Hereditary polycystic kidney disease is characterized by lymphopenia across all stages of kidney dysfunction: an observational study

Hereditary polycystic kidney disease is characterized by lymphopenia across all stages of kidney... Abstract Background Polycystic kidney disease (PKD) is characterized by urinary tract infections and extrarenal abnormalities such as an increased risk of cancer. As mutations in polycystin-1 and -2 are associated with decreased proliferation of immortalized lymphoblastoid cells in PKD, we investigated whether lymphopenia could be an unrecognized trait of PKD. Methods We studied 700 kidney transplant recipients with (n = 126) or without PKD at the time of kidney transplantation between 1 January 2003 and 31 December 2014 at Ghent University Hospital. We also studied 204 patients with chronic kidney disease (CKD) with PKD and 204 matched CKD patients without PKD across comparable CKD strata with assessment between 1 January 1999 and 1 February 2016 at three renal outpatient clinics. We compared lymphocyte counts with multiple linear regression analysis to adjust for potential confounders. We analysed flow cytometric immunophenotyping data and other haematological parameters. Results Lymphocyte counts were 264/µL [95% confidence interval (CI) 144–384] and 345/µL (95% CI 245–445) (both P < 0.001) lower in the end-stage kidney disease (ESKD) and CKD cohort, respectively, after adjustment for age, sex, ln(C-reactive protein) and estimated glomerular filtration rate (in the CKD cohort only). In particular, CD8+ T and B lymphocytes were significantly lower in transplant recipients with versus without PKD (P < 0.001 for both). Thrombocyte and monocyte counts were lower in patients with versus without PKD in both cohorts (P < 0.001 for all analyses except P = 0.01 for monocytes in the ESKD cohort). Conclusion PKD is characterized by distinct cytopenias and especially lymphopenia, independent of kidney function. This finding has the potential to alter our therapeutic approach to patients with PKD. apoptosis, lymphopenia, polycystic kidney, proliferation, uremia INTRODUCTION Hereditary autosomal dominant polycystic kidney disease (PKD) is caused by mutations in genes coding for polycystin-1 and more rarely polycystin-2 and is the most common monogenic disorder, affecting ∼1 in 500 people. It is the primary cause of end-stage kidney disease (ESKD) in 5–15% of those treated with renal replacement therapy [1, 2]. The disease has a particular phenotype that includes a tendency to develop aneurysms, cardiac valve disorders, abdominal wall hernias and hypertension. Patients with PKD also tend to easily contract ascending urinary tract infections, develop bronchiectasis more readily than chronic kidney disease (CKD) patients without PKD, with an adjusted odds ratio (AOR) of 2.78, and have a higher incidence of pneumonia and nonmelanoma skin cancer, with an AOR of 1.65, than kidney transplant recipients without PKD [2, 3–6]. This suggests an impaired immune response may be at play. The proteins polycystin-1 and -2, which form a functional complex, are present in renal tubular epithelial cells. If one of these proteins is dysfunctional, this causes disturbed calcium influx, increased apoptosis and cystogenesis, resulting in the PKD type 1 and 2 phenotypes, respectively [1]. Polycystin-1 and polycystin-2 are not only present in renal tubular cells, but also in cardiomyocytes [7] and endothelial cilia [8], and their dysfunction could explain the early development of cardiac hypertrophy and endothelial dysfunction in PKD, even with preserved kidney function and at a very young age [9, 10]. It is uncertain whether decreased proliferation of lymphoblastoid cells in PKD in vitro translates into fewer circulating lymphocytes in vivo [11]. We hypothesized that lymphopenia could be more common in PKD, which could explain a higher incidence of infections and skin cancer in this particular population. In 2002, Banerjee et al. [12] conducted a small unadjusted cross-sectional case–control study and found that among patients treated with haemodialysis, those with PKD (n = 11) had 610/µL (40%) fewer lymphocytes compared with the age- and sex-matched controls without PKD (n = 33). Aim We wanted to determine whether lymphocyte counts were lower in patients with PKD than in controls, independent of the effects of kidney function and after adjustment for confounders such as age and inflammation. We also wanted to verify whether differences in lymphocyte counts were confined to specific lymphocyte subpopulations. In secondary analyses, numbers of other leucocyte subtypes and platelets were analysed. MATERIALS AND METHODS Study settings In the first study, we compared lymphocyte counts of all patients with ESKD with versus without PKD who were admitted to the Ghent University Hospital for kidney transplantation (ESKD cohort). In a second analysis, we gathered data on lymphocyte counts from stable outpatient clinic patients with CKD and PKD from three Belgian hospitals: Ghent University Hospital, AZ Delta Hospital Roeselare and OLV Hospital Aalst (CKD cohort). We compared these participants with age- and sex-matched controls across the same categories of CKD (stage G1–5) but with an alternative renal diagnosis. The institutional review board and ethics committee of Ghent University Hospital reviewed and approved the research protocol (EC/2015/0655). Time of assessment In the people with ESKD and waitlisted for transplantation, we measured the lymphocyte count immediately before transplantation. In the second analysis, we retrieved data from the last measurement at the outpatient clinic from January 1999 to 1 February 2016. Study population The study population included: First cohort: all patients >18 years of age with ESKD who received a first or subsequent kidney transplant in Ghent University Hospital between 1 January 2003 and 31 December 2014. Second cohort: all patients >18 years of age with hereditary PKD and outpatient follow-up and free of active or recent (<1 month) infection at the time of assessment, intake of immunosuppressive drugs, active cancer or renal replacement therapy next to age- and sex-matched controls with comparable kidney function (1:1). Excluded were patients with acquired cystic disease, renal dysplasia, von Hippel–Lindau disease or autosomal recessive polycystic disease. Measurements We selected a priori variables considered for analysis on their theoretical association with lymphocyte count: age, sex, estimated glomerular filtration rate (eGFR) and C-reactive protein (CRP). We extracted the following baseline clinical characteristics of the ESKD cohort from the transplant database of the Ghent University Hospital, which contains prospectively collected data on baseline characteristics and outcomes of all kidney transplant recipients: continuous variables included age (years) and time on the waiting list (days) and categorical variables included sex (coded as male = 1, female = 0); diabetes at the time of assessment (yes/no), defined according to the American Diabetes Association criteria [13] or as intake of glucose-lowering drugs according to patient records; ethnicity (coded as Caucasian = 0, non-Caucasian = 1); dialysis modality (coded as pre-emptive transplantation = 1, haemodialysis = 2, peritoneal dialysis = 3); retransplantation (yes/no) and intake of immunosuppressive drugs before transplantation (yes/no). We extracted from the medical files the following baseline biochemical variables: lymphocytes, haemoglobin, leucocytes, thrombocytes, neutrophils, monocytes, serum albumin, CRP and body mass index (BMI), calculated as weight in kilograms divided by the square of height in metres. If data on flow cytometric immunophenotyping analysis were available at the time of transplantation, we included counts of CD4+ T lymphocytes, CD8+ T lymphocytes, CD19+ B lymphocytes and natural killer (NK) cells. Lymphopenia was defined as <1133 lymphocytes/μL of blood according to the laboratory reference values. All participating centres use automated cell counters (Sysmex, Kobe, Japan). For the CKD cohort, we recorded serum creatinine and calculated eGFR based on the Chronic Kidney Disease Epidemiology Collaboration formula [14] and gathered data on haemoglobin, haematocrit, thrombocytes, leucocytes, monocytes, neutrophils, lymphocytes, erythrocyte sedimentation rate and CRP from the respective databases of each participating centre. In additional analyses, we examined whether there were differences between leucocyte subpopulations and thrombocyte counts in patients with versus without PKD. Statistical analysis We summarized baseline characteristics as means and standard deviations (SDs) or medians with interquartile range (IQR) for continuous and percentages for categorical variables. We analysed differences between groups with independent sample t-tests, Mann–Whitney U-test, analysis of variance, Kruskal–Wallis test and chi-square test depending on the distribution. If appropriate, we used post hoc Scheffé testing to account for multiple comparisons. We used multiple linear regression analysis to study the relation between PKD and lymphocyte concentration as a dependent variable. CRP was natural log transformed (lnCRP) because of its skewed distribution. We constructed models adjusting for a priori specified covariates based on previous literature: age, sex, lnCRP and eGFR (CKD cohort). Additionally, we adjusted for covariates that were statistically significant (P < 0.1) in each separate univariate model. We tested for interactions between PKD and other covariates such as age and eGFR. We calculated the variance inflation factor to evaluate potential multicollinearity. In the CKD cohort, we matched each case of PKD for age, sex and CKD stage with an individual with an alternative renal diagnosis. We additionally analysed the mean lymphocyte count of the last two consultations as internal validation. We performed similar analyses to evaluate the association between PKD and leucocyte, monocyte and neutrophil counts, but refrained from multivariable modeling for each haematological covariate considering our main study question and also taking into account the risk of committing type I error due to multiple testing. A two-sided P-value <0.05 was considered to indicate statistical significance for all analyses. We conducted all statistical analyses using SPSS Statistics (version 22.0; IBM, Armonk, NY, USA). RESULTS ESKD cohort Between January 2003 and December 2014, 949 patients (mean age of 52 ± 12 years, 63% male, 17% diabetes, 95% Caucasian and 12% on immunosuppressive drugs at the time of transplantation) received a kidney transplant at Ghent University Hospital. We excluded 249 patients because the lymphocyte count was unavailable, leaving 700 included in the study. Flow cytometric analysis data were available in 484 participants. The proportion of participants with PKD was similar in the 249 people with missing data (Figure 1). There were no missing data on any other covariate. Participants with versus without PKD were 4 years older, were less frequently male and had a comparable CRP concentration. PKD patients had a lower BMI, less frequently had diabetes and only ∼5% were taking immunosuppressive drugs versus 14% in those without PKD (Table 1). Table 1 Population characteristics of the subpopulation with end-stage kidney failure Characteristics  Polycystic kidney disease (n = 169)  No polycystic kidney disease (n = 780)  P-value  Age (years)  55.3 ± 8.2  51.6 ± 13.3  0.001  Men  96 (57.0)  509 (65.2)  0.012  BMI (kg/m2)  24.7 ± 3.6  25.5 ± 4.6  0.035  Diabetes  5 (3.0)  159 (20.4)  <0.0001  Non-Caucasian  3 (1.8)  46 (5.9)  0.028  Serum albumin concentration (g/L)  43.9 ± 4.9  42.9 ± 5.9  0.038  Serum CRP (mg/L)  3.0 (5.0)  3.0 (6.0)  0.81  Previous kidney transplantation  5 (3.0)  52 (6.7)  0.07  Living donation  13 (7.7)  59 (7.6)  0.98  Dialysis vintage (days)  627 (848)  566 (835)  0.32  Immunosuppression  9 (5.3)  107 (13.7)  0.002  Dialysis modality      0.16  Pre-emptive  23 (13.6)  69 (8.9)    Haemodialysis  112 (66.3)  536 (68.7)    Peritoneal dialysis  34 (20.1)  175 (22.4)    Haemoglobin (g/dL)  13.0 ± 1.7  12.8 ± 1.5  0.07  Leucocytes (/µL)  6250 (2050)  6700 (2855)  <0.001  Thrombocytes (×103/µL)  199 (89)  222 (89)  <0.001  Neutrophils (/µL)  3800 (1700)  4060 (2350)  0.05  Lymphopenia (%)  38.1  24.0  0.001  Lymphocytes (/µL)a  1330 (760)  1550 (750)  <0.001  CD4+ T lymphocytes (/µL)b  690 (391)  738 (405)  0.06  CD8+ T lymphocytes (/µL)b  274 (187)  304 (248)  <0.001  CD19 lymphocytes (/µL)b  92 (70)  143 (130)  <0.001  NK cells (/µL)c  234 (143)  211 (202)  0.39  Monocytes (/µL)b  550 (250)  590 (285)  0.01  Characteristics  Polycystic kidney disease (n = 169)  No polycystic kidney disease (n = 780)  P-value  Age (years)  55.3 ± 8.2  51.6 ± 13.3  0.001  Men  96 (57.0)  509 (65.2)  0.012  BMI (kg/m2)  24.7 ± 3.6  25.5 ± 4.6  0.035  Diabetes  5 (3.0)  159 (20.4)  <0.0001  Non-Caucasian  3 (1.8)  46 (5.9)  0.028  Serum albumin concentration (g/L)  43.9 ± 4.9  42.9 ± 5.9  0.038  Serum CRP (mg/L)  3.0 (5.0)  3.0 (6.0)  0.81  Previous kidney transplantation  5 (3.0)  52 (6.7)  0.07  Living donation  13 (7.7)  59 (7.6)  0.98  Dialysis vintage (days)  627 (848)  566 (835)  0.32  Immunosuppression  9 (5.3)  107 (13.7)  0.002  Dialysis modality      0.16  Pre-emptive  23 (13.6)  69 (8.9)    Haemodialysis  112 (66.3)  536 (68.7)    Peritoneal dialysis  34 (20.1)  175 (22.4)    Haemoglobin (g/dL)  13.0 ± 1.7  12.8 ± 1.5  0.07  Leucocytes (/µL)  6250 (2050)  6700 (2855)  <0.001  Thrombocytes (×103/µL)  199 (89)  222 (89)  <0.001  Neutrophils (/µL)  3800 (1700)  4060 (2350)  0.05  Lymphopenia (%)  38.1  24.0  0.001  Lymphocytes (/µL)a  1330 (760)  1550 (750)  <0.001  CD4+ T lymphocytes (/µL)b  690 (391)  738 (405)  0.06  CD8+ T lymphocytes (/µL)b  274 (187)  304 (248)  <0.001  CD19 lymphocytes (/µL)b  92 (70)  143 (130)  <0.001  NK cells (/µL)c  234 (143)  211 (202)  0.39  Monocytes (/µL)b  550 (250)  590 (285)  0.01  Data are presented as number (%) or mean ± SD. a Results were available in 700 patients. b Results were available in 484 patients. c Results were available in 346 patients. Participants with PKD had on average 269 (95% CI 150–388; P < 0.001; 17%) fewer lymphocytes/μl than those without PKD. When defined as <1133 lymphocytes/μl, lymphopenia was ∼1.6 times as common for those with PKD. Blood concentrations of CD8+ T lymphocytes and CD19+ B lymphocytes were 25% and 31% lower, respectively, in people with versus without PKD (both P < 0.001). Patients with PKD also had lower total leucocyte and thrombocyte counts (P < 0.001 for both) (Table 1, Figure 2, Supplementary Figure S1). FIGURE 1 View largeDownload slide Flow diagram. There was a comparable proportion of polycystic disease in the original study population [169 participants (17.9%)] and in the study population [126 participants (18%)]. FIGURE 1 View largeDownload slide Flow diagram. There was a comparable proportion of polycystic disease in the original study population [169 participants (17.9%)] and in the study population [126 participants (18%)]. Regression model When adjusted for age, sex and lnCRP, patients with PKD had on average 264 fewer lymphocytes compared with those without PKD [adjusted mean difference −264 cells/µL (95% CI −384 to −144 cells/µL), P < 0.001] (Table 2, Model 2). After further adjustment for intake of immunosuppression and diabetes, participants with PKD had on average 310 fewer lymphocytes compared with those without PKD [adjusted mean difference −310 cells/µL (95% CI −431 to −188 cells/µL)] (Table 2, Model 3). Table 2 Multiple regression models of the associations of PKD with lymphocyte, neutrophil, monocyte, leucocyte and thrombocyte counts ESKD cohort   Dependent variable  Model 1     β (95% CI)  P-value  β (95% CI)  P-value  Lymphocyte (/µL)  −269 (−388 to − 150)  <0.001      CD4+ T lymphocyte (/µL)  −77 (−153 to − 1)  0.05      CD8+ T lymphocyte (/µL)  −95 (−145 to − 45)  <0.001      B lymphocyte (/µL)  −50 (−79 to − 22)  <0.001      NK cell (/µL)  5 (−34–44)  0.80      Neutrophil (/µL)  −432 (−797 to − 67)  0.02      Monocyte (/µL)  −58 (−103 to − 14)  0.01      Leucocyte (/µL)  −765 (−1132 to − 399)  <0.001      Thrombocyte (×103/µL)  −21 (-32 to − 9)  <0.001          Model 2        β (95% CI)  P-value      Lymphocytea (/µL)  −264 (−384 to − 144)  <0.001  −0.16 (−0.24 to − 0.09)  <0.001    Model 3      Lymphocyteb (/µL)  −310 (−431 to − 188)  <0.001  −0.19 (−0.27 to − 0.12)  <0.001    CKD cohort  Dependent variable  Model 1      β (95% CI)  P-value  Lymphocyte (/µL)  −364 (−466 to − 261)  <0.001      Neutrophil (/µL)  −121 (−413–170)  0.41      Monocyte (/µL)  −85 (−122 to − 47)  <0.001      Leucocyte (/µL)  −605 (−948 to − 261)  0.001      Thrombocyte (×103/µL)  −28 (-40 to − 16)  <0.001          Model 2        β (95% CI)  P-value      Lymphocytec  −345 (−445 to − 245)  <0.001  −0.31 (−0.40 to − 0.22)  <0.001    Model 3      Lymphocyted  −333 (−434 to − 233)  <0.001  −0.30 (−0.39 to − 0.21)  <0.001  ESKD cohort   Dependent variable  Model 1     β (95% CI)  P-value  β (95% CI)  P-value  Lymphocyte (/µL)  −269 (−388 to − 150)  <0.001      CD4+ T lymphocyte (/µL)  −77 (−153 to − 1)  0.05      CD8+ T lymphocyte (/µL)  −95 (−145 to − 45)  <0.001      B lymphocyte (/µL)  −50 (−79 to − 22)  <0.001      NK cell (/µL)  5 (−34–44)  0.80      Neutrophil (/µL)  −432 (−797 to − 67)  0.02      Monocyte (/µL)  −58 (−103 to − 14)  0.01      Leucocyte (/µL)  −765 (−1132 to − 399)  <0.001      Thrombocyte (×103/µL)  −21 (-32 to − 9)  <0.001          Model 2        β (95% CI)  P-value      Lymphocytea (/µL)  −264 (−384 to − 144)  <0.001  −0.16 (−0.24 to − 0.09)  <0.001    Model 3      Lymphocyteb (/µL)  −310 (−431 to − 188)  <0.001  −0.19 (−0.27 to − 0.12)  <0.001    CKD cohort  Dependent variable  Model 1      β (95% CI)  P-value  Lymphocyte (/µL)  −364 (−466 to − 261)  <0.001      Neutrophil (/µL)  −121 (−413–170)  0.41      Monocyte (/µL)  −85 (−122 to − 47)  <0.001      Leucocyte (/µL)  −605 (−948 to − 261)  0.001      Thrombocyte (×103/µL)  −28 (-40 to − 16)  <0.001          Model 2        β (95% CI)  P-value      Lymphocytec  −345 (−445 to − 245)  <0.001  −0.31 (−0.40 to − 0.22)  <0.001    Model 3      Lymphocyted  −333 (−434 to − 233)  <0.001  −0.30 (−0.39 to − 0.21)  <0.001  Depicted are unstandardized and standardized beta coefficients with 95% CIs and with PKD as an independent variable. Model 1 is unadjusted. a Model 2 = Model 1, adjusted for age, sex and lnCRP. b Model 3 = Model 1, adjusted for age, sex, lnCRP, diabetes and immunosuppressive drugs. c Model 2 = Model 1, adjusted for age, sex, lnCRP and eGFR. d Model 3 = Model 1, adjusted for age, sex, lnCRP, eGFR and diabetes. PKD patients had lower thrombocytes [mean difference −21.103/µL (95% CI −32.103 to −9.103), P < 0.001] and monocytes [mean difference −58/µL (95% CI −103 to −14), P = 0.01], while for neutrophils the difference was borderline significant (Table 2, Model 1). CKD cohort The population characteristics of the CKD cohort (n = 408) of age- and sex-matched individuals with CKD across comparable stages are presented in Table 3. The mean age was 49 ± 13 years and 43% were male, 7% had diabetes and the median eGFR was 35 (IQR 4–9) mL/min. There were no differences in CRP concentration and ethnicity. The prevalence of diabetes was ∼3% in the participants with versus 11% in those without PKD. Table 3 Characteristics of the CKD cohort Characteristics  PKD (n = 204)  No PKD (n = 204)  P-value  Age (years)  49.2 ± 13.2  49.0 ± 13.3  0.89  Male  88 (43.1)  87 (42.6)  0.92  Diabetes  7 (3.4)  23 (11.3)  0.002  Non-Caucasian race  1 (0.5)  4 (2.0)  0.18  Sedimentation (mm)  8 (12)  12 (19)  0.09  Serum CRP (mg/L)  2 (4.1)  2 (4.0)  0.96  Creatinine (mg/dL)  1.89 (2.5)  1.80 (1.95)  0.23  eGFR (mL/min)  33.3 (51.0)  38.1 (53.0)  0.23  Haemoglobin (g/dL)  13.3 ± 1.5  13.1 ± 1.6  0.47  Haematocrit (%)  39.9 ± 4.4  39.5 ± 4.6  0.44  Leucocytes (/µL)  6330 (2435)  6650 (2065)  0.001  Thrombocytes (×103/µL)  213 (70)  236 (74)  <0.001  Neutrophils (/µL)  3910 (2143)  4090 (2205)  0.63  Lymphocytes (/µL)  1610 (506)  1880 (710)  <0.001  Lymphopenia (%)  19.6  8.9  <0.001  Monocytes (/µL)  430 (235)  510 (225)  <0.001  Characteristics  PKD (n = 204)  No PKD (n = 204)  P-value  Age (years)  49.2 ± 13.2  49.0 ± 13.3  0.89  Male  88 (43.1)  87 (42.6)  0.92  Diabetes  7 (3.4)  23 (11.3)  0.002  Non-Caucasian race  1 (0.5)  4 (2.0)  0.18  Sedimentation (mm)  8 (12)  12 (19)  0.09  Serum CRP (mg/L)  2 (4.1)  2 (4.0)  0.96  Creatinine (mg/dL)  1.89 (2.5)  1.80 (1.95)  0.23  eGFR (mL/min)  33.3 (51.0)  38.1 (53.0)  0.23  Haemoglobin (g/dL)  13.3 ± 1.5  13.1 ± 1.6  0.47  Haematocrit (%)  39.9 ± 4.4  39.5 ± 4.6  0.44  Leucocytes (/µL)  6330 (2435)  6650 (2065)  0.001  Thrombocytes (×103/µL)  213 (70)  236 (74)  <0.001  Neutrophils (/µL)  3910 (2143)  4090 (2205)  0.63  Lymphocytes (/µL)  1610 (506)  1880 (710)  <0.001  Lymphopenia (%)  19.6  8.9  <0.001  Monocytes (/µL)  430 (235)  510 (225)  <0.001  Data are presented as number (%) or mean  ±  SD. The mean lymphocyte count in individuals with PKD was 364/µl lower than in those without PKD and lymphopenia was about twice as common in participants with PKD. Using the mean lymphocyte count of the last two visits, the difference in lymphocytes between patients with and without PKD remained comparable [360/µL (95% CI 260–460), P < 0.001]. Two measurements of lymphocyte counts were available in 90% of the patients. Similar to the findings in the ESKD cohort, lymphocyte, monocyte and thrombocyte counts were lower in the people with versus without PKD (P < 0.001 for all), whereas there was no difference in neutrophil count. Although the lymphocyte numbers were lower at more severe stages of CKD for all patients, the difference between lymphocyte counts between the two groups remained significant after stratification for CKD category (Figure 3, Supplementary Table S1, Figure S2), whereas the differences in leucocyte, thrombocyte and monocyte counts were not uniform across the stages of CKD (Supplementary Table S1). Haemoglobin concentrations were significantly lower in patients with more severe stages of kidney disease (P < 0.001), but concentrations were not different between people with and without PKD (Supplementary Table S1). FIGURE 2 View largeDownload slide Box plot of haematological differences between patients with ESKD with versus without polycystic kidney disease. Differences between patients with end-stage kidney failure with (n = 169) versus without (n = 780) polycystic kidney disease assessed at the time of transplantation. Bars represent the median number of cells per microlitre with IQR. (A) WBC, white blood cells; Neutro, neutrophils; Lympho, lymphocytes; Mono, monocytes. (B) Subset of lymphocytes: CD4, CD4+ T lymphocytes; CD8, CD8+ T lymphocytes; B, B lymphocytes; NK, natural killer cells. Differences for haematological tests between people with and without PKD; *P <0.05; ***P < 0.001. FIGURE 2 View largeDownload slide Box plot of haematological differences between patients with ESKD with versus without polycystic kidney disease. Differences between patients with end-stage kidney failure with (n = 169) versus without (n = 780) polycystic kidney disease assessed at the time of transplantation. Bars represent the median number of cells per microlitre with IQR. (A) WBC, white blood cells; Neutro, neutrophils; Lympho, lymphocytes; Mono, monocytes. (B) Subset of lymphocytes: CD4, CD4+ T lymphocytes; CD8, CD8+ T lymphocytes; B, B lymphocytes; NK, natural killer cells. Differences for haematological tests between people with and without PKD; *P <0.05; ***P < 0.001. FIGURE 3 View largeDownload slide Box plot of lymphocytes across categories of CKD in patients with and without PKD. Depicted are counts of lymphocytes across categories of CKD in individuals with versus without polycystic kidney disease. Bars represent median numbers of cells per microlitre with IQR. Differences for hematological tests between people with and without PKD; *P <0.05; **P<0.01; ***P < 0.001. FIGURE 3 View largeDownload slide Box plot of lymphocytes across categories of CKD in patients with and without PKD. Depicted are counts of lymphocytes across categories of CKD in individuals with versus without polycystic kidney disease. Bars represent median numbers of cells per microlitre with IQR. Differences for hematological tests between people with and without PKD; *P <0.05; **P<0.01; ***P < 0.001. Regression model When adjusted for age, sex, lnCRP and eGFR, PKD remained significantly associated with lymphocyte concentration [adjusted mean difference −345/µL (95% CI −445 to −245), P < 0.001] (Table 2). After further adjustment for diabetes, the difference between PKD and non-PKD was −333/µL (95% CI −434 to −233; P < 0.001). We did not observe any clinically meaningful significant interaction between PKD and the other covariates. Also, male sex and lower eGFR were associated with lower lymphocyte counts (Table 4). Using the mean lymphocyte count of the last two visits, the difference in lymphocytes between patients with and without PKD remained comparable [adjusted mean difference 290/µL (95% CI 185–395); P < 0.001]. Table 4 Multiple regression analysis of the association of PKD with lymphocyte counts ESKD cohort (dependent variable = lymphocyte count)            Coefficient  95% CI  β  95% CI  P-value   Age (per year)  −3  −6–1  −0.06  −0.13–0.02  0.16   Sex (male = 1, female = 0)  −83  −171–17  −0.07  −0.14–0.01  0.08   lnCRP  −21  −66–18  −0.04  −0.11–0.04  0.34   PKD (yes = 1, no  =  0)  −310  −431 to − 188  −0.19  −0.27 to − 0.12  <0.001   Diabetes (yes = 1, no  =  0)  −149  −276 to − 21  −0.09  −0.17 to − 0.01  0.022   Immunosuppression (yes = 1, no  =  0)  −206  −341 to − 70  −0.11  −0.18 to − 0.04  0.003  CKD cohort (dependent variable = lymphocyte count)             Age (per year)  1  −4–6  0.02  −0.10–0.13  0.76   Sex (male = 1, female = 0)  −147  −249 to − 46  −0.13  −0.22 to − 0.04  0.004   eGFR (/mL/min)  5  3–7  0.32  0.20–0.43  <0.001   lnCRP  4  −33–53  0.01  −0.08–0.10  0.84   PKD (yes = 1, no  =  0)  −333  −434 to − 233  −0.30  −0.39 to − 0.21  <0.001   Diabetes (yes = 1, no  =  0)  149  −53–350  0.07  −0.03–0.16  0.15  ESKD cohort (dependent variable = lymphocyte count)            Coefficient  95% CI  β  95% CI  P-value   Age (per year)  −3  −6–1  −0.06  −0.13–0.02  0.16   Sex (male = 1, female = 0)  −83  −171–17  −0.07  −0.14–0.01  0.08   lnCRP  −21  −66–18  −0.04  −0.11–0.04  0.34   PKD (yes = 1, no  =  0)  −310  −431 to − 188  −0.19  −0.27 to − 0.12  <0.001   Diabetes (yes = 1, no  =  0)  −149  −276 to − 21  −0.09  −0.17 to − 0.01  0.022   Immunosuppression (yes = 1, no  =  0)  −206  −341 to − 70  −0.11  −0.18 to − 0.04  0.003  CKD cohort (dependent variable = lymphocyte count)             Age (per year)  1  −4–6  0.02  −0.10–0.13  0.76   Sex (male = 1, female = 0)  −147  −249 to − 46  −0.13  −0.22 to − 0.04  0.004   eGFR (/mL/min)  5  3–7  0.32  0.20–0.43  <0.001   lnCRP  4  −33–53  0.01  −0.08–0.10  0.84   PKD (yes = 1, no  =  0)  −333  −434 to − 233  −0.30  −0.39 to − 0.21  <0.001   Diabetes (yes = 1, no  =  0)  149  −53–350  0.07  −0.03–0.16  0.15  Patients with PKD had lower thrombocyte [PKD versus non-PKD −28.103/µL (95% CI −40.103 to −16.103), P < 0.001] and monocyte counts [PKD versus non-PKD −85/µL (95% CI −122 to −47), P < 0.001], while neutrophil counts were not significantly lower (Table 2). DISCUSSION Patients with PKD and ESKD or CKD not on dialysis had lower lymphocyte counts than patients with other aetiologies of kidney failure. The strength of the association was comparable in both study groups and independent of confounders such as age, sex and CRP. In an age- and sex-matched multicentre cohort of patients stratified for CKD category, the association was consistent across all strata, further increasing the external validity of our findings. Since lymphocyte counts and especially of CD4+ T cells decrease progressively with declining kidney function in line with the previously reported incremental apoptosis with advancing CKD [15–18], lymphocyte numbers in patients with and without PKD can only be compared after adjustment for this confounder. The lower number of lymphocytes in patients with PKD is most likely clinically relevant. Lower numbers of lymphocytes in kidney transplant recipients were previously associated with mortality, the development of solid cancer, skin cancer and lymphoma [19–21]. According to Ducloux et al. [20], after adjustment for confounders, for each increase in 100/mm3 of CD4+ T cells there was a 27% lower risk of developing cancer, with a relative risk (RR) of 0.73 (95% CI 0.62–0.89). The difference in CD4+ counts at the time of transplantation in our analysis was 77 (95% CI 1–153). For each increase in B cells of 10/mm3 there was a 13% lower risk of cancer, with a RR of 0.87 (95% CI 0.59–1.02) while the difference in B cells between patients with and without PKD was 50 (95% CI 22–79) in our analysis. PKD increases the risk of squamous cell cutaneous, colon, liver and renal carcinoma in transplant recipients or the general population [2, 6, 22]. Unfortunately, data on lymphocyte counts are lacking in these analyses [2, 6, 22]. Lymphopenia is a well-established risk factor for infections in different study populations. Among kidney transplant recipients it is associated with the development of Pneumocystis jiroveci pneumonia (PJP) [23, 24] and in liver transplant recipients, pre-transplant lymphopenia is associated with an increased risk of cytomegalovirus (CMV) and non-CMV invasive infections after transplantation [25]. In another analysis, in liver transplant recipients, pre-transplant lymphocytes <1000/µL versus  ≥1000/µL were associated with an increased risk of infections after transplantation in a multivariate model, with an odds ratio of 10.1 (95% CI 1.9–39; P = 0.005) [26]. Prospective analyses should determine whether these baseline differences in lymphocyte counts translate into unfavorable outcomes in PKD patients before or after kidney transplantation. The differences in cellular differentiation were most pronounced in CD8+ T lymphocytes and B lymphocytes in the ESKD cohort (Figure 2), although we cannot exclude that the relative distribution of differences in lymphocyte counts is coincidental. In particular, the blood concentration of CD4+ T lymphocytes is expected to decrease with more advanced CKD, while concentrations of CD8+ T lymphocytes change less [18]. We can speculate on the causal mechanism of our findings. Aberrant expression of polycystin-1 and/or polycystin-2 could lead to decreased proliferation of lymphocytes, as has been shown in a study with Epstein–Barr virus immortalized lymphoblastoid cells [11]. An increased apoptosis rate of lymphoid and myeloid cells remains an interesting alternative hypothesis in line with increased apoptosis of renal tubular epithelial cells in PKD [1, 27]. A targeted deletion of the anti-apoptotic B-cell lymphoma 2 (Bcl-2) in mice was both associated with PKD and lymphopenia, especially of CD8+ lymphocytes, due to an increased rate of apoptosis [28]. Apoptotic pathways, including the intrinsic Bcl-2 pathway, determine the final concentration of circulating lymphocytes [29] and affect the generation and survival of thrombocytes as well [30]. Kidney cells with silenced polycystin-1 expression have increased degradation of Bcl-2, leading to more apoptosis [27]. Disturbed apoptosis could be the common denominator in PKD explaining the renal and haematological phenotype. As such, particularly lymphopenia might reveal itself as a novel biomarker of a more severe phenotype with a faster decline of kidney function. Obviously prospective analyses are needed. Already in the early stages of kidney failure, lymphopenia might characterize PKD. Of note, lymphopenia in our CKD cohort was more common in males, whose kidney function in PKD was recently shown to decline faster than in females [31]. Since not only lymphocyte but also monocyte and thrombocyte counts were consistently lower in both groups of patients with versus without PKD, we speculate that a distinct genetic defect affects the common progenitor cells, leading to decreased proliferation or increased apoptosis in an early phase of haematopoiesis. In a recent analysis, platelet counts were lower in 290 patients with PKD and ESKD than in age- and sex-matched controls (215 versus 238 × 103/µL; P < 0.01), whereas the difference was not significant in individuals with CKD and after kidney transplantation [32]. Also, thrombocytopenia in PKD seems to be associated with a more severe renal phenotype with a larger total kidney volume and faster increase of kidney volume and decline of kidney function [33]. Obviously, PKD is characterized by a higher incidence of erythrocytosis [34, 35], and a lower need for erythropoiesis-stimulating agents [36]. This suggests that other mechanisms (for instance, increased expression of hypoxia-inducible factor-1 and/or erythropoietin) might be at play. Our analysis has some limitations. First, the analysis is cross-sectional and as such it is impossible to make causal inferences. Yet, the data from our study align with existing in vitro data on the role of polycystin in lymphocyte proliferation in PKD. Second, data on immunophenotypic analysis in the ESKD cohort were incomplete within the constraints of retrospective data analysis. Also, the lack of a substantial number of lymphocyte subsets precluded imputations. Yet, we believe selection bias is unlikely, as all patients suitable for transplantation were included for analysis. Moreover, the proportion of patients with PKD in the ESKD cohort did not change after excluding patients without available data on lymphocyte counts (Figure 1). Other relevant characteristics did not differ between the groups, suggesting that there was no informed missing data. Third, we cannot exclude that hidden confounders are still present. And while CRP concentrations were not different between groups with and without PKD, we cannot exclude that patients with PKD have some degree of chronic inflammation that was not entirely captured by CRP. By excluding patients with active infection and malignancy, we believe we avoided this confounder to a large extent. CONCLUSION PKD is associated with lymphopenia and especially with lower blood concentrations of CD8+ T and B lymphocytes. This finding could explain the increased risk of pneumonia, urinary tract infections and non-melanoma skin cancer after transplantation in these patients, which might warrant less aggressive immunosuppression in PKD patients after transplantation. This finding can potentially aid in risk stratification of people with PKD. If mutations in polycystin-1 and/or -2 lead to decreased proliferation of lymphocytes or increased apoptosis of renal tubular epithelial cells and lymphocytes via a common pathway, the degree of lymphopenia might reflect the genotype–phenotype interaction in PKD. As such, lymphopenia could serve as an inexpensive and easily accessible biomarker of a more severe phenotype with a faster decline of kidney function, aneurysm formation or cardiovascular involvement. This hypothesis is amenable to further exploration by prospective observational studies with wide application of disease markers including renal imaging. Also, our study asks for the clarification of a potential role of polycystin in the pathophysiology of haematological disorders and in different health conditions including PKD. SUPPLEMENTARY DATA Supplementary data are available online at http://ndt.oxfordjournals.org. CONFLICT OF INTEREST STATEMENT None declared. REFERENCES 1 Wilson PD. Polycystic kidney disease. N Engl J Med  2004; 350: 151– 164 Google Scholar CrossRef Search ADS PubMed  2 Bretagnol A, Halimi JM, Roland M et al.   Autosomal dominant polycystic kidney disease: risk factor for nonmelanoma skin cancer following kidney transplantation. Transpl Int  2010; 23: 878– 886 Google Scholar PubMed  3 Sallée M, Rafat C, Zahar JR et al.   Cyst infections in patients with autosomal dominant polycystic kidney disease. Clin J Am Soc Nephrol  2009; 4: 1183– 1189 Google Scholar CrossRef Search ADS PubMed  4 Nielsen LH, Jensen-Fangel S, Jespersen B et al.   Risk and prognosis of hospitalization for pneumonia among individuals with and without functioning renal transplants in Denmark: a population-based study. Clin Infect Dis  2012; 55: 679– 686 Google Scholar CrossRef Search ADS PubMed  5 Moua T, Zand L, Hartman RP et al.   Radiologic and clinical bronchiectasis associated with autosomal dominant polycystic kidney disease. PLoS One  2014; 9: e93674 Google Scholar CrossRef Search ADS PubMed  6 Otley CC, Cherikh WS, Salasche SJ et al.   Skin cancer in organ transplant recipients: effect of pretransplant end-organ disease. J Am Acad Dermatol  2005; 53: 783– 790 Google Scholar CrossRef Search ADS PubMed  7 Pedrozo Z, Criollo A, Battiprolu P et al.   Polycystin-1 Is a cardiomyocyte mechanosensor that governs L-type Ca2+ channel protein stability. Circulation  2015; 131: 2141– 2142 Google Scholar CrossRef Search ADS   8 Nauli SM, Kawanabe Y, Kaminski JJ et al.   Endothelial cilia are fluid shear sensors that regulate calcium signaling and nitric oxide production through polycystin-1. Circulation  2008; 117: 1161– 1171 Google Scholar CrossRef Search ADS PubMed  9 Klawitter J, Reed-Gitomer BY, McFann K et al.   Endothelial dysfunction and oxidative stress in polycystic kidney disease. Am J Physiol Renal Physiol  2015; 307: F1198– F1206 Google Scholar CrossRef Search ADS   10 Nowak KL, Farmer H, Cadnapaphornchai MA et al.   Vascular dysfunction in children and young adults with autosomal dominant polycystic kidney disease. Nephrol Dial Transplant  2017; 32: 342– 347 Google Scholar CrossRef Search ADS PubMed  11 Aguiari G, Banzi M, Gessi S et al.   Deficiency of polycystin-2 reduces Ca2+ channel activity and cell proliferation in ADPKD lymphoblastoid cells. FASEB J  2004; 18: 884– 886 Google Scholar CrossRef Search ADS PubMed  12 Banerjee A, Chandna S, Jayasena D et al.   Leucopenia in adult polycystic kidney disease patients on haemodialysis. Nephron  2002; 91: 175– 176 Google Scholar CrossRef Search ADS PubMed  13 Chamberlain JJ, Rhinehart AS, Shaefer JCF et al.   Diagnosis and management of diabetes: synopsis of the 2016 American Diabetes Association Standards of Medical Care in Diabetes Synopsis of the 2016 ADA Standards of Medical Care in Diabetes. Ann Intern Med  2016; 19: 542– 552 Google Scholar CrossRef Search ADS   14 Levey AS, Stevens LA, Schmid CH et al.   A new equation to estimate glomerular filtration rate. Ann Intern Med  2009; 150: 604– 612 Google Scholar CrossRef Search ADS PubMed  15 Yoon JW, Gollapudi S, Pahl MV et al.   Naïve and central memory T-cell lymphopenia in end-stage renal disease. Kidney Int  2006; 70: 371– 376 Google Scholar CrossRef Search ADS PubMed  16 Agarwal R, Light RP. Patterns and prognostic value of total and differential leukocyte count in chronic kidney disease. Clin J Am Soc Nephrol  2011; 6: 1393– 1398 Google Scholar CrossRef Search ADS PubMed  17 Saad K, Elsayh KI, Zahran AM et al.   Lymphocyte populations and apoptosis of peripheral blood B and T lymphocytes in children with end stage renal disease. Ren Fail  2014; 36: 502– 507 Google Scholar CrossRef Search ADS PubMed  18 Litjens NH, van Druningen CJ, Betjes MG. Progressive loss of renal function is associated with activation and depletion of naive T lymphocytes. Clin Immunol  2006; 118: 83– 91 Google Scholar CrossRef Search ADS PubMed  19 Ducloux D, Courivaud C, Bamoulid J et al.   Prolonged CD4 T cell lymphopenia increases morbidity and mortality after renal transplantation. J Am Soc Nephrol  2010; 21: 868– 875 Google Scholar CrossRef Search ADS PubMed  20 Ducloux D, Carron PL, Motte G et al.   Lymphocyte subsets and assessment of cancer risk in renal transplant recipients. Transpl Int  2002; 15: 393– 396 Google Scholar CrossRef Search ADS PubMed  21 Ducloux D, Carron PL, Rebibou JM et al.   CD4 lymphocyteopenia as a risk factor for skin cancers in renal transplant recipients. Transplantation  1998; 65: 1270– 1272 Google Scholar CrossRef Search ADS PubMed  22 Yu TM, Chuang YW, Yu MC et al.   Risk of cancer in patients with polycystic kidney disease: a propensity-score matched analysis of a nationwide, population-based cohort study. Lancet Oncol  2016; 17: 1419– 1425 Google Scholar CrossRef Search ADS PubMed  23 Mulpuru S, Knoll G, Weir C et al.   Pneumocystis pneumonia outbreak among renal transplant recipients at a North American transplant center: risk factors and implications for infection control. Am J Infect Control  2016; 44: 425– 431 Google Scholar CrossRef Search ADS PubMed  24 Brakemeier S, Dürr P, Bachmann F et al.   Risk evaluation and outcome of Pneumocystis jirovecii pneumonia in kidney transplant patients. Transplant Proc  2016; 48: 2924– 2930 Google Scholar CrossRef Search ADS PubMed  25 Nierenberg NE, Poutsiaka DD, Chow JK et al.   Pretransplant lymphopenia is a novel prognostic factor in cytomegalovirus and noncytomegalovirus invasive infections after liver transplantation. Liver Transplant  2014; 20: 1497– 1507 26 Fernandez-Ruiz M, Lopez-Medrano F, Romo EM et al.   Pretransplant lymphocyte counts predicts the incidence of infection during the first two years after liver transplantation. Liver Transplant  2009; 15: 1209– 1216 Google Scholar CrossRef Search ADS   27 Yu W, Kong T, Beaudry S et al.   Polycystin-1 protein level determines activity of the Gα12/JNK apoptosis pathway. J Biol Chem  2010; 285: 10243– 10251 Google Scholar CrossRef Search ADS PubMed  28 Nakayama K, Nakayama K, Negishi I et al.   Targeted disruption of Bcl-2 alpha beta in mice: occurrence of gray hair, polycystic kidney disease, and lymphocytopenia. Proc Natl Acad Sci USA  1994; 91: 1300– 1304 Google Scholar CrossRef Search ADS   29 Dunkle A, He YW. Apoptosis and autophagy in the regulation of T lymphocyte function. Immunol Res  2011; 49: 70– 86 Google Scholar CrossRef Search ADS PubMed  30 Kile BT. The role of apoptosis in megakaryocytes and platelets. B J Haematol  2014; 165: 217– 226 Google Scholar CrossRef Search ADS   31 Cornec-Le Gall E, Audrézet M-P, Rousseau A et al.   The PROPKD score: a new algorithm to predict renal survival in autosomal dominant polycystic kidney disease. J Am Soc Nephrol  2016; 27: 942– 951 Google Scholar CrossRef Search ADS PubMed  32 Setyapranata S, Holt SG. Platelet counts in autosomal dominant polycystic kidney disease. Platelets  2016; 27: 262– 263 Google Scholar CrossRef Search ADS PubMed  33 Chen D, Ma Y, Wang X et al.   Clinical characteristics and disease predictors of a large Chinese cohort of patients with autosomal dominant polycystic kidney disease. PLoS One  2014; 20: e92232 Google Scholar CrossRef Search ADS   34 Nankivell BJ, Allen RD, O'Connell PJ et al.   Erythrocytosis after renal transplantation: risk factors and relationship with GFR. Clin Transplant  1995; 9: 375– 382 Google Scholar PubMed  35 Hadimeri H, Nordén G, Friman S et al.   Autosomal dominant polycystic kidney disease in a kidney transplant population. Nephrol Dial Transplant  1997; 12: 1431– 1436 Google Scholar CrossRef Search ADS PubMed  36 Shah A, Molnar MZ, Lukowsky LR et al.   Hemoglobin level and survival in hemodialysis patients with polycystic kidney disease and the role of administered erythropoietin. Am J Hematol  2012; 87: 833– 836 Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nephrology Dialysis Transplantation Oxford University Press

Hereditary polycystic kidney disease is characterized by lymphopenia across all stages of kidney dysfunction: an observational study

Loading next page...
 
/lp/ou_press/hereditary-polycystic-kidney-disease-is-characterized-by-lymphopenia-fK9evA6Ijy
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
ISSN
0931-0509
eISSN
1460-2385
D.O.I.
10.1093/ndt/gfx040
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Polycystic kidney disease (PKD) is characterized by urinary tract infections and extrarenal abnormalities such as an increased risk of cancer. As mutations in polycystin-1 and -2 are associated with decreased proliferation of immortalized lymphoblastoid cells in PKD, we investigated whether lymphopenia could be an unrecognized trait of PKD. Methods We studied 700 kidney transplant recipients with (n = 126) or without PKD at the time of kidney transplantation between 1 January 2003 and 31 December 2014 at Ghent University Hospital. We also studied 204 patients with chronic kidney disease (CKD) with PKD and 204 matched CKD patients without PKD across comparable CKD strata with assessment between 1 January 1999 and 1 February 2016 at three renal outpatient clinics. We compared lymphocyte counts with multiple linear regression analysis to adjust for potential confounders. We analysed flow cytometric immunophenotyping data and other haematological parameters. Results Lymphocyte counts were 264/µL [95% confidence interval (CI) 144–384] and 345/µL (95% CI 245–445) (both P < 0.001) lower in the end-stage kidney disease (ESKD) and CKD cohort, respectively, after adjustment for age, sex, ln(C-reactive protein) and estimated glomerular filtration rate (in the CKD cohort only). In particular, CD8+ T and B lymphocytes were significantly lower in transplant recipients with versus without PKD (P < 0.001 for both). Thrombocyte and monocyte counts were lower in patients with versus without PKD in both cohorts (P < 0.001 for all analyses except P = 0.01 for monocytes in the ESKD cohort). Conclusion PKD is characterized by distinct cytopenias and especially lymphopenia, independent of kidney function. This finding has the potential to alter our therapeutic approach to patients with PKD. apoptosis, lymphopenia, polycystic kidney, proliferation, uremia INTRODUCTION Hereditary autosomal dominant polycystic kidney disease (PKD) is caused by mutations in genes coding for polycystin-1 and more rarely polycystin-2 and is the most common monogenic disorder, affecting ∼1 in 500 people. It is the primary cause of end-stage kidney disease (ESKD) in 5–15% of those treated with renal replacement therapy [1, 2]. The disease has a particular phenotype that includes a tendency to develop aneurysms, cardiac valve disorders, abdominal wall hernias and hypertension. Patients with PKD also tend to easily contract ascending urinary tract infections, develop bronchiectasis more readily than chronic kidney disease (CKD) patients without PKD, with an adjusted odds ratio (AOR) of 2.78, and have a higher incidence of pneumonia and nonmelanoma skin cancer, with an AOR of 1.65, than kidney transplant recipients without PKD [2, 3–6]. This suggests an impaired immune response may be at play. The proteins polycystin-1 and -2, which form a functional complex, are present in renal tubular epithelial cells. If one of these proteins is dysfunctional, this causes disturbed calcium influx, increased apoptosis and cystogenesis, resulting in the PKD type 1 and 2 phenotypes, respectively [1]. Polycystin-1 and polycystin-2 are not only present in renal tubular cells, but also in cardiomyocytes [7] and endothelial cilia [8], and their dysfunction could explain the early development of cardiac hypertrophy and endothelial dysfunction in PKD, even with preserved kidney function and at a very young age [9, 10]. It is uncertain whether decreased proliferation of lymphoblastoid cells in PKD in vitro translates into fewer circulating lymphocytes in vivo [11]. We hypothesized that lymphopenia could be more common in PKD, which could explain a higher incidence of infections and skin cancer in this particular population. In 2002, Banerjee et al. [12] conducted a small unadjusted cross-sectional case–control study and found that among patients treated with haemodialysis, those with PKD (n = 11) had 610/µL (40%) fewer lymphocytes compared with the age- and sex-matched controls without PKD (n = 33). Aim We wanted to determine whether lymphocyte counts were lower in patients with PKD than in controls, independent of the effects of kidney function and after adjustment for confounders such as age and inflammation. We also wanted to verify whether differences in lymphocyte counts were confined to specific lymphocyte subpopulations. In secondary analyses, numbers of other leucocyte subtypes and platelets were analysed. MATERIALS AND METHODS Study settings In the first study, we compared lymphocyte counts of all patients with ESKD with versus without PKD who were admitted to the Ghent University Hospital for kidney transplantation (ESKD cohort). In a second analysis, we gathered data on lymphocyte counts from stable outpatient clinic patients with CKD and PKD from three Belgian hospitals: Ghent University Hospital, AZ Delta Hospital Roeselare and OLV Hospital Aalst (CKD cohort). We compared these participants with age- and sex-matched controls across the same categories of CKD (stage G1–5) but with an alternative renal diagnosis. The institutional review board and ethics committee of Ghent University Hospital reviewed and approved the research protocol (EC/2015/0655). Time of assessment In the people with ESKD and waitlisted for transplantation, we measured the lymphocyte count immediately before transplantation. In the second analysis, we retrieved data from the last measurement at the outpatient clinic from January 1999 to 1 February 2016. Study population The study population included: First cohort: all patients >18 years of age with ESKD who received a first or subsequent kidney transplant in Ghent University Hospital between 1 January 2003 and 31 December 2014. Second cohort: all patients >18 years of age with hereditary PKD and outpatient follow-up and free of active or recent (<1 month) infection at the time of assessment, intake of immunosuppressive drugs, active cancer or renal replacement therapy next to age- and sex-matched controls with comparable kidney function (1:1). Excluded were patients with acquired cystic disease, renal dysplasia, von Hippel–Lindau disease or autosomal recessive polycystic disease. Measurements We selected a priori variables considered for analysis on their theoretical association with lymphocyte count: age, sex, estimated glomerular filtration rate (eGFR) and C-reactive protein (CRP). We extracted the following baseline clinical characteristics of the ESKD cohort from the transplant database of the Ghent University Hospital, which contains prospectively collected data on baseline characteristics and outcomes of all kidney transplant recipients: continuous variables included age (years) and time on the waiting list (days) and categorical variables included sex (coded as male = 1, female = 0); diabetes at the time of assessment (yes/no), defined according to the American Diabetes Association criteria [13] or as intake of glucose-lowering drugs according to patient records; ethnicity (coded as Caucasian = 0, non-Caucasian = 1); dialysis modality (coded as pre-emptive transplantation = 1, haemodialysis = 2, peritoneal dialysis = 3); retransplantation (yes/no) and intake of immunosuppressive drugs before transplantation (yes/no). We extracted from the medical files the following baseline biochemical variables: lymphocytes, haemoglobin, leucocytes, thrombocytes, neutrophils, monocytes, serum albumin, CRP and body mass index (BMI), calculated as weight in kilograms divided by the square of height in metres. If data on flow cytometric immunophenotyping analysis were available at the time of transplantation, we included counts of CD4+ T lymphocytes, CD8+ T lymphocytes, CD19+ B lymphocytes and natural killer (NK) cells. Lymphopenia was defined as <1133 lymphocytes/μL of blood according to the laboratory reference values. All participating centres use automated cell counters (Sysmex, Kobe, Japan). For the CKD cohort, we recorded serum creatinine and calculated eGFR based on the Chronic Kidney Disease Epidemiology Collaboration formula [14] and gathered data on haemoglobin, haematocrit, thrombocytes, leucocytes, monocytes, neutrophils, lymphocytes, erythrocyte sedimentation rate and CRP from the respective databases of each participating centre. In additional analyses, we examined whether there were differences between leucocyte subpopulations and thrombocyte counts in patients with versus without PKD. Statistical analysis We summarized baseline characteristics as means and standard deviations (SDs) or medians with interquartile range (IQR) for continuous and percentages for categorical variables. We analysed differences between groups with independent sample t-tests, Mann–Whitney U-test, analysis of variance, Kruskal–Wallis test and chi-square test depending on the distribution. If appropriate, we used post hoc Scheffé testing to account for multiple comparisons. We used multiple linear regression analysis to study the relation between PKD and lymphocyte concentration as a dependent variable. CRP was natural log transformed (lnCRP) because of its skewed distribution. We constructed models adjusting for a priori specified covariates based on previous literature: age, sex, lnCRP and eGFR (CKD cohort). Additionally, we adjusted for covariates that were statistically significant (P < 0.1) in each separate univariate model. We tested for interactions between PKD and other covariates such as age and eGFR. We calculated the variance inflation factor to evaluate potential multicollinearity. In the CKD cohort, we matched each case of PKD for age, sex and CKD stage with an individual with an alternative renal diagnosis. We additionally analysed the mean lymphocyte count of the last two consultations as internal validation. We performed similar analyses to evaluate the association between PKD and leucocyte, monocyte and neutrophil counts, but refrained from multivariable modeling for each haematological covariate considering our main study question and also taking into account the risk of committing type I error due to multiple testing. A two-sided P-value <0.05 was considered to indicate statistical significance for all analyses. We conducted all statistical analyses using SPSS Statistics (version 22.0; IBM, Armonk, NY, USA). RESULTS ESKD cohort Between January 2003 and December 2014, 949 patients (mean age of 52 ± 12 years, 63% male, 17% diabetes, 95% Caucasian and 12% on immunosuppressive drugs at the time of transplantation) received a kidney transplant at Ghent University Hospital. We excluded 249 patients because the lymphocyte count was unavailable, leaving 700 included in the study. Flow cytometric analysis data were available in 484 participants. The proportion of participants with PKD was similar in the 249 people with missing data (Figure 1). There were no missing data on any other covariate. Participants with versus without PKD were 4 years older, were less frequently male and had a comparable CRP concentration. PKD patients had a lower BMI, less frequently had diabetes and only ∼5% were taking immunosuppressive drugs versus 14% in those without PKD (Table 1). Table 1 Population characteristics of the subpopulation with end-stage kidney failure Characteristics  Polycystic kidney disease (n = 169)  No polycystic kidney disease (n = 780)  P-value  Age (years)  55.3 ± 8.2  51.6 ± 13.3  0.001  Men  96 (57.0)  509 (65.2)  0.012  BMI (kg/m2)  24.7 ± 3.6  25.5 ± 4.6  0.035  Diabetes  5 (3.0)  159 (20.4)  <0.0001  Non-Caucasian  3 (1.8)  46 (5.9)  0.028  Serum albumin concentration (g/L)  43.9 ± 4.9  42.9 ± 5.9  0.038  Serum CRP (mg/L)  3.0 (5.0)  3.0 (6.0)  0.81  Previous kidney transplantation  5 (3.0)  52 (6.7)  0.07  Living donation  13 (7.7)  59 (7.6)  0.98  Dialysis vintage (days)  627 (848)  566 (835)  0.32  Immunosuppression  9 (5.3)  107 (13.7)  0.002  Dialysis modality      0.16  Pre-emptive  23 (13.6)  69 (8.9)    Haemodialysis  112 (66.3)  536 (68.7)    Peritoneal dialysis  34 (20.1)  175 (22.4)    Haemoglobin (g/dL)  13.0 ± 1.7  12.8 ± 1.5  0.07  Leucocytes (/µL)  6250 (2050)  6700 (2855)  <0.001  Thrombocytes (×103/µL)  199 (89)  222 (89)  <0.001  Neutrophils (/µL)  3800 (1700)  4060 (2350)  0.05  Lymphopenia (%)  38.1  24.0  0.001  Lymphocytes (/µL)a  1330 (760)  1550 (750)  <0.001  CD4+ T lymphocytes (/µL)b  690 (391)  738 (405)  0.06  CD8+ T lymphocytes (/µL)b  274 (187)  304 (248)  <0.001  CD19 lymphocytes (/µL)b  92 (70)  143 (130)  <0.001  NK cells (/µL)c  234 (143)  211 (202)  0.39  Monocytes (/µL)b  550 (250)  590 (285)  0.01  Characteristics  Polycystic kidney disease (n = 169)  No polycystic kidney disease (n = 780)  P-value  Age (years)  55.3 ± 8.2  51.6 ± 13.3  0.001  Men  96 (57.0)  509 (65.2)  0.012  BMI (kg/m2)  24.7 ± 3.6  25.5 ± 4.6  0.035  Diabetes  5 (3.0)  159 (20.4)  <0.0001  Non-Caucasian  3 (1.8)  46 (5.9)  0.028  Serum albumin concentration (g/L)  43.9 ± 4.9  42.9 ± 5.9  0.038  Serum CRP (mg/L)  3.0 (5.0)  3.0 (6.0)  0.81  Previous kidney transplantation  5 (3.0)  52 (6.7)  0.07  Living donation  13 (7.7)  59 (7.6)  0.98  Dialysis vintage (days)  627 (848)  566 (835)  0.32  Immunosuppression  9 (5.3)  107 (13.7)  0.002  Dialysis modality      0.16  Pre-emptive  23 (13.6)  69 (8.9)    Haemodialysis  112 (66.3)  536 (68.7)    Peritoneal dialysis  34 (20.1)  175 (22.4)    Haemoglobin (g/dL)  13.0 ± 1.7  12.8 ± 1.5  0.07  Leucocytes (/µL)  6250 (2050)  6700 (2855)  <0.001  Thrombocytes (×103/µL)  199 (89)  222 (89)  <0.001  Neutrophils (/µL)  3800 (1700)  4060 (2350)  0.05  Lymphopenia (%)  38.1  24.0  0.001  Lymphocytes (/µL)a  1330 (760)  1550 (750)  <0.001  CD4+ T lymphocytes (/µL)b  690 (391)  738 (405)  0.06  CD8+ T lymphocytes (/µL)b  274 (187)  304 (248)  <0.001  CD19 lymphocytes (/µL)b  92 (70)  143 (130)  <0.001  NK cells (/µL)c  234 (143)  211 (202)  0.39  Monocytes (/µL)b  550 (250)  590 (285)  0.01  Data are presented as number (%) or mean ± SD. a Results were available in 700 patients. b Results were available in 484 patients. c Results were available in 346 patients. Participants with PKD had on average 269 (95% CI 150–388; P < 0.001; 17%) fewer lymphocytes/μl than those without PKD. When defined as <1133 lymphocytes/μl, lymphopenia was ∼1.6 times as common for those with PKD. Blood concentrations of CD8+ T lymphocytes and CD19+ B lymphocytes were 25% and 31% lower, respectively, in people with versus without PKD (both P < 0.001). Patients with PKD also had lower total leucocyte and thrombocyte counts (P < 0.001 for both) (Table 1, Figure 2, Supplementary Figure S1). FIGURE 1 View largeDownload slide Flow diagram. There was a comparable proportion of polycystic disease in the original study population [169 participants (17.9%)] and in the study population [126 participants (18%)]. FIGURE 1 View largeDownload slide Flow diagram. There was a comparable proportion of polycystic disease in the original study population [169 participants (17.9%)] and in the study population [126 participants (18%)]. Regression model When adjusted for age, sex and lnCRP, patients with PKD had on average 264 fewer lymphocytes compared with those without PKD [adjusted mean difference −264 cells/µL (95% CI −384 to −144 cells/µL), P < 0.001] (Table 2, Model 2). After further adjustment for intake of immunosuppression and diabetes, participants with PKD had on average 310 fewer lymphocytes compared with those without PKD [adjusted mean difference −310 cells/µL (95% CI −431 to −188 cells/µL)] (Table 2, Model 3). Table 2 Multiple regression models of the associations of PKD with lymphocyte, neutrophil, monocyte, leucocyte and thrombocyte counts ESKD cohort   Dependent variable  Model 1     β (95% CI)  P-value  β (95% CI)  P-value  Lymphocyte (/µL)  −269 (−388 to − 150)  <0.001      CD4+ T lymphocyte (/µL)  −77 (−153 to − 1)  0.05      CD8+ T lymphocyte (/µL)  −95 (−145 to − 45)  <0.001      B lymphocyte (/µL)  −50 (−79 to − 22)  <0.001      NK cell (/µL)  5 (−34–44)  0.80      Neutrophil (/µL)  −432 (−797 to − 67)  0.02      Monocyte (/µL)  −58 (−103 to − 14)  0.01      Leucocyte (/µL)  −765 (−1132 to − 399)  <0.001      Thrombocyte (×103/µL)  −21 (-32 to − 9)  <0.001          Model 2        β (95% CI)  P-value      Lymphocytea (/µL)  −264 (−384 to − 144)  <0.001  −0.16 (−0.24 to − 0.09)  <0.001    Model 3      Lymphocyteb (/µL)  −310 (−431 to − 188)  <0.001  −0.19 (−0.27 to − 0.12)  <0.001    CKD cohort  Dependent variable  Model 1      β (95% CI)  P-value  Lymphocyte (/µL)  −364 (−466 to − 261)  <0.001      Neutrophil (/µL)  −121 (−413–170)  0.41      Monocyte (/µL)  −85 (−122 to − 47)  <0.001      Leucocyte (/µL)  −605 (−948 to − 261)  0.001      Thrombocyte (×103/µL)  −28 (-40 to − 16)  <0.001          Model 2        β (95% CI)  P-value      Lymphocytec  −345 (−445 to − 245)  <0.001  −0.31 (−0.40 to − 0.22)  <0.001    Model 3      Lymphocyted  −333 (−434 to − 233)  <0.001  −0.30 (−0.39 to − 0.21)  <0.001  ESKD cohort   Dependent variable  Model 1     β (95% CI)  P-value  β (95% CI)  P-value  Lymphocyte (/µL)  −269 (−388 to − 150)  <0.001      CD4+ T lymphocyte (/µL)  −77 (−153 to − 1)  0.05      CD8+ T lymphocyte (/µL)  −95 (−145 to − 45)  <0.001      B lymphocyte (/µL)  −50 (−79 to − 22)  <0.001      NK cell (/µL)  5 (−34–44)  0.80      Neutrophil (/µL)  −432 (−797 to − 67)  0.02      Monocyte (/µL)  −58 (−103 to − 14)  0.01      Leucocyte (/µL)  −765 (−1132 to − 399)  <0.001      Thrombocyte (×103/µL)  −21 (-32 to − 9)  <0.001          Model 2        β (95% CI)  P-value      Lymphocytea (/µL)  −264 (−384 to − 144)  <0.001  −0.16 (−0.24 to − 0.09)  <0.001    Model 3      Lymphocyteb (/µL)  −310 (−431 to − 188)  <0.001  −0.19 (−0.27 to − 0.12)  <0.001    CKD cohort  Dependent variable  Model 1      β (95% CI)  P-value  Lymphocyte (/µL)  −364 (−466 to − 261)  <0.001      Neutrophil (/µL)  −121 (−413–170)  0.41      Monocyte (/µL)  −85 (−122 to − 47)  <0.001      Leucocyte (/µL)  −605 (−948 to − 261)  0.001      Thrombocyte (×103/µL)  −28 (-40 to − 16)  <0.001          Model 2        β (95% CI)  P-value      Lymphocytec  −345 (−445 to − 245)  <0.001  −0.31 (−0.40 to − 0.22)  <0.001    Model 3      Lymphocyted  −333 (−434 to − 233)  <0.001  −0.30 (−0.39 to − 0.21)  <0.001  Depicted are unstandardized and standardized beta coefficients with 95% CIs and with PKD as an independent variable. Model 1 is unadjusted. a Model 2 = Model 1, adjusted for age, sex and lnCRP. b Model 3 = Model 1, adjusted for age, sex, lnCRP, diabetes and immunosuppressive drugs. c Model 2 = Model 1, adjusted for age, sex, lnCRP and eGFR. d Model 3 = Model 1, adjusted for age, sex, lnCRP, eGFR and diabetes. PKD patients had lower thrombocytes [mean difference −21.103/µL (95% CI −32.103 to −9.103), P < 0.001] and monocytes [mean difference −58/µL (95% CI −103 to −14), P = 0.01], while for neutrophils the difference was borderline significant (Table 2, Model 1). CKD cohort The population characteristics of the CKD cohort (n = 408) of age- and sex-matched individuals with CKD across comparable stages are presented in Table 3. The mean age was 49 ± 13 years and 43% were male, 7% had diabetes and the median eGFR was 35 (IQR 4–9) mL/min. There were no differences in CRP concentration and ethnicity. The prevalence of diabetes was ∼3% in the participants with versus 11% in those without PKD. Table 3 Characteristics of the CKD cohort Characteristics  PKD (n = 204)  No PKD (n = 204)  P-value  Age (years)  49.2 ± 13.2  49.0 ± 13.3  0.89  Male  88 (43.1)  87 (42.6)  0.92  Diabetes  7 (3.4)  23 (11.3)  0.002  Non-Caucasian race  1 (0.5)  4 (2.0)  0.18  Sedimentation (mm)  8 (12)  12 (19)  0.09  Serum CRP (mg/L)  2 (4.1)  2 (4.0)  0.96  Creatinine (mg/dL)  1.89 (2.5)  1.80 (1.95)  0.23  eGFR (mL/min)  33.3 (51.0)  38.1 (53.0)  0.23  Haemoglobin (g/dL)  13.3 ± 1.5  13.1 ± 1.6  0.47  Haematocrit (%)  39.9 ± 4.4  39.5 ± 4.6  0.44  Leucocytes (/µL)  6330 (2435)  6650 (2065)  0.001  Thrombocytes (×103/µL)  213 (70)  236 (74)  <0.001  Neutrophils (/µL)  3910 (2143)  4090 (2205)  0.63  Lymphocytes (/µL)  1610 (506)  1880 (710)  <0.001  Lymphopenia (%)  19.6  8.9  <0.001  Monocytes (/µL)  430 (235)  510 (225)  <0.001  Characteristics  PKD (n = 204)  No PKD (n = 204)  P-value  Age (years)  49.2 ± 13.2  49.0 ± 13.3  0.89  Male  88 (43.1)  87 (42.6)  0.92  Diabetes  7 (3.4)  23 (11.3)  0.002  Non-Caucasian race  1 (0.5)  4 (2.0)  0.18  Sedimentation (mm)  8 (12)  12 (19)  0.09  Serum CRP (mg/L)  2 (4.1)  2 (4.0)  0.96  Creatinine (mg/dL)  1.89 (2.5)  1.80 (1.95)  0.23  eGFR (mL/min)  33.3 (51.0)  38.1 (53.0)  0.23  Haemoglobin (g/dL)  13.3 ± 1.5  13.1 ± 1.6  0.47  Haematocrit (%)  39.9 ± 4.4  39.5 ± 4.6  0.44  Leucocytes (/µL)  6330 (2435)  6650 (2065)  0.001  Thrombocytes (×103/µL)  213 (70)  236 (74)  <0.001  Neutrophils (/µL)  3910 (2143)  4090 (2205)  0.63  Lymphocytes (/µL)  1610 (506)  1880 (710)  <0.001  Lymphopenia (%)  19.6  8.9  <0.001  Monocytes (/µL)  430 (235)  510 (225)  <0.001  Data are presented as number (%) or mean  ±  SD. The mean lymphocyte count in individuals with PKD was 364/µl lower than in those without PKD and lymphopenia was about twice as common in participants with PKD. Using the mean lymphocyte count of the last two visits, the difference in lymphocytes between patients with and without PKD remained comparable [360/µL (95% CI 260–460), P < 0.001]. Two measurements of lymphocyte counts were available in 90% of the patients. Similar to the findings in the ESKD cohort, lymphocyte, monocyte and thrombocyte counts were lower in the people with versus without PKD (P < 0.001 for all), whereas there was no difference in neutrophil count. Although the lymphocyte numbers were lower at more severe stages of CKD for all patients, the difference between lymphocyte counts between the two groups remained significant after stratification for CKD category (Figure 3, Supplementary Table S1, Figure S2), whereas the differences in leucocyte, thrombocyte and monocyte counts were not uniform across the stages of CKD (Supplementary Table S1). Haemoglobin concentrations were significantly lower in patients with more severe stages of kidney disease (P < 0.001), but concentrations were not different between people with and without PKD (Supplementary Table S1). FIGURE 2 View largeDownload slide Box plot of haematological differences between patients with ESKD with versus without polycystic kidney disease. Differences between patients with end-stage kidney failure with (n = 169) versus without (n = 780) polycystic kidney disease assessed at the time of transplantation. Bars represent the median number of cells per microlitre with IQR. (A) WBC, white blood cells; Neutro, neutrophils; Lympho, lymphocytes; Mono, monocytes. (B) Subset of lymphocytes: CD4, CD4+ T lymphocytes; CD8, CD8+ T lymphocytes; B, B lymphocytes; NK, natural killer cells. Differences for haematological tests between people with and without PKD; *P <0.05; ***P < 0.001. FIGURE 2 View largeDownload slide Box plot of haematological differences between patients with ESKD with versus without polycystic kidney disease. Differences between patients with end-stage kidney failure with (n = 169) versus without (n = 780) polycystic kidney disease assessed at the time of transplantation. Bars represent the median number of cells per microlitre with IQR. (A) WBC, white blood cells; Neutro, neutrophils; Lympho, lymphocytes; Mono, monocytes. (B) Subset of lymphocytes: CD4, CD4+ T lymphocytes; CD8, CD8+ T lymphocytes; B, B lymphocytes; NK, natural killer cells. Differences for haematological tests between people with and without PKD; *P <0.05; ***P < 0.001. FIGURE 3 View largeDownload slide Box plot of lymphocytes across categories of CKD in patients with and without PKD. Depicted are counts of lymphocytes across categories of CKD in individuals with versus without polycystic kidney disease. Bars represent median numbers of cells per microlitre with IQR. Differences for hematological tests between people with and without PKD; *P <0.05; **P<0.01; ***P < 0.001. FIGURE 3 View largeDownload slide Box plot of lymphocytes across categories of CKD in patients with and without PKD. Depicted are counts of lymphocytes across categories of CKD in individuals with versus without polycystic kidney disease. Bars represent median numbers of cells per microlitre with IQR. Differences for hematological tests between people with and without PKD; *P <0.05; **P<0.01; ***P < 0.001. Regression model When adjusted for age, sex, lnCRP and eGFR, PKD remained significantly associated with lymphocyte concentration [adjusted mean difference −345/µL (95% CI −445 to −245), P < 0.001] (Table 2). After further adjustment for diabetes, the difference between PKD and non-PKD was −333/µL (95% CI −434 to −233; P < 0.001). We did not observe any clinically meaningful significant interaction between PKD and the other covariates. Also, male sex and lower eGFR were associated with lower lymphocyte counts (Table 4). Using the mean lymphocyte count of the last two visits, the difference in lymphocytes between patients with and without PKD remained comparable [adjusted mean difference 290/µL (95% CI 185–395); P < 0.001]. Table 4 Multiple regression analysis of the association of PKD with lymphocyte counts ESKD cohort (dependent variable = lymphocyte count)            Coefficient  95% CI  β  95% CI  P-value   Age (per year)  −3  −6–1  −0.06  −0.13–0.02  0.16   Sex (male = 1, female = 0)  −83  −171–17  −0.07  −0.14–0.01  0.08   lnCRP  −21  −66–18  −0.04  −0.11–0.04  0.34   PKD (yes = 1, no  =  0)  −310  −431 to − 188  −0.19  −0.27 to − 0.12  <0.001   Diabetes (yes = 1, no  =  0)  −149  −276 to − 21  −0.09  −0.17 to − 0.01  0.022   Immunosuppression (yes = 1, no  =  0)  −206  −341 to − 70  −0.11  −0.18 to − 0.04  0.003  CKD cohort (dependent variable = lymphocyte count)             Age (per year)  1  −4–6  0.02  −0.10–0.13  0.76   Sex (male = 1, female = 0)  −147  −249 to − 46  −0.13  −0.22 to − 0.04  0.004   eGFR (/mL/min)  5  3–7  0.32  0.20–0.43  <0.001   lnCRP  4  −33–53  0.01  −0.08–0.10  0.84   PKD (yes = 1, no  =  0)  −333  −434 to − 233  −0.30  −0.39 to − 0.21  <0.001   Diabetes (yes = 1, no  =  0)  149  −53–350  0.07  −0.03–0.16  0.15  ESKD cohort (dependent variable = lymphocyte count)            Coefficient  95% CI  β  95% CI  P-value   Age (per year)  −3  −6–1  −0.06  −0.13–0.02  0.16   Sex (male = 1, female = 0)  −83  −171–17  −0.07  −0.14–0.01  0.08   lnCRP  −21  −66–18  −0.04  −0.11–0.04  0.34   PKD (yes = 1, no  =  0)  −310  −431 to − 188  −0.19  −0.27 to − 0.12  <0.001   Diabetes (yes = 1, no  =  0)  −149  −276 to − 21  −0.09  −0.17 to − 0.01  0.022   Immunosuppression (yes = 1, no  =  0)  −206  −341 to − 70  −0.11  −0.18 to − 0.04  0.003  CKD cohort (dependent variable = lymphocyte count)             Age (per year)  1  −4–6  0.02  −0.10–0.13  0.76   Sex (male = 1, female = 0)  −147  −249 to − 46  −0.13  −0.22 to − 0.04  0.004   eGFR (/mL/min)  5  3–7  0.32  0.20–0.43  <0.001   lnCRP  4  −33–53  0.01  −0.08–0.10  0.84   PKD (yes = 1, no  =  0)  −333  −434 to − 233  −0.30  −0.39 to − 0.21  <0.001   Diabetes (yes = 1, no  =  0)  149  −53–350  0.07  −0.03–0.16  0.15  Patients with PKD had lower thrombocyte [PKD versus non-PKD −28.103/µL (95% CI −40.103 to −16.103), P < 0.001] and monocyte counts [PKD versus non-PKD −85/µL (95% CI −122 to −47), P < 0.001], while neutrophil counts were not significantly lower (Table 2). DISCUSSION Patients with PKD and ESKD or CKD not on dialysis had lower lymphocyte counts than patients with other aetiologies of kidney failure. The strength of the association was comparable in both study groups and independent of confounders such as age, sex and CRP. In an age- and sex-matched multicentre cohort of patients stratified for CKD category, the association was consistent across all strata, further increasing the external validity of our findings. Since lymphocyte counts and especially of CD4+ T cells decrease progressively with declining kidney function in line with the previously reported incremental apoptosis with advancing CKD [15–18], lymphocyte numbers in patients with and without PKD can only be compared after adjustment for this confounder. The lower number of lymphocytes in patients with PKD is most likely clinically relevant. Lower numbers of lymphocytes in kidney transplant recipients were previously associated with mortality, the development of solid cancer, skin cancer and lymphoma [19–21]. According to Ducloux et al. [20], after adjustment for confounders, for each increase in 100/mm3 of CD4+ T cells there was a 27% lower risk of developing cancer, with a relative risk (RR) of 0.73 (95% CI 0.62–0.89). The difference in CD4+ counts at the time of transplantation in our analysis was 77 (95% CI 1–153). For each increase in B cells of 10/mm3 there was a 13% lower risk of cancer, with a RR of 0.87 (95% CI 0.59–1.02) while the difference in B cells between patients with and without PKD was 50 (95% CI 22–79) in our analysis. PKD increases the risk of squamous cell cutaneous, colon, liver and renal carcinoma in transplant recipients or the general population [2, 6, 22]. Unfortunately, data on lymphocyte counts are lacking in these analyses [2, 6, 22]. Lymphopenia is a well-established risk factor for infections in different study populations. Among kidney transplant recipients it is associated with the development of Pneumocystis jiroveci pneumonia (PJP) [23, 24] and in liver transplant recipients, pre-transplant lymphopenia is associated with an increased risk of cytomegalovirus (CMV) and non-CMV invasive infections after transplantation [25]. In another analysis, in liver transplant recipients, pre-transplant lymphocytes <1000/µL versus  ≥1000/µL were associated with an increased risk of infections after transplantation in a multivariate model, with an odds ratio of 10.1 (95% CI 1.9–39; P = 0.005) [26]. Prospective analyses should determine whether these baseline differences in lymphocyte counts translate into unfavorable outcomes in PKD patients before or after kidney transplantation. The differences in cellular differentiation were most pronounced in CD8+ T lymphocytes and B lymphocytes in the ESKD cohort (Figure 2), although we cannot exclude that the relative distribution of differences in lymphocyte counts is coincidental. In particular, the blood concentration of CD4+ T lymphocytes is expected to decrease with more advanced CKD, while concentrations of CD8+ T lymphocytes change less [18]. We can speculate on the causal mechanism of our findings. Aberrant expression of polycystin-1 and/or polycystin-2 could lead to decreased proliferation of lymphocytes, as has been shown in a study with Epstein–Barr virus immortalized lymphoblastoid cells [11]. An increased apoptosis rate of lymphoid and myeloid cells remains an interesting alternative hypothesis in line with increased apoptosis of renal tubular epithelial cells in PKD [1, 27]. A targeted deletion of the anti-apoptotic B-cell lymphoma 2 (Bcl-2) in mice was both associated with PKD and lymphopenia, especially of CD8+ lymphocytes, due to an increased rate of apoptosis [28]. Apoptotic pathways, including the intrinsic Bcl-2 pathway, determine the final concentration of circulating lymphocytes [29] and affect the generation and survival of thrombocytes as well [30]. Kidney cells with silenced polycystin-1 expression have increased degradation of Bcl-2, leading to more apoptosis [27]. Disturbed apoptosis could be the common denominator in PKD explaining the renal and haematological phenotype. As such, particularly lymphopenia might reveal itself as a novel biomarker of a more severe phenotype with a faster decline of kidney function. Obviously prospective analyses are needed. Already in the early stages of kidney failure, lymphopenia might characterize PKD. Of note, lymphopenia in our CKD cohort was more common in males, whose kidney function in PKD was recently shown to decline faster than in females [31]. Since not only lymphocyte but also monocyte and thrombocyte counts were consistently lower in both groups of patients with versus without PKD, we speculate that a distinct genetic defect affects the common progenitor cells, leading to decreased proliferation or increased apoptosis in an early phase of haematopoiesis. In a recent analysis, platelet counts were lower in 290 patients with PKD and ESKD than in age- and sex-matched controls (215 versus 238 × 103/µL; P < 0.01), whereas the difference was not significant in individuals with CKD and after kidney transplantation [32]. Also, thrombocytopenia in PKD seems to be associated with a more severe renal phenotype with a larger total kidney volume and faster increase of kidney volume and decline of kidney function [33]. Obviously, PKD is characterized by a higher incidence of erythrocytosis [34, 35], and a lower need for erythropoiesis-stimulating agents [36]. This suggests that other mechanisms (for instance, increased expression of hypoxia-inducible factor-1 and/or erythropoietin) might be at play. Our analysis has some limitations. First, the analysis is cross-sectional and as such it is impossible to make causal inferences. Yet, the data from our study align with existing in vitro data on the role of polycystin in lymphocyte proliferation in PKD. Second, data on immunophenotypic analysis in the ESKD cohort were incomplete within the constraints of retrospective data analysis. Also, the lack of a substantial number of lymphocyte subsets precluded imputations. Yet, we believe selection bias is unlikely, as all patients suitable for transplantation were included for analysis. Moreover, the proportion of patients with PKD in the ESKD cohort did not change after excluding patients without available data on lymphocyte counts (Figure 1). Other relevant characteristics did not differ between the groups, suggesting that there was no informed missing data. Third, we cannot exclude that hidden confounders are still present. And while CRP concentrations were not different between groups with and without PKD, we cannot exclude that patients with PKD have some degree of chronic inflammation that was not entirely captured by CRP. By excluding patients with active infection and malignancy, we believe we avoided this confounder to a large extent. CONCLUSION PKD is associated with lymphopenia and especially with lower blood concentrations of CD8+ T and B lymphocytes. This finding could explain the increased risk of pneumonia, urinary tract infections and non-melanoma skin cancer after transplantation in these patients, which might warrant less aggressive immunosuppression in PKD patients after transplantation. This finding can potentially aid in risk stratification of people with PKD. If mutations in polycystin-1 and/or -2 lead to decreased proliferation of lymphocytes or increased apoptosis of renal tubular epithelial cells and lymphocytes via a common pathway, the degree of lymphopenia might reflect the genotype–phenotype interaction in PKD. As such, lymphopenia could serve as an inexpensive and easily accessible biomarker of a more severe phenotype with a faster decline of kidney function, aneurysm formation or cardiovascular involvement. This hypothesis is amenable to further exploration by prospective observational studies with wide application of disease markers including renal imaging. Also, our study asks for the clarification of a potential role of polycystin in the pathophysiology of haematological disorders and in different health conditions including PKD. SUPPLEMENTARY DATA Supplementary data are available online at http://ndt.oxfordjournals.org. CONFLICT OF INTEREST STATEMENT None declared. REFERENCES 1 Wilson PD. Polycystic kidney disease. N Engl J Med  2004; 350: 151– 164 Google Scholar CrossRef Search ADS PubMed  2 Bretagnol A, Halimi JM, Roland M et al.   Autosomal dominant polycystic kidney disease: risk factor for nonmelanoma skin cancer following kidney transplantation. Transpl Int  2010; 23: 878– 886 Google Scholar PubMed  3 Sallée M, Rafat C, Zahar JR et al.   Cyst infections in patients with autosomal dominant polycystic kidney disease. Clin J Am Soc Nephrol  2009; 4: 1183– 1189 Google Scholar CrossRef Search ADS PubMed  4 Nielsen LH, Jensen-Fangel S, Jespersen B et al.   Risk and prognosis of hospitalization for pneumonia among individuals with and without functioning renal transplants in Denmark: a population-based study. Clin Infect Dis  2012; 55: 679– 686 Google Scholar CrossRef Search ADS PubMed  5 Moua T, Zand L, Hartman RP et al.   Radiologic and clinical bronchiectasis associated with autosomal dominant polycystic kidney disease. PLoS One  2014; 9: e93674 Google Scholar CrossRef Search ADS PubMed  6 Otley CC, Cherikh WS, Salasche SJ et al.   Skin cancer in organ transplant recipients: effect of pretransplant end-organ disease. J Am Acad Dermatol  2005; 53: 783– 790 Google Scholar CrossRef Search ADS PubMed  7 Pedrozo Z, Criollo A, Battiprolu P et al.   Polycystin-1 Is a cardiomyocyte mechanosensor that governs L-type Ca2+ channel protein stability. Circulation  2015; 131: 2141– 2142 Google Scholar CrossRef Search ADS   8 Nauli SM, Kawanabe Y, Kaminski JJ et al.   Endothelial cilia are fluid shear sensors that regulate calcium signaling and nitric oxide production through polycystin-1. Circulation  2008; 117: 1161– 1171 Google Scholar CrossRef Search ADS PubMed  9 Klawitter J, Reed-Gitomer BY, McFann K et al.   Endothelial dysfunction and oxidative stress in polycystic kidney disease. Am J Physiol Renal Physiol  2015; 307: F1198– F1206 Google Scholar CrossRef Search ADS   10 Nowak KL, Farmer H, Cadnapaphornchai MA et al.   Vascular dysfunction in children and young adults with autosomal dominant polycystic kidney disease. Nephrol Dial Transplant  2017; 32: 342– 347 Google Scholar CrossRef Search ADS PubMed  11 Aguiari G, Banzi M, Gessi S et al.   Deficiency of polycystin-2 reduces Ca2+ channel activity and cell proliferation in ADPKD lymphoblastoid cells. FASEB J  2004; 18: 884– 886 Google Scholar CrossRef Search ADS PubMed  12 Banerjee A, Chandna S, Jayasena D et al.   Leucopenia in adult polycystic kidney disease patients on haemodialysis. Nephron  2002; 91: 175– 176 Google Scholar CrossRef Search ADS PubMed  13 Chamberlain JJ, Rhinehart AS, Shaefer JCF et al.   Diagnosis and management of diabetes: synopsis of the 2016 American Diabetes Association Standards of Medical Care in Diabetes Synopsis of the 2016 ADA Standards of Medical Care in Diabetes. Ann Intern Med  2016; 19: 542– 552 Google Scholar CrossRef Search ADS   14 Levey AS, Stevens LA, Schmid CH et al.   A new equation to estimate glomerular filtration rate. Ann Intern Med  2009; 150: 604– 612 Google Scholar CrossRef Search ADS PubMed  15 Yoon JW, Gollapudi S, Pahl MV et al.   Naïve and central memory T-cell lymphopenia in end-stage renal disease. Kidney Int  2006; 70: 371– 376 Google Scholar CrossRef Search ADS PubMed  16 Agarwal R, Light RP. Patterns and prognostic value of total and differential leukocyte count in chronic kidney disease. Clin J Am Soc Nephrol  2011; 6: 1393– 1398 Google Scholar CrossRef Search ADS PubMed  17 Saad K, Elsayh KI, Zahran AM et al.   Lymphocyte populations and apoptosis of peripheral blood B and T lymphocytes in children with end stage renal disease. Ren Fail  2014; 36: 502– 507 Google Scholar CrossRef Search ADS PubMed  18 Litjens NH, van Druningen CJ, Betjes MG. Progressive loss of renal function is associated with activation and depletion of naive T lymphocytes. Clin Immunol  2006; 118: 83– 91 Google Scholar CrossRef Search ADS PubMed  19 Ducloux D, Courivaud C, Bamoulid J et al.   Prolonged CD4 T cell lymphopenia increases morbidity and mortality after renal transplantation. J Am Soc Nephrol  2010; 21: 868– 875 Google Scholar CrossRef Search ADS PubMed  20 Ducloux D, Carron PL, Motte G et al.   Lymphocyte subsets and assessment of cancer risk in renal transplant recipients. Transpl Int  2002; 15: 393– 396 Google Scholar CrossRef Search ADS PubMed  21 Ducloux D, Carron PL, Rebibou JM et al.   CD4 lymphocyteopenia as a risk factor for skin cancers in renal transplant recipients. Transplantation  1998; 65: 1270– 1272 Google Scholar CrossRef Search ADS PubMed  22 Yu TM, Chuang YW, Yu MC et al.   Risk of cancer in patients with polycystic kidney disease: a propensity-score matched analysis of a nationwide, population-based cohort study. Lancet Oncol  2016; 17: 1419– 1425 Google Scholar CrossRef Search ADS PubMed  23 Mulpuru S, Knoll G, Weir C et al.   Pneumocystis pneumonia outbreak among renal transplant recipients at a North American transplant center: risk factors and implications for infection control. Am J Infect Control  2016; 44: 425– 431 Google Scholar CrossRef Search ADS PubMed  24 Brakemeier S, Dürr P, Bachmann F et al.   Risk evaluation and outcome of Pneumocystis jirovecii pneumonia in kidney transplant patients. Transplant Proc  2016; 48: 2924– 2930 Google Scholar CrossRef Search ADS PubMed  25 Nierenberg NE, Poutsiaka DD, Chow JK et al.   Pretransplant lymphopenia is a novel prognostic factor in cytomegalovirus and noncytomegalovirus invasive infections after liver transplantation. Liver Transplant  2014; 20: 1497– 1507 26 Fernandez-Ruiz M, Lopez-Medrano F, Romo EM et al.   Pretransplant lymphocyte counts predicts the incidence of infection during the first two years after liver transplantation. Liver Transplant  2009; 15: 1209– 1216 Google Scholar CrossRef Search ADS   27 Yu W, Kong T, Beaudry S et al.   Polycystin-1 protein level determines activity of the Gα12/JNK apoptosis pathway. J Biol Chem  2010; 285: 10243– 10251 Google Scholar CrossRef Search ADS PubMed  28 Nakayama K, Nakayama K, Negishi I et al.   Targeted disruption of Bcl-2 alpha beta in mice: occurrence of gray hair, polycystic kidney disease, and lymphocytopenia. Proc Natl Acad Sci USA  1994; 91: 1300– 1304 Google Scholar CrossRef Search ADS   29 Dunkle A, He YW. Apoptosis and autophagy in the regulation of T lymphocyte function. Immunol Res  2011; 49: 70– 86 Google Scholar CrossRef Search ADS PubMed  30 Kile BT. The role of apoptosis in megakaryocytes and platelets. B J Haematol  2014; 165: 217– 226 Google Scholar CrossRef Search ADS   31 Cornec-Le Gall E, Audrézet M-P, Rousseau A et al.   The PROPKD score: a new algorithm to predict renal survival in autosomal dominant polycystic kidney disease. J Am Soc Nephrol  2016; 27: 942– 951 Google Scholar CrossRef Search ADS PubMed  32 Setyapranata S, Holt SG. Platelet counts in autosomal dominant polycystic kidney disease. Platelets  2016; 27: 262– 263 Google Scholar CrossRef Search ADS PubMed  33 Chen D, Ma Y, Wang X et al.   Clinical characteristics and disease predictors of a large Chinese cohort of patients with autosomal dominant polycystic kidney disease. PLoS One  2014; 20: e92232 Google Scholar CrossRef Search ADS   34 Nankivell BJ, Allen RD, O'Connell PJ et al.   Erythrocytosis after renal transplantation: risk factors and relationship with GFR. Clin Transplant  1995; 9: 375– 382 Google Scholar PubMed  35 Hadimeri H, Nordén G, Friman S et al.   Autosomal dominant polycystic kidney disease in a kidney transplant population. Nephrol Dial Transplant  1997; 12: 1431– 1436 Google Scholar CrossRef Search ADS PubMed  36 Shah A, Molnar MZ, Lukowsky LR et al.   Hemoglobin level and survival in hemodialysis patients with polycystic kidney disease and the role of administered erythropoietin. Am J Hematol  2012; 87: 833– 836 Google Scholar CrossRef Search ADS PubMed  © The Author 2017. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.

Journal

Nephrology Dialysis TransplantationOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial