with CKDu from the Andhra Pradesh region. Table S3: Interstitial changes detected by light microscopy in kidney biopsies from 13 participants with CKDu from the Andhra Pradesh region. Table S4: Vascular changes observed by light microscopy in kidney biopsies from 13 participants with CKDu from the Andhra Pradesh region. Table S5: Electron microscopy findings in kidney biopsies from 13 participants with CKDu from the Andhra Pradesh region. Figure S1: Light microscopy showing focal dysplasia in a few tubular cross sections in a kidney biopsy from one study participant. (Patient 3, light microscopy, Hematoxylin-Eosin staining, Bar = 20 μm).