Predictors of Autosomal Dominant Polycystic Kidney Disease Progression

Predictors of Autosomal Dominant Polycystic Kidney Disease Progression Robert W. Schrier,* Godela Brosnahan,* Melissa A. Cadnapaphornchai,* Michel Chonchol,* Keith Friend, Berenice Gitomer,* and Sandro Rossetti *Division of Renal Diseases and Hypertension, University of Colorado School of Medicine, Aurora, Colorado; and Cardiorenal Department, Medical Affairs, Otsuka America Pharmaceutical, Inc., Princeton, New Jersey ABSTRACT Autosomal dominant polycystic kidney disease is a genetic disorder associated with substantial variability in its natural course within and between affected families. Understanding predictors for rapid progression of this disease has become increasingly important with the emergence of potential new treatments. This systematic review of the literature since 1988 evaluates factors that may predict and/or effect autosomal dominant polycystic kidney disease progression. Predicting factors associated with early adverse structural and/or functional outcomes are considered. These factors include PKD1 mutation (particularly truncating mutation), men, early onset of hypertension, early and frequent gross hematuria, and among women, three or more pregnancies. Increases in total kidney volume and decreases in GFR and renal blood flow greater than expected for a given age also signify rapid disease progression. Concerning laboratory markers include overt proteinuria, macroalbuminuria, and perhaps, elevated serum copeptin levels in affected adults. These factors and others may help to identify patients with autosomal dominant polycystic kidney disease who are most likely to benefit from early intervention with novel treatments. J Am Soc Nephrol 25: ccc ccc, 2014. doi: 10.1681/ASN.2013111184 Autosomal dominant polycystic kidney disease (ADPKD) is a genetic disorder caused by mutations in the PKD1 gene located on chromosome 16p13.3 or the PKD2 gene located on chromosome 4q21. 1 PKD1 and PKD2 encode the proteins polycystin- 1 and -2, respectively; mutations disrupt the function of these proteins on the primary cilium, forming fluid-filled cysts that progressively increase in size, leading to gross enlargement of the kidneys and distortion of the renal architecture. 2 Glomerular hyperfiltration compensates for the progressive loss of healthy glomeruli, and therefore, by the time GFR decline becomes detectable, as much as one half of the original functional glomeruli are irreversibly lost. 3,4 The majority of patients with ADPKD ultimately progress to ESRD. 3 The natural course of ADPKD varies significantly, with onset of ESRD reported from childhood to age.80 years, and a median age of 58 years was reported recently for PKD1, which is much more commonthanpkd2. 5 Rapid ADPKD progression may be defined as onset of ESRD at age,55 years, development of stage 3 CKD at,40 years old, onset of hypertension at,18 years old, presence of total kidney volume (TKV) greater than that expected for a given age, or presence of multiple complications. Identification of patients at high risk for rapid progression has become increasingly important given the emergence of potential new treatments. These treatments are most likely to be beneficial when started early in the disease course. This systematic review evaluates markers of and factors contributing to ADPKD progression. METHODS A comprehensive systematic review of the literature was undertaken using several structured steps (Table 1). All authors participated in each step of the review process and ranking of final articles for inclusion. RENAL DISEASE PROGRESSION IN ADPKD Genetic Factors A summary of the main studies examining genetic predictors of rapid renal disease progression is provided in Table 2. Locus Heterogeneity Several large studies have shown that patients with mutations in PKD1 generally have a more severe form of ADPKD than patients with PKD2 mutations, with a younger age at diagnosis, a higher number of cysts, earlier onset of hypertension, Published online ahead of print. Publication date available at. Correspondence: Dr. Robert W. Schrier, Division of Renal Diseases and Hypertension, University of Colorado School of Medicine, Box C281, 12700 East 19th Avenue, Research 2, Room 7001, Aurora, CO 80045. Email: robert.schrier@ucdenver.edu Copyright 2014 by the American Society of Nephrology J Am Soc Nephrol 25: ccc ccc, 2014 ISSN : 1046-6673/2511-ccc 1

Table 1. Summary of systematic review process Step Procedure Result 1 Systematic literature review through Medline, EMBASE, and 2056 citations Biosis for articles published from January of 1988 to May of 2013 a 2 Review abstracts; omit articles not dealing with factors affecting 1078 citations ADPKD severity or progression, reviews, case reports, editorials, commentaries, and letters 3 Omit non-english language articles and articles with focus on 863 citations transplantation, dialysis, or diabetes, with ESRD used as a cutoff for inclusion 4 Omit duplicate references; results separated into four categories: randomized clinical trials; other clinical trials; observational, retrospective, or epidemiologic studies; and preclinical studies 666 articles ranked for inclusion b a The searches comprised ADPKD or autosomal dominant polycystic kidney disease combined with the following terms: angiogenic growth factor; anxiety, depression, or quality of life; child or pediatric; cyst (number or area) or liver fibrosis index; cyst segmentation; depression and dialysis; diabetes mellitus; epidemiology, hospitalization, or cost; ESRD or renal disease; family history, genetic or allelic heterogeneity, genotype phenotype, mutation type or position, polycystin, vasopressin, or camp; glomerular hyperfiltration; HALT PKD; hematuria; hypertension; intracellular calcium, cell proliferation, or fluid secretion; kidney, liver, or renal disease progression; left ventricular mass hypertrophy; molecular diagnostics; proteinuria or microalbuminuria; psychonephrology; renal blood flow, urine osmolality, biomarker, proteome, copeptin, exosome, or serum uric acid; renal or kidney ultrasonography, computed tomography, or MRI; TKV or height-adjusted TKV; total liver volume; UTI; and vitamin D. b Additional references were identified by the authors on the basis of their personal knowledge of the literature that reported specific information important to ADPKD progression. and faster progression to ESRD. 5 11 In the European PKD1-PKD2 study cohort of 624 patients, median ages at death or ESRD onset in patients with PKD1 and PKD2 were 53 and 69 years, respectively (P,0.001). 8 Similarly, in the recent large Genkys study of 741 patients with ADPKD from northwestern France, median ages at onset of ESRD in PKD1 and PKD2 carriers were 58 and 79 years, respectively. 5 Some PKD1 families, however, have mild disease that is indistinguishable from PKD2 families, whereas other PKD1 families have a particularly severe course. 3,5 In contrast to different gene loci, this observation may be caused by allelic heterogeneity within the PKD1 locus. Allele Heterogeneity: Mutation Type and Location Results from the Genkys study showed that truncating PKD1 mutations (frameshift, nonsense, splice mutations, and large rearrangements), affecting approximately two thirds of PKD1 families, were associated with a significantly younger median age of ESRD onset than nontruncating mutations (in-frame and missense mutations; 55.6 versus 67.9 years [P,0.001]). 5 In another study, mutations toward the 59 end of PKD1 were associated with a vascular phenotype andrupturedintracranialaneurysms. 12 Mutation location within the PKD2 gene has also been suggested to influence outcome in a study of 22 PKD2 families 13 ; however, this result was not confirmed in a larger study of 461 subjects from 71 PKD2 families. 14 In the latter study, after adjustment for sex, patients with splice mutations tended to have better renal survival than patients with other types of mutations. Hypomorphic Alleles Several families have been described with a mild or atypical disease presentation, in which ADPKD is not explained by the dominant inheritance of a single PKD1 or PKD2 mutation. 15 18 In these families, sequence variants function as hypomorphic alleles, because they do not result in the full phenotype when occurring as the only abnormality in heterozygous subjects, but they do result in severe disease in homozygous (for this variant) individuals or when occurring together with a truncating mutation on the second allele. 16 18 For example, a unique ADPKD family has been described in which 30 affected individuals had mild disease but only 14 individuals carried a PKD2 mutation, whereas the other 16 individuals did not have this mutation but rather a sequence (Y528C missense) variant in PKD1. 15 Expression of this variant in cultured cells confirmed its pathogenicity. Two family members carried both the PKD2 mutation and PKD1 variant, resulting in more severe disease than observed in carriers of the PKD2 mutation alone. Mosaicism for a PKD mutation may also account for an atypical mild phenotype. 19 Modifier Genes Highly variable disease severity occurs even within families, which cannot be explained by locus or allelic heterogeneity. 3,14,20 Intrafamilial variability may be caused by other genes that modify disease severity. 20 Several gene polymorphisms (e.g., in the angiotensin-converting enzyme, endothelial nitric oxide synthase, and a8 integrin genes) have been associated with age at ESRD onset, although usually in men and not women. 21 24 Many of these reports were, however, based on small study populations with lack of correction for population stratification. Additional insight regarding the identity of genes that modify the severity of ADPKD will be provided by the outcome of ongoing large genome-wide association studies. It is also significant that PKD1 mutations maymodifythediseasecourseinpkd2 carriers with resultant earlier onset of ESRD, which was described in a case report of a de novo PKD1 splice mutation on the background of a homozygous PKD2 variant. 25 Demographic and Familial Factors A summary of the main studies examining demographic and familial predictors of rapid renal disease progression is provided in Table 3. Sex Several studies reported more severe disease in men thanwomen, with earlier onset of hypertension, more severe hypertension, and earlier onset of ESRD. 5 7,14,26,27 In one large retrospective cohort study of 580 patients with ADPKD who were followed for 25 years, men had worse renal function at a given age. 26 In a second retrospective study of 1215 patients with 2 Journal of the American Society of Nephrology J Am Soc Nephrol 25: ccc ccc, 2014

BRIEF REVIEW Table 2. Genetic predictors of rapid ADPKD progression: phenotypic differences between PKD1 and PKD2 mutations Ref./Findings PKD1 PKD2 P Value 5 Patients, n 392 95 Mean age at hypertension 39 49,0.001 diagnosis, yr Median age at ESRD onset, yr 58 79,0.001 6 Patients, n 146 20 Mean age at diagnosis 27 41,0.001 of ADPKD, yr Mean age at diagnosis 35 50 0.001 of hypertension, yr Mean age at onset of ESRD, yr 54 73,0.001 7 Patients, n 287 34 Median age at onset of ESRD, yr 53 68,0.001 8 Patients, n 333 291 Median age at diagnosis of 42 56 NA ADPKD caused by symptoms, yr Median age at 53 69,0.001 death/esrd, yr Median age at ESRD, yr 54 74 9 Patients, n 136 60 Median age at hypertension 46 51 NA treatment, yr Median age at stage 3 CKD 50 66 NA onset, yr Median age at ESRD 53 Only 21% had ESRD by NA onset, yr age 70 yr Median age at death, yr 67 71 NA 10 Patients, n 185 34 Findings Patients with PKD1 had significantly larger kidneys and more cysts versus patients with PKD2 mutations; they also had significantly higher frequency of hypertension/higher urinary albumin excretion 11 Patients, n 50 10 Findings Children with PKD1 had significantly larger kidneys with more and larger cysts than children with PKD2 (mean age of 8.6 versus 8.9 yr); they also had significantly higher daytime/nighttime systolic BP by ambulatory monitoring over 24 h NA, not available. ADPKD, median age at onset of ESRD was 4 years younger in men versus women (52 versus 56 years). 7 Becausenoneofthese studies were population-based, the true sex effect remains unknown. Race ADPKD occurs in all races, but differences in severity between races have not been well studied. In one small study, African Americans with ADPKD had significantly earlier onset of ESRD than whites (age 43.2 versus 55.4 years), and African Americans with both ADPKD and sickle cell trait reached ESRD even earlier than African Americans with ADPKD alone (38 versus 48 years). 28 More studies are needed to define the effect of race and ethnicity on ADPKD progression. Parental History of Hypertension Two studies have evaluated the effects of parental hypertension on ADPKD severity in offspring. In a European study of 162 subjects, hypertension in the non-adpkd parent (versus normotension) was associated with ESRD at a significantly younger age in both male and female offspring (49 versus 54 years). 29 Another study from the University of Colorado reported that hypertension in the parent with ADPKD was associated with a higher frequency of hypertension in affected children as well as younger age at diagnosis of hypertension, independent of kidney volume and renal function. 30 Effects on ESRD onset in offspring were not examined. Family History of Early ESRD One study has explored whether family history of ESRD predicts the mutated gene and hence, the risk of disease progression. Barua et al. 31 used a large population of 484 patients with ADPKD with known genotype and renal function data (including egfr and age at ESRD) from 90 ADPKD pedigrees. The investigators found that the presence of one or more affected family members who developed ESRD at age #55 years was highly predictive of the PKD1 genotype (positive predictive value=100%; sensitivity=72%). In contrast, the presence of one or more affected family member who had adequate kidney function or developed ESRD at age $70 years was highly predictive of the PKD2 genotype (positive predictive value=100%; sensitivity=74%). The study provided a simple but effective prognostic tool by investigating carefully the history of renal function in the pedigree under analysis. Limitations of the study included the small population (only 90 pedigrees); the need to have a detailed family history, which limited the use of the approach for small pedigrees or de novo cases; and the J Am Soc Nephrol 25: ccc ccc, 2014 Rapid ADPKD Progression 3

impossibility of predicting age at ESRD in other family members given the overall high intrafamilial variability observed in ADPKD pedigrees. Clinical Factors The major studies reporting clinical and imaging characteristics predictive of rapid renal disease progression are summarized in Table 4. Hyperfiltration in Childhood The effect of glomerular hyperfiltration (creatinine clearance [CrCl]$140 ml/min per 1.73 m 2 ) was evaluated in a cohort of 140 children with ADPKD followed longitudinally for a median of 5.8 years. 32 The subset of 32 children with glomerular hyperfiltration showed a higher rate of growth in TKV (corrected for body surface area) over 5 years and a faster decline in renal function than children without glomerular hyperfiltration. Because the final CrCl measures were similar in the two groups of children, it is unclear whether hyperfiltration translates into an earlier onset of ESRD. Gross Hematuria Early or frequent episodes of gross hematuria may be signs of more severe disease and contributors to renal function deterioration, perhaps by causing AKI and/or chronic iron toxicity. Cyst ruptures with gross hematuria may lead to the release and deposition of free iron and heme, promoting generation of reactive oxygen species and proinflammatory cytokines. 33,34 An observational study of 191 patients with ADPKD found that patients who had one or more episode of gross hematuria had significantly larger kidneys and were more often hypertensive than patients without an episode. 35 Patients who reported more episodes of gross hematuria had worse renal function than patients who had only one episode. Renal survival was significantly worse in 128 patients with ADPKD who had gross hematuria at age,30 years than in 448 patients without gross hematuria or whose first episode occurred at age.30 years (ESRD at 49 versus 59 years). 7 Similar findings were reported in other studies. 36 Although these observations are derived Table 3. Demographic, familial, clinical, and imaging predictors of rapid renal disease progression in ADPKD Predictor/Cohort Size (No. of Patients) Type of Observational Study Major Supporting Refs. Sex 1215 7 580 26 100 Prospective 40 Race 72 28 Hyperfiltration in childhood 180 32 Gross hematuria 1215 7 580 26 191 35 180 36 100 Prospective 40 Pregnancy 235 37 Multiple UTIs 256 39 180 36 94 38 Cyst rupture 100 Prospective 40 Hypertension 2085 70 1215 7 580 26 543 73 323 Prospective 48 235 37 TKV 312 Prospective 54 241 Prospective 4 229 Prospective 55 RBF 241 Prospective 58 59 P values for all studies were,0.01. from retrospective studies, they indicate that gross hematuria is associated with more severe disease. Pregnancy No prospective study examining the effects of pregnancy on ADPKD renal outcome has been reported. An analysis of historical data from 235 women (605 pregnancies), however, showed that pregnancy did not seem to affect renal function in normotensive women, but hypertensive women with more than three pregnancies had significantly worse renal function than ageadjusted women with fewer pregnancies. 37 Survival analysis of 236 women with ADPKD showed significantly shorter renal survival among women with three or more pregnancies. 7 Because the latter analysis relied on historical data, it is unclear whether the shorter renal survival was because of complications of pregnancy, such as development or exacerbation of hypertension and preeclampsia, or other factors. Multiple Urinary Tract Infections Several small retrospective ADPKD studies have suggested that multiple urinary tract infections (UTIs) are associated with a more rapid decline in renal function. An 4 Journal of the American Society of Nephrology J Am Soc Nephrol 25: ccc ccc, 2014

BRIEF REVIEW Table 4. Summary of the most relevant prospective and cross-sectional/retrospective studies where predictors and outcomes of rapid ADPKD progression were analyzed and validated Patients, N Baseline or Major Patient Characteristics Prospective observational studies 241 Age 15 45 yr; CrCl.70 ml/min; other conditions affecting renal function (e.g., diabetes) excluded 196 Mean (SD) age (PKD1 and PKD2): 27614 and 40614 yr 100 Mean (SD) age 31.266.4 yr; mean (SD) estimated CrCl (Cockcroft Gault): 109.8625.5 ml/min 140 Children, mean age 11 yr (range=5 mo to 18 yr) Follow-Up Duration, Other Study Features, or Study Type/Summary Predictors and Outcome (OM) Mean follow-up: 3, 4,57 TKV, RBF, serum copeptin, total cyst 7.9, 55 8.5, 79 and 6 yr 82 volume, genotype, percent cystic volume; OM: decline in egfr (P,0.001) Follow-up: 22 yr Follow-up: 6 mo Follow-up: 2 5 yr 323 Mean (SD) age 53615 yr Mean (SD) follow-up 100638 mo (n=198) 312 Children (,18 yr) Mean follow-up 5.7 yr (range=3 15) for 115 of 185 affected children 229 Mean (SD) age 37611 yr; mean (SD) egfr=71622 ml/min per 1.73 m 2 (range=20 152) 154 Children; mean (SEM) age 9.660.5 yr (range=8 mo to 17 yr) 200 (ADPKD subset of MDRD) Measured GFR=13 55 ml/min per 1.73 m 2 209 (CRISP) Mean age 32.1 yr; mean egfr=89.9 ml/min per 1.73 m 2 19 Postmenopausal women; Cr,2.5 mg/dl; lack of other serious medical conditions Mean follow-up 7.8 yr (range=2.6 15.1) Mean 11 BP measurements, routine echocardiography Secondary subgroup analysis of randomized controlled trial; mean follow-up 2.2 yr Baseline and follow-up measurement at 1, 2, and 3 yr Postmenopausal estrogen use (n=11) versus no use (n=8); follow-up 1 yr (observational) PKD1 and PKD2; OM: age at onset of stage 3 CKD TKV and TCV change detectable at 6 mo Increased cysts and KV; OM: disease severity measured by flank, back, or abdominal pain, HTN, inguinal hernia, palpitation, frequency, maximum urine osmolality, GFR (P,0.05) Age, HTN, urinary stones, proteinuria, smoking; OM: disease progression, egfr decline (P,0.05) Baseline and follow-up KV (mean of two kidneys), cysts; OM: disease severity (P,0.01) Increased TKV; OM: egfr decline (P,0.001) Refs. 4,10,56 59,83,86 (CRISP) HTN; OM: LVMI (P,0.001) 69 Greater scr, proteinuria, HTN, young age, low protein diet; OM: renal function decline NGAL, IL-8; OM: egfr and TKV 93 Postmenopausal estrogen use; OM: liver volume by CT scan (P,0.03) 9 40 43 48 54 55 78 109 J Am Soc Nephrol 25: ccc ccc, 2014 Rapid ADPKD Progression 5

Table 4. Continued Patients, N Baseline or Major Patient Characteristics Randomized interventional studies 72 HTN and LVH; age 20 60 yr; CrCl.30 ml/min per 1.73 m 2 24 Mean age: 41 (amlodipine) to 43 yr (enalapril); CrCl.50 ml/min per 1.73 m 2 46 ADPKD with HTN; age range=18 65 yr; scr#4 mg/dl 8 Mean (SD) age 38610 yr; mean (SD) scr=0.760.1 mg/dl; mean (SD) egfr=113631 ml/min per 1.73 m 2 / retrospective studies 741 Mean (SD) age 53.4614.8 yr; stages 1 4CKD; patients with no molecular data excluded 336 (267 at risk) Mean age: 35.4 (PKD1) and 44.5 yr (PKD2); 42 patients had CrCl,70 ml/min or ESRD 1215 622 alive without ESRD; 205 died before ESRD; patients with no specific event or younger than cutoff age excluded from analysis for specific predictor 624 44% of PKD1 and 29% of PKD2 had reached ESRD 60 Children with PKD1 (n=50) or PKD2 (n=10); patients without molecular data excluded 51 PKD1 families Families with evidence of vascular phenotype in unaffected relatives excluded Follow-Up Duration, Other Study Features, or Study Type/Summary Prospective; strict versus standard BP control; follow-up 7 yr Follow-up 5 yr Double-blind 3-yr trial comparing ramipril with metoprolol Conducted over a 2-wk interval Genotype phenotype Genotype phenotype Genotype phenotype Genotype phenotype Genotype phenotype 234 PKD2 only Genotype phenotype 461 PKD2 only Genotype phenotype Predictors and Outcome (OM) Refs. HTN control; OM: LVM (P,0.01) 49 HTN control; OM: decreased urinary albumin excretion (P,0.05) BP control; OM: renal function, LVM 51 Water intake; OM: urine osmolality,285 mosm/kg PKD1 disease, PKD1 truncating mutation; OM: kidney function (P,0.001) 50 107 PKD1; OM: kidney function (P,0.001) 6 Sex, PKD1, young age at diagnosis, HTN, gross hematuria; OM: renal survival (P,0.001) PKD1; OM: kidney function (P,0.001) 8 PKD1; OM: bilateral renal cysts (P,0.001) PKD1 mutation position 59 to 39; OM: vascular phenotype (P,0.001) PKD2 mutation location; OM: phenotypic score variability (P=0.002) Sex, PKD2 mutation other than splice mutations; OM: renal function (P,0.001) 5 7 11 12 13 14 6 Journal of the American Society of Nephrology J Am Soc Nephrol 25: ccc ccc, 2014

BRIEF REVIEW Table 4. Continued Patients, N Baseline or Major Patient Characteristics Follow-Up Duration, Other Study Features, or Study Type/Summary / retrospective studies 704 PKD1 only; 46% had ESRD Genotype phenotype 155 PKD1 only; mean (SD) age 37.4617.29 yr; ESRD: 31.1%; patients with no molecular data excluded 173 PKD1 only; patients with linkage to PKD2 or no molecular data excluded 294 PKD1 and PKD2; 41% reached ESRD 580 71% alive without ESRD at age 50 yr, 53% at 58 yr, 22% at 70 yr 157 Study A: from CrCl=30 50 ml/min per 1.73 m 2 to ESRD; study B: CrCl=50 60 ml/min per 1.73 m 2, followed for 4 yr Genotype phenotype Genotype phenotype Genotype phenotype Predictors and Outcome (OM) PKD1, modifier genes; OM: phenotype heritability (P,0.05) ACE DD polymorphism; OM: ESRD,50 yr (P=0.02) ENOS Glu298Asp polymorphism in men; OM: age at ESRD (P=0.02) ITGA8 polymorphism; OM: age at ESRD (P=0.03) PKD1, younger age at diagnosis, men, HTN, increased LVM, hepatic cysts in women, three or more pregnancies, gross hematuria, UTIs in men, KV; OM: renal function (P,0.05) Sex, MAP, age; OM: renal function, age at ESRD (P,0.001) 58 ESRD Race; OM: ESRD (P,0.001) 28 162 ESRD (dialysis or renal HTN in nonaffected parent; OM: age 29 transplant) at renal death (P,0.03) 475 Patients mean (SD) HTN in affected parent; OM: HTN in 30 age 40611 yr offspring (P,0.05) 484 Mean age (PKD1 and PKD2): 41 yr (range=2 67) and 56 yr (range=17 88); ESRD: PKD1, n=140 (mean age 48.9 yr) and PKD2, n=24 (mean age 70.2 yr) Family history of ADPKD-related ESRD; OM: mutated gene 31 180 Children; age 4 18 yr with normal renal function 191 Mean (SEM) age 4161 yr (range=17 81) 108 Patients with history of two or more UTIs 235 Mean (SEM) age 4261 yr; 91 women excluded because of pregnancy at study visit, Cr excretion variability, different laboratory source for Cr data, or missing data Glomerular hyperfiltration; OM: total renal volume, renal function (P,0.001) Gross hematuria; OM: total renal volume, renal function (P,0.03) Efficacy of antibiotic prophylaxis in reducing UTIs and gross hematuria and delaying disease progression, gross hematuria at age,30 yr; OM: renal function (P,0.001) Preexisting HTN; OM: maternal complications in pregnancy (P,0.001) Refs. 20 21 23 24 26 27 32 35 36 37 J Am Soc Nephrol 25: ccc ccc, 2014 Rapid ADPKD Progression 7

Table 4. Continued Patients, N Baseline or Major Patient Characteristics / retrospective studies 94 Patients who doubled scr in,36 mo (rapid progressors) or.36 mo (slow progressors) 256 Stages 1 4 CKD; patients with,6 mo follow-up excluded 32 Children; mean (SD) age 12.364.7 yr 62 Children; mean (SD) age 12.364.3 yr Follow-Up Duration, Other Study Features, or Study Type/Summary Predictors and Outcome (OM) HTN, UTI, hematuria, overt proteinuria; OM: renal function (P=0.01) Refs. Recurrent, persistent pyuria; OM: renal function (P=0.01) 39 HTN onset in early-stage ADPKD; 41 OM: renal volume, length (P,0.05) HTN; OM: renal volume (P,0.01) 42 199 Children with ADPKD Very early onset of ADPKD, ADPKD 44 signs, symptoms; OM: KV, HTN, ESRD (P,0.005) 147 Age 16 45 yr; CrCl=75 150 HTN; OM: renal volume (P,0.005) 45 ml/min per 1.73 m 2 134 Mean (SD) age 42.3613.1 yr HTN; OM: clinical severity index 46 (P,0.001) 30 Mean age 32 yr (range=19 45); CrCl.80 ml/min per 1.73 m 2 ; patients with less than six scr measurements or less than three scr measurements in 1 yr excluded HTN; OM: renal function (P,0.001) 47 1044 Studies A and B: age 15 49 and 18 64 yr; egfr$60 and 25 59 ml/min per 1.73 m 2 103 Mean (SD) age 40611 yr; mean (SD) GFR 92636 ml/min 116 Adults; mean (SEM) age 40.761.1 yr 26 Mean (SD) age 26.567.4 yr; normotensive by clinic BP; normal renal function 46 Mean (SD) age 25.968.1 yr; normotensive by clinic BP; normal renal function 18 Mean (SD) age 24.166 yr; mean (SD) scr=74.1610.6 mm/l comparison with 103 age- and sex-matched healthy controls (live kidney donors) Albuminuria, TKV, RBF; OM: renal function (P,0.001) ADPKD status, age quartile; OM: MAP, total renal volume, measured GFR, effective renal plasma flow, renal vascular resistance, filtration fraction (P,0.001) 38 52 (HALT PKD baseline data) HTN; OM: LVH frequency (P,0.05) 62 comparison with 35 age- and sex-matched healthy controls 24-h systolic BP by ambulatory monitoring; OM: LVMI (P=0.01) ADPKD; OM: LVMI, diastolic function (P,0.001) Exercise systolic and diastolic BP; OM: LVMI (P=0.001) 60 63 64 65 8 Journal of the American Society of Nephrology J Am Soc Nephrol 25: ccc ccc, 2014

BRIEF REVIEW Table 4. Continued Patients, N Baseline or Major Patient Characteristics / retrospective studies 20 Mean (SD) age 46.7613.3 yr; mean (SD) scr=112633 mm/l; patients with scr.177 mm/l, cardiovascular complications, diabetes, pulmonary or liver disease, or pregnancy excluded 24 Age range 5.7 24.9 yr; normal renal function 85 Children; mean (SEM) age 1261 yr; normal renal function Follow-Up Duration, Other Study Features, or Study Type/Summary Predictors and Outcome (OM) 24-h systolic and diastolic BP; OM: LVMI (P,0.001) Refs. HTN; OM: LVMI (P,0.002) 67 HTN; OM: LVMI (P,0.001) 68 2085 Mean age 50 yr HTN; OM: mortality (P,0.001) 70 513 Mean ages for women/men: HTN control; OM: time to ESRD, 71 37/38 yr (1985 1992) and 39/36 yr (1992 2001) ADPKD man6htn (P=0.02), ADPKD woman6htn (P=0.02 0.03) 693 Mean age 55.9 yr (1990 1995) HTN control; OM: patient survival, 72 and 60.6 yr (2000 2007) age at ESRD (P,0.001) 543 Age 15 49 yr; GFR$60 ml/min per 1.73 m 2 Systolic BP measured at office, sex; OM: LVMI (P,0.001) 73 (HALT study A baseline data) 31 Mean age 35.8 yr (normotensive) CrCl 39.6 yr (hypertensive), mean CrCl=91 106 ml/min per 1.73 m 2 ; patients with diabetes, cardiovascular disease, and other conditions affecting endothelial function excluded 15 Mean (SD) age 40.1610.3 yr; mean (SD) CrCl=87.8613.7 ml/min per 1.73 m 2 ; patients with diabetes, cardiovascular disease, hypercholesterolemia, or smokers excluded 144 n=52: HTN and egfr,60 ml/min per 1.73 m 2 ; n=50: HTN and egfr$60 ml/min per 1.73 m 2 ; n=42: normotensive and egfr$60 ml/min per 1.73 m 2 ;51 healthy controls 97 Proteinuria,300 and.300 mg/d: mean age 39 and 48 yr, mean CrCl=82 and 37 ml/min per 1.73 m 2 HTN; OM: endothelial-dependent dilation (P,0.05) HTN; OM: endothelial dysfunction (P,0.005) ADPKD versus control; HTN, renal function; OM: peripheral augmentation index, circulating inflammation biomarkers (P,0.05) Overt proteinuria (.300 mg/d), microalbuminuria; OM: renal function (P,0.001) 66 75 76 77 79 J Am Soc Nephrol 25: ccc ccc, 2014 Rapid ADPKD Progression 9

Table 4. Continued Patients, N Baseline or Major Patient Characteristics / retrospective studies 102 Mean (SD) age 40611 yr; mean (SD) measured GFR=77631 ml/min per 1.73 m 2 102 Mean (SD) age 40611 yr; mean (SD) measured GFR=77631 ml/min per 1.73 m 2 71 Mean (SD) age 1664 yr (range=8 26); mean (SD) CrCl=130.2629.4 ml/min per 1.73 m 2 680 Consecutive adult ADPKD patients not in ESRD 91 Adult normotensive ADPKD patients; normal renal function; no other chronic conditions 55 Age 22 79 yr; scr=0.5 11.9 mg/dl (median=1.2); patients with clinical renal infections, on dialysis, or with transplant excluded 144 n=50: HTN and egfr.60 ml/min per 1.73 m 2 ; n=52: HTN and egfr=25 60 ml/min per 1.73 m 2 ; n=42: normotensive and egfr.60 ml/min per 1.73 m 2 ; patients with vascular disease, diabetes, and severe heart failure excluded 100 Mean (SD) age 3166 yr; mean (SD) egfr=94618 ml/min per 1.73 m 2 ; CKD stages 1/2 Test/validation sets: 134/158 Test/validation sets: mean (SD) age 31.466.3/32.468.7 yr, mean (SD) egfr=86.46 15.5/88.1627.8 Follow-Up Duration, Other Study Features, or Study Type/Summary Predictors and Outcome (OM) Plasma copeptin; OM: disease severity by measured GFR, effective RBF, total renal volume, albuminuria (P,0.001) Urinary biomarkers; OM: disease severity by measured GFR, effective RBF, total renal volume (P,0.05) Serum angiogenic growth factors (particularly Ang-1); OM: renal function (P=0.01) Serum uric acid; OM: ESRD hazard ratio (P,0.001) Uric acid, serum ADMA, egfr; OM: endothelial dysfunction (P,0.001) Refs. Urine excretion of MCP-1; OM: scr 89 Disease severity defined by presence or absence of HTN and renal function (normal versus decreased); OM: levels of serum inflammatory and oxidative stress markers (P,0.001) FGF23; OM: increased renal phosphate excretion (P=0.07) Proteomic severity score made of 142 peptides; OM: htkv (P,0.001) 284 Mean (SD) age at ESRD 54.169.9 yr Low birth weight; OM: age at ESRD (P=0.01) 102 Mean (SD) age 39612 yr Coffee consumption; OM: caffeine in ADPKD patients versus controls (P,0.001) 82 84 85 87 88 90 91 92 95 97 10 Journal of the American Society of Nephrology J Am Soc Nephrol 25: ccc ccc, 2014

BRIEF REVIEW Table 4. Continued Patients, N Baseline or Major Patient Characteristics / retrospective studies 582 ESRD; patients with systemic disease involving kidney, immunosuppressive therapy, and age at renal death,21 yr excluded 239 Mean (SEM) age 38.660.9 yr (range=2 80); patients with primary PLD diagnosis excluded Follow-Up Duration, Other Study Features, or Study Type/Summary Predictors and Outcome (OM) Tobacco consumption; OM: ESRD risk (P=0.002) Sex; OM: massive hepatic cystic disease (P,0.04) OM, outcome; TCV, total cyst volume; KV, kidney volume; HTN, hypertension; scr, serum creatinine; Cr, creatinine; CT, computed tomography; LVM, left ventricular mass; ACE DD, angiotensin converting enzyme DD genotype; ENOS, endothelial nitric oxide synthase; ITGA8, integrina-8 gene; MAP, mean arterial pressure; Ang-1, angiotensin-1; ADMA, asymmetric dimethylarginine; MCP-1, monocyte chemoattractant protein-1; FGF23, fibroblast growth factor-23; htkv, height-adjusted TKV; PLD, polycystic liver disease. Refs. 98 108 uncontrolled study in 108 patients with a history of two or more UTIs reported that antibiotic prophylaxis significantly reduced the incidence of infection and limited the lossof renal function compared with no prophylaxis. 36 In another study, patients classified as fast progressors (defined by doubling of serum creatinine in #36 months) had a higher incidence of single and recurrent UTIs than slow progressors. 38 A third retrospective, singlecenter study suggested that asymptomatic pyuria and particularly, overt UTI were associated with faster decline in renal function, but it is unclear whether this result was independent of other factors, such as baseline GFR and TKV. 39 In addition, overt UTI was uncommon in that study, occurring in only 33 of 256 patients during a mean observation period of 81 months. It remains, therefore, uncertain whether UTIs play a significant role in the loss of renal function in ADPKD. Cyst Rupture One study assessed changes in renal volume over a 6-month period by magnetic resonance imaging (MRI) in a cohort of 100 young patients with ADPKD and found direct evidence for cyst rupture, resulting in a decrease in renal volume in 6 patients. 40 Compared with the rest of the study population, these six patients had higher baseline TKV and reported more symptoms (flank pain, macrohematuria, infection, and hypertension), suggesting that cyst rupture is a marker of more severe disease. Hypertension and Kidney Function Hypertension occurs in up to 80% of patients with ADPKD before significant loss of renal function and seems to be both a marker of more severe disease and a contributor to renal function loss. 3 Hypertension is already present in a significant subset of children with ADPKD, and several cross-sectional studies have shown that the prevalence of hypertension in children correlates with kidney size and number of cysts. 41 43 Another observational study of children with ADPKD found that those who were diagnosed by ultrasound at age,18 months had larger kidneys, more cysts, and more frequent hypertension on follow-up than children diagnosed at age.18 months, confirming the relationship between renal structural severity and prevalence of hypertension, even in children with ADPKD. 44 Similarly, cross-sectional analysis of 147 adults with ADPKD revealed that renal volume was significantly greater in patients with hypertension versus normotension among both men and women, whereas serum creatinine and CrCl did not differ. 45 Furthermore, TKV assessed by MRI was significantly higher in young (mean age of 31 years) ADPKD patients with hypertension (n=67) versus normotension (n=33). 40 Another cross-sectional study of 134 patients found a strong and graded association between presence and severity of hypertension and decreased egfr. 46 Early onset of hypertension may lead to earlier onset of ESRD. A survival analysis using data from 506 adults with ADPKD revealed that patients with a diagnosis of hypertension at age,35 years developed ESRD 14 years earlier (at 51 versus 65 years old) than patients who were normotensive until.35 years old. 7 Another small (n=30) retrospective study found that the decline in mean standardized CrCl slope over time was significantly greater in hypertensive versus normotensive patients. 47 Similarly, in an earlier retrospective study of 157 patients with decreased renal function, patients with higher mean arterial pressure had a faster CrCl decline than patients with lower BP. 27 In a prospective observational study of 198 patients with ADPKD followed for a mean of 100 months, hypertension was an independent risk factor of CKD progression. 48 These data show that early-onset hypertension and severity of hypertension are associated with larger J Am Soc Nephrol 25: ccc ccc, 2014 Rapid ADPKD Progression 11

kidneys and progression to ESRD, and thus, they are markers of more severe disease. To date, no randomized clinical trial has been able to show an effect of BP control on progression to ESRD, but these trials were all limited by their small size and short duration. 49 51 The effect of BP control on renal structural and functional disease progression is currently being examined in a large multicenter prospective trial (HALT PKD) involving 1044 patients over 5 8 years; results are expected by the end of 2014. 52 Episodes of AKI Although there is strong evidence in animal models that AKI accelerates the progression of PKD, it has not been studied in humans. Only one case-control study reported that patients with ADPKD hospitalized for pneumonia were significantly more likely to develop AKI (serum creatinine elevation$0.3 mg/dl) than matched patients without ADPKD with similar severity of pneumonia. 53 Renal Volume above Age-Corrected Mean Two longitudinal observational studies in adults and one study in children showed that greater renal volume is predictive of more rapid ADPKD progression. 4,54 56 In the children s study (n=108), those patients with early renal enlargement had a significantly faster rate of renal growth (assessed over a mean followup of 5.7 years) than other children with ADPKD (26 versus 11 ml/kidney per year adjusted for age). 54 In a study of 229 adults with ADPKD, sequential ultrasound examinations performed over a mean interval of 7.8 years showed that initial renal volume as well as renal volume growth rate were correlated with therateofegfrdeclineovertime. 55 The Consortium for Radiologic Imaging Studies in Polycystic Kidney Disease (CRISP) enrolled patients with ADPKD with CrCl.70ml/minandusedMRIto determine TKV. 4,56 Age-adjusted renal volume was inversely related to GFR and albuminuria at baseline. 56 During repeated assessments over 8 years, the strength of the between increase in TKV and decrease in measured (iothalamate clearance) GFR increased. 57 Receiver-operator characteristic curve analysis indicated that a baseline height-adjusted TKV$600 cm 3 /m predicted progression to stage 3 CKD within 8 years with 74% sensitivity and 75% specificity. Decreased Renal Blood Flow In CRISP, mean renal blood flow (RBF) measured by MRI decreased progressively over a 3-year period and preceded the decline in GFR. 58,59 Lower baseline RBF was associated with a larger increase in TKV and total cyst volume over 3 years and a decline in measured GFR. On multivariate analysis, RBF was an independent predictor of renal structural and functional disease progression. 59 Adifferent group of investigators using iothalamate and hippuran clearances Figure 1. Diagram depicting the spectrum of predictors for rapid renal disease progression in ADPKD. Shaded ovals represent the most established predictors. 12 Journal of the American Society of Nephrology J Am Soc Nephrol 25: ccc ccc, 2014

BRIEF REVIEW in a cross-sectional analysis of 103 subjects also reported a marked decrease in effective renal plasma flow, whereas GFR was preserved in young adult patients with ADPKD compared with matched healthy control subjects. 60 Although RBF measurements are currently not available for routine clinical practice, they may be useful as outcome parameters for future interventional studies. The early decrease in RBF is also important for elucidating the pathogenesis of renal function loss in ADPKD. It remains to be determined whether it is mainly because of mechanical compression of blood vessels by the cysts, intrinsic vascular abnormalities caused by PKD mutations, or a secondary effect of oxidative stress and inflammation causing vascular dysfunction. In summary, the best evidence from carefully designed prospective studies and multiple cross-sectional and retrospective studies identifies PKD1 mutations (particularly if truncating), early onset of hypertension, early and multiple episodes of gross hematuria, large kidney volumes, and decrease in RBF orgfratayoungageasthemostreliable predictors of rapid renal disease progression (Figure 1). CARDIOVASCULAR DISEASE PROGRESSION IN ADPKD Hypertension and Left Ventricular Hypertrophy Hypertension is the most commonly reported initial sign of ADPKD and a well recognized risk factor for left ventricular hypertrophy (LVH). 61 In an earlier study, LVH was common in adults with ADPKD, which was shown in an echocardiographic study of 116 consecutive patients, where 41% had LVH at a mean age of only 44 years. 62 Patients with ADPKD with versus without LVH had a significantly higher prevalence (83% versus 61%) and duration(7.5versus3.9years)ofhypertension and worse renal function (serum creatinine of 2.0 versus 1.4 mg/dl). Another study reported higher left ventricular mass index (LVMI) in young normotensive patients with ADPKD than healthy control subjects and found a between 24-hour systolic BP on ambulatory monitoring and LVMI. 63 These young patients also had signs of diastolic dysfunction by echocardiogram and higher BP during exercise stress testing than healthy control subjects. 64,65 Levels of neurohormones were not correlated with LVMI, but 24-hour systolic and diastolic BPs were correlated. 66 These observations highlight the importance of 24-hour BP load on the development of LVH in ADPKD. Early Increase in LVMI Even in children and young adults (ages,25 years) with ADPKD, a significantly higher LVMI than in healthy control subjects has been reported, although most LVMI values were in the normal range. 67 These young adults had higher BP by ambulatory monitoring than control subjects but not children with ADPKD. In another study, children with ADPKD and hypertension or borderline hypertension had significantly higher LVMI than children with ADPKD and normotension. 68 Similarly, children with ADPKD, but not their unaffected siblings, showed a significant between systolic BP and LVMI; children with hypertension had significantly higher LVMI than children with normotension. 69 Children with ADPKD also had a significantly higher incidence of mitral valve prolapse than their unaffected siblings. BP Control There is indirect evidence from three recent epidemiologic studies that BP control improves cardiovascular and renal outcomes. First, a population-based study using the United Kingdom General Practice Research Database (n=2085) found that antihypertensive drug use increased in frequency in patients with ADPKD over the period from 1991 to 2008, particularly for agents blocking the renin-angiotensin system (RAS). 70 After adjusting for potential confounders, mortality decreased over the same time period. Second, a study that compared two ADPKD cohorts (n=513) over the 1985 1992 versus 1992 2001 periods found that patients in the later cohort had lower mean arterial pressure and increased use of angiotensin-converting enzyme inhibitors. 71 Cardiovascular outcomes were not determined, but patients in the later cohort had longer renal and overall survival. Third, a study of patients with ADPKD in Denmark (n=693) found improved survival and an increase in age (by 4.7 years) at onset of ESRD between 1990 and 2007, coinciding with increased use of RAS-blocking drugs. 72 Small randomized trials of hypertension control showed benefits of strict BP control for LVMI and LVH but not renal function. 49,51 In contrast to earlier results, 62 the prevalence of LVH was assessed by cardiac MRI in 543 hypertensive patients with ADPKD (mean age of 36 years) with normal renal function who were participating in the HALT PKD trial. 73 Baseline data revealed that the prevalence of LVH was only 3.9% by nonindexed left ventricular mass and 0.9% by LVMI, despite a mean duration of diagnosed hypertension of 5.8 years. This low prevalence of LVH may be attributed to the fact that BP was well controlled (mean of approximately 124/82 mmhg at baseline) and that 61% of patients were using RAS blockers.thereis,therefore,strong epidemiologic evidence that strict BP control using RAS-blocking drugs ameliorates cardiovascular disease progression in ADPKD. Aneurysm Rupture Intracranial aneurysm (ICA) rupture is one of the most dramatic cardiovascular disease manifestations of ADPKD. The only predictor that has been consistently identified is family history of ICA rupture, which was shown in a retrospective study of 608 adults from 199 ADPKD families. 74 The frequency of ruptured ICAs was significantly higher in patients with ADPKD than unaffected family members, but ruptured ICAs were not randomly distributed; rather, clustering was evident within certain ADPKD families. As mentioned previously, mutations toward the 59 end of PKD1 may be more likely to cause ICA and other vascular complications. 12 J Am Soc Nephrol 25: ccc ccc, 2014 Rapid ADPKD Progression 13

Endothelial Dysfunction Endothelial dysfunction is an early and common finding in patients with ADPKD. In one study, patients with ADPKD with and without hypertension exhibited greater endothelial dysfunction(measured by endothelium-dependent dilation of the brachial artery after transient ischemia) than healthy subjects with and without hypertension, respectively. 75 Carotid intima media thickness was increased in patients with ADPKD and normotension compared with healthy subjects and further increased in patients with ADPKD and hypertension. Other studies have also shown evidence of endothelial dysfunction in early ADPKD before the onset of hypertension and renal dysfunction when measured in the cutaneous microcirculation by laser Doppler examination or peripheral artery tone recordings using finger plethysmography. 76,77 Presence of endothelial dysfunction in early ADPKD likely confers increased atherosclerosis risk, consistent with the observed increase in carotid intima media thickness, and it may contribute to the early decline in RBF described previously. Endothelial dysfunction and carotid intima media thickness are not routinely assessed in Table 5. Laboratory predictors associated with renal and cardiovascular disease progression in ADPKD Predictor Association/Change Refs. Urine predictors Urine-concentrating capacity Reduced in children with.10 versus 43 #10 cysts MCP-1 Increased in ADPKD; associated with 57,89 TKV; predictor of stage 3 CKD Increased in ADPKD; associated 82 with TKV NGAL and IL-8 NGAL increased in ADPKD; associated 82 with GFR, TKV, effective RBF Both increased in ADPKD 93 Kidney injury molecule-1 Increased in ADPKD; associated with 82 GFR, TKV Heart-type-fatty-acid binding protein Increased in ADPKD; associated with 82 GFR, RBF Collagen-derived peptides and Increased in ADPKD 92 other biomarkers Serum or other laboratory predictors HDL cholesterol Decreased HDL levels associated with 78,86 GFR decline Low levels associated with TKV increase 86 Vascular endothelial growth factor In children, correlated with TKV and 85 LVMI; negatively correlated with CrCl Angiopoietin 1 In children, correlated with TKV and 85 LVMI; negatively correlated with urinary protein excretion Serum uric acid Higher level associated with earlier 87 hypertension onset, larger TKV, ESRD risk Higher level associated with endothelial 88 dysfunction FGF23 Increased in ADPKD patients versus 91 controls; associated with increased renal phosphate excretion Inflammation-associated biomarkers (MCP-1 and TNF-a) Overexpression of inflammation factors drives interstitial inflammation and fibrosis 94 MCP-1, monocyte chemoattractant protein-1; FGF23, fibroblast growth factor-23. clinical practice, but they can serve as markers or treatment targets in research settings. ASSOCIATION OF LABORATORY FACTORS WITH RENAL AND CARDIOVASCULAR DISEASE PROGRESSION A summary of the main studies examining laboratory predictors of rapid renal and cardiovascular disease progression is provided in Table 5. Proteinuria and Microalbuminuria The amount of proteinuria correlates with the risk for progression in many renal diseases, including ADPKD. Among 200 patients with ADPKD participating in the Modification of Diet in Renal Disease (MDRD) Study, higher levels of proteinuria were associated with a steeper decline in GFR. 78 Another study of 270 consecutive adults with ADPKD identified overt proteinuria (.300 mg/d) in 48 patients (18%), which was associated with worse renal function, higher BP, and larger renal volume. 79 Patients with versus without overt proteinuria reached a serum creatinine concentration of 1.5 mg/dl at a significantly younger age (44 versus 58 years old). Thesamestudyfoundmicroalbuminuria in 20 of 49 patients with hypertension and LVH but not overt proteinuria. BP, renal volume, and filtration fraction were significantly higher in patients with microalbuminuria. Similarly, in a study of 100 young patients (mean age of 31 years) with ADPKD and preserved renal function, the degree of albuminuria was correlated with TKV and kidney volume growth rate. 40 Among 1044 adult participants in the HALT PKD trials at baseline, urine albumin excretion was positively correlated with TKV and negatively correlated with egfr. 52 Because TKV is usually not available in clinical practice, increased levels of urinary albumin excretion in adults may be a valuable marker of ADPKD severity before a decline in egfr is evident. In children, microalbuminuria is not consistently correlated with disease severity. 44,54,68 14 Journal of the American Society of Nephrology J Am Soc Nephrol 25: ccc ccc, 2014