European Journal of Obstetrics & Gynecology and Reproductive Biology 180 (2014) 12–15

Neutrophil-to-lymphocyte ratio as a diagnostic marker of intrahepaticcholestasis of pregnancy

Ayse Kirbas *, Ebru Biberoglu, Korkut Daglar, Cantekin _Iskender, Salim Erkaya,Hülya Dede, Dilek Uygur, Nuri DanismanDepartment of Perinatology, Zekai Tahir Burak Women's Health Education and Research Hospital, Ankara, Turkey

A R T I C L E I N F O

Article history:Received 4 April 2014Received in revised form 23 May 2014Accepted 28 May 2014

Keywords:InflammationAminotransferaseHepatocellular injuryCholestasis of pregnancy

A B S T R A C T

Introduction: In this study, we aimed to investigate the relationship between neutrophil-to-lymphocyteratio (NLR) and total bile acid (TBA) concentration in pregnant women with intrahepatic cholestasis ofpregnancy (ICP).Materials and methods: Fasting and postprandial TBA, NLR, and aminotransferase (AST/ALT) levels in theblood samples of 65 pregnant women with intrahepatic cholestasis were examined in this prospectivecase–control study. Thirty-three of the patients had mild disease and 32 had severe disease; 70 healthywomen in uncomplicated pregnancies served as the control group.Results: Not only was the mean NLR elevated in the pregnant women with cholestasis when compared tothe controls, but it also predicted the severity of the cholestasis. The correlation between fasting TBA andNLR was significant.Comments: Although TBA is still the diagnostic standard, NLR can be used as an initial screening tool dueto its high specificity.

ã 2014 Elsevier Ireland Ltd. All rights reserved.

Contents lists available at ScienceDirect

European Journal of Obstetrics & Gynecology andReproductive Biology

journal homepage: www.else vie r .com/ locat e/e jogrb

Background and objectives

Intrahepatic cholestasis of pregnancy (ICP) is a medicalcondition characterized by elevated liver enzyme and serum bileacid levels, as well as intense itching localized to the abdomen,legs, palms, and soles [1–3]. ICP can lead to complications for bothmother and fetus, such as fetal distress, perinatal mortality,stillbirth, and preterm labor [4–6].

The etiology of ICP is still not fully understood. Genetic,endocrinologic, nutritional, and environmental factors are likely tobe important in the pathogenesis of the disease. The associationbetween inflammation and cholestatic liver disease has beenpreviously described [7–9]. It has been reported that elevated bileacid levels trigger an inflammatory response, causing hepatocel-lular injury. Systemic inflammation can be measured by using avariety of biochemical and hematological markers. Although noveldisease-specific biomarkers have been identified, most of them,including the measurement of total bile acid (TBA) concentration,

* Corresponding author. Tel.: +90 533 646 9213; fax: +90 312 312 4931.E-mail address: ayseozdemirkirbas@hotmail.com (A. Kirbas).

http://dx.doi.org/10.1016/j.ejogrb.2014.05.0420301-2115/ã 2014 Elsevier Ireland Ltd. All rights reserved.

are time consuming and expensive. Recent evidence indicates thatmeasuring the ratio of blood cell subtypes, such as the neutrophil-to-lymphocyte ratio (NLR), might have prognostic significance fordiseases related to chronic low-grade inflammation [10,11]. Inaddition, because it is readily available and easily calculated, thismethod might be a promising alternative diagnostic tool fordiseases associated with chronic low-grade inflammation, such asICP [10,11]. However, little is known and has been published aboutNLR and its relationship with ICP; therefore, the present study wasconducted to evaluate the efficacy of NLR for the diagnosis andmanagement of ICP.

Materials and methods

This prospective case–control study was conducted at theperinatology department of Dr. Zekai Tahir Burak Women’s HealthCare, Education and Research Hospital, Ankara, Turkey. The studywas approved by the Institutional Review Board of Dr. Zekai TahirBurak Women’s Hospital.

Sixty-five pregnant women with ICP, 33 with mild disease and32 with severe disease, were recruited for the study between June2013 and January 2014.

A. Kirbas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 180 (2014) 12–15 13

The control group consisted of 70 healthy women uncompli-cated pregnancies in third trimester, matched for maternal ages,during the same period.

Inclusion criteria for the study group included itching withoutrash in different parts of the body, severe enough to requirehospital admission, and elevated total bile acid (TBA) (�10 mmol/L)and/or aminotransferase levels in the blood sample.

All the pregnants with ICP were hospitalized and tested forhepatitis A, B and C virus as part of initial evaluation. Routine urinecultures were obtained and daily body temperature weremeasured for all the patients. However, further investigations ofinfection markers are made on clinical ground. Patient with signsand symptoms of active infection are excluded from the study. Inaddition, patients with multiple gestation, with chronic inflam-matory diseases like ulcerative colitis, Crohn's diseases, rheuma-toid arthritis, vasculitis, familial Mediterranean fever or a history ofsystemic disease other than ICP, or in active labor were excludedfrom the study.

Blood samples were obtained from the antecubital vein early inthe morning, following ten hours of fasting, and two-hourpostprandial and TBA levels were determined by enzymatic assay.The patients were classified as having mild or severe cholestasisbased on TBA concentrations of 10–40 or �40 mmol/L, respectively[1,2]. NLR was calculated as the ratio of absolute neutrophil countto absolute lymphocyte count, both obtained from the same bloodsample. All the other blood analyses were carried out within twohours of blood sampling using a Beckman-Coulter Gen-S systemdevice at the central laboratories of Dr. Zekai Tahir Burak Hospital,Ankara. The following clinical and demographic data wereobtained: maternal age, obstetric history, and presence of maternaland neonatal complications.

The statistical analyses were conducted using the StatisticalPackage for the Social Sciences version 12.0 for Windows (SPSS,Chicago, IL). The data were summarized as mean � standarddeviation and median (minimum–maximum). Proportions werecompared with Fisher’s exact test or the chi-square test whereappropriate. For parametric variables, one-way ANOVA was usedfor comparing three independent groups, and post hoc compar-isons were performed with Tukey’s HSD test. The Mann–WhitneyU test was used for nonparametric variables in two independentgroups. Spearman’s rank correlation was used to assess therelationship between quantitative variables.

The dependence of fasting and postprandial serum TBA levelsand NLR was evaluated by linear regression analysis. Receiveroperating characteristic (ROC) analysis was used to evaluate theperformance of NLR in the diagnosis of ICP and in the detection of

Table 1Demographic, obstetric, and laboratory characteristics of intrahepatic cholestasis (ICP)

Clinical and laboratoryparameters

Normal pregnancyn = 70

Mn

Age (years) 28.1 � 4.7 2Gestational week 38.7 � 1.1b,c 3BMI 28.9 � 3.8 2Gravidity 1.6 � 0.7 1.Parity 0.6 � 0.6 0AST (U/L) 22.4 � 7.1c 4ALT (U/L) 23.0 � 9.5c 6WBC � 109/L 8.32 � 1485 8Neutrophils � 109/L 5.2 � 1.6b,c 6Lymphocytes � 109/L 2.4 � 0.8c 2NLR 2.32 � 0.77b,c 3

Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, bopregnancy; WBC, white blood cell.

a Different from the control group.b Different from the mild ICP group.c Different from the severe ICP group.

severe ICP based on fasting serum TBA concentrations. Statisticalsignificance was set at p � 0.05.

Results

The demographic and obstetric data of the ICP and controlgroups are shown in Table 1. While mean white blood cell countwas significantly higher in the group of women with severecholestasis compared to the control group, their lymphocyte countwas significantly lower. The mean gestational age, neutrophilcounts, neutrophil lymphocyte ratios, alanine aminotransferase,and aspartate aminotransferase levels were significantly differentwithin each group. The mean NLR was significantly different ineach group. NLR levels were higher in the mild ICP group than inthe control group and higher in the severe ICP group than in both ofthe other groups (Table 1).

Serum TBA concentrations in the mild and severe ICP groups arepresented in Table 2.

The ROC analysis of the diagnostic performance of NLR for ICP isshown in Fig. 1. The area under the curve was 0.92 � 0.03 (95%confidence interval (CI) 0.861–0.969). The best NLR cutoff valuewas 2.93. The ratio above this value had 91% sensitivity, 84%specificity, 83% positive predictive value, and 91% negativepredictive value for the diagnosis of ICP.

The ROC curve analysis was also performed to assess thepredictive value of NLR in differentiating minimal and severecholestasis. The area under the curve was 0.75 � 0.060 (95% CI0.638–0.872) (Fig. 2). The best NLR cutoff value in predicting theseverity of ICP was 4.05, above which the sensitivity and specificitywere 78% and 67% and the positive and negative predictive valueswere 69% and 76%, respectively.

Multivariable logistic regression was used to verify theassociation between serum TBA levels and NLR further, and itdemonstrated a positive correlation between the two. Regressionequality was obtained at NLR = 3.352 + (0.034 � fasting bile acid)(Fig. 3). Only the correlation between fasting TBA (not postprandialTBA) and NLR was significant (regression coefficient 0.034;p < 0.001).

Comments

Because ICP is also an inflammatory process, in the search of acost-effective, readily available laboratory test, we decided toconduct the current study, using NLR to investigate its role in thediagnosis of ICP. We aimed to study the possible relationship

and control study populations (mean � SD).

ild cholestasis = 33

Severe cholestasisn = 32

P value

8.8 � 4.8 28.9 � 5.3 0.6867.2 � 1.7a,c 35.6 � 2.7a,b <0.0018.5 � 3.9 29.6 � 4.2 0.9158 � 0.9 1.6 � 0.7 0.530.7 � 0.8 0.5 � 0.6 0.4654.8 � 38.3c 176.0 � 94.7a,b <0.0011.4 � 59.4c 225.7 � 166.7a,b <0.001.63 � 1696.8 9.12 � 1403.5 0.44.7 � 2a,c 7.991 � 2a,b <0.001 � 1.4 1.6 � 0.4a <0.001.97 � 1.5a,c 5.59 � 1.98a,b <0.001

dy mass index; NLR, neutrophil/lymphocyte ratio; ICP, intrahepatic cholestasis of

Table 2Fasting and postprandial TBA concentrations in pregnant women with mild andsevere intrahepatic cholestasis (mean � SD).

TBA mmol/L Mild cholestasisn = 33

Severe cholestasisn = 32

Pvalue

Fasting 18 � 5.9 57.3 � 28.6 <0.001Postprandial 23.7 � 9.7 69.3 � 36.1 <0.001

TBA, total bile acid concentration.

Fig. 2. Area under the receiver operating characteristic curve for NLR measure-ments in differentiating mild and severe intrahepatic cholestasis. The area underthe curve was 0.755.

14 A. Kirbas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 180 (2014) 12–15

between NLR and TBA concentration in the blood samples ofwomen with ICP.

Although the etiology of ICP is heterogeneous and not fullyunderstood, inflammatory processes are implicated in thepathogenesis in addition to the known cholestatic effects ofpregnancy hormones and nutritional and environmental factors ingenetically predisposed women (7–9). However, the mechanismby which cholestasis initiates an inflammatory response in theliver is not known [7]. It is evident that cholestasis takes placewhen the liver rapidly reduces bile formation as a negative acutephase response in adaptation to inflammation [12]. Recent studieshave suggested that inflammatory cells such as neutrophils areactivated and recruited into the liver during obstructive cholesta-sis, causing significant liver injury [13]. Cholestatic inflammatoryresponses of the liver are not restricted to hepatocytes, as bile ductinflammatory responses also occur. Therefore, the clinical courseand outcome for these patients depends on the balance of theliver’s secreted, local, and intracellular responses to inflammationand cholestasis [8].

The most frequent and, therefore, diagnostic laboratoryabnormality noted in ICP is elevated serum TBA concentration[14]. However, there is a wide variability in the suggesteddiagnostic cutoff values of TBA, depending on the measurementmethod used, fasting status, population studied, and gestationalage at diagnosis [3]. In this study, we considered the cutoff value ofTBA as 10 mmol/L for the diagnosis of ICP and classified thepatients’ disease as mild or severe based on levels of 10–40 mmol/Lor >40 mmol/L, respectively, as suggested in the literature [15].Although the levels of liver enzymes such as alanine aminotrans-ferase and aspartate aminotransferase are significantly elevated inwomen with ICP, as demonstrated in the present study, there is noconsensus on the diagnostic cutoff level of liver enzymes [15]. Inaddition, TBA measurements may not be readily available,especially in developing countries with limited resources, andrepeated testing of serum liver enzymes and TBA can be timeconsuming and expensive. NLR has already been identified as a

Fig. 1. Area under the receiver operating characteristic curve for NLR measure-ments in the diagnosis of intrahepatic cholestasis. The area under the curve was0.915.

potentially useful marker of clinical outcome in disease states withan inflammatory component.

Recent studies have reported its association with poor out-comes in patients with cardiovascular diseases, malignancies,cystic fibrosis, and familial Mediterranean fever [10,11]. Althoughsystemic inflammation has also been suggested to play a role in theetiology of certain obstetric complications such as preeclampsia,preterm delivery, and miscarriage [16–19]; the clinical relevance ofNLR has not been yet established in these disorders.

As described above, the present study demonstrated for the firsttime in the medical literature that blood NLR is significantly higherin pregnant women with intrahepatic cholestasis, even in mildcases, compared to healthy controls. The relationship betweenfasting TBA levels and NLR was also significant. Therefore, wesuggest that NLR may be considered a useful marker for diagnosingICP and for predicting the severity of the disease.

There have not been any clear guidelines for the mostappropriate gestational age for the delivery of patients with ICP.The main objective is to reduce symptoms in the mother and todecrease the occurrence of sudden fetal death and pretermdelivery. Serum TBA levels play an important role in decidingmanagement options, as elevated levels are linked to increased riskof fetal complications [20]. However, there is still considerablecontroversy regarding the management of ICP. While expectantmanagement carries with it the risk of adverse pregnancyoutcome, active management, may cause unnecessary prematurity[1–3].

Our results suggest that NLR is a viable diagnostic marker forICP. Moreover, it is correlated with TBA and valuable in the

Fig. 3. Relationship between fasting and postprandial TBA levels and NLR inintrahepatic cholestasis of pregnancy (fasting TBA/NLR, r = 0.034; p < 0.001).

A. Kirbas et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 180 (2014) 12–15 15

differentiation of mild and severe ICP. Therefore, it can be assumedthat NLR could also be a potential marker in managing inflamma-tion-related pregnancy complications. However, this conclusionshould be confirmed with further research, as our relatively lowsample size may depotentiate these findings.

Conflict of interest

None.

References

[1] Geenes V, Williamson C. Intrahepatic cholestasis of pregnancy. World JGastroenterol 2009;15:2049–66.

[2] Diken Z, Usta IM, Nassar AH. A clinical approach to intrahepatic cholestasis ofpregnancy. Am J Perinatol 2013(Jan 28) [Epub ahead of print].

[3] Pathak B, Sheibani L, Lee RH. Cholestasis of pregnancy. Obstet Gynecol ClinNorth Am 2010;37(2)269–82 Review.

[4] Heinonen S, Kirkinen P. Pregnancy outcome with intrahepatic cholestasis.Obstet Gynecol 1999;94:189–93.

[5] Wikström Shemer E, Marschall HU, Ludvigsson JF, Stephansson O. Intrahepaticcholestasis of pregnancy and associated adverse pregnancy and fetaloutcomes: a 12-year population-based cohort study. British J Obstet Gynaecol2013;120(6)717–23 [Epub 2013 Feb 19].

[6] Rook M, Vargas J, Caughey A, Bacchetti P, Rosenthal P, Bull L. Fetal outcomes inpregnancies complicated by intrahepatic cholestasis of pregnancy in aNorthern California cohort. PLoS One 2012;7(3)e28343 [Epub 2012 Mar 5].

[7] Allen K, Jaeschke H, Copple BL. Bile acids induce inflammatory genes inhepatocytes: a novel mechanism of inflammation during obstructivecholestasis. Am J Pathol 2011;178(1)175–86 [Epub 2010 Dec 23].

[8] Kosters A, Karpen SJ. The role of inflammation in cholestasis–clinical andbasic aspects. Semin Liver Dis 2010;30(2):186–94.

[9] Woolbright BL, Jaeschke H. Novel insight into mechanisms of cholestatic liverinjury. World J Gastroenterol 2012;18(36)4985–93 Review.

[10] Nunez J, Nunez E, Bodi V, et al. Usefulness of the neutrophil to lymphocyteratio in predicting long-term mortality in ST segment elevation myocardialinfarction. Am J Cardiol 2008;101:747–52.

[11] Balta S, Demirkol S, Celik T, et al. Association between coronary artery ectasiaand neutrophil-lymphocyte ratio. Angiology 2013;64(Nov (8))627–32, doi:http://dx.doi.org/10.1177/0003319713480424 [Epub 2013 Mar 7].

[12] Moshage H. Cytokines and the hepatic acute phase response. J Pathol 1997;181(3):257–66.

[13] Gujral JS, Farhood A, Bajt ML, Jaeschke H. Neutrophils aggravate acute liverinjury during obstructive cholestasis in bile duct-ligated mice. Hepatology2003;38:355–63.

[14] Brites D, Rodrigues CM, van-Zeller H, Brito A, Silva R. Relevance of serum bileacid profile in the diagnosis of intrahepatic cholestasis of pregnancy in an highincidence area: Portugal. Eur J Obstet Gynecol Reprod Biol 1998;80:31–8.

[15] Mays JK. The active management of intrahepatic cholestasis of pregnancy. CurrOpin Obstet Gynecol 2010;22:100–3.

[16] Borzychowski AM, Sargent IL, Redman CW. Inflammation and preeclampsia.Semin Fetal Neonatal Med 2006;11:309–16.

[17] Sacks GP, Studena K, Sargent K, Redman CW. Normal pregnancy andpreeclampsia both produce inflammatory changes in peripheral bloodleukocytes akin to those of sepsis. Am J Obstet Gynecol 1998;179(1):80–6.

[18] Christiansen OB, Nielsen HS, Kolte AM. Inflammation and miscarriage. SeminFetal Neonatal Med 2006;11(5)302–8 [Epub 2006 May 6].

[19] Romero R, Espinoza J, Gonçalves LF, Kusanovic JP, Friel LA, Nien JK.Inflammation in preterm and term labour and delivery. Semin Fetal NeonatalMed 2006;11(5)317–26 [Epub 2006 Jul 12], Review.

[20] Lee RH, Kwok KM, Ingles S, et al. Pregnancy outcomes during an era ofaggressive management for intrahepatic cholestasis of pregnancy. Am JPerinatol 2008;25:341–5.