Preimplantation genetic screening is a diagnostic approach dedicated to patients undergoing IVF with the proper indications (advanced maternal age, recurrent implantation failure, recurrent pregnancy loss) in order to increase pregnancy rates per transfer via euploid embryo selection. This strategy, and all the associated techniques, are in constant evolution and will shed more light on unexplored aspects of embryology, such as female meiosis or chromosomal mosaicism, creating new criteria for embryo selection.

by Dr D. Cimadomo, Dr A. Capalbo, Dr L. Rienzi and Dr  F. M. Ubaldi

Background
Preimplantation genetic diagnosis (PGD) and preimplantation genetic screening (PGS) are two diagnostic approaches increasingly exploited in recent decades within assisted reproduction facilities in the presence of specific indications. PGD is used to identify unaffected embryos in couples at high reproductive risk of a hereditary disease. Usually, these couples conceive naturally and undergo prenatal genetic testing, i.e. villocentesis or amniocentesis; procedures that are invasive and carry a high risk of subsequent miscarriage. The ultimate aim of PGD is, therefore, to prevent the conception of a fetus affected specifically and uniquely by a pathology whose causative mutations have been identified and characterized in the parental genomes before conception. Consequently, PGD depends on a preliminary ad hoc work-up for each couple approaching to an IVF cycle. PGS, instead, is meant to identify only chromosomally normal embryos, thus looking for the presence of chromosomal abnormalities. Since the development of aneuploidies is a de novo event directly linked to maternal age, this diagnostic method is independent from any specific preliminary set-up, thus being identical for each PGS cycle. The indications for this analysis are mainly advanced reproductive maternal age (more than 35 years old; AMA), recurrent implantation failure (more than three failed IVF attempts; RIF) and recurrent pregnancy loss (more than three miscarriages; RPL). From an embryological perspective there is no difference between PGD and PGS. Indeed, strategy and planning of the cycle and biopsy techniques are similar, whereas the genetic technical aspects are significantly different.

Testing for aneuploidy
Interestingly, the data collected by the ESHRE PGD consortium IX showed a constant increase in the number of the PGD cycles approached uniquely for euploid embryo selection. In particular, more than 60% of PGD cycles were actually PGS for AMA, RIF or RPL patients, and this percentage is still currently increasing. There is, in fact, a striking impact of aneuploidies on human reproduction. In particular, their incidence in newborns is around 0.3%, mostly represented by trisomies of chromosomes 13, 18 and 21 and sex chromosome aneuploidies. However, tracking backwards through the developmental stages sees this incidence sharply increase, involving other chromosomes and reaching an incidence of up to 60% in preimplantation embryos and 70% in eggs or polar bodies [1]. On the contrary, this incidence in sperm is definitely less severe, as it is never greater than 3–4%. Moreover, a significant number of spontaneous abortions are linked to aneuploidies (more than 60% of products of conception follow chromosomal abnormalities), both increase exponentially with maternal age and fertility rate collapses (Fig. 1) [2].

From a biological perspective, the origin of high trisomy rates found in clinically recognized pregnancies (which sharply increases in patients older than 35 years) resides mainly in maternal meiosis I and II [3]. Recent data obtained through array comparative genomic hybridization (aCGH) on polar bodies (PBs) showed that chromatid errors in female meiosis, such as premature separation of sister chromatids, definitely outnumber impairments involving whole chromosomes as previously thought [4, 5]. Capalbo et al. [5] performed analyses on biopsies at sequential stages of development, in particular the two PBs, a single blastomere at day 3 of embryo development and also a trophectoderm (TE) sample at the blastocyst stage (Fig. 2). This study design allowed the determination of PB analysis accuracy and the impact of male and mitotic errors as well as the evaluation of the occurrence of correction mechanisms throughout preimplantation development. It came to light that 76 out of 78 (97.4%) abnormal meiotic segregations concerned errors involving chromatids rather than whole chromosomes at meiosis I. Furthermore, it unveiled not only a false positive rate in PB biopsy analysis of 20.5%, as just 79.5% (62/78) of meiotic segregation errors identified in PB biopsies were confirmed in blastomeres, but also a false negative rate of 47.6%, as 10 out of 21 embryos showed mitotic or male-derived aneuploidies confirmed at day 3 and at the blastocyst stage of development, which are, obviously, not observable in PBs. This evidence subverts our previous scenario of chromosomal aneuploidy genesis, as well as undermining the reliability of the PB analysis strategy.

Chromosomal mosaicism
From a diagnostic perspective in PGS, post-zygotic mitotic segregation errors are definitely more troubling than meiotic ones, as, whereas the latter involve the same aberrant chromosomal layout in the whole developing embryo, the former entail the phenomenon of chromosomal mosaicism. In the last decade several publications focused on the problem of mosaicism and its influence on PGD/PGS, claiming an incidence fluctuating between 25% [6, 7] and up to more than 70% [8]. Even when these data are analysed with a critical approach, it still emerges that mosaicism is a substantial source of misdiagnosis when the embryo is biopsied at day 3 post-fertilization. This evidence encouraged a shift of the biopsy strategy toward the blastocyst stage and, to this end, different studies were conducted in order to thoroughly describe its cytogenetic constitution and the impact of biopsy itself on embryonic developmental competence. In particular, Capalbo et al. [9] published data outlining the impact of chromosomal mosaicism on a diagnosis at day 5/6 of embryo development as well as the aneuploid cells setting between inner cell mass (ICM) and TE. To this end, a novel method of ICM biopsy was conceived [as described in 9], characterized via KRT18 staining [as described in 10] and its efficiency tested. It led to the absence of TE contamination in 85.7% of the ICM biopsy products, and a low TE contamination rate (only 2% of TE cells) in the rest of them. These data attest the reliability of this biopsy procedure to test the influence of mosaicism at the blastocyst stage. The study design entailed a preliminary aCGH analysis on a TE biopsy during blastocyst-stage PGS clinical cycles, followed by FISH re-analysis of three further fragments of TE and of the ICM from those blastocysts found to be carriers of copy-number chromosomal errors as well as euploid embryos. This revealed that at the blastocyst stage of development, 79.1% of the aneuploidies were constitutional, while 20.9% of them were mosaic. However, only 4% of the blastocysts were found to be mosaic diploid/aneuploid, being at risk of misdiagnosis due to mosaicism when testing at the blastocyst stage. These data strengthen the theory that the impact of mosaicism could be critical at day 3 of embryo development, but it has definitely less influence at the blastocyst stage, thus strongly presenting the latter as the most reliable candidate biopsy stage to perform PGS. Importantly, in the same paper, Capalbo et al. demonstrated that, after excluding low grade mosaicism (<20% of aneuploid cells) and mosaicism confined to one or two TE sections, in 97.1% of cases concordance for all chromosomes re-analysed by FISH between ICM and TE was observed. On a per embryo analysis, instead, complete concordance in TE-based prediction of ICM chromosomal complement was reported (Fig. 3) [9]. Northrop et al. [11] conducted a similar analysis exploiting a single nucleotide polymorphism (SNP) array, which is a comprehensive chromosomal screening technique. This method was found to detect aneuploidy in samples possessing more than 25% aneuploidy, thus when as few as 2 of the 5 cells within a TE biopsy contain the same chromosomal error. Their data showed no preferential aneuploid cell migration to the TE layer, as aneuploidy was observed in 31% of ICM samples (15 out of 48 ICM products) and 32% of TE ones (46 of 144 TE products). Furthermore, a mosaicism rate of 24% was attested, since 12 out of 50 blastocysts screened showed more than a single diagnosis in all of the multiple sections that were re-analysed.

Does the biopsy procedure affect embryo reproductive competence?
One  concern about PGS is that biopsy could affect embryo reproductive competence. To investigate this possibility, Scott et al. [12] designed a randomized and paired clinical trial. They selected two of the best quality embryos from the same patient to be transferred and randomized them, one to undergo biopsy, either at day 3 or at day 5 of embryo development, and the other as a control. The biopsy was submitted to SNP array analysis. If only one embryo implanted, buccal DNA obtained from the neonate after delivery was analysed by SNP array to determine whether the implanted embryo was the control one or not. The data collected clearly showed that conducting the biopsy at the cleavage stage affects the clinical outcome, as an absolute reduction in implantation rate of 19.6% with respect to the control was reported. On the contrary, blastocyst biopsy led to a non-significant overall reduction of implantation of 3%; thus an implantation rate equivalent to the control. It is still unclear whether this is due to a smaller proportion of the embryo’s total number of cells being removed, or to the fact that only extra-embryonic cells are involved, or to a higher stress-tolerance of the blastocyst; however, it is still additional important evidence supporting TE biopsy as the ‘gold standard’ for PGS. From a clinical perspective, the same authors also published a randomized controlled trial [13] comparing the clinical outcomes of single euploid blastocyst transfer versus double untested blastocyst transfer. Ongoing pregnancy rates per randomized patient were similar between the two groups (60.7% in the study group vs 65.1% in the control group), whereas a higher multiple pregnancy rate in the control group was recorded (54% vs 0% in the study group). Ultimately then, PGS on TE biopsy associated with a single euploid blastocyst transfer elicits the same clinical outcomes as conventional IVF, but reduces its risks.

Conclusion
In conclusion, PGS is an important diagnostic approach for patients with the proper indications (AMA, RIF or RPL), performed in order to boost implantation rate per transfer. Euploid embryo selection prevents useless and potentially detrimental embryo transfers. Consequently, further advantages of performing PGS are a lower time-to-pregnancy and a higher cost-effectiveness of each single treatment. Moreover, by adopting a biopsy strategy at day 5/6, it is possible to take advantage of a more robust genetic analysis, a high clinical predictive value, the absence of impact of the biopsy on embryo quality, a low influence of mosaicism, as well as a reduced number of embryos to analyse per cycle, as only developmentally competent ones would reach the blastocyst stage. These last aspects will help in reducing costs, thus extending the patients population that can benefit from this technology. Finally, novel comprehensive chromosomal screening techniques, i.e. aCGH, SNP array and quantitative real-time PCR (qPCR), provide us with reliable, sensible and accurate analysis methods, making of PGS also a technically solid approach.

References
1. Nagaoka SI, Hassold TJ, Hunt PA. Human aneuploidy: mechanisms and new insights into an age-old problem. Nat Rev Genet. 2012; 13(7): 493-504.
2. Heffner LJ. Advanced maternal age–how old is too old? N Engl J Med. 2004; 351(19): 1927-1929.
3. Hassold T, Hunt P. To err (meiotically) is human: the genesis of human aneuploidy. Nat Rev Genet. 2001; 2(4): 280-291.
4. Handyside AH, Montag M, Magli MC, Repping S, et al. Multiple meiotic errors caused by predivision of chromatids in women of advanced maternal age undergoing in vitro fertilisation. Eur J Hum Genet. 2012; 20(7): 742-747.
5. Capalbo A, Bono S, Spizzichino L, Biricik A, et al. Sequential comprehensive chromosome analysis on polar bodies, blastomeres and trophoblast: insights into female meiotic errors and chromosomal segregation in the preimplantation window of embryo development. Hum Reprod. 2013; 28(2): 509-518.
6. Voullaire L, Slater H, Williamson R, Wilton L. Chromosome analysis of blastomeres from human embryos by using comparative genomic hybridization. Hum Genet. 2000; 106(2): 210-217.
7. Wells D, Delhanty JD. Comprehensive chromosomal analysis of human preimplantation embryos using whole genome amplification and single cell comparative genomic hybridization. Mol Hum Reprod. 2000; 6(11): 1055-1062.
8. Mertzanidou A, Wilton L, Cheng J, Spits C, et al. Microarray analysis reveals abnormal chromosomal complements in over 70% of 14 normally developing human embryos. Hum Reprod. 2013; 28(1): 256-264.
9. Capalbo A, Wright G, Elliott T, Ubaldi FM, et al. FISH reanalysis of inner cell mass and trophectoderm samples of previously array-CGH screened blastocysts shows high accuracy of diagnosis and no major diagnostic impact of mosaicism at the blastocyst stage. Hum Reprod. 2013; 28(8): 2298-2307.
10. Cauffman G, De Rycke M, Sermon K, Liebaers I, Van de Velde H. Markers that define stemness in ESC are unable to identify the totipotent cells in human preimplantation embryos. Hum Reprod. 2009; 24(1): 63-70.
11. Northrop LE, Treff NR, Levy B, Scott RT Jr. SNP microarray-based 24 chromosome aneuploidy screening demonstrates that cleavage-stage FISH poorly predicts aneuploidy in embryos that develop to morphologically normal blastocysts. Mol Hum Reprod. 2010; 16(8): 590-600.
12. Scott RT Jr, Upham KM, Forman EJ, Zhao T, Treff NR. Cleavage-stage biopsy significantly impairs human embryonic implantation potential while blastocyst biopsy does not: a randomized and paired clinical trial. Fertil Steril. 2013; 100(3): 624-630.
13. Forman EJ, Hong KH, Ferry KM, Tao X, et al. In vitro fertilization with single euploid blastocyst transfer: a randomized controlled trial. Fertil Steril. 2013; 100(1): 100-107.

The authors
Danilo Cimadomo BSc, Antonio Capalbo* PhD, Laura Rienzi MS, Filippo Maria Ubaldi MS
G.EN.E.R.A. Centre for Reproductive
Medicine, Clinica Valle Giulia, Via G. De Notaris 2b, 00197 Rome, Italy
*Corresponding author
E-mail: capalbo@generaroma.it

Prostate specific antigen (PSA) has significantly improved the early detection of prostate cancer (PCa), reducing the related mortality rate. However, PSA has a low specificity, being affected by many benign conditions. [-2]proPSA, a PSA precursor, is a more specific and accurate biomarker indicating prostate biopsy in men at real risk of PCa.

by Dr A. Abrate, Dr M. Lazzeri and Prof. G. Guazzoni

PSA as a marker for prostate cancer
Prostate specific antigen (PSA) is a serum marker widely used for the early detection of prostate cancer (PCa). Its introduction into clinical practice in the early 1990s had an extraordinary impact on the diagnosis and management of PCa. In fact, 20 years after its introduction, the PSA-based PCa opportunistic or systematic screening has resulted in a stage migration to more organ-confined tumours at the time of diagnosis, and consequently to a consistent reduction in PCa related mortality [1, 2]. However, PSA is not a perfect marker for the detection of PCa because of its low specificity and sensitivity. Its levels may increase as a result of benign conditions, such as benign prostatic hyperplasia (BPH) and chronic prostatitis. Moreover, PSA levels are also affected by biologic variability, which may be related to differences in androgen levels, prostate manipulation or ejaculation. Finally, alterations in PSA levels may be related to sample handling, laboratory processing, or assay standardization. All these factors made it difficult to find an appropriate PSA cut-off point diagnostic for PCa (for many years considered to be 4 ng/ml).

Thus, prostate biopsy is still mandatory to confirm the diagnosis. However, this is positive in only approximately 30% of patients [3], and the European Association of Urology suggests a repeat biopsy if PSA is persistently elevated, the digital rectal examination (DRE) is suspicious, or there is a pathological diagnosis of atypical small acinar proliferation (ASAP) or high-grade prostatic intraepithelial neoplasia (HG-PIN) [4]. Finally, PCa (also high-grade cancer) is not rare (approximately 15.2%) among men with PSA levels lower than 4 ng/ml, the previously widely accepted cut-off point [5].

Considering all these observations, it is clear that PSA is an organ-specific rather than an ideal cancer-specific marker.

The introduction into clinical practice of measuring the levels of several derivatives of PSA (free PSA, percentage of free PSA, PSA density, PSA velocity) improved the accuracy of total PSA (tPSA) in detecting PCa. Recently, free PSA (fPSA) has been found to include several subforms, such as proPSA. In particular [-2]proPSA seems to be specific for PCa, opening new ways for early cancer detection.

Biological basis of proPSA
The currently measurable serum tPSA consists of either a complexed form (cPSA, 70–90%), bound by protease inhibitors (primarily alpha1-antichymotrypsin), and a non-complexed form (fPSA). Recently fPSA has been discovered to exist in at least three molecular forms: proPSA, benign PSA (BPSA), and inactive intact PSA (iPSA), covering approximately 33%, 28%, and 39% of fPSA respectively (Fig. 1) [6]. In particular, proPSA is a proenzyme (precursor) of PSA, which is associated with PCa [7].

PSA is synthesized with a 17-amino acid leader sequence (preproPSA) that is cleaved co-translationally to generate an inactive 244-amino acid precursor protein (proPSA, with seven additional amino acids compared to mature PSA). proPSA is normally secreted from the prostate luminal epithelial cells. Immediately after its release into the lumen, the pro-leader part is removed, creating the active form, by the effect of human kallikrein (hK)-2 and hK-4, which have a trypsin-like activity and are expressed predominantly by prostate secretory epithelium. Other kallikreins, localized in the prostate, such as hK216 or prostin17, are involved in the conversion and activation of proPSA. Cleavage of the N-terminal seven amino acids from proPSA generates the active enzyme, which has a mass of 33 kDa.

The partial removal of this leader sequence leads to other truncated forms of proPSA. Thus, theoretically seven isoforms of proPSA could exist, although only [-1], [-2], [-4], [-5], [-7]proPSA were found;  there is still no evidence of [-3], [-6]proPSA. However, all forms of proPSA are enzymatically inactive [8]. It is possible to detect three truncated forms of proPSA in serum: [-5/-7], [-4] and [-2]proPSA, which is the most stable form (Fig. 1).

Notably, in vitro experiments showed that the [-2]proPSA form cannot be activated by either hK2 or trypsin; thus, once it is formed, [-2]proPSA is resistant to activation into the mature PSA form and consequently this is the most reliable test.

Mikolajczyk et al. [7], using a monoclonal antibody recognizing [-2]proPSA, found increased staining in the secretions from malignant prostate glands. In particular [-2]proPSA is differentially elevated in peripheral gland cancer tissue; conversely transition zone tissue contains little or no proPSA.

The increased serum tPSA concentrations in patients with PCa do not result from increased expression but rather from an increased release of PSA into the bloodstream, due to disruption of the epithelial architecture. fPSA is catalytically inactive because of internal cleavages, occurring in seminal plasma, and does not form complexes with protease inhibitors or other proteins: in PCa %fPSA is lower presumably because, consequently to an increased release of PSA into the bloodstream, a very low part is still degraded into the ducts.

In another later study [9], Mikolajczyk et al. found that [-2]proPSA was specifically higher in patients with PCa. Analysing a small number of patients with biopsy positive for PCa and tPSA between 6 and 24 ng/ml, they found that [-2]proPSA constituted a high fraction of fPSA (25% to 95%), which was greater than in patients with a negative biopsy. However, the molecular basis for the proPSA elevation in PCa is uncertain, although a decreased cleavage by hK2 could be the cause.

Clinical utility of proPSA
Sokoll et al. [10] were the first to study the role of proPSA in the early detection of PCa. The study involved archival serum from 119 men (31 PCa, 88 non-cancer), obtained before biopsy and in the tPSA range of 2.5–4.0 ng/ml. The serum levels of tPSA, fPSA, proPSA, and proPSA/fPSA ratio (%proPSA) were analysed: PSA and %fPSA values were similar between the non-cancer and PCa groups, and %proPSA was relatively higher in the PCa group (50.1±4.4%) compared to the non-cancer group (35.5±6.7%; P=0.07). Concerning the clinical utility, the area under the curve (AUC) for %proPSA was 0.688 compared to 0.567 for %fPSA. At fixed sensitivity of 75%, the specificity was significantly greater for %proPSA at 59% compared with %fPSA at 33% (P<0.0001). Afterwards, the Prostate Health Index (PHI) has been proposed as a mathematical algorithm combining tPSA, fPSA and [-2]proPSA according to the formula: ([-2]proPSA/fPSA) × √tPSA. A large American prospective trial [11], involving 892 men who had tPSA levels of 2–10 ng/ml and negative digital rectal examination results, showed that PHI had greater predictive accuracy for prostate biopsy outcome (AUC 0.703) than [-2]proPSA (AUC 0.557), %fPSA (AUC 0.648) and PSA (AUC 0.525), directly correlating with Gleason score (GS) (P=0.013), with an AUC of 0.724 for GS ≥4+3 disease. Moreover, men with PHI >55 had a 42% likelihood of being diagnosed with high-grade disease on biopsy compared to 26% of men with PHI 0–24.9.
Accordingly, an observational European multicenter cohort study involved 646 men with tPSA levels of 2–10 ng/ml, who had undergone prostate biopsy [12]. [-2]proPSA and PHI improved the predictive accuracy for the detection of overall PCa (and also GS ≥7 disease) compared to PSA and derivatives. In fact, at 90% sensitivity, the PHI cut-off of 27.6 could avoid 100 (15.5%) biopsies, missing 26 (9.8%) cancers (23 with GS 6, three with GS 3+4).

Moreover, a PHI based nomogram to predict PCa at extended prostate biopsy was developed and validated over 729 patients [13]. Including PHI in a multivariable logistic regression model, based on patient age, prostate volume, digital rectal examination and biopsy history, significantly increased predictive accuracy by 7% from 0.73 to 0.80 (P<0.001). Decision curve analysis showed that using the PHI based nomogram resulted in the highest net benefit. Recently, it was demonstrated that PHI might have a role in screening patients at high risk of PCa [14]. Specifically, the study involved 158 men with a positive family history undergoing prostate biopsy within the multicentre European PROMEtheuS cohort. Similarly to previous studies in the general population, PHI outperformed tPSA and %fPSA for PCa detection on biopsy (AUC 0.73, 0.55 and 0.60, respectively). In addition, both [-2]proPSA and PHI were directly associated with GS in men with a positive family history. Overall, the authors reported that using a PHI cutoff value of 25.5 would have avoided 17.2% of biopsies while missing only two GS 7 cancers. On decision curve analysis, the addition of PHI to a base predictive model that included age, prostate volume, tPSA, fPSA and %fPSA resulted in net benefit at threshold probabilities of 35–65%. This result suggests that PHI should be incorporated into a multivariable risk assessment for high-risk patients because it offers improved performance for PCa detection. Conclusions
[-2]proPSA and PHI are more accurate than the currently used tests (PSA and derivatives) in predicting the presence of PCa at biopsy. Their implementation in clinical practice has the potential to significantly increase physicians’ ability to detect PCa and avoid unnecessary biopsies. Further work is needed to confirm and generalize these data on wider populations.

References
1. Hoffman RM, Stone SN, Espey D, Potosky AL. Differences between men with screening-detected versus clinically diagnosed prostate cancers in the USA. BMC Cancer 2005; 5: 27.
2. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin. 2013; 63: 11–30
3. Vickers AJ, Cronin AM, Roobol MJ, et al. The relationship between prostate-specific antigen and prostate cancer risk: the Prostate Biopsy Collaborative Group. Clin Cancer Res. 2010; 16: 4374–4381.
4. Heidenreich A, Bellmunt J, Bolla M, et al. EAU guidelines on prostate cancer. Part 1: screening, diagnosis, and treatment of clinically localised disease. Eur Urol. 2011; 59: 61–71.
5. Thompson IM, Pauler DK, Goodman PJ, et al. Prevalence of prostate cancer among men with a prostate-specific antigen level < or =4.0 ng per milliliter. N Engl J Med. 2004; 350: 2239–2246.
6. Mikolajczyk SD, Rittenhouse HG. Pro PSA: a more cancer specific form of prostate specific antigen for the early detection of prostate cancer. Keio J Med. 2003; 52: 86–91.
7. Mikolajczyk SD, Millar LS, Wang TJ, et al. A precursor form of prostate-specific antigen is more highly elevated in prostate cancer compared with benign transition zone prostate tissue. Cancer Res. 2000; 60: 756–759.
8. Jansen FH, Roobol M, Jenster G, Schroder FH, Bangma CH. Screening for prostate cancer in 2008 II: the importance of molecular subforms of prostate-specific antigen and tissue kallikreins. Eur Urol. 2009; 55: 563–74.
9. Mikolajczyk SD, Marker KM, Millar LS, et al. A truncated precursor form of prostate-specific antigen is a more specific serum marker of prostate cancer. Cancer Res. 2001; 61: 6958–6963
10. Sokoll LJ, Chan DW, Mikolajczyk SD, et al. Proenzyme psa for the early detection of prostate cancer in the 2.5–4.0 ng/ml total psa range: preliminary analysis. Urology 2003; 61: 274–276.
11. Catalona WJ, Partin AW, Sanda MG, et al. A multicenter study of [-2]pro-prostate specific antigen combined with prostate specific antigen and free prostate specific antigen for prostate cancer detection in the 2.0 to 10.0 ng/ml prostate specific antigen range. J Urol. 2011; 185: 1650–1655.
12. Lazzeri M, Haese A, de la Taille A, et al. Serum isoform [-2]proPSA derivatives significantly improve prediction of prostate cancer at initial biopsy in a total PSA range of 2-10 ng/ml: a multicentric European study. Eur Urol. 2013; 63: 986–994.
13. Lughezzani G, Lazzeri M, Larcher A, et al. Development and internal validation of a Prostate Health Index based nomogram for predicting prostate cancer at extended biopsy. J Urol. 2012; 188: 1144–1150.
14. Lazzeri M, Haese A, Abrate A, et al. Clinical performance of serum prostate-specific antigen isoform [-2]proPSA (p2PSA) and its derivatives, %p2PSA and the prostate health index (PHI), in men with a family history of prostate cancer: results from a multicentre European study, the PROMEtheuS project. BJU Int. 2013; 112: 313–321.

The author
Alberto Abrate* MD; Massimo Lazzeri MD, PhD; and Giorgio Guazzoni MD
Dept of Urology, Ospedale San Raffaele Turro, San Raffaele
Scientific Institute, Milan, Italy
*Corresponding author
E-mail: alberto.abrate@gmail.com

Establishing metrological traceability of measurement is essential to improve the accuracy and comparability of measurement results. With increasing recognition of the importance of traceability, some regulatory policies have been applied to enforce its implementation. Technology advancement also provides more tools for improving measurement traceability.  During the assay development on the Mindray CL-2000i Chemiluminescence Immunoassay System,  well recognized highest reference methods or reference materials were used in assigning the values of master calibrators; the accuracy of product calibrators was guaranteed through an unbroken metrological traceability chain.

by Xiang Yu and Ke Li

Introduction
With the advancement in automation over the past 20 years, most of the immunoassays have been shifted from traditional manual assays to fully automatic systems leading to an overall improvement of the quality of measurements. The accuracy and comparability of testing results have been emphasized, since they are the keys to defining and using common clinical decision values and reference intervals, following constant standards and practice guidelines, pooling data from different studies based on different analytical systems to facilitate clinical research.

One critical mechanism to improve the accuracy and comparability of clinical testing results is to make the testing results traceable to higher reference materials or methods in calibration hierarchy. Briefly, the testing results should have metrological traceability. The general principles and features have been described in the document of the International Organization for Standardization (ISO) 17511:2003 [1].

Ideally, results produced by different routine methods for the same measurand should be metrologically traceable to the highest level of calibration hierarchy – the International System of Units (SI units), with an estimated measurement uncertainty. However, only a limited number of analytes, including some metabolites, electrolytes, steroid hormones, has reference materials available with traceability to the SI unit. Most of the clinical analytes still have no primary and secondary reference measurement procedures and are not traceable to the SI unit. They are not well defined and have only traceability to an international conventional standard or manufacturers’ internal standard, such as tumour markers and viral antigens [2].

The EU directive on in vitro diagnostic devices (IVDD) enacted in 1998 stated “The traceability of values assigned to calibrators and/or control materials must be assured through available reference measurement procedures and/or available reference materials of a higher order” [3]. Therefore, for all the IVD analytical system (reagents), manufacturers must ensure their products are standardized against available reference materials or methods in order to be distributed in the EU market.

Traceability chain and value assignment procedure on Mindray CL-2000i System
Mindray CL-2000i system is a closed system composed of a fully automatic immunoanalyser, related reagents and calibrators. The calibration hierarchy was established and documented strictly based on EN ISO 17511:2003 [1]. Mindray’s traceability procedure is indicated in figure 1, ensuring the establishment of metrological traceability between the testing results and the highest standard available. Based on the characteristics of different analytes, three major traceability chains have been used: traceable to an SI unit, traceable to an international conventional calibrator, and traceable to manufacturers’ selected procedure.

Measurements traceable to the SI unit
If the chemical and physical properties of an analyte are well defined, there should be a primary reference measurement procedure with the measurement traceable to the SI unit (mole). CL-2000i total T3, total T4, progesterone, testosterone and estradiol are traceable to this highest level of calibration hierarchy. Mindray has performed the traceability of the above measurements in collaboration with the Reference Institute for Bioanalytics (RfB), a German reference laboratory certified by the Joint Committee for Traceability in Laboratory Medicine (JCTLM) [4]. Thirty Mindray master calibrators at different levels covering the whole detection range were assigned values for each analyte at RfB with the reference measurement procedure of Isotope dilution mass spectrometry (ID-MS). The calibrator values with uncertainty were then applied to define the values of Mindray working calibrators and product calibrators, and the metrological traceability between the testing results of CL-2000i end-users’ routine measurement procedure and the SI unit was finally established. The assays that are traceable to the SI unit are indicated in Table 1.

Measurements traceable to an international conventional calibrator
The reference materials, such as WHO standards and some national standard materials are defined by convention or consensus, without traceability to the SI unit; the assigned values are in arbitrary units (e.g. WHO international unit). Most of assays for tumour markers, hormones, and viral antigen/antibody of the CL-2000i system are traceable to this kind of reference materials, indicated in Table 1.

Measurements traceable to manufacturers’ selected procedure
For analytes that are either not traceable to the SI unit, or for which no reference method and reference material are available, a commercial certified measurement procedure with traceability, high accuracy and analytical specificity was selected for Mindray master calibrator value assignment; the measurement accuracy of the Mindray routine measurement procedure is ensured and also indicated in Table 1.

Principle of the traceability of Mindray CL-2000i end-user’s measurement results
The immunoanalyser is calibrated by measuring three levels of product calibrators and relative light units (RLUs) generated. The corresponding concentration of each calibrator was used to adjust the master calibration curve stored in the barcode of each lot of reagents.

The value of end-user’s product calibrators and the master curve stored in the barcode are both defined by the Mindray routine measurement procedure that is calibrated by Mindray working calibrators in the manufacturer’s laboratory. The working calibrators have roughly 12 concentration levels and have the same matrix as the end-user’s product calibrators.

It is the Mindray standard measurement procedure that determines the values of Mindray working calibrators. The Mindray standard measurement procedure makes use of the Mindray standard CL-2000i automatic immunoassay analyser, standard reagents, and Mindray master calibrators. Mindray master calibrators are composed of a series of human serum at different concentration levels. They are stored at -70°C and represent the highest accepted standard available.

The values of the Mindray master calibrators are fixed, and the measurement standard established by the Mindray standard measurement procedure is preferably not variable and should be kept as consistent as possible. On the other hand, the value of working calibrators and end-user’s product calibrators can be flexible within a certain range. The assigned values of calibrators will be adjusted according to the results of internal QC and method comparison so as to ensure the traceability between the reference and end-user’s routine measurement procedures.

Discussion
We have made our best efforts for the traceability of the Mindray CL-2000i system, eventhough the implementation of traceability is challenging, especially the traceability in immunoassays.

Firstly, majority of analytes lack a primary reference measurement procedure and thus are not traceable to the SI units. The chemistry and physical properties of these analytes still require more accurate definition.

Secondly, the international conventional calibrators have played an important role in harmonizing testing results. However, there are still some issues with using the international standards, such as the long term stability of WHO standards, the matrix effect, the difference between different generations of the standards, and difference between the source of the standards and the real sample in the clinic.

Thirdly, some of the analytes have neither reference materials nor reference methods available, and are only traceable to manufacturers’ in-house standards. The harmonization of clinical results could not be fully implemented [5].

References
1. ISO 17511:2003. In vitro diagnostic medical devices –measurement of quantities in biological samples – metrological traceability of values assigned to calibrators and control materials. Geneva, Switzerland: ISO
2. Database of higher-order reference materials and reference measurement methods/procedures. http://www.bipm.org/en/committees/jc/jctlm/jctlm-db
3. Directive 98/79/EC of the European Parliament and of the Council of 27 October 1998 on in vitro diagnostic medical devices. Off J Eur Union 7 December 1998; L 331:1–37.
4. JCTLM: Joint Committee for Traceability in Laboratory Medicine. http://www.bipm.org/en/committees/jc/jctlm/
5. Danni L. Meany and Daniel W. Chan Comparability of tumor marker immunoassays: still an important issue for clinical diagnostics? Clin Chem Lab Med 2008; 46(5):575–576.

The authors
Xiang Yu*, MSc and Ke Li, PhD
Immunoassay Department, Shenzhen
Mindray Bio-Medical Electronics Co. Ltd., Nanshan, Shenzhen, 518057 China

*Corresponding author
Email: yuxiang@mindray.com