Clinical Gastroenterology and Hepatology 2014;-:-–-

Liver Involvement in Early Autosomal-Dominant PolycysticKidney Disease

Marie C. Hogan,* Kaleab Abebe,‡ Vicente E. Torres,* Arlene B. Chapman,§ Kyongtae T. Bae,‡

Cheng Tao,‡ Hongliang Sun,‡ Ronald D. Perrone,k Theodore I. Steinman,¶ William Braun,#

Franz T. Winklhofer,** Dana C. Miskulin,k Frederic Rahbari-Oskoui,§ Godela Brosnahan,‡‡

Amirali Masoumi,‡‡ Irina O. Karpov,‡ Susan Spillane,‡ Michael Flessner,§§ Charity G. Moore,‡

and Robert W. Schrier‡‡

*Mayo Clinic College of Medicine, Rochester, Minnesota; ‡University of Pittsburgh School of Medicine, Pittsburgh,Pennsylvania; §Emory University School of Medicine, Atlanta, Georgia; kTufts Medical Center, Boston, Massachusetts;¶Beth Israel Deaconess Medical Center, Boston, Massachusetts; #Cleveland Clinic, Cleveland, Ohio; **University of KansasMedical Center, Kansas City, Kansas; ‡‡University of Colorado Anschutz Medical Campus, Aurora, Colorado; §§NationalInstitute of Diabetes and Digestive Diseases, National Institutes of Health, Bethesda, Maryland

BACKGROUND & AIMS:

Abbreviations used in this papekidney disease; ALP, alkalineferase; BMI, body mass index; Barea (corrected for height); eGHALT-PKD, Halt Progression ofPKD-A, Halt Progression of Poheight-adjusted liver cyst volumhtTKV, height-adjusted total ki

Polycystic liver disease (PLD), the most common extrarenal manifestation of autosomal-dominant polycystic kidney disease (ADPKD), has become more prevalent as a result ofincreased life expectancy, improved renal survival, reduced cardiovascular mortality, and renalreplacement therapy. No studies have fully characterized PLD in large cohorts. We investigatedwhether liver and cyst volumes are associated with volume of the hepatic parenchyma, resultsfrom liver laboratory tests, and patient-reported outcomes.

METHODS:

We performed a cross-sectional analysis of baseline liver volumes, measured by magneticresonance imaging, and their association with demographics, results from liver laboratory andother tests, and quality of life. The data were collected from a randomized, placebo-controlledtrial underway at 7 tertiary-care medical centers to determine whether the combination of anangiotensin I–converting enzyme inhibitor and angiotensin II–receptor blocker was superior tothe inhibitor alone, and whether low blood pressure (<110/75 mm Hg) was superior to stan-dard blood pressure (120–130/70–80 mm Hg), in delaying renal cystic progression in 558patients with ADPKD, stages 1 and 2 chronic kidney disease, and hypertension (age, 15–49 y).

RESULTS:

We found hepatomegaly to be common among patients with ADPKD. Cysts and parenchymacontributed to hepatomegaly. Cysts were more common and liver and cyst volumes weregreater in women, increasing with age. Patients with advanced disease had a relative loss ofliver parenchyma. We observed small abnormalities in results from liver laboratory tests, andthat splenomegaly and hypersplenism were associated with PLD severity. Higher liver volumeswere associated with a lower quality of life.

CONCLUSIONS:

Hepatomegaly is common even in early stage ADPKD and is not accounted for by cysts alone.Parenchymal volumes were larger, compared with liver volumes of patients without ADPKD orwith those predicted by standardized equations, even among patients without cysts. Theseverity of PLD was associated with altered biochemical and hematologic features, as well asquality of life. ClinicalTrials.gov identifier: NCT00283686.

Keywords: HALT-PKD-A; Hepatic Cyst; CKD; MRI Analysis.

r: ADPKD, autosomal-dominant polycysticphosphatase; ALT, alanine aminotrans-P, blood pressure; BSA(m2), body surfaceFR, estimated glomerular filtration rate;Polycystic Kidney Disease studies; HALT-lycystic Kidney Disease Study A; htLCV,e; htLV, height-adjusted total liver volume;dney volume; htLPV, height-adjusted liver

parenchymal volume; LCV, liver cyst volume; LPV, liver parenchymal vol-ume; LV, total liver volume; MRI, magnetic resonance imaging; PLD,polycystic liver disease; QOL, quality of life; SF-36, Short-Form 36.

© 2014 by the AGA Institute1542-3565/$36.00

http://dx.doi.org/10.1016/j.cgh.2014.07.051

2 Hogan et al Clinical Gastroenterology and Hepatology Vol. -, No. -

Autosomal-dominant polycystic kidney disease(ADPKD) is the most common monogenic kidney

disease and the fourth leading cause of end-stage kidneydisease worldwide.1,2 Polycystic liver disease (PLD),arbitrarily defined as the presence of any liver cyst, isits most common extrarenal expression.3,4 Liver cystsoften remain asymptomatic, but some individuals experi-ence chronic manifestations related to progressive liverenlargement, leading to disability and severely impactingquality of life.5 Symptoms of PLD relate to effects fromthe volume of hepatic cysts or complications such ascyst hemorrhage, rupture, or infection. Massive hepato-megaly can cause compression of the adjacent gastroin-testinal tract, vasculature, and diaphragm. Highlysymptomatic PLD has become more common as a resultof reduced cardiovascular mortality and extended renalsurvival and life expectancy on renal replacementtherapy.6–11 Some individuals require interventionsincluding cyst aspiration and/or sclerosis, fenestrationand/or combined liver resection/cyst fenestration, he-patic artery embolization, or, in extreme cases, liver orcombined liver-kidney transplantation. Several interven-tional clinical trials with the objective of delaying theprogression of PLD have been published.12–17

No studies have fully characterized PLD burden inlarge cohorts with ADPKD. This characterization isneeded because its full clinical impact is not well un-derstood. The HALT Progression of Polycystic KidneyDisease Study A (HALT-PKD-A) is the largest clinicalcohort of ADPKD patients with liver and kidney imagingstudied to date and comprised 558 individuals with earlydisease (estimated glomerular filtration rate [eGFR], >60mL/min/1.73 m2) and hypertension, prospectively fol-lowed up for up to 8 years.18,19 The objective of thisstudy was to systematically characterize PLD-associatedmorbidities in this cohort and correlate the extent ofPLD assessed by magnetic resonance imaging (MRI) withrelevant baseline clinical and laboratory data.

Methods

The HALT-PKD-A trial is multicenter, randomized,placebo-controlled trial to investigate the effect ofblockade of the renin-angiotensin-aldosterone system ontotal kidney volume and renal function in adults withADPKD; eligibility criteria and protocol design have beenpublished previously.18 A full description of methods canbe found in the Supplementary Methods.

Results

Total liver volume (LV) (available for 534 individuals)(Figure 1A and B; 267 males; 267 females) was higher inmen (2022 � 740 mL) than in women (1905 � 859 mL)(P ¼ .0003). After height adjustment there was nogender difference: 1115 � 405 mL/m (males) vs 1141 �518 mL/m (females) (Table 1, Figure 1A).

Higher Prevalence and Earlier Development ofLiver Cysts in Women

Liver cyst volume (LCV) (unadjusted for height) wassignificantly higher in women (311 � 763 mL) than inmen (122 � 648 mL) (P < .0001) (Table 1, Figure 1A).Females had a higher liver cyst prevalence (78.7%) thanmales (69.3%) (P ¼ .0137); the age-adjusted odds ratiowas 1.54 (P ¼ .032) (Table 1). Older men and womenhad a higher liver cyst prevalence (44%, 68%, and 84%of women age �24, 25–34, and �35 y, compared with17%, 57%, and 79% of men in the same age groups).Eighty-three percent (20 patients) of those with height-adjusted LCV (htLCV) greater than 700 mL/m werewomen, as compared with 49% (275 of 558) in theentire cohort (Figure 1B).

Both Liver Cyst Volume and Liver ParenchymalVolume Contribute to Hepatomegaly

In both sexes, LVs were significantly larger in theADPKD patients than those reported in a group of po-tential liver donors and for the general population.20 Byusing a standardized equation for Caucasians, LVs alsowere higher than those predicted for the same HALT-PKD ADPKD patients (Table 2).20–22 When adjusted forheight, height-adjusted liver parenchymal volume(htLPV) was higher in men and htLCV was higher inwomen (Table 1). To ascertain whether cyst develop-ment fully accounts for hepatomegaly, we compared liverparenchymal volume (LPV) in the cohort with observedand expected normal LVs in the general population andfound that LPV in our cohort was significantly largerthan the LVs in liver transplant living donors (P <.0001), LVs reported in healthy controls (P < .0001), andpredicted LVs in our participants using a standardizedequation. Even in those without liver cysts, this wassignificant (P < .0001).22

Relative Reduction in Liver ParenchymalVolume in Severe Polycystic Liver Disease

As expected, htLCV correlated positively with height-adjusted total liver volume (htLV), particularly in severePLD (Figure 1A) (females: r ¼ 0.69; males: r ¼ 0.48).After reaching an htLCV threshold of approximately700 mL/m, htLPV does not continue to increase(Figures 1B, 2A, and 2B). Because most of these caseswere women, the trends of htLV, htLCV, and htLPV (inFigure 1B) accounted for the lower LPVs in womencompared with those without liver cysts (1555 � 289 vs1735 � 448 mL; P ¼ .0003) (Table 2). Thus, althoughLPV was normal or larger compared with healthy nor-mals, in the majority of ADPKD patients, relative re-ductions in htLPV occurred in a small subset of patientswith severe PLD (Figure 1B).

Figure 1. Effects of sex on HtLV and relationship of HtLV to HtLPV. (A) Distribution of HtLCVs in relation to HtLVs. A sub-population of women have more severe PLD with higher HtLCVs. (B) Plot of HtLPV and HtLCV vs HtLV. As HtLV increases,HtLPV also increases, but then plateaus. The interrupted diagonal line (slope ¼ 1) denotes the maximum values for HtLPV andHtLCV, indicating that HtLPV closely parallels HtLV for the majority of individuals until HtLV reaches approximately 1800 mL/m(vertical interrupted line). To illustrate this, the patient with liver cysts and mildest disease (lower left) had HtLV of 475 mL/m(mild disease; HtLV<1000 mL/m, vertical interrupted line), HtLCV of 1 mL/m, and HtLPV of 475 mL/m, and the most severecase (points farthest right) had HtLV of 6382 mL/m, HtLCV of 5287 mL/m, and HtLPV of 1095 mL/m, but their HtLPV is nodifferent than less severe PLD cases with lower HtLVs. X and Y axes are log scale in both figures. The correlation between logHtLV and log HtLPV was r ¼ 0.65 (P < .001). Note the second vertical interrupted line (HtLV > 1800 mL/m), the cut-off point forHtLV that delineated group 3 or the group with the highest htLVs (Table 3). To the right of the graph are 2 additional verticallines delineating HtLVs �2 � those of normal female (1� ¼ 1858 mL/m) and male (1� ¼ 1986 mL/m) controls.

- 2014 Liver Disease Burden in Early ADPKD 3

Classification of Polycystic Liver Disease IntoMild, Moderate, and Severe

We used the data shown in Figure 1B to classify theextent of PLD into mild, moderate, and severe. We

arbitrarily defined mild PLD as htLV less than 1000mL/m, and considered htLV between 1000 and 1800mL/m to represent moderate PLD. We defined severePLD as htLV greater than 1800 mL/m (women, 81%),concordant with the plateauing of htLPV at this cut-off

Table 1. Baseline Characteristics of the Study Population

Variable

Male (N ¼ 283; 51%) Female (N ¼ 275; 49%)

P valuen Mean � SD n Mean � SD

Vital statisticsAge, y 283 35.7 � 8.1 275 37.6 � 8.4 .006a

Weight, kg 275 90.5 � 15.5 273 74.3 � 16.4 <.0001a

Height, cm 275 181.0 � 7.8 271 166.5 � 6.5 <.0001a

BMI, kg/m2 274 27.7 � 4.5 271 26.8 � 5.7 .044a

BSA, m2 274 2.1 � 0.2 271 1.8 � 0.2 <.0001a

SBP home 192 126.6 � 9.3 180 121.3 � 8.9 <.0001a

DBP home 192 82.2 � 8.1 180 83.0 � 7.4 .2956Image analysis

Patients with liver cysts, % 185 69.3 210 78.7 .0137a

LV, mL 267 2021.8 � 748.5 267 1904.6 � 859.5 .0003a,b

htLV, mL/m 259 1115.2 � 405.5 263 1141.2 � 517.6 .7251b

LCV, mL 267 121.6 � 647.6 267 310.8 � 763 <.0001a,b

htLCV, mL/m 259 64.6 � 351.3 263 184.3 � 459.3 <.0001a,b

LPV 267 1900.2 � 343.9 267 1593.8 � 336.7 <.0001a,b

htLPV, mL/m 259 1050.6 � 184.5 263 956.9 � 196.8 <.0001a,b

TKV, mL 269 1403.7 � 752.6 269 1014.6 � 618.8 <.0001a,b

htTKV 261 773 � 408 265 610 � 368 <.0001a,b

Spleen volume, mL 273 339.7 � 129.2 272 252.6 � 91.4 <.0001a,b

htSV, mL/m 265 187.3 � 69.1 268 151.3 � 53.7 <.0001a,b

LaboratoryeGFR, mL/min 282 90.4 � 17.8 275 92.7 � 17.1 .12Serum albumin level, mg/dL 280 4.3 � 0.7 274 4.1 � 0.5 <.0001a,b

Serum sodium level, mEq/L 283 139.3 � 2.1 275 138.6 � 7.8 .1371Hemoglobin level, g/dL 281 15.0 � 2.1 270 13.8 � 7.1 .005a

WBC, cells/mL 281 6.24 � 1.8 271 6.6 � 5.9 .295Platelets, cells/mL 281 248.2 � 58.5 272 260.4 � 60.4 .017a

AST level, u/L 282 26.4 � 13.1 275 21.8 � 6.9 <.0001a

ALT level, u/L 282 29.3 � 17 275 19.1 � 9.0 <.0001a

Alkaline phosphatase level, u/L 282 64.7 � 28.7 275 54.5 � 17.3 <.0001a

Bilirubin level, mg/dL 282 0.85 � 0.41 275 0.66 � 0.3 <.0001a

SF-36 QOLPhysical functioning 278 91.5 � 18.2 270 90.5 � 17.5 .49Physical role 277 91.6 � 19.9 269 88.3 � 21.5 .058Bodily pain 277 80.8 � 20.5 268 74.4 � 22.3 .0006a

General health 278 67.0 � 19.7 270 64.5 � 20.0 .13Vitality 277 64.7 � 18.2 269 59.8 � 20.9 .003a

Social functioning 277 90.3 � 2 269 85.5 � 21.6 .004a

Role emotional 276 91.9 � 2 267 88.6 � 20.6 .04a

Mental health 278 91.5 � 18.2 270 90.5 � 17.5 .49Standardized mental component 277 51.7 � 7.9 268 50 � 9.8 .021a

Standardized physical component 277 52.8 � 7.3 268 51.5 � 7.8 .0503

AST, aspartate aminotransferase; DBP home, home diastolic blood pressure; htSV, height-adjusted spleen volume; SBP home, home systolic blood pressure; SD,standard deviation; WBC, peripheral white blood cell count.aSignificant P value.bP value calculated using log transformed variables.

4 Hogan et al Clinical Gastroenterology and Hepatology Vol. -, No. -

point, which corresponds to htLCV greater thanapproximately 700 mL/m. Table 3 compares de-mographic, laboratory, and imaging characteristics ofpatients according to PLD severity. Alternatively, arbi-trary htLCV cut-off values can be used to define disease(Supplementary Table 5). Patients with severe PLDwere older (women in their fourth decade), with alower platelet count, albumin level, eGFR, quality of life,and higher alkaline phosphatase (ALP) level, alanineaminotransferase (ALT) level, and height-adjustedspleen volume (Figure 2C).

Determinants of Total Liver Volume, Liver CystVolume, and Liver Parenchymal Volume

In bivariate analyses, age, eGFR, height-adjusted totalkidney volume (htTKV) (women only), body weight, BSAand BMI, blood pressure (BP), and caffeine intake wasassociated with htLV (Supplementary Table 1 and datanot shown). Variables associated with htLCV weresimilar, including age, eGFR, htTKV (women only), bodyweight, BSA, and BMI (women only) (SupplementaryTable 1). Baseline variables associated with htLPV were

Table 2. Comparison of Liver and Liver Parenchymal Volumes and Predicted Liver Volumes Using a Standardized EquationVersus Liver Volumes for Healthy Controls

Liver volume, mL Liver parenchymal volume, mL

Normal living livertransplant donors

HALT(predicted)

HALT(observed)

HALT (allpatients)

HALT (patientswith cysts)

HALT (patientswithout cysts)

MenMean � SD, n 1772 � 393 (11) 1875 � 238 (274) 2022 � 749 (267) 1900 � 344 (267) 1892 � 331 (185) 1918 � 372 (82)P value vs

normal donors———— ———— <.0001a <.0001a <.0001a <.0001a

P value vspredicted LV

———— ———— .0004b .6046b .7931b .5969b

WomenMean � SD, n 1495 � 357 (21) 1511 � 235 (271) 1905 � 860 (267) 1594 � 337 (267) 1555 � 289 (210) 1735 � 448 (57)P value vs

normal donors———— ———— <.0001a <.0001a <.0001a <.0001a

P value vspredicted LV

———— ———— <.0001b <.0001b .0001b <.0663b

NOTE. To determine predicted LV (reported as “HALT predicted”) we used the Vauthey formula, a validated equation for the estimation of standard liver volume inthe Caucasian population (LV in mL ¼ -794.41 þ 1267.28 � BSA in m2).22 Note values in this table are unadjusted for height to permit between-group com-parisons. Normal liver transplant donor MRI-derived LVs were obtained at the University of Pittsburgh from patients considering liver donation. Similar results wereobtained using LVs from a computerized tomography-based study from another published healthy population (males, 1710 � 288 mL (n ¼ 41) and females, 1411� 263 mL (n ¼ 36)).20aOne-sample t test compared with healthy controls.bPaired t test compared with predicted LV.

- 2014 Liver Disease Burden in Early ADPKD 5

body weight, BSA, BMI, and BPs (SupplementaryTable 1). In multivariable analyses (not shown), sex(P ¼ .004), age (P ¼ .001), weight (P < .001), and systolicBP (P ¼ .010) were associated independently with htLV;sex (P < .001), age (P < .001), and lower eGFR(P ¼ .046) were associated with htLCV; weight(P < .001) and systolic BP (P ¼ .002) were associatedindependently with htLPV.

Figure 2.MR images. (A)Hepatomegaly (TLV 2677mL) with a larger contribu-tion of LPV (2576 mL)compared with LCV (101mL). (B) Hepatomegaly(11,834 mL) with a largercontribution of LCV (9806mL) compared with LPV(2028 mL). (C) Splenomeg-aly (601 mL) associatedwith severe PLD (3388 mL)with LCV (1044 mL). (D)Splenomegaly (542 mL)associated with moderatePLD (LV 2082 mL, LCV 110mL). (E) Pancreatic cyst(arrow) (diameter, 21.8mm). (F) Splenic cysts (ar-row) (diameter, 12.2 mm).

Subtle Liver Laboratory Test Abnormalities inWomen With Polycystic Liver Disease

Liver laboratory test results were mostly within thenormal range (Table 1). Only 2.3%, 2.7%, and 4.5% ofpatients had levels of serum ALP, aspartate amino-transferase, and ALT above the upper limit of laboratorynormal, respectively, and 5.1% had a serum albumin

Table 3. Demographics, Laboratory Values, Image Analysis, and SF-36 by htLV Category: Mild, Moderate, and SeverePolycystic Liver Disease

Variable

htLV (1)<1000 mL/m(mild PLD)(n ¼ 230)

htLV (2) 1000–1800 mL/m(moderate PLD) (n ¼ 264)

htLV (3) >1800 mL/m(severe PLD)

(n ¼ 28) P value

Vital statisticsAge, mean � SD, y 35.2 � 9.0 37.2 � 7.4 42.5 � 4.5 <.0001a

1 vs 2, .01601 vs 3, <.00012 vs 3, .0026

Sex, n (% female) 123 (43.16%) 106 (54.36%) 34 (80.95%) <.0001a

1 vs 2, .01581 vs 3, <.00012 vs 3, .0015

Laboratory values, mean � SDeGFR 95.5 � 17.8 89.1 � 16.7 84.2 � 16.7 <.0001a

1 vs 2, .00101 vs 3, .0029

Serum albumin level, mg/dL, n 4.3 � 0.8 (229) 4.1 � 0.4 (262) 4.0 � 0.7 .0042a

1 vs 2, .01351 vs 3, .046

Bilirubin level, mg/dL 0.8 � 0.4 0.8 � 0.3 0.6 � 0.2 .1451Alkaline phosphatase, U/L level 57.8 � 22.2 59.9 � 26.5 74.0 � 22.7 .0042a

1 vs 3, .00282 vs 3, .0108

AST level, u/L 23.2 � 8.8 24.8 � 12.6 26.4 � 8.3 .1531ALT level, u/L 21.1 � 13.9 27.4 � 15.3 26.2 � 11.4 <.0001a

1 vs 2, <.0001s Sodium level, mEq/L 139.2 � 2.2 138.6 � 8.0 140.1 � 2.0 .2638Hemoglobin, g/dL, n 14.0 � 2.5 (226) 14.8 � 7.2 (262) 13.7 � 0.8 .2072Platelets, cells/uL, n 254.3 � 57.5 (227) 259.7 � 58.9 (263) 219.2 � 45.6 .0020a

1 vs 3, .00702 vs 3, .0013

WCC, cells/uL, n 6.5 � 6.4 (226) 6.4 � 1.8 (262) 6.4 � 1.7 .9372Image analysis, mean � SD

LCV, mL 33.4 � 73.6 140.8 � 263.3 2358.9 � 2004.8 <.0001a,b

1 vs 2, <.00011 vs 3, <.00012 vs 3, <.0001

htLCV, mL/m, n 19.8 � 44.3 (230) 82.4 � 156.0 1389.9 � 1127.4 <.0001a,b

1 vs 2, <.00011 vs 3, <.00012 vs 3, <.0001

TKV, mL, n 1197.3 � 748.7 (226) 1204.1 � 672.9 (259) 1303.6 � 744.8 .4181b

htTKV, mL/m, n 685.9 � 412.3 (226) 683.6 � 370.5 (259) 766.0 � 430.6 .3409b

LPV, mL 1488.7 � 205.4 1948.4 � 329.2 1972.8 � 517.7 <.0001a,b

1 vs 2, <.00011 vs 3, <.0001

htLPV, mL/m 858.9 � 92.2 1112.0 � 165.2 1165.5 � 292.0 <.0001a,b

1 vs 2, <.00011 vs 3, <.0001

Spleen volume, mL, n 255.4 � 97.9 (224) 328.5 � 125.8 318.0 � 126.7 <.0001a,b

1 vs 2, <.00011 vs 3, .0148

htSV, mL/m, n 147.0 � 52.5 (224) 186.5 � 66.3 186.7 � 70.6 <.0001a,b

1 vs 2, <.00011 vs 3, .0032

SF-36 QOL, mean � SDPhysical functioning, n 90.4 � 19.3 (226) 91.8 � 15.7 (260) 89.8 � 18.1 .6325Physical role, n 92.0 � 17.0 (225) 88.8 � 23.2 (259) 82.4 � 22.3 .0358a

Bodily pain, n 78.4 � 21.5 (224) 77.8 � 21.7 (259) 68.9 � 23.5 .0881General health, n 66.9 � 20.4 (226) 65.0 � 19.5 (260) 59.8 � 20.4 .1699Vitality, n 63.1 � 19.5 (225) 61.7 � 20.3 (259) 58.2 � 22.8 .4266Social functioning, n 88.2 � 19.6 (225) 88.0 � 20.4 (259) 80.8 � 21.4 .1781

6 Hogan et al Clinical Gastroenterology and Hepatology Vol. -, No. -

Table 3. Continued

Variable

htLV (1)<1000 mL/m(mild PLD)(n ¼ 230)

htLV (2) 1000–1800 mL/m(moderate PLD) (n ¼ 264)

htLV (3) >1800 mL/m(severe PLD)

(n ¼ 28) P value

Role emotional, n 90.2 � 18.4 (223) 90.4 � 19.8 (258) 84.5 � 20.5 .3040Mental health, n 77.8 � 15.8 (225) 77.3 � 14.5 (259) 74.5 � 12.4 .5419Standardized mental component, n 51.0 � 9.6 (224) 50.7 � 8.6 (259) 48.4 � 9.4 .3542Standardized physical component, n 52.5 � 7.3 (224) 52.1 � 7.4 (259) 49.8 � 8.4 .1905

NOTE. Each column represents mild (1), moderate (2), and severe (3) PLD. This permits comparison of the groups in column 4.AST, aspartate aminotransferase; htSV, height-adjusted spleen volume; WCC, white cell count.aSignificant P value. For quality-of-life data, despite an overall P < .05, there were still no statistically significant pairwise comparisons values (probably owing tothe marginal P value ¼ .0358).bP value was calculated using log-transformed variables.

- 2014 Liver Disease Burden in Early ADPKD 7

concentration below the lower limit of normal. Serumalbumin and blood platelet levels were lower, whereasserum ALP level, spleen volume, and physical role werehigher in the severe PLD group (Table 3). In both sexes,ALT level correlated with htLV and htLPV (Table 3,Supplementary Table 2, and data not shown). ALP level(both sexes) and aspartate aminotransferase level(women only) correlated positively with htLV (P ¼ .0042and P ¼ .0471, respectively), whereas serum albuminlevel correlated negatively with htLV (P ¼ .0042) andhtLPV (P < .0001) (Supplementary Table 2). Correlationsbetween ALP, ALT, or serum albumin level and htLV inwomen remained significant after adjusting for age (P ¼.001, P ¼ .001, and P ¼ .003, respectively). In men, thecorrelation between ALT and htLV or htLPV persistedafter adjusting for age (P ¼ .006).

Total Liver Volume, Quality of Life, and Pain

Scores on several health-related quality of life ques-tionnaire subdomains (Short-Form 36 [SF-36]) includingphysical role, general health, vitality, mental health, andstandardized component, were lower in womencompared with men (Table 1). There was a weak asso-ciation between htLV and SF-36 role-physical (P ¼.0017), and the physical component of the health sum-mary score in both sexes (P ¼ .0053) with general heath(P ¼ .0048), vitality (P ¼ .0359), and mental health (P ¼.0450) in men (Table 3). When analyzed by htLVseverity, there was an association with decreased phys-ical role (P ¼ .0358) (Table 3). The associations betweenquality of life (QOL) parameters and htLV did not persistafter adjusting for age, BMI, and TKV. When we consid-ered additive effects of height-adjusted spleen volume þhtLV þ htTKV (data not shown), we did not see anycorrelation with QOL measurements in the multivariateanalysis, after adjusting for age and BMI for either sex.There were no significant differences across each sub-domain according to LCV burden (SupplementaryTable 5).

Mild Splenomegaly and Hypersplenism

See the Supplementary Results section for moredetail.

Impact of Estrogen Use and PregnancyNumber on Polycystic Liver Disease Severity

See the Supplementary Results section for moredetail.

Prevalence of Pancreatic and SplenicCysts in Autosomal-Dominant PolycysticKidney Disease

See the Supplementary Results section for moredetail.

Discussion

Despite its significance in ADPKD, few studies haveinvestigated the impact of PLD.23–26 HALT-PKD-A pro-vides an opportunity to correlate MR imaging with clin-ical, laboratory, and other parameters in a large cohort ofpatients with ADPKD and chronic kidney disease stages 1and 2. Patient selection criteria did not include thepresence of or the extent of PLD.18 Therefore, althoughthis cohort represents a highly selected ADPKD popula-tion, recruitment was unaffected by biases favoring in-clusion or exclusion of patients with severe PLD. Beyondconfirming the recognized higher prevalence andseverity of PLD in women compared with men, this studyshows a number of novel observations that advance theunderstanding of PLD.26 Observed LVs (in both malesand females) were significantly higher than in the gen-eral population, and, in women, LPVs were significantlylarger than LVs predicted for the same patients. Thefinding of increased LPV compared with normals in-dicates that hepatomegaly cannot be accounted forentirely by macroscopic cysts detectable by MRI. It is

8 Hogan et al Clinical Gastroenterology and Hepatology Vol. -, No. -

possible that microscopic cysts or bile duct proliferationcontribute to LPV.27,28 Alternatively, hepatocyte hyper-plasia or hypertrophy also may contribute to hepato-megaly.29 Indeed, this novel observation is consistentwith the recent description of hepatocyte hypertrophy ina mouse model of ADPKD and PLD generated by theinsertion of a human PKD1 mutation, and the demon-stration that polycystin 1 and polycystin 2 are expressedin hepatocytes, where their reduced expression/functionmay cause the hepatocyte hypertrophy observed byHopp et al.30 Interestingly, many cytokine and growthfactors implicated in hepatic regeneration after livermass reduction are up-regulated in PLD.30,32–36

A previous small study showed that hepatic paren-chyma and function remain normal even in the presenceof hepatomegaly.26 Our study shows a relative reductionin LPV in patients with severe PLD. Furthermore, hepa-tomegaly is accompanied by subtle changes in liver lab-oratory tests (reflected in positive correlations with liverenzyme levels and negative correlations with serum al-bumin levels) and mild hypersplenism (reflected in thepositive correlation with spleen volume and negativecorrelation with platelet count). Therefore, mild in-creases in liver laboratory tests and mild thrombocyto-penia are associated with hepatomegaly in PLD.28

Although the association of pancreatic cysts is wellrecognized, splenic cysts have been reported only anec-dotally.31–33 In the current study, the prevalence ofsplenic cysts was 2.7%, but their origin (epithelial,lymphatic, or other) is unknown. Mild common bile ductdilatation occurs frequently in ADPKD.34 Common bileduct diameter correlated positively with htLV and htLPV,possibly owing to shared pathogenic mechanisms.Awareness of this association is important to avoid un-necessary invasive diagnostic procedures.

The correlation between systolic BP and htLPV wasunexpected. Angiotensinogen is synthesized in both liverand proximal renal tubules, and overexpression at eithersite might increase BP, potentially accounting for thiscorrelation. Hormone overproduction at either site isassociated with hypertension in mouse transgenicmodels, and there is some basis for this underlyingmechanism in human beings.35,36

Another notable finding was the significant impact ofLVonquality of life. Consistentwith greater PLDseverity inwomen, the impact also was more pronounced in females.Although life-threatening hepatic dysfunction is rare inADPKD, the effect of massive organ enlargement per se onQOL provides a strong rationale for medical interventionsto retard liver growth. Liver cyst prevalence and LVs werehigher in women with a history of pregnancy comparedwith nulliparous women, regardless of prior birth controlpill use. This difference, however, disappeared afteradjusting for age. Therefore, we could not show an inde-pendent effect of pregnancy or exogenous estrogens on LV,likely owing to the relatively young age and low PLDseverity of the patients in our study and the lower estrogencontent of contemporary contraceptives.37,38

The strengths of this study were the cohort design,large size, prospective data collection, including QOLassessments, high-quality imaging, and analysis. LCV ismore difficult than total LV measurement.39

There were several study limitations. Cases wereselected for early chronic kidney disease and likelyunderestimated the burden of PLD because the HALT-PKD-A cohort was limited by its narrow age inclusioncriteria (age, 15–49 y). In fact, the patients in this studywere more likely to represent the more general PLDpopulation seeking specialty care (for hypertensionmanagement, disease-specific complications, and coun-seling) rather than individuals with severe symptomaticPLD and hepatomegaly or other liver-specific complica-tions, which tend to be over-represented in the medicalliterature and often are referred to gastroenterologists.Additional clinical information will emerge from a lon-gitudinal analysis of this cohort. Because MR protocolsdid not use gadolinium, the portal veins, hepatic veins,and arteries could not be assessed accurately.

Although we do not have a clear explanation for theobserved reductions in QOL, they most likely were owingto the impact of progressive liver mass effects, which aremore frequent and severe in women than in men, as seenin other studies.25,40 The SF-36 questionnaire has notbeen validated in PLD and better patient-reportedoutcome tools are needed.

In summary, we have used data from the HALT-PKD-A cohort to better define PLD burden. Previous studieshave proven MRI to be a useful tool in assessing kidneyvolume in ADPKD, and kidney volume is being consid-ered as a biomarker of disease severity. We show therelationship between MRI-measured LVs with baselinebiochemical/hepatologic perturbations and QOL, thusshowing their potential for validation as outcome mea-sures for clinical trials in PLD. Over the course of thisstudy, sequential longitudinal imaging will provide livergrowth rates and a deeper understanding of the diseaseseverity spectrum in PLD and its complications.

Supplementary Material

Note: To access the supplementary material accom-panying this article, visit the online version of ClinicalGastroenterology and Hepatology at www.cghjournal.org,and at http://dx.doi.org/10.1016/j.cgh.2014.07.051.

References

1. Gabow PA. Autosomal dominant polycystic kidney disease.

N Engl J Med 1993;329:332–342.

2. US Renal Data System, USRDS 2013 Annual Data Report: Atlasof Chronic Kidney Disease and End-Stage Renal Disease in theUnited States, National Institutes of Health, National Institute ofDiabetes and Digestive and Kidney Diseases, Bethesda, MD,2013. Accessed: May 8, 2013.

3. Everson GT, Taylor MR, Doctor RB. Polycystic disease of theliver. Hepatology 2004;40:774–782.

- 2014 Liver Disease Burden in Early ADPKD 9

4. Comfort MW, Gray HK, Dahlin DC, et al. Polycystic disease ofthe liver: a study of 24 cases. Gastroenterology 1952;20:60–78.

5. Chauveau D, Fakhouri F, Grunfeld JP. Liver involvement inautosomal-dominant polycystic kidney disease: therapeuticdilemma. J Am Soc Nephrol 2000;11:1767–1775.

6. Everson GT. Hepatic cysts in autosomal dominant polycystickidney disease. Mayo Clin Proc 1990;65:1020–1025.

7. Patch C, Charlton J, Roderick PJ, et al. Use of antihypertensivemedications and mortality of patients with autosomal dominantpolycystic kidney disease: a population-based study. Am JKidney Dis 2011;57:856–862.

8. Orskov B, Romming Sorensen V, Feldt-Rasmussen B, et al.Improved prognosis in patients with autosomal dominant poly-cystic kidney disease in Denmark. Clin J Am Soc Nephrol 2010;5:2034–2039.

9. Orskov B, Sorensen VR, Feldt-Rasmussen B, et al. Changes incauses of death and risk of cancer in Danish patients withautosomal dominant polycystic kidney disease and end-stagerenal disease. Nephrol Dial Transplant 2012;27:1607–1613.

10. Perrone RD, Abebe KZ, Schrier RW, et al. Cardiac magneticresonance assessment of left ventricular mass in autosomaldominant polycystic kidney disease. Clin J Am Soc Nephrol2011;6:2508–2515.

11. Ecder T, Edelstein CL, Chapman AB, et al. Reversal of leftventricular hypertrophy with angiotensin converting enzyme in-hibition in hypertensive patients with autosomal dominantpolycystic kidney disease. Nephrol Dial Transplant 1999;14:1113–1116.

12. Keimpema LV, Nevens F, Vanslembrouck R, et al. Lanreotide re-duces the volume of polycystic liver: a randomized, double-blind,placebo-controlled trial. Gastroenterology 2009;137:1661–1668,e1–2.

13. Caroli A, Antiga L, Cafaro M, et al. Reducing polycystic livervolume in ADPKD: effects of somatostatin analogue octreotide.Clin J Am Soc Nephrol 2010;5:783–789.

14. Hogan MC, Masyuk TV, Page LJ, et al. Randomized clinical trialof long-acting somatostatin for autosomal dominant polycystickidney and liver disease. J Am Soc Nephrol 2010;21:1052–1061.

15. Hogan MC, Masyuk TV, Page L, et al. Somatostatin analogtherapy for severe polycystic liver disease: results after 2 years.Nephrol Dial Transplant 2012;27:3532–3539.

16. Gevers TJG, IntHout J, Caroli A, et al. Young women withpolycystic liver disease respond best to somatostatin ana-logues: a pooled analysis of individual patient data. Gastroen-terology 2013;145:357–365, e1–2.

17. Chrispijn M, Gevers TJ, Hol JC, et al. Everolimus does notfurther reduce polycystic liver volume when added to longacting octreotide: results from a randomized controlled trial.J Hepatol 2013;59:153–159.

18. Chapman AB, Torres VE, Perrone RD, et al. The HALT polycystickidney disease trials: design and implementation. Clin J Am SocNephrol 2010;5:102–109.

19. Torres VE, Chapman AB, Perrone RD, et al. Analysis of baselineparameters in the HALT polycystic kidney disease trials. KidneyInt 2012;81:577–585.

20. Geraghty EM, Boone JM, McGahan JP, et al. Normal organvolume assessment from abdominal CT. Abdom Imaging 2004;29:482–490.

21. Heinemann A, Wischhusen F, Püschel K, et al. Standard livervolume in the Caucasian population. Liver Transplant Surg 1999;5:366–368.

22. Vauthey JN, Abdalla EK, Doherty DA, et al. Body surface areaand body weight predict total liver volume in Western adults.Liver Transpl 2002;8:233–240.

23. Perrone RD, Ruthazer R, Terrin NC. Survival after end-stagerenal disease in autosomal dominant polycystic kidney dis-ease: contribution of extrarenal complications to mortality. Am JKidney Dis 2001;38:777–784.

24. Bae KT, Zhu F, Chapman AB, et al. Magnetic resonance imagingevaluation of hepatic cysts in early autosomal-dominant poly-cystic kidney disease: the Consortium for Radiologic ImagingStudies of Polycystic Kidney Disease cohort. Clin J Am SocNephrol 2006;1:64–69.

25. Schnelldorfer T, Torres VE, Zakaria S, et al. Polycystic liverdisease: a critical appraisal of hepatic resection, cyst fenestra-tion, and liver transplantation. Ann Surg 2009;250:112–118.

26. Everson GT, Scherzinger A, Berger-Leff N, et al. Polycystic liverdisease: quantitation of parenchymal and cyst volumes fromcomputed tomography images and clinical correlates of hepaticcysts. Hepatology 1988;8:1627–1634.

27. Itai Y, Ebihara R, Eguchi N, et al. Hepatobiliary cysts in patientswith autosomal dominant polycystic kidney disease: preva-lence and CT findings. AJR Am J Roentgenol 1995;164:339–342.

28. Gupta S, Seith A, Dhiman RK, et al. CT of liver cysts in patientswith autosomal dominant polycystic kidney disease. Acta Radiol1999;40:444–448.

29. Nagy P, Teramoto T, Factor VM, et al. Reconstitution of livermass via cellular hypertrophy in the rat. Hepatology 2001;33:339–345.

30. Hopp K, Hommerding CJ, Wang X, et al. Tolvaptan plus pasir-eotide shows enhanced efficacy in a PKD1 model. J Am SocNephrol 2014:Epub ahead of print.

31. Torra R, Nicolau C, Badenas C, et al. Ultrasonographic study ofpancreatic cysts in autosomal dominant polycystic kidney dis-ease. Clin Nephrol 1997;47:19–22.

32. Malka D, Hammel P, Vilgrain V, et al. Chronic obstructivepancreatitis due to a pancreatic cyst in a patient with autosomaldominant polycystic kidney disease. Gut 1998;42:131–134.

33. Mosetti MA, Leonardou P, Motohara T, et al. Autosomaldominant polycystic kidney disease: MR imaging evaluationusing current techniques. J Magn Reson Imaging 2003;18:210–215.

34. Ishikawa I, Chikamoto E, Nakamura M, et al. High incidenceof common bile duct dilatation in autosomal dominant poly-cystic kidney disease patients. Am J Kidney Dis 1996;27:321–326.

35. Gordon DB. The role of renin substrate in hypertension. Hy-pertension 1983;5:353–362.

36. Ramkumar N, Stuart D, Ying J, et al. A possible interactionbetween systemic and renal angiotensinogen in the control ofblood pressure. Am J Hypertens 2013;26:473–480.

37. Gabow PA, Johnson AM, Kaehny WD, et al. Risk factors for thedevelopment of hepatic cysts in autosomal dominant polycystickidney disease. Hepatology 1990;11:1033–1037.

38. Sherstha R, McKinley C, Russ P, et al. Postmenopausal estro-gen therapy selectively stimulates hepatic enlargement inwomen with autosomal dominant polycystic kidney disease.Hepatology 1997;26:1282–1286.

39. Zollner FG, Svarstad E, Munthe-Kaas AZ, et al. Assessment ofkidney volumes from MRI: acquisition and segmentation tech-niques. AJR Am J Roentgenol 2012;199:1060–1069.

10 Hogan et al Clinical Gastroenterology and Hepatology Vol. -, No. -

40. Wijnands TF, Neijenhuis MK, Kievit W, et al. Evaluating health-related quality of life in patients with polycystic liver diseaseand determining the impact of symptoms and liver volume. LiverInt 2014;34:1578–1583.

Reprint requestsAddress requests for reprints to: Marie C. Hogan, MD, PhD, Division ofNephrology, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905.e-mail: hogan.marie@mayo.edu; fax: (507) 266-7891.

AcknowledgmentsThe authors are grateful to all of the Halt Progression of Polycystic KidneyDisease study patients across the United States who participate in theseongoing studies. The authors thank Kristi J. Simmons, Research and Aca-demic Support Services, Mayo Clinic, for editing and proofing of themanuscript.

Conflicts of interestThese authors disclose the following: Marie Hogan and Vicente Torres areconducting a clinical trial sponsored by Novartis in polycystic liver disease;Marie Hogan receives funding support from Novartis; Vicente Torres receivesresearch funding from Otsuka Pharmaceuticals; Robert Schrier sits on theadvisory board of Otsuka, Janssen Pharmaceutical, and Ikaria; Ronald Perronereceives research support from Otsuka, and is a consultant for Otsuka andSanofi-Genzyme; and Kyongtae Bae is a consultant for Otsuka. The remainingauthors disclose no conflicts.

FundingSupported by funding from the National Institutes of Health (U01 DK082230)and supported in part by Boehringer Ingelheim Pharmaceuticals, Inc. Boeh-ringer Ingelheim Pharmaceuticals, Inc, had no role in the design, analysis, orinterpretation of the results in this study and were given the opportunity toreview the manuscript for medical and scientific accuracy as it relates to theirrespective substances as well as intellectual property considerations.

Current address of A.M.: Columbia University College of Surgeons and Phy-sicians, New York, New York.

Supplementary Methods

Only data from HALT-PKD-A participants (withavailable MRIs) were used in this study (age, 15–49 y,with eGFR > 60 mL/min/1.73 m2 and BP � 130/80 orreceiving treatment for hypertension before enrollment).Study design and all protocols were approved by theinstitutional review boards at each participating site, andall enrollees provided informed consent. Data from thehealth-related quality-of-life questionnaire SF-36, andfrom imaging and central laboratory parameters (per-formed at Cleveland Clinic) were extracted and analyzedat the University of Pittsburgh Data Coordination Center.

The following baseline measurements were of inter-est for this analysis: liver cyst prevalence, LV, LCV, LPV,common bile duct diameters, spleen volume, presence ofcysts, presence of pancreatic cysts, laboratory values, SF-36 QOL measurement, and exogenous estrogen use (infemales).

Radiologic Assessments

Each participating clinical center obtainedMR imagingto determine LV, LCV, and TKV using a protocol developedby the HALT-PKD Imaging Subcommittee.1 Just as in anyimaging-based volume measurement study, our LV andLCV measurements were subjected to image acquisitionparameters (eg, slice thickness, volume averaging, andsignal intensity variations). To minimize measurementvariation, we used a rigorous quality control and stan-dardization protocol for image acquisition and the anal-ysis process.2 When only a few cysts were present, theycould be readily counted manually. However, because inmost cases liver cysts were too numerous and touchedeach other, accurate counts were very difficult and labo-rious. In addition, although there have been several re-ports of the occurrence of hepatic fibrosis in ADPKD, ourimaging methods could not differentiate or quantitatehepatic fibrosis as distinct from normal hepatic paren-chyma. Also, with the current limitations of image reso-lution, microcysts in the liver could not be discerned fromthe liver parenchyma and could not be detected ormeasured.3,4 Any visible extrahepatic vessels (ie, portalvein or hepatic artery) were excluded from LV measure-ments. However, when these vessels branched off andbecame encompassed by the hepatic parenchyma andcysts, they were included in the LV. We estimated LPV bysubtracting LCV from LV. After acquisition, images werereviewed locally, de-identified, and transferred securely,via the internet, to the Image Analysis Center at the Uni-versity of Pittsburgh.5,6 Small differences in LV and LCVcompared with those reported in the baseline analysiswere explained by rescan measurements performedafter the publication of Torres’s et al.6 The common bileduct was considered dilated if larger than 7 mm.7 Weused the reported spleen volume values (exceeding 2SDs) to define splenomegaly.8 MRI data collected from

prospective living donor liver transplant candidates (atthe University of Pittsburgh) were used for comparativeLV and LPV analyses.

Statistical Analysis

All data were analyzed using SAS (SAS Institute, Inc,Cary, NC) and STATA/SE 12 (StataCorp LP, College Sta-tion, TX). Descriptive statistics were used to quantifydemographic and clinical characteristics across andwithin sexes. Sample means and SDs were used forcontinuous variables, while sample proportions werecalculated for categoric variables. Because it is wellestablished that PLD is more severe in women than inmen, data were stratified by sex.2,9,10 To assess between-sex differences, 2-sample t tests and chi-square tests ofindependence were performed. For TKV, adjustment byheight did not alleviate anthropometric sex differences.For LV, adjusting for height-alleviated sex differencesperformed better than adjusting for BMI or body surfaceor any anthropometric adjustment (data not shown).Analysis of variance or chi-square tests were conductedto assess differences between htLV groups. If theomnibus test was significant, multiple comparisons wereperformed with Bonferroni-adjusted P values.

Because of the skewed nature of the volume vari-ables, each was log-transformed for normality (naturallog, and for data display in Figures 1 and 2). Pearsoncorrelation coefficients (r) were calculated between eachof log-transformed and htLV, htLCV, htLPV, and baselineclinical, laboratory, and QOL variables. Univariate asso-ciations that were significant at the 0.10 significancelevel were examined further by fitting a multiple linearregression to htLV, htLCV, and htLPV outcomes adjustingfor age, BMI, and htTKV in males and females, currentbirth control use (a composite variable that categorizedexposure at baseline to any estrogen-containing contra-ceptive pills/patch, injection/implant, or intrauterinedevice), and number of pregnancies in females. Becauseof the exploratory nature of these analyses, all signifi-cance testing was performed with 5% type I error withno adjustments for multiplicity.

Supplementary Results

Mild Splenomegaly and Hypersplenism inPolycystic Liver Disease

Splenomegaly was seen in 1.8% of patients (10 cases:7 men, 3 women) and spleen volumes (296 � 120 mL)were significantly higher than those reported in studies ofhealthy American (201 � 77 mL), European (219 � 76mL), Indian (162 � 90 mL), Chinese (191 � 70 mL), andJapanese (127� 63 mL) cohorts (Figure 2C and D).8,11–14

Similar to the general population, height-adjusted spleenvolume correlated negatively with age (SupplementaryTable 3). In both sexes, unadjusted and height-adjusted

- 2014 Liver Disease Burden in Early ADPKD 10.e1

spleen volumes correlated with htLV and were differentonly between the mild and moderate and mild and severePLD groups (P ¼ .0001) (Table 3). HtSVs correlated withhtLPV, but not with htLCV (Supplementary Table 2).HtSVs correlated positively with ALT and ALP levels inwomen, but not in men (Supplementary Table 3). Plateletcounts were within the normal range in most patients,with thrombocytopenia (<150 � 109/L) present in only0.7%. However, platelet counts negatively correlated withhtLVs (P ¼ .0020) (Table 3), htLPVs, and height-adjustedspleen volumes in both sexes (Supplementary Tables 2and 3), and the decrease in platelets correlated withboth htLV severity (P ¼ .0020) (Table 3) and LCV(P ¼ .0003; data not shown).

Influence of Estrogen Use and PregnancyNumber on Polycystic Liver Disease Severity

Forty-seven of the 275 women had not been pregnantor taken birth control pills (Supplementary Table 4). TheirhtLVs (975 � 261 mL/m) and htLCVs (72 � 187 mL/m)did not differ from those of patients never pregnant withexposure to birth control pills (1055� 188, P¼ .0935; and44 � 119 mL/m, P ¼ .3134), but were significantly lowerthan those with a history of pregnancy (1193 � 580, P ¼.0046; and 229 � 566 mL/m, P ¼ .0052) (SupplementaryTable 4). However, after adjusting for age, these differ-ences were not statistically significant. Furthermore,85.3% of women with an htLCV greater than 400 mL/mhad a prior pregnancy comparedwith 71.8% of those withan htLV less than 400 mL/min. The mean numbers ofpregnancies in these 2 groups with and without severePLD were 3.07 � 1.79 and 2.73 � 1.49, respectively.

Prevalence of Pancreatic and SplenicCysts in Autosomal-Dominant PolycysticKidney Disease

Thirteen (1.8%) of 560 patients had pancreatic cysts; 8(1.4%) of these patients had only 1 cyst (Figure 2E). Oneindividual had 2 cysts and another had 4 cysts. Diametersranged from 4 to 22 mm, with only 1 cyst greater than 20mm. Splenic cysts (Figure 2F) were seen in 2.6% (n¼ 14)of patients, with no significant difference between sexes.Common bile duct dilatation, seen in 17.3%, was morecommon in women (P ¼ .03) and correlated with htLV,htLCV, and htLPV (Supplementary Table 2).

Supplementary References1. King BF, Torres VE, Brummer ME, et al. Magnetic resonance

measurements of renal blood flow as a marker of diseaseseverity in autosomal-dominant polycystic kidney disease.Kidney Int 2003;64:2214–2221.

2. Bae KT, Zhu F, Chapman AB, et al. Magnetic resonance imagingevaluation of hepatic cysts in early autosomal-dominant poly-cystic kidney disease: the Consortium for Radiologic ImagingStudies of Polycystic Kidney Disease cohort. Clin J Am SocNephrol 2006;1:64–69.

3. Bae KT, Grantham JJ. Imaging for the prognosis of autosomaldominant polycystic kidney disease. Nat Rev Nephrol 2010;6:96–106.

4. Grantham JJ, Mulamalla S, Grantham CJ, et al. Detected renalcysts are tips of the iceberg in adults with ADPKD. Clin J AmSoc Nephrol 2012;7:1087–1093.

5. Moore SM, Maffitt DR, Blaine GJ, et al. Workstation acquisitionnode for multicenter imaging studies. SPIE 2001;4323:271–277.

6. Torres VE, Chapman AB, Perrone RD, et al. Analysis of baselineparameters in the HALT polycystic kidney disease trials. KidneyInt 2012;81:577–585.

7. Fukukura Y, Fujiyoshi F, Sasaki M, et al. Pancreatic duct:morphologic evaluation with MR cholangiopancreatography af-ter secretin stimulation. Radiology 2002;222:674–680.

8. Geraghty EM, Boone JM, McGahan JP, et al. Normal organvolume assessment from abdominal CT. Abdom Imaging 2004;29:482–490.

9. Everson GT, Scherzinger A, Berger-Leff N, et al. Polycystic liverdisease: quantitation of parenchymal and cyst volumes fromcomputed tomography images and clinical correlates of hepaticcysts. Hepatology 1988;8:1627–1634.

10. Sherstha R, McKinley C, Russ P, et al. Postmenopausal estro-gen therapy selectively stimulates hepatic enlargement inwomen with autosomal dominant polycystic kidney disease.Hepatology 1997;26:1282–1286.

11. Prassopoulos P, Daskalogiannaki M, Raissaki M, et al. Deter-mination of normal splenic volume on computed tomography inrelation to age, gender and body habitus. Eur Radiol 1997;7:246–248.

12. Asghar A, Agrawal D, Yunus SM, et al. Standard splenic volumeestimation in North Indian adult population: using 3D recon-struction of abdominal CT scan images. Anat Res Int 2011;2011:707325.

13. Harris A, Kamishima T, Hao HY, et al. Splenic volume mea-surements on computed tomography utilizing automaticallycontouring software and its relationship with age, gender, andanthropometric parameters. Eur J Radiol 2010;75:e97–e101.

14. Kaneko J, Sugawara Y, Matsui Y, et al. Normal splenic volume inadults by computed tomography. Hepatogastroenterology 2002;49:1726–1727.

10.e2 Hogan et al Clinical Gastroenterology and Hepatology Vol. -, No. -

Supplementary Table 1. Correlations Between Baseline Parameters and Liver Cyst and Parenchymal Volumes in HALT-PKD-A Male and Female Participants

Variable

lnHtLCV lnHtLPV

N r P value n r P value

AgeMales 259 0.47 <.0001a 259 0.06 .3247Females 263 0.42 <.0001a 263 �0.09 .1386

Height, cmMales 259 �0.02 .7934 259 0.005 .9289Females 263 �0.04 .4910 263 0.013 .0400a

Weight, kgMales 258 �0.006 .9207 258 0.50 <.0001a

Females 263 �0.29 <.0001a 263 0.54 <.0001a

BSA, m2

Males 258 0.0008 .9903 258 0.41 <.0001a

Females 263 �0.27 <.0001a 263 0.55 <.0001a

BMI, kg/m2

Males 258 �0.004 .9446 258 0.50 <.0001a

Females 263 �0.28 <.0001a 263 0.56 <.0001a

lnHtTKVMales 254 0.12 .0589 254 0.05 .3900Females 259 0.21 .0006a 259 0.06 .3255

eGFRMales 259 �0.29 <.0001a 259 �0.09 .1626Females 263 �0.31 <.0001a 263 �0.06 .3303

SBP homeMales 178 0.01 .8899 178 0.16 .0288a

Females 172 �0.0006 .9936 172 0.30 <.0001a

DBP homeMales 117 0.11 .1619 178 0.10 .1710Females 172 0.05 .5286 172 0.25 .0011a

Caffeine, cups/dMales 240 0.1170 .0704 240 0.1709 .0080a

Females 254 0.0115 .8548 254 0.2029 .0012a

DBP home, home diastolic blood pressure; lnHtTKV, natural log of height-adjusted total kidney volume.aSignificant P value.

- 2014 Liver Disease Burden in Early ADPKD 10.e3

Supplementary Table 2. Correlations of Liver Cyst and Parenchymal Volumes With Liver Function and Imaging Parameters inMales and Females

Variable

ln(htLCV) ln(htLPV)

n R P value n r P value

AST level, U/LMales 259 0.03 .6270 259 0.09 .1590Females 263 0.08 .1801 263 0.04 .5473

ALT level, U/LMales 259 0.0006 .9925 259 0.22 .0006a

Females 263 0.05 .4445 263 0.23 .0002a

ALP level, U/LMales 259 �0.04 .5024 259 0.006 .9162Females 263 0.05 .4065 263 0.11 .0839

Serum albumin level, mg/dLMales 257 �0.05 .4011 257 �0.09 .1290Females 262 �0.12 .0557 262 �0.27 <.0001a

Bilirubin level, mg/dLMales 259 �0.05 .3841 259 �0.06 .2938Females 263 0.13 �.0385 263 �0.11 .0617

Platelets, cells/uLMales 258 �0.003 .9553 258 0.03 .5897Females 260 �0.19 .0023a 260 0.07 .2301

WBC, cells/mLMales 257 �0.065 .2925 257 0.16 .0085a

Females 259 �0.03 .6553 259 0.04 .5119Serum sodium, mEq/L

Males 259 0.01 .8121 259 �0.06 .3229Females 263 0.10 .0934 263 �0.09 .1409

ln(htSV)Males 254 0.002 .9716 254 0.41 <.0001a

Females 262 �0.07 .2276 262 0.48 <.0001a

Common bile duct diameterMales 254 0.06 .3133 254 �0.03 .6696Females 262 0.14 .0227a 262 0.16 .0105a

AST, aspartate aminotransferase; ln(htLCV), natural log of height-adjusted liver cyst volume; ln(htLPV), natural log of height-adjusted liver parenchymal volume;ln(htSV), natural log of height-adjusted spleen volume; htSV, height-adjusted spleen volume; WBC, white blood cell.aSignificant P value.

Supplementary Table 3. Correlations of Height-Adjusted Spleen Volumes With Age and Laboratory Parameters

Variable

Males Females

N r P value n r P value

Age 265 �0.140 .0227a 268 �0.176 .0038a

AST level, U/L 265 �0.061 .3224 268 �0.046 .4565ALT level, U/L 265 0.106 .0856 268 0.131 .0320a

Alkaline phosphatase level, U/L 265 0.099 .1087 268 0.133 .0291a

Total bilirubin level, mg/dL 265 0.013 .8270 268 �0.048 .4365Serum albumin level, mg/dL 263 �0.06003 .332 267 0.048 .4302White blood cells, cells/mL 264 0.002 .9724 264 0.072 .2425Hemoglobin level, g/dL 264 0.058 .3506 263 0.039 .5321Platelets, cells/mL 264 �0.333 <.0001a 265 �0.140 .0225a

AST, aspartate aminotransferase.aSignificant P value.

10.e4 Hogan et al Clinical Gastroenterology and Hepatology Vol. -, No. -

Supplementary Table 4. Effect of Estrogen Exposure on Liver Cystic Disease in ADPKD Patients

With liver cysts HtLV HtLCV HtLPV

Never pregnant, no history of oralcontraceptive use

47 (97.9%) 975 � 261 (n ¼ 47) 72.3 � 186.8 (n ¼ 47) 903 � 199 (n ¼ 47)

Never pregnant, history of oralcontraceptive use

22 (91.7%) 1055 � 188 (n ¼ 22) 43.5 � 118.8 (n ¼ 22) 1012 � 136 (n ¼ 22)

History of pregnancy 196 (97.5%) 1193 � 580 (n ¼ 192) 229 � 521 (n ¼ 192) 963 � 201 (n ¼ 192)1 pregnancy 39 (100%) 1159 � 459 (n ¼ 39) 185 � 405 (n ¼ 39) 974 � 187 (n ¼ 39)�2 pregnancies 157 (96.9%) 1201 � 608 (n ¼ 153) 241 � 548 (n ¼ 153) 961 � 205 (n ¼ 153)

- 2014 Liver Disease Burden in Early ADPKD 10.e5

Supplementary Table 5. Analysis of Demographics, Laboratory Parameters, Image Analysis, and SF-36 Data by htLCVCategory

Variable

htLCV<10 mL(mild cystic disease)

(n ¼ 285) (1)

htLCV 10-400 mL(mild-moderatecystic disease)(n ¼ 195) (2)

htLCV > 400 mL(severe cystic disease)

(n ¼ 42) (3) P value

Age, y 33.6 � 8.6 39.8 � 6.1 41.8 � 5.7 <.0001a

1 vs 2, <.00011 vs 3, <.0001

Sex, n (% female) 113 (43.46) 107 (50.47) 47 (75.81) <.0001a

1 vs 3, <.00012 vs 3, .0004

LaboratoryeGFR 94.9 � 18.3 89.2 � 15.6 80.8 � 13.9 <.0001a

1 vs 2, .00101 vs 3, <.00012 vs 3, .0116

Serum albumin level, mg/dL (n) 4.3 � 0.7 (283) 4.1 � 0.5 (194) 4.0 � 0.6 .0048a

1 vs 2, .02921 vs 3, .0261

Bilirubin level, mg/dL 0.7 � 0.4 0.8 � 0.3 0.7 � 0.3 .1606Alkaline phosphatase level, U/L 61.1 � 28.2 57.2 � 18.0 62.3 � 25.4 .1870AST level, u/L 24.1 � 9.0 24.5 � 13.9 23.9 � 5.0 .9096ALT level, u/L 25.2 � 15.3 24.2 � 15.4 22.3 � 7.0 .4543Serum sodium, mEq/L 138.7 � 7.7 139.3 � 2.1 139.3 � 2.4 .5025Hemoglobin level, g/dL (n) 14.4 � 2.3 (282) 14.6 � 8.4 13.7 � 1.0 .6361Platelets, cells/uL (n) 256.2 � 60.9 (283) 260.9 � 55.1 (193) 221.5 � 42.2 .0003a

1 vs 3, .00092 vs 3, .0002

WCC, cells/uL (n) 6.5 � 1.9 (282) 6.5 � 6.9 (192) 6.0 � 1.5 (42) .7856Image analysis

LV, mL (n) 1778.0 � 386.2 (285) 1851.3 � 378.5 3695.0 � 1807.5 (42) <.0001a,b

1 vs 3, <.00012 vs 3, <.0001

HtLV, mL/m 1014.7 � 205.5 1066.5 � 193.4 2186.5 � 1012.4 <.0001a,b

1 vs 2, .01661 vs 3, <.00012 vs 3, <.0001

TKV, mL (n) 1141.7 � 692.5 (281) 1288.4 � 706.1 (191) 1269.2 � 814.9 (41) .0481b

1 vs 2, .0469HtTKV, mL/m (n) 647.0 � 376.3 (281) 738.4 � 391.5 (191) 748.0 � 473.6 (41) .0174a,b

1 vs 2, .0225LPV, mL 1775.0 � 386.6 1708.0 � 362.3 1740.3 � 328.7 .1578b

HtLPV, mL/m 1013.0 � 205.8 983.1 � 183.1 1032.4 � 184.9 .1513b

Spleen volume, mL (n) 306.4 � 123.0 (281) 283.7 � 115.6 (193) 285.1 � 113.6 .0654b

HtSV, mL/m (n) 174.2 � 66.1 (281) 162.6 � 61.0 (193) 167.7 � 60.5 .1065b

SF-36 QOLPhysical functioning (n) 89.5 � 19.6 (282) 93.5 � 13.7 (191) 90.7 � 16.9 (41) .0507Physical role (n) 90.5 � 20.2 (281) 89.6 � 21.0 (191) 86.4 � 23.0 (40) .4961Bodily pain (n) 78.0 � 22.4 (281) 77.7 � 20.5 (190) 74.3 � 23.1 (40) .5981General health (n) 64.9 � 20.5 (282) 67.1 � 19.2 (191) 62.7 � 19.3 (41) .3090Vitality (n) 61.4 � 20.9 (281) 63.5 � 18.9 (191) 60.2 � 19.9 (40) .4360Social functioning (n) 88.3 � 20.3 (281) 87.6 � 19.3 (191) 84.1 � 22.8 (40) .4671Role emotional (n) 89.9 � 19.5 (280) 90.6 � 18.4 (189) 87.9 � 21.5 (40) .7194Mental health (n) 77.3 � 15.2 (281) 77.5 � 15.2 (191) 76.5 � 12.8 (40) .9264Standardized mental component (n) 50.8 � 9.4 (281) 50.9 � 8.5 (190) 49.7 � 9.2 (40) .7596Standardized physical component 51.9 � 7.6 (281) 52.7 � 7.0 (190) 51.1 � 7.9 (40) .3332

NOTE. The htLCV categories were as follows: mild (1), mild–moderate (2), and severe cystic disease (3). Data are presented as means � SD unless otherwiseindicated.AST, aspartate aminotransferase; HtSV, height-adjusted spleen volume.aSignificant P value.bP value was calculated using log-transformed variables.

10.e6 Hogan et al Clinical Gastroenterology and Hepatology Vol. -, No. -