The phenotypic and genetic assessment of protein C deficiency


Mr Peter Cooper, Sheffield Haemophilia and Thrombosis Centre, Royal Hallamshire Hospital, Glossop Road, Sheffield, S10 2JF, UK. Tel.: +44(0)114 2713022; Fax: +44(0)114 2712149; E-mail:


This paper outlines the methods and approaches used for the laboratory detection and investigation of protein C (PC) deficiency. It does not make recommendations as to which patients should have thrombophilia testing performed; this should be done in line with local guidance. Interpretation of PC level is complicated because level varies with age, and many conditions can cause acquired deficiency. Protein C is most usually measured by chromogenic assay as a part of the thrombophilia screen. There exists, however, a very small group of individuals with significant PC deficiency, in whom the chromogenic PC assay is normal. The coagulometric assay of PC is more sensitive to these rare defects, but these assays may lack specificity. Genetic analysis allows definitive diagnosis and may be useful in confirming that deficiency is inherited and not acquired and is particularly valuable in families with severe PC deficiency.


Protein C (PC) is a vitamin K-dependent plasma protein that is synthesized in the liver and has a half-life of around 6 h, which is similar to that of FVII, but significantly shorter than that of the other vitamin K-dependent clotting factors, FII, FIX and FX. In vivo, protein C is activated to the serine protease, activated protein C (APC), by thrombin that is bound to thrombomodulin (Tm) on the vascular endothelium (Kottke-Marchant & Comp, 2002). Endothelial cell protein C receptor (EPCR) further increases this rate of PC activation, and the APC bound to EPCR has important cytoprotective properties (Crawley, 2007). The anticoagulant activity of APC is expressed when it cleaves bonds in activated factors FV and FVIII and destroys their activities. APC cleaves at arginine 506, 306 and 679 in FVa and at arginine 336 and 562 in FVIIIa (Castoldi & Rosing, 2010). For the full expression of anticoagulant activity, APC requires the plasma cofactors: protein S, FV, calcium ions and phospholipids. Factor V Leiden lacks the Arg506 cleavage site (Arg506Gln), and FVa destruction by APC is slowed and achieved only through cleavage at Arg306. Unlike factor V, factor V Leiden is not a cofactor for APC-mediated destruction of FVIIIa. Protein C deficiency results in impairment of the ability to control coagulation through destruction of FVa and FVIIIa and an increased risk of venous thrombosis. Familial protein C deficiency was first associated with thrombosis in 1981 and is present in around 3% of subjects with history of venous thromboembolism (Walker, Greaves & Preston, 2001).

The protein C gene (PROC) is positioned at Ch2q13-14; the genomic sequence is approximately 10802 bp long and consists of nine exons (1790 bp), eight of which encode the protein (, Online Mendelian Inheritance in Man, PROC – OMIM no. 612283; Foster, Yoshitake & Davie, 1985; Plutzky et al., 1986).

Protein C deficiency has been associated with numerous different mutations throughout PROC, with no noted ethnic variation and no significant recurrence of specific mutations (although some recurrence owing to founder effects have been noted in some populations (Reitsma et al., 1991; Kuismanen et al., 2006). To date, over 270 different variants have been listed on the Human Gene Mutation Database (HGMD;, The Human Gene Mutation Database at the Institute of Medical Genetics in Cardiff; ed. Millar et al., 2000) and the locus-specific ProCMD protein C mutation database (, ProCMD: 3D protein C mutations database). Most individuals with an inherited protein C deficiency will have a mild deficiency with an autosomal dominant pattern of inheritance, associated with heterozygosity for a mutation in the PROC gene. Severe hereditary protein C deficiency is inherited in an autosomal recessive manner and has been associated with homozygosity or compound heterozygosity for the same spectrum of mutations (Millar et al., 2000).

The majority of mutations associated with protein C deficiency (approximately 75%) are because of single base changes resulting in missense or nonsense mutations; however, gross gene deletions and mutations affecting splicing and gene regulation, small insertions and deletions have been reported (

Protein C deficiency may be quantitative (type I) and qualitative (type II) and may be further classified as IIA to IIE, depending on the molecular defect (Marlar, Adcock & Madden, 1990). The majority of cases are type I deficiencies with an equivalent reduction in both protein C antigen and activity. Type II deficiencies are associated with a dysfunctional variant, with reduced activity in relation to antigenic levels (Reitsma et al., 1995), and these variants are more likely to be associated with missense mutations (

Genotype–phenotype correlations

The prevalence of mild (heterozygous) protein C deficiency in the general population is estimated at 1/200-500 (Tait et al., 1995), and this has been associated with a mild increase risk of venous thrombosis (Dahlback, 2008; Goldenberg & Manco-Johnson, 2008). Only a small proportion of these individuals have a clinical significant disorder, which may be associated with co-inheritance of additional thrombotic risk factors, such as the factor V Leiden variant (Koeleman et al., 1994; Dahlback, 2008; Goldenberg & Manco-Johnson, 2008). Plasma protein C activity level has not been found to be a good prognostic indicator of clinical severity of the disorder (Henkens et al., 1993). In addition, there is little evidence to support the association of specific mutations with clinical outcome as genetic analysis of individuals/families with clinically significant or asymptomatic protein C deficiency has identified the same spectrum of PROC mutations (Millar et al., 2000).

Although several PROC promoter polymorphisms have been identified, which influence protein C level variability (Spek et al., 1995; Millar et al., 2000), they do not appear to influence clinical severity in protein C deficiency (Millar et al., 2000). The possibility of regions outside of the PROC gene that may affect protein C levels has been explored by genome-wide screening. (Buil et al., 2004).

Severe protein C deficiency is rare and usually associated with a severe prothrombotic diathesis, which can present as purpura fulminans and disseminated intravascular coagulation (DIC) in newborns. (Goldenberg & Manco-Johnson, 2008). Based on the prevalence of partial PC deficiency, it has been calculated that the occurrence of severe PC deficiency (homozygous or compound heterozygous) should be one in 40 000–250 000 individuals, but as there are only eight persons in the UK receiving long-term treatment for the severe deficiency, it is probable that the deficiency results in high foetal loss and perinatal mortality (Chalmers et al., 2011). Purpura fulminans can also be caused by severe acute infections or be due to an autoimmune response with acquired PC or PS deficiency or antiphospholipid syndrome; early detection of the cause with suitable treatment can reduce mortality and prevent major long-term health sequelae (Chalmers et al., 2011). Interestingly, a number of patients with a clinically milder form of homozygous protein C deficiency have been reported, and these subjects were characterized as having very low but measurable PC level. The authors described one such patient with PC level of <10% who had a negative history of thrombosis at age 38 years, despite exposure to thrombotic risk factors (Tripodi et al., 1990).

Methods for detecting protein C deficiency

Protein C is measured in plasma prepared from blood that has been anticoagulated with one part of 0.109 m sodium citrate to nine parts blood and should be measured against a standard that is calibrated in International Units (IU)/mL or IU/dL. The first international standard for PC was established in 1988 (Hubbard, 1988), and the second international standard (02/342) was established in 2006, it may be purchased from The National Institute for Biological Standards and Control (NIBSC) to calibrate in-house reference material. The SSC/ISTH Secondary Coagulation Standard (currently Lot#3) is also calibrated against the international standard and used by manufacturers to calibrate their reference materials.

Early functional assays required activation of PC by thrombin or by thrombin in complex with thrombomodulin, then detection of APC by means of clotting or chromogenic end point. As these functional assays were complex and expensive, haemostasis laboratories often measured PC antigen by the more familiar method of Laurell immunoelectrophoresis. However, PC antigenic assays do not detect functional defects.

Nowadays, protein C activity is usually measured in assays that use Protac® (Pentapharm, Basel, Switzerland). Protac® is a single-chain glycoprotein protease that is purified from the venom of the Southern copperhead viper (Agkistrodon contortrix contortrix). Protac rapidly activates PC to APC (Gempeler-Messina et al., 2001), so minimizing the influence of PC inhibitors, and it does not have unwanted activity that adversely affects chromogenic or clot-based assays. In clotting-based assays, APC destruction of FVa and FVIIIa or FVa alone is detected, whereas chromogenic assays rely on amidolysis of a small synthetic chromogenic substrate. The recommended method for screening for protein C deficiency is the chromogenic assay because it is less prone to interference and is more precise than clot-based assays (Baglin et al., 2010). However, chromogenic assays will not detect all cases of PC deficiency, and a range of protein C assays may be required if severe PC deficiency is suspected, but PC level is not <5%, for example, when identifying the cause of purpura fulminans (Baglin et al., 2010).

Protein C deficiency can also be detected by global screening tests for the protein C anticoagulant pathway (Gardiner et al., 2002). Global assays include ProC Global (Siemens, Marburg, Germany), the chromogenic HemosIL ThromboPath assay (Instrumentation Laboratory, Orangeburg, NY, USA) and a modification of the thrombin generation test. In a European multicentre study, ProC Global detected all carriers of factor V Leiden (FVL), and most (91.3%) patients with protein C deficiency, however, 20% of control subjects and, interestingly, 40.5% patients without known PC pathway defect had abnormal results (Toulon et al., 2000). In a multicentre study of the HemosIL ThromboPath assay, the test detected 95% of patients with PC deficiency (Toulon et al., 2009). Global screening tests will not be considered further in this review.

Clotting-based protein C assays

Protein C from plasma dilutions is activated to APC by Protac®, and then, this APC prolongs the clotting time of PC-deficient plasma. Clotting is generally determined using the activated partial thromboplastin time (APTT) or by Russell’s viper venom clotting time (RVV-X). Prolongation of clotting time is proportional to PC concentration, and patient’s PC is determined from a calibration curve, which is tested alongside the test samples. Type of assay can influence the specificity of the assay, for example, with APTT-based assay, and high levels of FVIII can cause underestimation of PC (de Moerloose, Reber & Bouvier, 1988; Flanders & Rodgers, 2004). Lupus anticoagulant (LA) can also cause inaccuracy in PC assay, either by adding to the anticoagulant effect of APC and causing overestimation or by interfering with the activity of PC and causing underestimation of PC. Factor V Leiden frequently causes underestimation of PC (Jennings et al., 2000, 2005). These influences may be detected by comparing chromogenic or antigenic level to clot-based PC assay and by testing plasma at multiple dilutions, although this is not normally the manufacturer’s recommended procedure. The influence of FVL can be significantly reduced by diluting affected plasma one part in four parts PC-deficient plasma and then assaying this mixture (Jennings et al., 2000), but this would be anticipated to reduce assay sensitivity when PC levels are low. Where a protein C assay result may be falsely reduced by APC resistance or affected by LA, it has been suggested that a comment might be included in the report to warn of the possible interference (Favaloro, Lippi & Adcock, 2008). In a sample from a patient with homozygous FVL that was distributed to participants in UK NEQAS Blood Coagulation, clotting PC level was consistent with PC deficiency, but chromogenic PC level was correctly reported as normal (Figure 1, personal communication from Dr Ian Jennings, UK NEQAS Blood Coagulation). Method type and reagent composition can have an influence on assay specificity, as illustrated by a DRVVT-based PC assay (Precision BioLogic, Dartmouth, NS, Canada), which, being an RVV-based assay, is insensitive to high levels of FVIII, but in addition, LA and FVL were shown to have minimal influence on the assay (Cooper et al., 2008). Heparin up to 1–2 u/mL may not influence clot-based assays, but protein C level has been reported to be falsely elevated in coagulometric assays where samples contain direct thrombin inhibitors, including argatroban, hirudin and bivalirudin (Khor & Van Cott, 2010).

Figure 1.

 Protein C level was determined in a sample from a patient with homozygous factor V Leiden and no known protein C defect. The sample was distributed by UK NEQAS Blood Coagulation, and participants assayed protein C level by their usual assay. The graph shows the number of participants reporting different ranges of protein C by clotting-based (inline image) and chromogenic (inline image) assay. Figure courtesy of Dr Ian Jennings (UK NEQAS Blood Coagulation).

Chromogenic protein C assays

Chromogenic substrates are small oligopeptide often attached to paranitroanilide (pNA) by a peptide bond that mimics the cleavage site in the natural substrate. APC cleaves pNA from the substrate and creates a yellow colour that is detected photometrically at 405 nm. Calibration curves can be stored and used to measure PC levels. Chromogenic substrates for APC lack specificity and can be cleaved by several enzymes; for example, S-2366 (Quadratech, Epsom, UK) is cleaved by thrombin, FXa, kallikrein that may be present in activated or clotted blood specimen. S-2366 is also cleaved by tissue plasminogen activator (tPA) and plasmin that may be found in blood samples from patients who have hyperfibrinolysis such as with tPA infusion. Figure 2 is an illustration of how enzymes in clotted blood cleave S-2366. In a study of cold-activated plasma and plasma from women taking combined oral contraceptives, high background amidolytic activity was found that made subtraction of blank optical density (OD) from test OD necessary to make the chromogenic PC assay specific (Mackie, Gallimore & Machin, 1992). The influence of these interfering enzymes can vary according to the assay conditions when clinical samples are assayed. We have noted that a 5- to 10-min reaction time of APC with chromogenic substrate in a PC end point assay often results in significant nonspecific substrate amidolysis, whereas a short (30-s) incubation with substrate, using the same reagents, normally results in very little nonspecific amidolysis. With the former assay, we would always test calibrator, quality controls and patient samples with Protac® and a blank (water in place of Protac®) and subtract the blank OD from the test OD, whereas blank ODs are generally very low and normally insignificant with the assay using a short acquisition time. We measured PC in 898 samples referred to us for thrombophilia screening using an in-house two-point kinetic chromogenic protein C assay with and without a water blank; without the blank and compared to the PC value using the blank: one sample had PC 15 IU/mL higher, four samples had PC 4–6 IU/mL higher, seven samples PC levels were 3–5 IU/mL higher and the remaining 886 sample results differed by no more than 3 IU/dL (Figure 3). In the author’s laboratory, with the shorter detection times we now use, we only include the blank if the presence of interfering enzymes is suspected, such as when testing a sample from a patient with DIC, patients on tissue plasminogen activator, or when activation of the sample might be likely, as when taken from a baby or small child when the blood may be drawn slowly from a difficult venepuncture. Heparin up to 1–2 u/mL has been shown to have no influence on routine chromogenic PC assays, and these assays are not affected by direct thrombin inhibitors (Khor & Van Cott, 2010).

Figure 2.

 Optical densities after chromogenic substrate S-2366 was added in a protein C assay where plasma was reacted with Protac® (inline image) and a water blank (inline image), and serum was also tested with Protac® (inline image) and water blank (inline image). The optical densities when plasma was reacted with Protac® are representative of normal protein C levels that were determined on a Sysmex CA-6000 analyser (Sysmex UK, Milton Keynes, UK).

Figure 3.

 Bland Altman plot of protein C levels of 898 sequential patient samples measured on Sysmex CA-6000 analyser (Sysmex UK, Milton Keynes, UK) with and without inclusion of a water blank.

Assay of protein C antigen

Protein C antigen assay is insensitive to type II PC defects but is necessary to determine whether a patient has type I or type II PC deficiency, and a low ratio of activity to antigen may indicate that a patient is a carrier of a type II defect even if PC activity level is normal. However, the 2001 UK guidelines for investigation and management of heritable thrombophilia state that there is no clinical value in determining whether a patient has type I or type II protein C deficiency (Walker, Greaves & Preston, 2001). PC antigen level may be used in conjunction with FVII and FX antigen levels to help identify type I protein C deficiency in patients who are on warfarin therapy. For diagnosis of deficiency in these patients, the INR must be stable within the therapeutic range to allow the PC antigen level to be compared to the antigenic level of two vitamin K-dependent clotting factors by ratio, alternatively, activity assays have been used by some authors. However, detection of PC deficiency in patients on vitamin K antagonists by comparison of PC with other vitamin K-dependent clotting factors by ratio has been considered to be unreliable and is not recommended (Kottke-Marchant & Comp, 2002). Protein C antigen can be measured by radial immunodiffusion, Laurell immunolectrophoresis and enzyme-linked immunosorbent assay (ELISA). Electrophoretic assays require specialist apparatus and are time-consuming and also require inclusion of a calcium chelator in the electrophoresis buffer (Mikami & Tuddenham, 1986). ELISA can give spuriously high antigen values because of rheumatoid factor affecting the assays; however, use of suitable blocking agents or other strategies can remove this effect (de Jager et al., 2005). We use polyethylene glycol precipitation to remove nonspecific binding that can cause falsely raised antigen assay in ELISA and confirm that type II deficiency is not falsely reported (unpublished observation). ELISA is very sensitive and can accurately measure levels below 5 IU/dL if appropriate dilutions are used so that the patient value can be read directly from the calibration curve without extrapolation.

Choice of assay

The chromogenic assay is recommended for screening for PC deficiency because the assay is less subject to interference and is therefore more specific than coagulometric assays (Baglin et al., 2010). In addition, chromogenic assays are generally more precise than clotting-based assays. In the majority of cases, the chromogenic PC assay is sufficient, but this assay will fail to detect the rare type II defects where PC cleaves chromogenic substrate normally, but anticoagulant activity is defective (Marlar, Adcock & Madden, 1990). Where there is a strong suspicion of a thrombophilia and all tests are normal, it may be appropriate to carry out further testing including assay of protein C by coagulometric assay. Genetic analysis of PC is indicated in the investigation if the clinical phenotype is suggestive of homozygous protein C deficiency because not all carriers of PC deficiency have PC levels below the reference range (Pabinger et al., 1992). Protein C Vermont was the first reported case of a mutation in the GLA domain of protein C that resulted in a type II PC deficiency with normal amounts of PC by antigenic assay and venom-based chromogenic assay, but significantly reduced PC level by clotting-based assay (Bovill et al., 1992). Since then, many kindred’s have been described with similar patterns of type II PC deficiency. One kindred’s PC results are shown, as an illustration, in Table 1. The authors recommended that the clotting-based assay of PC be carried out when a chromogenic assay does not reflect the clinical phenotype (Lyall et al., 2008). Different APTT reagents have been shown to have different sensitivities to the cofactor effect of protein S; phospholipid composition appears to be the main variable, and dilution of phospholipid may increase the sensitivity of the reagent to the cofactor effect of PS; the authors suggested that their results questioned the validity of APTT-based PC assays (D’Angelo, Gilardoni & D’Angelo, 1989). Recently, a type II protein C defect with normal chromogenic PC activity (Asn2Ile) was detected by one APTT-based assay, but a second APTT-based assay completely failed to detect the defect, and an RVV-X-based assay gave borderline low levels. The authors concluded that coagulometric assays are not equally sensitive to clinically important functional defects of PC and that multiple assays may be required to identify all variants (Cooper et al., 2011).

Table 1.   Variable pattern of PC levels in a family affected by two PC mutations, the mother has a type II defect detected by both chromogenic and clotting-based assay, and the father has a type II defect that was only detected by clot-based assay, and the child has symptoms of severe protein C deficiency and is affected by both mutations. (Lyall et al., 2008). Clearly abnormal results are shown in bold
 PC defectClotting-based PC (IU/dL)Chromogenic PC (IU/dL)PC antigen (IU/dL)
ChildHis253Gln & Arg57Trp    <51171
Adult reference range 73–15571–15175–131

Interpretation of protein C level

Preanalytical errors are a major cause of errors in laboratory diagnosis (Lippi et al., 2009). Samples must be taken into the correct anticoagulant, and plasma must not be contaminated or stored in wrongly labelled secondary tubes. We have observed false high levels of PC in serum from clotted blood samples (Figure 2), and others have seen falsely reduced PC activity levels in serum and heparinized plasma, along with increased PC levels in EDTA plasma (Favaloro, Lippi & Adcock, 2008).

As with all diagnostic assays, interpretation of PC level can only be carried out after assay validation and establishment of normal reference ranges that take into account physiological variation and method variables. When deriving a normal reference range for PC, the log-normal distribution of PC must be taken into account by log-transforming the data for the statistical analysis (Dolan et al., 1994). Carriers of PC deficiency may have a PC level within the normal reference range, and in an attempt to detect these individuals, a ratio of [PC/((FII antigen + FX antigen)/2)] improved detection rate, whereas use of a single comparator failed, but FX antigen helped detect PC deficiency if used in a quadratic discriminant analysis (Pabinger et al., 1992). Protein C results should be validated by the use of normal and abnormal internal quality control material and by ongoing quality assurance through participation in an external quality assurance programme, such as that run by the U.K. National External Quality Assessment Scheme for Blood Coagulation (Jennings et al., 2005). In a long-term study by the ECAT Foundation, chromogenic PC assays were shown to be significantly more precise than clotting-based PC assays; a linear regression model was established to determine within-laboratory variability of protein C assays. The authors considered that within-laboratory performance may vary due to many factors that influence imprecision, for example, reagent stability, pipetting errors, analyser maintenance, environmental factors and staff training. In addition, bias results in systematic error that can often be improved by using standards calibrated against in international units (Meijer et al., 2003; Meijer, Haverkate & Kluft, 2006).

Protein C and the vitamin K-dependent clotting proteins are at significantly lower levels in babies and children than in adults, and this has an important impact on interpreting results (Andrew et al., 1987, 1992). Clearly, adult-derived reference ranges must not be used to interpret PC levels from babies and children (Table 2). In another study, levels in teenagers were shown to be lower than in adults and continued to rise until the age of 25–30 years (Tait et al., 1991). Diagnosis of protein C deficiency has been shown to be problematic in pregnancy because protein C level rises from early pregnancy, by more than 20% between 6 and 20 weeks gestation (Said et al., 2010), and PC level is still raised in the postpartum period (Malm, Laurell & Dahlbäck, 1988). Many conditions reduce PC level (Table 3), with vitamin K deficiency, vitamin K antagonist therapy and liver disease, and the greatest reduction is in the clotting-based assay. Patients suffering from DIC may have significantly reduced protein C level, which complicates the diagnosis of PC deficiency when homozygous PC deficiency is suspected. It is good practise to carry out a coagulation screen when carrying out thrombophilia testing because it will help identify clotted or contaminated samples as well as subjects with liver disease or on anticoagulation where the clinical details on request forms are inadequate. Understanding the significance of a low PC level is only achieved through considering all causes of reduced PC level. To help understand the cause of PC deficiency, we often measure FVII and FX. Diagnosis of PC deficiency in a patient with fresh thrombosis may be unreliable; in one study, PC was measured within 24 h and prior to anticoagulation of a confirmed unprovoked VTE, and patients with abnormal PC level were retested following completion of anticoagulant therapy. The authors identified low protein C levels in 10 of the 254 patients on initial testing, but six of these patients had normal PC on repeat testing, and three had PC levels of <0.50 U/mL on initial testing (Kovacs et al., 2006). Guidelines for thrombophilia testing state that testing at the time of acute thrombosis is not indicated as the utility and implications of testing need to be considered and that treatment of acute venous thrombosis is not influenced by test results (Baglin et al., 2010).

Table 2.   Reference ranges for protein C antigen were estimated using pooled plasma standard, as mean level ±2 standard deviations; subjects were healthy full-term infants up to 6 months age (Andrew et al., 1987) and healthy children (Andrew et al., 1992). No differences were seen between antigenic and chromogenic PC levels in the children (Andrew et al., 1992)
Age range of healthy subjects studied1 day5 days30 days90 days180 days1–5 years6–10 years11–16 yearsAdult
Calculated reference range for PC antigen (U/dL)17–5320–6421–6528–8037–8140–9245–9355–11064–128
Table 3.   Conditions that have been associated with reduced protein C level
Congenital deficiency (heterozygous and homozygous states)
Young age
Vitamin K deficiency
Vitamin K antagonists
Liver disease
Kidney disease
Disseminated intravascular coagulation
Venous thrombosis
Asparaginase therapy

To detect homozygous PC deficiency, the PC assay must reliably detect PC levels of less that 5% normal. The lower limit of sensitivity can be determined by assaying a zero-level sample several times and defining the lower cut-off of sensitivity as being two to three standard deviations above the mean level (Saah & Hoover, 1997), and it is better to assay the matrix (plasma in this case) that lacks the analyte, rather than buffer (Stamey, 1996). The lower limit of sensitivity is different for the different types of protein C assay; for example, in the author’s experience, protein C antigen levels of <5 IU/dL can be reliably measured by ELISA. For clotting-based protein C assay, the standard dilution is usually 1 in 5 or 1 in 10, giving scope for using stronger-than-normal dilutions, such as 1 in 2 or 1 in 3 to increase assay sensitivity. The chromogenic protein C assays may have the least sensitivity to low PC levels because plasma is often introduced into the assay without prior dilution; the background amidolytic activity must be taken into account when measuring very low PC level to avoid overestimation. Very low levels of protein C may be detected in babies with DIC and suspected homozygous PC deficiency. It is with these subjects that assay sensitivity needs to be greatest, so that PC level can be accurately reported. When investigating protein C deficiency in neonates and children, it is important to measure protein C levels in both parents because results will be simpler to interpret than in the affected child. In such cases, antigen assay may identify type I deficiency, but type II deficiency must be investigated by measuring clotting, chromogenic and antigenic protein C levels, a discrepancy between the PC levels may indicate type II PC deficiency, and genetic analysis may be invaluable. Clinical interpretation should be carried out by clinicians with expertise in the management and laboratory investigation of protein C deficiency.

Role of genetic analysis in protein C deficiency

Genetic analysis is particularly valuable in families with severe protein C deficiency. Identification of the underlying genetic disorder facilitates counselling and allows definitive diagnosis (Takagi et al., 2009) and the provision of antenatal diagnostic testing or preimplantation diagnosis. Antenatal testing may also facilitate treatment during late gestation or immediately prior to delivery. Genetic analysis also provides a useful tool for carrier analysis. Most heterozygous carriers will be asymptomatic and may not be easily identified by reduced protein C activity alone owing to the significant overlap between plasma levels in normal individuals and those heterozygous for PROC mutations (Goldenberg & Manco-Johnson, 2008). In some cases, detection of a well-characterized PROC variant may also assist in confirming a diagnosis of hereditary as opposed to acquired protein C deficiency (Kottke-Marchant & Comp, 2002).

Identification of the underlying genetic basis of protein C deficiency cannot provide definitive clinically valuable information concerning thrombotic risk. However, carrier analysis in association with thrombophilia risk has been used within a family to evaluate attitudes towards genetic testing (van Korlaar et al., 2005). Study of protein C genetic variants and their effect has played an important role in furthering our understanding of the structure and function of protein C (Millar et al., 2000).

Approaches to Genetic Analysis

Genetic investigation must include the provision of appropriate counselling and support for the patient and should be carried out after obtaining informed consent. An accurate family pedigree is essential, and it may be necessary to test several family members to establish inheritance patterns.

Early studies suggested that genetic analysis of the PROC gene did not reveal a causative mutation in 10-30% of cases (Koeleman, Reitsma & Bertina, 1997), although this figure does not reflect more recent advances in molecular diagnostic techniques. The genetic analysis of additional thrombophilia risk factors such as factor V Leiden may also be informative.

Testing should be provided by accredited laboratories, which adhere to international quality standards [, OECD (2007)]. The UK Haemophilia Centre Doctors Organisation (UKHCDO) provides a directory of NHS laboratories who offer genetic analysis for inherited haemostatic disorders, including protein C deficiency (

As most of the mutations underlying protein C deficiency are single base changes, genetic investigation of an index case, usually involves initial nucleotide sequence analysis, extending across the entire exonic sequence including splice junctions and promoter regions of the PROC gene. Where sequence analysis does not reveal a causative mutation, further quantitative analysis may reveal genetic deletions/duplications, and possible gene rearrangements may be detected using Southern blotting techniques (Millar et al., 2000). Further laboratory characterization of a novel variant may be beyond the resources of many diagnostic laboratories, but may include expression studies, evaluation of promoter activity, and in vitro splice analysis or molecular modelling (Millar et al., 2000; Rovida et al., 2007). Once the familial mutation has been identified, genetic analysis of potential carriers and prenatal diagnostic testing is available.


The Human Genome Variation Society (HGVS) have recommended a standardized approach to the description of genetic variation at DNA, RNA and protein level, which is becoming widely used (, nomenclature for the description of sequence variants.).

This has resulted in changes in the numbering used to describe protein C mutations; for example, in common with many coagulation proteins, the amino acids of the protein C molecule have been traditionally numbered according to the mature protein (Foster, Yoshitake & Davie, 1985), whereas HGVS guidelines recommend that numbering starts from the earlier, initiator methionine ( It is important to be aware of this when using previously determined genetic information to test further family members.

Current reference sequences for PROC available through the National Centre for Biotechnology Information (NCBI) ( are NG_016323.1 (genomic), NM_000312.3 (mRNA) and NP_000303.1 (protein).