Lower interleukin-2 and higher serum tumor necrosis factor-a levels are associated with perimenstrual, recurrent, facial Herpes simplex infection in young women

ARTICLE

The time around menstruation has been recognised as conferring a susceptibility to several infectious diseases. Higher, clinically symptomatic incidences of vulvovaginal candidiasis [1, 2] chlamydia trachomatis infections [3], increased vaginal transmission of simian immunodeficiency virus [4] have already been noted as related to a late luteal phase and menstruation.

The recurrent Herpes simplex type I infection belongs to the perimenstrual dermatoses ­ diseases affecting young women at around the time of menstruation [5]. Immunity against Herpes simplex virus-infected cells is conferred by the natural killer cells (NK) [6] and T lymphocytes [7] which are both able to exert a direct cytotoxic effect upon the infected cells. This effect is, however, possible only when these cells are prestimulated with interleukin-2 (IL-2) [8] ­ a pivotal factor for the maturation and function of NK and T cells [9]. It has been documented that the impaired production of IL-2 is associated with a recurrence of viral infections [10]. In the course of viral infections, there is a high stimulation of tumor necrosis factor-alpha (TNF-alpha) production, coinciding with a peak of clinical symptoms [11]. TNF-alpha as a critical mediator of host defence mechanisms against infection regulates the extent and duration of the inflammatory process [12]. The association of the Herpes simplex type I infection with the perimenstrual period implies that the systemic blood levels of IL-2 and TNF-alpha may shift during the menstrual cycle in such a way as to facilitate the recurrence of the symptoms.

The fact that we have not come across data on the systemic levels of IL-2 and TNF-alpha in women suffering from perimenstrual dermatoses, prompted us to carry out the present work. The aim of this work was to look at a possible relationship between the recurrent perimenstrual facial Herpes simplex infection and the serum concentrations of IL-2 and TNF-alpha.

MATERIALS AND METHODS

Twenty one, healthy, young (19-29 year old) women medical students were examined. All the volunteers had regular menstrual cycles lasting twenty eight days. They had neither received hormonal treatment, including contraception, nor any other medication. The health status of the women had been regularly checked each year in The Students' Outpatient Department. Moreover, the volunteers underwent scrupulous medical examination (clinical and laboratory) immediately before joining our study. These examinations revealed no pathological symptoms. They had not suffered from any disease during the 12 months, prior to the stydy, with the exception of the complaint described below. A detailed interview disclosed that ten women frequently (a minimum five times a year, for at least two years) suffered from recurrent facial Herpes simplex infection around the time of menstruation. The information about the time and frequency of this complaint was gathered from a questionnaire. The remaining eleven women were free from Herpes simplex infection. All volunteers signed an agreement for blood analysis. This work was approved by The Ethics Committee of The Medical University of Gdanesk. The levels of sex hormones: 17-beta estradiol and progesterone, as well as those of IL-2 and TNF-alpha were determined in all the women, on the 5th, 8th, 14h, 18th, and 25th day of the menstrual cycle. When the blood was taken for the analysis, the group of affected women was free from clinical symptoms.

Specimen collection and preparation

A fasting venous blood sample (5 ml) was collected aseptically, without any additives, between 8.00 and 9.00 a.m., on 5th, 8th, 14h, 18th and 25th day of the menstrual cycle. The blood was allowed to clot for 15 min at room temperature and was then transferred to the refrigerator (+ 4° C) for an additional 30-45 min. The serum was separated by centrifugation. The samples were stored at ­ 80° C, no longer than 30 days.

Coated-tube radioimmunoassay for progesterone

The radioimmunological assay was based on coated-tube technology ("Spectria ­ Progesterone¹²⁵/Orion Diagnostica). Fifty mul of standard (ready-to-use, progesterone samples in human serum containing: 0, 0.25, 1.25, 3.14, 12.56, 31.4 ng/ml respectively, with 0.1% NaN₃ and 0.1% Kathon), 50 mul of control human serum, which contained progesterone as a one of a number of analytes, and 50 mul of volunteer serum were added to the appropriate tubes. The tubes were already coated with a second antibody to which a primary rabbit polyclonal progesterone antibody had been bound. 500 mul of diluted ¹²⁵I-labelled progesterone were added to all coated tubes and accessory uncoated tubes (for total counts). All samples were quickly mixed and then incubated for 2 hours at room temperature. The labelled and unlabelled progesterone was allowed to compete for the limited number of high affinity binding sites of the antibody during this time. After incubation, each tube, except the uncoated tubes (for total counts) were decanted and washed with 1 ml of washing solution. Radioactivity were measured using a gamma counter for at least one minute, or until 10,000 counts per tube had been accumulated. A standard curve was produced by calculating the binding of the serum standards.

The sensitivity of the method, defined as the detectable concentration equivalent to three times the standard deviation of the zero-binding value, was < 0.094 ng/ml. Recoveries were in the range 95.1 ­ 103.7% with a mean value of 98.2% and SD 4.8%. The analysis of control sera in both the lower and upper portions of the normal range acted as the quality control for monitoring the performance of the procedure. The values obtained by the controls fell within the manufacturer's stated acceptable range.

Coated-tube radioimmunoassay for 17-beta-estradiol

The radioimmunological assay was based on coated-tube technology ('Spectria-Estradiol ¹²⁵/Orion Diagnostica). One hundred mul of standard (ready-to-use estradiol samples containing: 0, 13.62, 40.86, 136.2, 408.6, 4086.0 pg/ml respectively, with 0.1% NaN₃ and 0.1% Kathon as preservative), 100 mul of control human serum, containing estradiol as one of a number of analytes and 100 mul of volunteer serum were added to the appropriate tubes. The tubes were already coated with an anti-rabbit second antibody raised in goat, to which a primary polyclonal estradiol antibody raised in rabbit had been bound. 500 mul of diluted ¹²⁵I-labelled estradiol were added to all coated tubes and accessory, uncoated tubes (for total counts). All samples were quickly mixed on a vortex mixer and then covered with plastic film and incubated for 2 hours at 37° C in a water bath. The labelled and unlabelled estradiol was allowed to compete for the limited number of high affinity binding sites of the antibody during this time. After incubation, each tube, except uncoated tubes (tubes for total counts), were decanted and washed with 1 ml washing solution. Radioactivity was measured using a gamma counter for at least one minute or until 10 000 counts per tube had been accumulated. A standard curve was produced by calculating the binding of 6 serum standards. Then, estradiol concentrations of the unknowns were read from the standard curve.

The sensitivity of the method, defined as the detectable concentration equivalent to twice the standard deviation of the zero-binding value, was better than 5.45 pg/ml. Recoveries were in the range 85.6-108.9% with a mean value of 99.5%. The analysis of control sera in both the lower and upper portions of the normal range was the quality control for monitoring the performance of the procedure. The values obtained for the controls fell within the manufacture's stated acceptable range.

Bioassay for tumour necrosis factor (TNF-alpha) activity

TNF-alpha-sensitive WEHI 164 cells (T. Espevick, Institute of Cancer Research, University of Trodheim, Trodheim, Norway) were cultured for 24 hours on 96-well plastic plates (Corning, Science Products, Rochester NY, USA) at 37° C in an atmosphere containing 5% CO₂ and at a concentration of 20 x 10³/cell in RPMI medium (Gibco, BRL Life Technologies, Gaithersburg) containing 10% fetal calf serum (Gibco, BRL Life Technologies, Gaithersburg) 2 mM L-glutamine, gentamycin and actinomycin-D (Sigma Chemical Co., St. Louis, USA) (1.0 mug/ml). Ten mul of the sera were added, in triplicate, to each well of the plate. Cell viability was measured by colorimetric MTT. The optical densities obtained from experimental wells were fitted with titration standards of rTNF-alpha (Genzyme, Cambridge MA, USA). Neutralising rabbit anti-TNF-alpha antibody (Genzyme, Cambridge MA, USA) (1:10, 1:20, and 1:50) was added to the sera in order to confirm the specificity of the test. The parallel control samples received normal rabbit serum (1:50). The anti-TNF-alpha antibody completely blocked the proliferative effect of the sera on WEHI cells. This assay had a detection limit of 1.0 pg/ml. The intra-assay coefficient of variation ranged between 5% and 12%. The inter-assay coefficient of variation ranged between 15 and 23%.

Bioassay for interleukin-2 (IL-2)

IL-2-dependent CTLL cells (from Institute of Immunology, Wroclaw, Poland) were cultured for 48 hours on 96-well plastic plates (Corning, Science Products, Rochester NY, USA) at 37° C, in an atmosphere containing 5% CO₂ and at a concentration of 20 x 10³ well in RPMI medium (Gibco, BRL Life Technologies, Gaithersburg) containing 5% fetal calf serum (Gibco, BRL Life Technologies, Gaithersburg), 2mM L-glutamate and gentamycin. Ten mul of the sera were added, in triplicate, to each well of the plate. Cell viability was measured by colorimetric MTT. The optical densities obtained from experimental wells were fitted with titration standards for IL-2 (Genzyme, Cambridge MA, USA). Neutralising rabbit anti-IL2 antibody (Genzyme, Cambridge MA, USA) (1:10, 1:20, and 1:50) was added to the sera in order to confirm the specificity of the test. The parallel control samples received normal rabbit serum (1:50). The anti-IL-2 antibody completely blocked the proliferative effect of the sera on CTLL cells. This assay had a detection limit of 12.5 pg/ml. The intra-assay coefficient of variation ranged between 8.5 and 12.8%. The inter-assay coefficient of variation ranged between 16 and 25%.

Bioassay for interleukin-6 (IL-6)

Fifty mul of IL-6-dependent murine hybridoma cell line B9 (from dr L.Aarden, Netherland Red Cross, Amsterdam, Netherlands) cells were cultured for 48 hours on 96-well plastic plates (Corning, Science Products, Rochester NY, USA) at a concentration of 10 x 10³ well in IMDM medium containing 10% fetal calf serum (Gibco, BRL Life Technologies, Gaithersburg), 2mM L-glutamine and gentamycin. Ten mul of the supernatants were added in triplicate to each well. Cell viability was measured by colorimetric MTT assay identically to that described for TNF-alpha determination. The optical densities obtained from the experimental wells were fitted with titration standards of rIL6 (Genzyme, Cambridge MA, USA). Polyclonal anti-IL6 antibody (Genzyme, Cambridge MA, USA) was added (1:10, 1:20, 1:50) to the sera in order to confirm the specificity of the test. The parallel control samples received (1:50) normal rabbit serum. The anti-IL-6 antibody completely blocked the effect of the supernatants on the B9 cells. This assay had a detection limit of 10 pg/ml. The intra-assay coefficient of variation ranged between 9.5 and 10.9%. The inter-assay coefficient of variation ranged between 17.7 and 25%.

Colorimetric MTT assay

After incubation of the cells, 20 mul MTT [(3-4.5 dimethylthiazol-2-yl)-2.5 diphenyl-tetrazolium bromide] (Sigma Chemical Co., St. Louis, USA) was added to the 96-well plastic plates and incubated at 37° C in an atmosphere containing 5% CO₂ for 4 hours; then, 100 mul of isopropanol was added. Optical density was read at 570 nm on the automated plate reader (Multiscan MCC/340, Labsystems, Helsinki, Finland). The optical densities obtained from the experimental wells were fitted with titration standards of TNF or IL-2 (Genzyme, Cambridge MA, USA), respectively.

Statistics

All results were computed using the program: Statistica (version 5). Student's t-test was used to calculate the differences between the values for IL-2 and TNF-alpha in the two groups on the same days. The Mann-Whitney U test was used to assess the differences in the cytokine levels between different groups at the 18th and 25th day. The ANOVA test for repeated measures was used for checking the significance of the differences between the concentration of progesterone and estradiol as well as that of IL-2, IL-6 and TNF-alpha in two groups of women, during the menstrual cycle. P < 0.05 was considered as statistically significant.

RESULTS

Occurrence of the perimenstrual symptoms during the menstrual cycle

Twenty-one young women were divided into two groups. Eleven women did not experience perimenstrual symptoms. Ten young women suffered from the perimenstrual, facial symptoms associated with the Herpes simplex virus. In seven out of the ten volunteers, the symptoms started on the 18th while in the other three they appeared on the 25th day of the cycle. Usually the symptoms persisted until few days after menstruation. At the time of our examination, the group with recurrent infection was free of the clinical symptoms.

The hormonal status of the women without the symptoms

The comparison of the blood levels of 17beta-estradiol and progesterone on the 5, 8, 14, 18 and 25th day of the menstrual cycle did not reveal any difference between the concentrations of either hormones. The levels of both hormones changed similarly in both groups. They remained in the range of the reference values characteristic of the physiological menstrual cycle (Table 1).

Thus, the hormonal status of the examined women in the perimenstrual period does not seem to be related to the recurrence of symptoms around menstruation.

Interleukin-2 in the women with and without the symptoms

The comparison of the blood levels of IL-2 in women with and without symptoms is presented in Figure 1, Figure 1a and Figure 1b.

The inter-group analysis of the IL-2 values showed that the level of this cytokine in the women with symptoms was significantly lower on all days, as compared to that of women without symptoms (t-independent test). The significance of this difference was confirmed by the ANOVA test between groups for repeated measures (F = 51.63; p = 0.000001) (Table 2).

The intra-group analysis by means of the ANOVA test for repeated measures revealed a significant decrease in the IL-2 level in the second phase of menstrual cycle (p = 0.03) in the group of women with symptoms. At the same time, the women without the symptoms had levels of IL-2 which remained unchanged during the course of the menstrual period (p = 0.27) (Table 3).

TNF-alpha in the women with and without the symptoms

The comparison of the blood levels of TNF-alpha in the women with and without symptoms is presented in Figure 2, Figures 2a and 2b.

The inter-group analysis of the values for TNF-alpha showed that the level of TNF-alpha in the women with symptoms was significantly higher, on the 18th day of the cycle, than that of women without symptoms (t-independent test).

The intra-group analysis by means of the ANOVA test for repeated measures revealed a significant increase of the TNF-alpha levels in the luteal phase of the menstrual period of women without symptoms (p = 0.013) as well as those with symptoms (p = 0.028) (Table 3).

Interleukin-6 in the women with and without symptoms

The comparison of the blood level of IL-6 in women with and without the symptoms is presented in Figure 3.

The inter-group analysis by the between groups ANOVA test for repeated measures revealed that the values for IL-6 in the women with symptoms were significantly higher than those in the women without symptoms (F = 1.48; p = 0.027) (Table 2).

The level of IL-6 did not fluctuate significantly during the menstrual cycle in either group. (The ANOVA test for repeated measures; F = 1.27, p = 0.299 for the women with symptoms and F = 0.77; p = 0.55 for the women without symptoms) (Table 3).

Perimenstrual infections are associated with lower IL-2 and higher TNF-alpha levels

The results illustrating the differences in the levels of IL-2 and TNF-alpha in women with and without symptoms on the 18th and 25th day of the cycle are presented in Figure 4 and Figure 5.

Analysis of the data by the Mann-Whitney U test revealed that the women suffering from perimenstrual infections were characterised by a significantly lower level of IL-2 on 18th (p = 0.006) and 25th (p = 0.02) day of the cycle as compared to those of women without symptoms. Simultaneously, the women with symptoms had a significantly higher level of TNF-alpha on 18th day (p = 0.0002) compared to that of the women without symptoms.

These results suggest that the perimenstrual facial Herpes simplex infection was associated with lower levels of IL-2 on the 18th and the 25th days and a higher level of TNF-alpha on the 18th day of the menstrual cycle.

DISCUSSION

The aim of this work was to look at a possible relationship between recurrent perimenstrual facial Herpes simplex infection and the serum concentrations of interleukin-2 and tumor necrosis factor alpha in young, otherwise healthy women, who were free of the clinical symptoms at the time of our examination. The results of the hormonal analysis showed that the levels of 17beta-estradiol and progesterone were similar during the menstrual period in the women with and without perimenstrual Herpes infection symptoms. The detailed analysis of the levels of IL-2, TNF-alpha and IL-6 during the menstrual cycle revealed that the women suffering from perimenstrual infections were characterised by a significantly lower level of IL-2 on the 18th and 25th day of the cycle as compared to those without symptoms. The affected women also had a significantly higher level of TNF-alpha on the 18th day, as compared to those without symptoms. Additionally, these women also had a lower level of IL-2 and a higher level of IL-6 during the whole menstrual period as compared to women without the recurrent infections.

The time around menstruation has been recognised as conferring a susceptibility to several infectious diseases [1-4]. The recurrent Herpes simplex type I infection is one of the most common perimenstrual dermatoses [5]. We can speculate that the lower level of IL-2 in the group with symptoms (although symptom free at the time of sampling), may be due to the prolonged repeated viral stimulation leading to a subclinically low-grade immune response during the symptom free time. Such a phenomenon has been documented in asymptomatic, seronegative HIV-infected patients [13, 14]. Moreover, it has been reported that while the primary Herpes virus infection promotes IL-2 production, the secondary one suppresses it [15]. Thus, after a long-term stimulation, the IL-2 level may be expected to be lower.

TNF-alpha has an essential homeostatic role in limiting the extent and duration of an inflammatory process [12]. A promotion of inflammatory reactions caused by TNF-alpha has already been noticed in Herpes infections [16, 17]. The only significant difference in the TNF-alpha level between the women with and without symptoms was, however, found on the 18th day of the cycle. Therefore, it would be difficult to suggest that the higher susceptibility to Herpes infection may be due to the TNF gene polymorphism that is associated with a distinct secretion of this cytokine [18]. In that case, one would expect that the level of TNF-alpha in the affected women would be higher throughout the entire menstrual cycle.

It is possible to suggest that the initiation of the viral infection during the menstrual cycle might be due to the low level of IL-2. The immune deficiency state associated with a low IL-2 level has been found as a phenomenon underlying impaired anti-tumor [19] and anti-viral [10, 15] immunity. The correlation between the elevated blood levels of TNF-alpha and IL-6 and the occurrence of the symptoms, in the affected women, may be the result of an active, though subclinical Herpes infection. In the course of viral infections, high stimulation of TNF-alpha production, coinciding with a peak in clinical symptoms, has been noted [11]. There is, however, no way of excluding another possible source of TNF-alpha and IL-6 in the blood of these young women. Brannstrom et al. [20] have found that the serum concentration of TNF-alpha increases in the mid- and late luteal phase of the cycle. They also documented the observation that cells from the late luteal phase-corpus luteum, taken from the women during abdominal surgery, produced, in in vitro culture higher levels of TNF-alpha as compared to cells from an early luteal phase. These authors interpreted their results as indicating the ovarian origin of the blood TNF-alpha. Thus, most probably, the final level of TNF-alpha seen in women with recurrent symptoms, may be a sum of that produced by the ovarian follicles [21, 22] and by the immune cells participating in the antiviral defence.

The group of unaffected women was characterised, by a similar tendency as that found in affected women, towards an increased concentration of TNF-alpha before menstruation. This increase was however, smaller than that seen in the affected ones. These results are in line with the finding of the Brannstrom et al. [20] document that an increase in TNF-alpha levels before menstruation is a real event.

The phenomenon of perimenstrual immunological suppression has been observed both in animals and women, but its mechanism was unclear. Female rats with breast cancer, operated at the end of the estrous phase, were characterised by a higher risk of metastases [23, 24]. A worse prognosis in operable breast cancer has also been observed in women undergoing surgery just before menstruation [25, 26]. Since in these papers IL-2 and TNF-alpha were not analysed, we cannot exclude that both cytokines contributed to the immune suppression described. The naturally occurring peak of TNF-alpha levels before menstruation, for example, may be an important factor, ­ as this cytokine exerts not only pro-inflammatory activity but also suppresses T-cell-dependent immune reactions [27]. There may also exist other reasons for the perimenstrual immunosuppression. The perimenstrual interval in healthy women is also characterised by a shift in the balance between the IFN-gamma and IL-10 blood concentrations [28]. The IL-10 concentration significantly increases during the perimenstrual time with a concomitant decrease in INF-gamma. It is possible that IL-10, as Th-2 cytokine, potentiates the IL-2 deficiency at that point of the menstrual cycle.

CONCLUSION

REFERENCES

2. Kalo-Klein A, Witkin S S. 1989. Candida albicans cellular immune system interactions during different stages of the menstrual cycle. Am. J. Obstet. Gynecol. 161: 1132.

3. Crowley T, Horner P, Hughes A, Berry J, Paul I, Caul O. 1997. Hormonal factors and the laboratory detection of Chlamydia trachomatis in women' implications for screening? Int. J. STD. AIDS 8: 25.

4. Sodora D L, Gettie A, Miller C J, Marx P A. 1998. Vaginal transmission of SIV assessing infectivity and hormonal influences in macaques inoculated with cell-free and cell-associated viral stocks. AIDS Res. Hum. Retroviruses. 14 Suppl. 1: S119.

5. Pierard-Franchimot C, Fraiture A L, Delvoye P, Dehevay J, Deleixhe-Mauhin F, Van Cauvenberge J R, Pierard G E. 1999. Les dermatoses catamenides: une facette fréquente de la chronobiologie. Rev. Med. Liege 54: 318.

6. Lopez C, Kirkpatrick D, Read S E, Fitzgerald P A, Pitt J, Pahwa S, Ching C Y, Smithwick E M. 1983. Correlation between low natural killing of fibroblasts infected with herpes simplex virus type 1 and susceptibility to herpes virus infections. J. Infect. Dis. 147: 1030.

7. Rinaldo C R Jr, Torpey D J. 3rd 1993. Cell-mediated immunity and immunosuppression in herpes simplex virus infection. (Review) Immunodeficiency 5: 33.

8. Kwiatkowski B, Artrip J H, Michler R E, Itescu S. 1997. Human NK cells demonstrate increased adherence and lysis of porcine endothelium following activation with IL-2. Ann. Transplant. 2: 21.

9. Kuo Y C, Lin C Y, Cheng S F, Lin C C, Liu W T. 1987. Impaired natural killer cytotoxicity during recrudescence of recurrent herpes simplex virus type 1 infection. Cancer Detect. Prev. Suppl 1: 51.

10. Kuo Y C, Lin C Y. 1990. Recurrent herpes simplex virus type 1 infection precipitated by the impaired production of interleukin-2, alpha-interferon, and cell-mediated cytotoxicity. J. Med. Virol. 31: 183.

11. van Reeth K, Labarque G, Nauwynck H, Pensaert M. 1999. Differential production of proinflammatory cytokines in the pig lung during different respiratory virus infections: correlations with pathogenicity. Res. Vet. Sci. 67: 47.

12. Marino M W, Dunn A, Grail D, Inglese M, Noguchi Y, Richards E, Jungbluth A, Wada H, Moore M, Williamson B, Basu S, Old L J. 1997. Characterization of tumor necrosis factor-deficient mice. Proc. Natl Acad. Sci. USA 94: 8093.

13. Nicastri E, Sarmati L, Ercoli L, Mancino G, d'Ambrosio E, d'Ettorre G, Mastroianni C M, Vullo V, Andreoni M. 1999. Reduction of IFN-gamma and IL-2 production by peripheral lymphocytes of HIV-exposed seronegative subjects. AIDS 13: 1333.

14. Zerhouni B, Sanhadji K, Kehrli L, Livrozet J M, Touraine J L. 1997. Interleukin (IL) 2 deficiency aggravates the defect of natural killer cell activity in AIDS patients. Thymus 24: 147.

15. Ongradi J, Ahmad A, Menezes J. 1999. In vitro cytokine induction in leukicytes by human Herpes virus 7. Orv. Hetil. 140: 1935.

16. Liu T, Tang Q, Hendricks R L. 1996. Inflammatory infiltration of the trigeminal ganglion after Herpes simplex virus type 1 corneal infection. J. Virol. 70: 264.

17. Sprecher E, Becker Y. 1992. Detection of IL-1 beta, TNF-alpha, and IL-6 gene transcription by the polymerase chain reaction in keratinocytes, Langerhans cells and peritoneal exudate cells during infection with Herpes simplex virus-1. Arch. Virol. 126: 253.

18. Knight J C, Kwiatkowski D. 1999. Inherited variability of tumor necrosis factor production and susceptibility to infectious disease. (Review) Proc. Assoc. Am. Physicians.111: 290.

19. Gough D B, Winstanley F P, Fearon K C, Carter D C. 1992. Regulation of tumor necrosis factor production in healthy humans and in patients with cancer. Arch. Surg. 127: 713.

20. Brannstrom M, Friden B E, Jasper M, Norman R J. 1999. Variations in peripheral blood levels of immunoreactive tumor necrosis factor-alpha (TNF-alpha) throughout the menstrual cycle and secretion of TNF-alpha from the human corpus luteum. Eur. J. Obstet. Gynecol. Reprod. Biol. 83: 213.

21. Murdoch W J, Colgin D C, Ellis J A. 1997. Role of tumor necrosis factor-alpha in the ovulatory mechanism of ewes. J. Anim. Sci. 75: 1601.

22. Brannstrom M, Norman R J. 1993. Involvement of leukocytes and cytokines in the ovulatory process and corpus luteum function. (Review) Hum. Reprod. 8: 1762.

23. Hrushesky W J, Gruber S A, Sothern R B, Hoffman R A, Lakatua D, Carlson A, Cerra F, Simmons R L. 1988. Natural killer cell activity age, estrous and circadian-stage dependence and inverse correlation with metastatic potential. J. Nat. Cancer Inst. 80: 1232.

24. Ratajczak H V, Sothern R B, Hrushesky W J. 1988. Estrous influence on surgical cure of mouse breast cancer. J. Exp. Med. 168: 73.

25. Nathan B, Bates T, Anbazhagan R, Norman A R. 1993. Timing of surgery for breast cancer in relation to the menstrual cycle and survival of premenopausal women. Br. J. Surg. 80: 43.

26. Badwe R A, Richard M A, Fentiman I S, Gregory W, Saad Z, Chaudary M A, Bentley A, Rubens R D. 1991. Surgical procedures, menstrual cycles phases and prognosis in operable breast cancer . (Letter) Lancet 338: 815.

27. Cope AP. 1999. Regulation of autoimmunity by proinflammatory cytokines. (Review) Curr. Opin. Immunol. 10: 669.

28. Agarwal S K, Marshall G D Jr. 1999. Perimenstrual alterations in type-1/type-2 cytokine balance of normal women. Ann. Allergy Asthma Immunol. 83:222.