Natural killer cells and reproductive failure—theory, practice and prejudice

Raj Rai1,3, Gavin Sacks1,2 and Geoffrey Trew2

1 Department of Obstetrics and Gynaecology, Faculty of Medicine, Imperial College London, St Mary's Hospital, and 2 Department of Reproductive Medicine, Hammersmith Hospital, London, UK

3 To whom correspondence should be addressed at: Department of Obstetrics and Gynaecology, Imperial College London, Mint Wing, South Wharf Road, London W2 1PG, UK. Email: r.rai{at}imperial.ac.uk


    Abstract
 Top
 Abstract
 Introduction
 NK cell biology
 Reproductive role for NK...
 NK cells and reproductive...
 Treatment of raised peripheral...
 Conclusions
 References
 
The relationship between peripheral blood natural killer (NK) cells and reproductive failure is one of the most controversial areas in reproductive medicine. Amidst much publicity, peripheral blood NK cell testing is being promoted as a useful diagnostic test to guide the initiation of a variety of immunosuppressive therapies amongst patients with either recurrent miscarriage or infertility. We contend (i) that at present there is no scientific basis for the introduction of NK cell testing into routine clinical practice, and (ii) that the use of immunosuppressant agents based on the results of such testing may potentially be harmful.

Key words: NK cells/peripheral blood/pregnancy/TNF-{alpha}


    Introduction
 Top
 Abstract
 Introduction
 NK cell biology
 Reproductive role for NK...
 NK cells and reproductive...
 Treatment of raised peripheral...
 Conclusions
 References
 
The investigation and treatment of couples with reproductive failure is characterized by historical beliefs, anecdotal evidence and the results of small uncontrolled studies. This has lead to the situation where women have been subjected, often at significant financial cost, to treatments such as lymphocyte immunization therapy (LIT), the efficacy of which are at best controversial despite initially promising reports which have not been substantiated in more recent studies (Mowbray et al., 1987Go; Rai et al., 1996Go; Scott, 2003Go).

The arguments supporting an aberrant immune response as being implicated in many cases of reproductive failure have been extensively rehearsed (Coulam et al., 1999Go; Gleicher, 2002Go). However, the evidence in favour of these various putative immune or autoimmune causes is scant (Regan et al., 2004Go). Natural killer (NK) cells are the latest potential immune cause of reproductive failure to enter the clinical arena.


    NK cell biology
 Top
 Abstract
 Introduction
 NK cell biology
 Reproductive role for NK...
 NK cells and reproductive...
 Treatment of raised peripheral...
 Conclusions
 References
 
The name ‘natural killer’ is evocative. It derives from an in-vitro assay used to detect these cells. Importantly, however, NK cells only kill trophoblast in vitro if activated by interleukin (IL)-2, which is not present in the endometrium at the time of implantation (King et al., 1995Go; Abadia-Molina et al., 1996Go).

NK cells are lymphocytes that are part of the innate immune system. They express the cell surface antigens CD16 and CD56. CD16 is a low-affinity receptor for IgG complexes and is expressed on the majority of NK cells. It is the receptor responsible for NK-mediated, antibody-dependent cellular cytotoxicity. Based on the intensity of CD56 expression, NK cells may be divided into two populations: CD56dim and CD56bright. NK cells that are CD56dim are cytotoxic in vitro. In contrast, those that are CD56bright have little cytotoxic ability, but produce immunoregulatory cytokines such as interferon-{gamma} and tumour necrosis factor-{alpha} (TNF-{alpha}).

NK cell function is tightly regulated by a network of specific activating and inhibitory receptors. In particular, CD69 is one of the earliest cell surface activation markers expressed by NK cells. NK cells also express a variety of killer inhibitory receptors (KIRs) and killing activating receptors (KARs), which recognize HLA-G expressed on extravillous trophoblast (Moffett-King, 2002Go).

NK cells are found in both peripheral blood and the uterine mucosa. There are, however, important phenotypic and functional differences between NK cells present at the two sites.

The majority (90%) of peripheral blood NK cells are CD56dim and express high levels of CD16; these levels do not fluctuate during the menstrual cycle, and during pregnancy peripheral NK cell numbers and functional activity are suppressed (Sacks et al., 1999Go; Yovel et al., 2001Go). In contrast, NK cells are the predominant leucocyte population in the endometrium, particularly in the decidua basalis at the implantation site. The number of uterine NK (uNK) cells varies during the menstrual cycle. They are sparse during the proliferative phase, increase significantly throughout the secretory phase, remain in high numbers during early gestation, decrease after 20 weeks gestation and are absent in term decidua (Bulmer and Sunderland, 1984Go; Bulmer et al., 1991Go; Trundley and Moffett, 2004Go). Microarray analysis combined with flow-cytometric and RT–PCR studies have demonstrated that the phenotype of uNK cells (CD56bright CD16) is different from that of NK cells in peripheral blood (CD56dim CD16+) (Koopman et al., 2003Go). In addition, Koopman et al. were able to demonstrate that uNK cells have an immunoregulatory potential that peripheral blood NK cells do not demonstrate. This suggests that uNK cells either represent a distinct subpopulation of circulating NK cells or that they have undergone some tissue-specific differentiation. Consequently, data derived from studies of peripheral blood NK cells may not reflect what is happening at the feto-maternal interface.


    Reproductive role for NK cells
 Top
 Abstract
 Introduction
 NK cell biology
 Reproductive role for NK...
 NK cells and reproductive...
 Treatment of raised peripheral...
 Conclusions
 References
 
The function of NK cells in governing reproductive outcome has not been demonstrated. Studies of transgenic mice deficient in NK cells demonstrate that uNK cells play a role in uterine spiral artery modification in the placenta (Croy et al., 2003Go). Several investigators have used murine NK knock-out models to study the effect of NK cells on pregnancy outcome. However, different knock outs have reported differing results. Using Terhost mutants, Croy's group have reported that NK deficient mice have an increased rate of fetal loss (Guimond et al., 1998Go). In contrast, the fetal resporption (miscarriage) rate is not increased in IL-15–/–, NK cell-deficient mice, nor amongst IL-2 receptor knock-outs, which also lack uNK cells (Miyazaki et al., 2002Go; Barber and Pollard, 2003Go). These differing results may be explained, at least in part, by the subset of NK cells that is affected by the knock-out.

In human pregnancy, the temporal and spatial distribution of uNK cells suggest they play a role in controlling trophoblast invasion. uNK cell numbers are increased at the time of implantation; they lie in close proximity to the extravillous trophoblast and they express receptors that interact with ligands expressed on trophoblast (Moffett-King, 2002Go).


    NK cells and reproductive outcome
 Top
 Abstract
 Introduction
 NK cell biology
 Reproductive role for NK...
 NK cells and reproductive...
 Treatment of raised peripheral...
 Conclusions
 References
 
The relationship between peripheral blood NK cells and pregnancy outcome after either IVF or spontaneous conception has been examined in several small observational studies (Aoki et al., 1995Go; Beer et al., 1996Go; Fukui et al., 1999Go; Emmer et al., 2000Go; Ntrivalas et al., 2001Go; Yamada et al., 2003Go; Putowski et al., 2004Go). Although the underlying aetiology of reproductive failure amongst women in individual studies may well be different, the reports are consistent in associating enhanced NK cell activity with subsequent failure to conceive or miscarriage. These studies have received extensive attention in all media channels. This has led to many patients demanding the performance of these tests and ‘treatment’ for ‘abnormal’ results. Is peripheral NK cell testing an important advance or, as we argue, poor science enthusiastically endorsed by its proponents?

There are fundamental flaws in the methodologies used in the published studies. The levels and activation of NK cells is dependent, amongst other variables, on whether whole blood or fractionated mononuclear cells is used, the time of day a sample is taken, whether any physical exercise has been performed, the parity of the patient and whether the samples have been previously frozen (Strong et al., 1982Go; Porzsolt et al., 1983Go; Pross and Maroun, 1984Go; Reichert et al., 1991Go; Plackett et al., 2004Go). Furthermore, it is unclear what an abnormal NK cell number is. Whilst traditionally a peripheral NK cell level >12% of all lymphocytes has been regarded as the cut-off between a raised and a normal level (Beer et al., 1996Go), this figure is well within the normal range (up to 29%) published by others (Eidukaite et al., 2004Go). Hence individuals with entirely normal results are being labelled as having raised NK cell numbers.

A recent study published in this Journal illustrates several of the important pitfalls in peripheral blood NK cell testing and the interpretation of the results (Thum et al., 2004Go). The authors have reported that an increase in the absolute count of activated NK cells (CD56dim CD16+ CD69+) is associated with a reduced implantation rate amongst women undergoing IVF. By inference it is suggested that (i) this is a useful test for women undergoing IVF, and (ii) therapeutic intervention of some sort to decrease peripheral blood NK cell activation will be of benefit to those women with an increased number of activated NK cells. Are these claims justified?

Peripheral blood NK cells are counted using flow-cytometric analysis. The count will be dependent on (i) the pre-analytical variables we have discussed and (ii) the setting of the lymphocyte gate on the flow cytometer, which is arbitrary and will vary from experiment to experiment. In their manuscript, the authors state that the count for the cell population of interest (CD56dim CD16+ CD69+) is 1.66 x 106 out of an overall count of 212 x 106 for CD56dim CD16+ (0.78%). With numbers so small, variations in the count induced by the settings of the fluorescence-activated cell-sorting parameters can be larger than the claimed differences between groups.

Moving beyond this fundamental issue, do the published data support the hypothesis that peripheral blood NK cell assays are useful tests? The area under a receiver–operating characteristic (ROC) graph is a measure of how well a test (levels of peripheral blood NK cells) divides a cohort of patients into those with a disease (implantation failure following IVF) and those without (implantation achieved). An area of 1.0 represents a perfect test, whilst a figure of 0.5 implies a worthless test. At the cut-off value used by the authors themselves, they have calculated an area of 0.63, demonstrating that the test performs poorly.


    Treatment of raised peripheral blood NK cells
 Top
 Abstract
 Introduction
 NK cell biology
 Reproductive role for NK...
 NK cells and reproductive...
 Treatment of raised peripheral...
 Conclusions
 References
 
Apart from the morbidity generated by NK cell testing, it is of concern that women with ‘raised’ NK cell levels are being treated with a variety of agents—intravenous immunoglobulin, anti-TNF-{alpha} drugs and glucocorticoids—in order to dampen an ‘excessive immune response’. Not only is there no evidence base for these interventions, which are potentially associated with significant morbidity, the rationale for their use may be false.

Hiby et al. (2004)Go recently reported a comparison of the genotypes of (i) maternal uNK cell killer inhibitory receptors and (ii) fetal HLA-C amongst women with and without pre-eclampsia. The genotype combination that results in a maximum inhibitory effect on uNK cell inhibitory receptors was found significantly more often amongst those pregnancies complicated by pre-eclampsia. This suggests that overly inhibited uNK cells cause trophoblast cells to prematurely cease the remodelling of the maternal uterine spiral arteries, which in turn predisposes to the development of pre-eclampsia.

As far as the pharmacological interventions are concerned, intravenous immunoglobulin is a pooled blood product and is associated with anaphylactic response, fever, flushing, muscle pains, nausea and headache (Sherer et al., 2001Go). Anti-TNF-{alpha} agents have been reported to be associated with the development of lymphoma, granulomatous diseases such as tuberculosis, systemic lupus erythematosus-like syndromes, demyelinating disease and congestive cardiac failure (Fleischmann et al., 2004Go). Of equal concern is that the whilst TNF-{alpha} has traditionally been viewed as a cytokine involved in triggering immunologically mediated pregnancy loss, it is also involved in anti-apoptotic signalling pathways and has a regulatory role in cell proliferation. Indeed, studies in TNF-{alpha} knock-out mice suggest that this cytokine may play an important role in embryo development and the prevention of structural abnormalities (Toder et al., 2003Go). These abnormalities may not, of course, be unmasked until later infant life. Glucocorticoids themselves during pregnancy are associated with an increased risk of pre-term delivery secondary to rupture of membranes and the development of pre-eclampsia and gestational diabetes (Laskin et al., 1997Go). Importantly, glucocorticoid receptors are present in the stromal compartment of the endometrium, thus suggesting they play an important role in decidualization. The effect of exogenous glucocorticoid therapy on the endometrial gene expression profile during decidualization has not been examined.


    Conclusions
 Top
 Abstract
 Introduction
 NK cell biology
 Reproductive role for NK...
 NK cells and reproductive...
 Treatment of raised peripheral...
 Conclusions
 References
 
Patient expectations for an answer and, even better, for a cure for their reproductive failure are ever increasing. This leaves them vulnerable to exploitation by those pedalling new tests and treatments that have little scientific validity. We contend that the present evidence does not support peripheral blood NK cell testing amongst those with reproductive failure. Moreover, despite the competitive environment in which assisted conception services are provided, it is incumbent upon us to lead by example and protect our patients from vocal advocates of tests and treatments that are largely based on pseudo-science.


    References
 Top
 Abstract
 Introduction
 NK cell biology
 Reproductive role for NK...
 NK cells and reproductive...
 Treatment of raised peripheral...
 Conclusions
 References
 
Abadia-Molina AC, Ruiz C, Montes MJ, King A, Loke YW and Olivares EG (1996) Immune phenotype and cytotoxic activity of lymphocytes from human term decidua against trophoblast. J Reprod Immunol 31, 109–123.[CrossRef][ISI][Medline]

Aoki K, Kajiura S, Matsumoto Y, Ogasawara M, Okada S, Yagami Y and Gleicher N (1995) Preconceptional natural-killer-cell activity as a predictor of miscarriage. Lancet 345, 1340–1342.[CrossRef][ISI][Medline]

Barber EM and Pollard JW (2003) The uterine NK cell population requires IL-15 but these cells are not required for pregnancy nor the resolution of a Listeria monocytogenes infection. J Immunol 171, 37–46.[Abstract/Free Full Text]

Beer AE, Kwak JY and Ruiz JE (1996) Immunophenotypic profiles of peripheral blood lymphocytes in women with recurrent pregnancy losses and in infertile women with multiple failed in vitro fertilization cycles. Am J Reprod Immunol 35, 376–382.[ISI][Medline]

Bulmer JN and Sunderland CA (1984) Immunohistological characterization of lymphoid cell populations in the early human placental bed. Immunology 52, 349–357.[ISI][Medline]

Bulmer JN, Morrison L, Longfellow M, Ritson A and Pace D (1991) Granulated lymphocytes in human endometrium: histochemical and immunohistochemical studies. Hum Reprod 6, 791–798.[Abstract]

Coulam CB, Branch DW, Clark DA, Gleicher N, Kutteh W, Lockshin MD and Rote NS (1999) American Society for Reproductive Immunology report of the Committee for Establishing Criteria for Diagnosis of Reproductive Autoimmune Syndrome. Am J Reprod Immunol 41, 121–132.[ISI][Medline]

Croy BA, He H, Esadeg S, Wei Q, McCartney D, Zhang J, Borzychowski A, Ashkar AA, Black GP, Evans SS et al. (2003) Uterine natural killer cells: insights into their cellular and molecular biology from mouse modelling. Reproduction 126, 149–160.[Abstract/Free Full Text]

Eidukaite A, Siaurys A and Tamosiunas V (2004) Differential expression of KIR/NKAT2 and CD94 molecules on decidual and peripheral blood CD56bright and CD56dim natural killer cell subsets. Fertil Steril 81 Suppl 1, 863–868.

Emmer PM, Nelen WL, Steegers EA, Hendriks JC, Veerhoek M and Joosten I (2000) Peripheral natural killer cytotoxicity and CD56(pos)CD16(pos) cells increase during early pregnancy in women with a history of recurrent spontaneous abortion. Hum Reprod 15, 1163–1169.[Abstract/Free Full Text]

Fleischmann RM, Iqbal I and Stern RL (2004) Considerations with the use of biological therapy in the treatment of rheumatoid arthritis. Expert Opin Drug Saf 3, 391–403.[CrossRef][Medline]

Fukui A, Fujii S, Yamaguchi E, Kimura H, Sato S and Saito Y (1999) Natural killer cell subpopulations and cytotoxicity for infertile patients undergoing in vitro fertilization. Am J Reprod Immunol 41, 413–422.[ISI][Medline]

Gleicher N (2002) Some thoughts on the reproductive autoimmune failure syndrome (RAFS) and Th-1 versus Th-2 immune responses. Am J Reprod Immunol 48, 252–254.[CrossRef][ISI][Medline]

Guimond MJ, Wang B and Croy BA (1998) Engraftment of bone marrow from severe combined immunodeficient (SCID) mice reverses the reproductive deficits in natural killer cell-deficient tg epsilon 26 mice. J Exp Med 187, 217–223.[Abstract/Free Full Text]

Hiby SE, Walker JJ, O'Shaughnessy KM, Redman CW, Carrington M, Trowsdale J and Moffett A (2004) Combinations of maternal KIR and fetal HLA-C genes influence the risk of preeclampsia and reproductive success. J Exp Med 200, 957–965.[Abstract/Free Full Text]

King A, Jokhi PP, Smith SK, Sharkey AM and Loke YW (1995) Screening for cytokine mRNA in human villous and extravillous trophoblasts using the reverse-transcriptase polymerase chain reaction (RT-PCR). Cytokine 7, 364–371.[CrossRef][ISI][Medline]

Koopman LA, Kopcow HD, Rybalov B, Boyson JE, Orange JS, Schatz F, Masch R, Lockwood CJ, Schachter AD, Park PJ et al. (2003) Human decidual natural killer cells are a unique NK cell subset with immunomodulatory potential. J Exp Med 198, 1201–1212.[Abstract/Free Full Text]

Laskin CA, Bombardier C, Hannah ME, Mandel FP, Ritchie JW, Farewell V, Farine D, Spitzer K, Fielding L, Soloninka CA et al. (1997) Prednisone and aspirin in women with autoantibodies and unexplained recurrent fetal loss. N Engl J Med 337, 148–153.[Abstract/Free Full Text]

Miyazaki S, Tanebe K, Sakai M, Michimata T, Tsuda H, Fujimura M, Nakamura M, Kiso Y and Saito S (2002) Interleukin 2 receptor gamma chain (gamma(c)) knockout mice show less regularity in estrous cycle but achieve normal pregnancy without fetal compromise. Am J Reprod Immunol 47, 222–230.[CrossRef][ISI][Medline]

Moffett-King A (2002) Natural killer cells and pregnancy. Nat Rev Immunol 2, 656–663.[CrossRef][ISI][Medline]

Mowbray JF, Underwood JL, Michel M, Forbes PB and Beard RW (1987) Immunisation with paternal lymphocytes in women with recurrent miscarriage. Lancet 2, 679–680.[Medline]

Ntrivalas EI, Kwak-Kim JY, Gilman-Sachs A, Chung-Bang H, Ng SC, Beaman KD, Mantouvalos HP and Beer AE (2001) Status of peripheral blood natural killer cells in women with recurrent spontaneous abortions and infertility of unknown aetiology. Hum Reprod 16, 855–861.[Abstract/Free Full Text]

Plackett TP, Boehmer ED, Faunce DE and Kovacs EJ (2004) Aging and innate immune cells. J Leukoc Biol 76, 291–299.[Abstract/Free Full Text]

Porzsolt F, Gaus W and Heimpel H (1983) The evaluation of serial measurements of the NK cell activity in man. Immunobiology 165, 475–484.[ISI][Medline]

Pross HF and Maroun JA (1984) The standardization of NK cell assays for use in studies of biological response modifiers. J Immunol Methods 68, 235–249.[CrossRef][ISI][Medline]

Putowski L, Darmochwal-Kolarz D, Rolinski J, Oleszczuk J and Jakowicki J (2004) The immunological profile of infertile women after repeated IVF failure (preliminary study). Eur J Obstet Gynecol Reprod Biol 112, 192–196.[CrossRef][ISI][Medline]

Rai R, Clifford K and Regan L (1996) The modern preventative treatment of recurrent miscarriage. Br J Obstet Gynaecol 103, 106–110.[ISI][Medline]

Regan L, Backos M, Rai R (2004) Immunological testing and interventions for reproductive failure. Royal College of Obstetricians and Gynaecologists. Available online. http://www.rcog.org.uk/resources/public/sac_opinion_paper_5.pdf (???, date last accessed).

Reichert T, DeBruyere M, Deneys V, Totterman T, Lydyard P, Yuksel F, Chapel H, Jewell D, Van Hove L, Linden J et al. (1991) Lymphocyte subset reference ranges in adult Caucasians. Clin Immunol Immunopathol 60, 190–208.[CrossRef][ISI][Medline]

Sacks G, Sargent I and Redman C (1999) An innate view of human pregnancy. Immunol Today 20, 114–118.[CrossRef][ISI][Medline]

Scott JR (2003) Immunotherapy for recurrent miscarriage. Cochrane Database Syst Rev 1, CD000112.[Medline]

Sherer Y, Levy Y, Langevitz P, Rauova L, Fabrizzi F and Shoenfeld Y (2001) Adverse effects of intravenous immunoglobulin therapy in 56 patients with autoimmune diseases. Pharmacology 62, 133–137.[CrossRef][ISI][Medline]

Strong DM, Ortaldo JR, Pandolfi F, Maluish A and Herberman RB (1982) Cryopreservation of human mononuclear cells for quality control in clinical immunology I. Correlations in recovery of K- and NK-cell functions, surface markers, and morphology. J Clin Immunol 2, 214–221.[CrossRef][ISI][Medline]

Thum MY, Bhaskaran S, Abdalla HI, Ford B, Sumar N, Shehata H and Bansal AS (2004) An increase in the absolute count of CD56dimCD16+CD69+ NK cells in the peripheral blood is associated with a poorer IVF treatment and pregnancy outcome. Hum Reprod 19, 2395–2400.[Abstract/Free Full Text]

Toder V, Fein A, Carp H and Torchinsky A (2003) TNF-alpha in pregnancy loss and embryo maldevelopment: a mediator of detrimental stimuli or a protector of the fetoplacental unit? J Assist Reprod Genet 20, 73–81.[CrossRef][ISI][Medline]

Trundley A and Moffett A (2004) Human uterine leukocytes and pregnancy. Tissue Antigens 63, 1–12.[CrossRef][ISI][Medline]

Yamada H, Morikawa M, Kato EH, Shimada S, Kobashi G and Minakami H (2003) Pre-conceptional natural killer cell activity and percentage as predictors of biochemical pregnancy and spontaneous abortion with normal chromosome karyotype. Am J Reprod Immunol 50, 351–354.[CrossRef][ISI][Medline]

Yovel G, Shakhar K and Ben Eliyahu S (2001) The effects of sex, menstrual cycle, and oral contraceptives on the number and activity of natural killer cells. Gynecol Oncol 81, 254–262.[CrossRef][ISI][Medline]

Submitted on October 25, 2004; resubmitted on January 16, 2005; accepted on January 20, 2005.





This Article
Abstract
Full Text (PDF )
All Versions of this Article:
20/5/1123    most recent
deh804v1
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Request Permissions
Google Scholar
Articles by Rai, R.
Articles by Trew, G.
PubMed
PubMed Citation
Articles by Rai, R.
Articles by Trew, G.