- Email: [email protected]
Serum vascular endothelial growth factor levels are poorly predictive of subsequent ovarian hyperstimulation syndrome in highly responsive women undergoing assisted conception
FERTILITY AND STERILITY威 VOL. 78, NO. 6, DECEMBER 2002
CONTROVERSY: VASCULAR ENDOTHELIAL GROWTH FACTOR
Copyright ©2002 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A.
Serum vascular endothelial growth factor levels are poorly predictive of subsequent ovarian hyperstimulation syndrome in highly responsive women undergoing assisted conception Raj Mathur, M.D.,a Grant Hayman, M.D.,b Amolak Bansal, M.D.,b and Julian Jenkins, D.M.a Bristol University Centre for Reproductive Medicine, Bristol, and St. Helier’s Hospital, Carshalton, United Kingdom
Objective: To determine whether serum vascular endothelial growth factor (VEGF) levels can distinguish highly responsive women who subsequently develop ovarian hyperstimulation syndrome (OHSS) from women with a similar ovarian response who do not. Design: Prospective controlled study. Setting: University IVF unit. Patient(s): Women undergoing IVF who met predetermined risk criteria for OHSS. Patients developing OHSS were compared with patients who did not develop OHSS. Intervention(s): Long-protocol pituitary down-regulation followed by FSH stimulation by a standard protocol without coasting. A maximum of three embryos was transferred. Vaginal progesterone was used for luteal support. Main Outcome Measure(s): Occurrence of OHSS; serum VEGF concentrations on the day of embryo transfer (ET) and at 5 and 10 days after ET. Result(s): Serum VEGF levels at any time point did not differ significantly between 9 OHSS cases and 9 controls. Vascular endothelial growth factor levels in samples collected from cases before the onset of OHSS were higher than levels in time-matched samples from controls (medians, 177.6 [range, 64.02–549.1] pg/mL vs. 100.7 [range, 37.59 –246] pg/mL, respectively), but the difference failed to reach statistical significance (P⫽.0518), and there was considerable overlap between cases and controls. Conclusion(s): Serum VEGF levels in the luteal phase do not distinguish between high-risk women who subsequently develop OHSS and women with a similar risk profile who do not develop OHSS. (Fertil Steril威 2002;78:1154 – 8. ©2002 by American Society for Reproductive Medicine.) Key Words: OHSS, VEGF, IVF, prediction, complications Received May 29, 2002; revised and accepted July 5, 2002. Reprint requests: Raj Mathur, M.D., Reproductive Medicine Unit, Rosie Hospital, Cambridge, CB2 2SW, United Kingdom (FAX: 00-44-1223-215327; E-mail: rajmathur@ doctors.org.uk). a Bristol University Centre for Reproductive Medicine. b Department of Immunology, St. Helier’s Hospital. 0015-0282/02/$22.00 PII S0015-0282(02)04243-7
1154
As subfertility treatment moves into the 21st century, OHSS remains a significant iatrogenic complication of ovarian stimulatory treatments, causing considerable morbidity and even mortality (1). Attempts to reduce the incidence of OHSS have foundered on the lack of certainty regarding its pathogenesis and the difficulties involved in the clinical prediction of significant OHSS (2). Evidence suggests that the pathogenesis of OHSS may involve the immune system and a complex network of interacting cytokines and
growth factors (3). Of these, VEGF has been implicated as having a major role in the development of OHSS. Follicular and peritoneal fluid from patients who responded to gonadotropin stimulation with ⱖ25 follicles was found to increase permeability across endothelial cell monolayers, whereas follicular fluid from natural cycles had no effect (4). Follicular fluid was shown to contain VEGF by Western blot (5). The vascular permeability– enhancing effect of ascitic fluid from patients of severe OHSS was largely blocked by antiserum directed against VEGF (6).
Cultured granulosa cells are known to produce VEGF, and this production is increased in a time- and dose-dependent fashion by human chorionic gonadotropin (7). In theory, the actions of VEGF on vascular endothelium could account for the pathophysiological changes seen in OHSS. These actions include increased capillary permeability and neovascularization (8) and up-regulation of procoagulant factors (9). Serum VEGF concentrations during the course of OHSS have been reported to reflect the clinical severity of the disease (10). It has been suggested (11) that measurement of serum VEGF concentrations may have a role in the clinical prediction of OHSS. This study was performed to investigate whether serum VEGF concentrations before the onset of OHSS can be used to distinguish high-risk women who develop OHSS from women who do not develop OHSS despite a similar response to ovarian stimulation. VEGF concentrations in women undergoing assisted conception may vary with ovarian response, patient age, and the presence or absence of polycystic ovaries (12, 13). Hence, unless efforts are made to control for these factors, they may act as potential confounders when VEGF levels are compared between populations of women with and without OHSS. To avoid this, this study recruited women who satisfied prospectively established criteria putting them at high risk of OHSS. The aim was to achieve a control group that was similar to the OHSS group in all relevant respects except the occurrence of OHSS. A previous report in the literature examined VEGF levels in highly responsive women on the days of hCG administration, oocyte retrieval (OR), and embryo transfer (ET) (14). However, clinically severe OHSS often does not present till several days after ET (2), and VEGF levels remote from onset may not reflect the likelihood of developing late-onset OHSS. Hence, we aimed to investigate VEGF levels at later points in the luteal phase as well.
MATERIALS AND METHODS This was a prospective observational study carried out on patients undergoing in vitro fertilization at the Bristol University Centre for Reproductive Medicine between 1 January 1998 and 31 March 1999. Pituitary down-regulation and ovarian stimulation were carried out by a standardized regime that has been described elsewhere (2). Coasting was not carried out, and all patients had embryos transferred in the treatment cycle. A maximum of three embryos were transferred. In all cases, luteal support was by progesterone suppositories (400 mg twice per day) for 14 days, starting on the day of ET. Prior approval for the study was obtained from the institutional ethics committee. Individual patients gave informed, written consent for their participation in the study. Patients were approached for recruitment if they were thought to be at high risk for the development of OHSS on the basis of any of the following criteria: FERTILITY & STERILITY威
1. History of moderate or severe OHSS 2. Polycystic ovaries (PCO), diagnosed by ultrasonographic criteria (15) 3. Presence of ⱖ30 follicles in total, or ⱖ10 follicles of ⱖ13 mm in diameter on the final ultrasound scan before administering hCG for follicular maturation 4. Peak estradiol level of ⱖ10,000 pmol/L
Patients were excluded from the study if they suffered from significant systemic disease, (e.g., diabetes, thyroid disease) or were receiving corticosteroid or antiinflammatory medication. Serum samples were collected from subjects at the following times: at the time of OR, immediately after ET, between 9 AM and 12 PM 5 days after ET, and between 9 AM and 12 PM 10 days after ET. Ten mililiters of blood was collected and immediately centrifuged at 600 RPM for 6 minutes. Serum was pipetted off and stored in sterile, nonpyrogenic polypropylene microtubes (Sarstedt Limited, Leicester, United Kingdom) at ⫺80°C. Patients were assessed to determine the presence and severity of OHSS on the day of OR, the day of ET, 5 days after ET, and 10 days after ET. In addition, patients reporting symptoms of OHSS at other times were seen within 24 hours for assessment, in most instances by the same person (R.M.). OHSS severity was graded according to the criteria of Golan et al. (16). Further management including hospital admission was arranged according to the clinical picture in the individual case. One patient was admitted with OHSS in another hospital before she could be assessed; however, sufficient information was available from the admitting unit to make an accurate diagnosis of severity. Serum VEGF levels in samples from subjects who developed OHSS (cases) were compared with levels in samples from subjects who did not develop OHSS (controls). One control was matched for each case from the subjects recruited. Controls were patients treated closest in time to the OHSS case for whom all the same serum samples were available as the case. Not all planned samples were available for all cases of OHSS: day-of-OR samples were available for eight cases, day-of-ET samples for five cases, and ET ⫹5 and ET ⫹10 samples, for seven cases each. The same samples were assayed for each control as the corresponding case. Serum samples from patients and controls were assayed in a single batch for VEGF 165 using an ELISA based on the sandwich method of R&D Systems (Abingdon, Oxon, United Kingdom). No cross-reactivity has been found between VEGF and other recombinant cytokines and growth factors using this assay. The immobilized anti-human VEGF antibody was monoclonal, and the second antibody, conjugated with horseradish peroxidase, was polyclonal. All samples were run in duplicate and standards were included in the assay. Serial dilutions were tested to confirm assay linearity. 1155
TABLE 1 Characteristics of patients developing OHSS. No. 1 2 3 4 5 6 7 8 9
Age (y)
PCO
FSH (no. of Ampoules)a
Follicles (n)
Peak E2 (pmol/L)
Oocytes (n)
Outcome (no. of Sacs)
Onset (d after OR)
Severity
34 33 28 29 39 34 34 36 36
N N Y N Y N N N Y
22 22 26 16.5 52 24 22 42 24
24 24 55 24 43 35 39 15 32
NM NM 13,440 NM 33,712 NM 13,000 14,563 9194
13 23 14 14 21 19 26 7 17
NP P (1) NP NP NP NP P (3) P (1) P (1)
3 4 3 5 2 3 11 10 10
Moderate Moderate Moderate Moderate Severe Moderate Moderate Severe Moderate
Note: Y ⫽ yes; N ⫽ no; NM ⫽ not measured; NP ⫽ not pregnant; P ⫽ pregnant. Total ampoules, each containing 75 IU of FSH.
a
Mathur. Serum VEGF is poorly predictive of OHSS. Fertil Steril 2002.
The intra-assay and interassay coefficients of variation were 9% and 12%, respectively. The assay range was from 31.25– 2000 pg/mL. Absorbance was read at 405 nm using a DPC plate reader (Diagnostics Products Corporation, Los Angeles, CA). A standard curve was generated by plotting the optical density of the standards against the concentration of the standards. The VEGF concentration of the samples was determined by comparing their optical densities to the standard curve. The statistical software employed for calculations and graphs was Arcus Biomedical (version 6.80, Medical Computing, Cambridge, United Kingdom). Patient characteristics and VEGF concentrations were compared between OHSS cases and controls using a nonparametric method (MannWhitney U test).
RESULTS Fifty-four women were recruited, of whom 9 (16.6%) developed OHSS. The patient details, treatment cycle characteristics, and the times of onset and severity of OHSS are shown in Table 1. There was no significant difference between women who developed OHSS and controls in terms of patient age, body mass index, infertility diagnosis, incidence of PCO, duration, and type of infertility (primary vs. secondary; Table 2). OHSS cases and controls did not differ significantly in regard to treatment cycle characteristics and outcomes (Table 3 ). Peak estradiol concentrations were measured in five cases of OHSS and in three controls and did not differ significantly between the two groups (13,440 pmol/L [range, 9194–33,712] vs. 7289 pmol/L [range, 2869–16,268]), respectively.
Serum VEGF Concentrations in Cases and Controls Serum VEGF concentrations on the days of OR and ET, 5 days after ET, and 10 days after ET did not differ significantly between OHSS cases and controls (Table 4 ). 1156 Mathur et al.
Serum VEGF is poorly predictive of OHSS
Serum VEGF Concentrations Before Onset of OHSS Fifteen serum samples were available before the onset of OHSS from the nine cases. Vascular endothelial growth factor concentrations in these overlapped considerably with concentrations in an equal number of samples from controls matched for time of collection (Fig. 1 ). Samples from OHSS cases showed higher VEGF concentrations before the onset of OHSS than did matched control samples (median, 177.6 [range, 64.02–549.1] pg/mL vs. 100.7 [range, 37.59 –246] pg/mL, respectively), but the difference failed to reach statistical significance (P⫽.0518 by Mann-Whitney U test).
DISCUSSION The present study set out to investigate whether lutealphase serum VEGF concentrations can distinguish between
TABLE 2 Patient characteristics of OHSS cases and controls. Characteristic Age in y, median Infertility diagnosis, n (%)a Sperm dysfunction Tubal Endometriosis Ovulatory Unexplained PCO, n (%) Primary infertility, n (%) Duration of infertility in y, median (range) Body mass index, median (range)
OHSS cases (n ⫽ 9)
Controls (n ⫽ 9)
33 (28–39)
33 (29–40)
2 (21.2) 1 (11.1) 2 (22.2) 3 (33.3) 1 (11.1) 3 (33.3) 5 (55.5) 3 (1–7) 23 (19–25)
3 (33.3) 3 (33.3) 1 (11.1) 1 (11.1) 2 (22.2) 1 (11.1) 5 (55.5) 4 (3–8) 22 (21–31)
Note: None of the comparisons were statistically significant. a Some patients had more than one infertility diagnosis, hence, total may exceed 100%. Mathur. Serum VEGF is poorly predictive of OHSS. Fertil Steril 2002.
Vol. 78, No. 6, December 2002
TABLE 3
FIGURE 1
Treatment cycle characteristics of OHSS cases and controls. Characteristics
OHSS cases (n ⫽ 9)
Controls (n ⫽ 9)
No. of ampoules of FSH No. of stimulation days No. of follicles ⱖ12 mm Follicles ⬎12 mm Oocytes Embryos transferred Pregnancies, n (%) Multiple pregnancies, n (%) Implantation rate (%)
23 (16.5–52) 11 (11–14) 10 (2–21) 17 (11–35) 15 (7–26) 3 (2–3) 4 (44.4) 1 (11.1) 22.7
26 (22–46) 13 (11–17) 10 (2–12) 16 (12–33) 15 (7–31) 3 (2–3) 5 (55.5) 2 (22.2) 31.8
Vascular endothelial growth factor concentrations in serum samples collected from OHSS cases before the onset of OHSS and in time-matched samples from controls.
Note: All data are median (range) unless otherwise specified. No comparisons were statistically significant. Mathur. Serum VEGF is poorly predictive of OHSS. Fertil Steril 2002.
women who develop OHSS and controls, once the confounding factors of patient age, ovarian response, and treatment cycle outcome are excluded. The study aimed to compare women with similar characteristics and similar magnitudes of ovarian response to stimulation who had similar treatment cycle outcomes, differing only in that some of them developed OHSS whereas others did not. As is apparent from Tables 2 and 3, this aim was achieved; the case and control groups were similar in respect to infertility characteristics, ovarian response, and cycle outcome. Other investigators have compared women with “excessive” ovarian responses with those with more modest responses (11, 17). In such a comparison, it may be difficult to decide whether any detected difference in VEGF activity relates to the magnitude of ovarian response or to the occurrence of OHSS. Agrawal et al. (11) observed higher VEGF concentrations throughout the IVF cycle in 20 women who developed OHSS, compared with 87 women who did not develop OHSS. However, the OHSS group differed significantly from the control group in their ovarian response, with significantly higher peak E2 concentrations and significantly
TABLE 4 Serum VEGF (pg/mL) at different points of the treatment cycle in OHSS cases and controls. Day of cycle OR ET ET ⫹ 5 ET ⫹ 10
OHSS cases
Controls
126.7 (64.02–549.1) 191 (92.75–449.2) 128.3 (72.87–376.8) 135.8 (107.8–265.6)
115.65 (37.59–252.8) 91.55 (64.02–124.3) 120.75 (64.02–206.9) 178.4 (49.33–291.7)
Note: Values are expressed as median (range. No comparisons were statistically significant. Mathur. Serum VEGF is poorly predictive of OHSS. Fertil Steril 2002.
FERTILITY & STERILITY威
Mathur. Serum VEGF is poorly predictive of OHSS. Fertil Steril 2002.
greater number of follicles and oocytes retrieved in the OHSS group. In addition, the prevalence of PCO was significantly higher in the OHSS group (85% vs. 29.8%). The observed difference in serum VEGF concentrations may relate to these differences, rather than to the occurrence of OHSS per se. Serum VEGF levels are higher in women with PCO than in women with ultrasonically normal ovaries (12). Artini et al. (17) observed higher plasma VEGF concentrations on the day of OR in 15 women who were considered to be at high risk of developing OHSS compared with women whose ovarian responses were more modest. Another study (18) found significantly higher free-serum VEGF concentrations on the days of hCG administration and ET, but not the day of OR, in 10 women who developed OHSS compared with 15 controls. It is possible that at least part of the difference observed in serum VEGF concentrations between the two groups in that study may be accounted for by the significantly higher prevalence of PCO in the OHSS group compared with in the control group (30% vs. 6.7%). In contrast, the findings of the present study are in agreement with other studies comparing women with similar ovarian responses. A pilot study (19) observed no difference in serum VEGF concentrations between four women who 1157
developed OHSS and four controls with a similar magnitude of ovarian response, until the actual onset of OHSS. No significant difference was found in serum VEGF concentrations between 12 women developing OHSS and 12 high-risk controls on the days of hCG administration, OR, or ET (14). Similar findings were noted in the present study, which also examined serum VEGF concentrations at later points in the luteal phase of the cycle (5 and 10 days after ET). This was considered important because the onset of clinically significant OHSS is often delayed till at least a week after ET (2). Serum VEGF concentrations have been suggested as having a possible role in the clinical prediction of OHSS (11). The present study found no significant difference in VEGF concentrations between samples collected from cases before the onset of OHSS and samples collected from controls at the same time in the treatment cycle. Although concentrations were higher in women who subsequently developed OHSS, the difference did not reach statistical significance. From the overlap between both sets of values (Fig. 1) it may be surmised that although it is possible that a bigger sample may show a statistically significant difference, the clinical utility of this in predicting the risk of OHSS is likely to be limited. A rise in total VEGF concentration between the day of hCG administration and OR was found in one study to have a positive predictive value of 40% in determining the risk of OHSS (11). Because the present study did not measure VEGF concentrations before OR, it cannot provide data regarding the usefulness of this parameter. However, a rise in free or total VEGF between the days of hCG and OR or ET was not found in another study in either the OHSS group or controls (18). It is possible that simple measurement of VEGF in serum is not an adequate reflection of the true VEGF status of the individual. Vascular endothelial growth factor release by platelets at clotting may give falsely high results, whereas soluble VEGF receptors and ␣2-macroglobulin may neutralize some circulating VEGF, making it difficult to interpret the significance of VEGF concentrations on their own (20). Awareness of these limitations strengthens further our conclusion of a lack of value, in present-day clinical practice, for serum VEGF measurement in predicting the risk of OHSS in highly responsive women. However, it cannot be ruled out that a more sophisticated assessment of VEGF status may in future turn out to be useful for prediction of OHSS. A limitation of the current study is the relatively small number of patients, which is in keeping with other studies of this nature on OHSS, a relatively uncommon condition that is difficult to predict. In addition, not all serum samples were available for all cases, as in some instances, patients were unable to attend at the correct time for the blood sample to be taken. This highlights the problems of performing clinical research in this area.
1158 Mathur et al.
Serum VEGF is poorly predictive of OHSS
In conclusion, this study did not show significant differences in serum VEGF levels between high-risk women who developed OHSS and controls with similar risk profiles. Serum VEGF concentrations before the onset of OHSS are unlikely to be clinically useful in distinguishing highly responsive women who subsequently develop OHSS from women with a similar ovarian response who do not develop OHSS. References 1. Semba S, Moriya T, Youssef EM, Sasano H. An autopsy case of ovarian hyperstimulation syndrome with massive pulmonary edema and pleural effusion. Pathol Int 2000;50:549 –52. 2. Mathur RS, Akande AV, Keay SD, Hunt LP, Jenkins JM. Distinction between early and late ovarian hyperstimulation syndrome. Fertil Steril 2000;73:901–12. 3. Elchalal U, Schenker JG. The pathophysiology of ovarian hyperstimulation syndrome—views and ideas. Hum Reprod 1997;12:1129 –37. 4. Goldsman MP, Pedram A, Dominguez CE, Ciuffardi I, Levin E, Asch RH. Increased capillary permeability induced by human follicular fluid: a hypothesis for an ovarian origin of the hyperstimulation syndrome. Fertil Steril 1995;63:268 –72. 5. Amato P, Zwain I, Yen SSC. Identification of vascular endothelial growth factor in human follicular fluid: a possible role in ovarian hyperstimulation syndrome. Proceedings of the 52nd Annual Meeting of the American Society of Reproductive Medicine; Birmingham, Alabama. Boston:1996. 6. McClure N, Healy DL, Rogers PA, Sullivan J, Beaton L, Haning RV, Jr, et al. Vascular endothelial growth factor as capillary permeability agent in ovarian hyperstimulation syndrome. Lancet 1994;344:235–6. 7. Neulen J, Yan Z, Raczek S, Weindel K, Keck C, Weich HA, et al. Human chorionic gonadotropin-dependent expression of vascular endothelial growth factor/vascular permeability factor in human granulosa cells: importance in ovarian hyperstimulation syndrome. J Clin Endocrinol Metab 1995;80:1967–71. 8. Ferrara N, Houck K, Jakeman L, Leung DW. Molecular and biological properties of the vascular endothelial growth factor family of proteins. Endocr Rev 1992;13:18 –32. 9. Brock TA, Dvorak H, Senger DR. Tumour secreted vascular permeability factor increases cytosolic Ca2⫹ and von Willebrand factor release in human endothelial cells. Am J Pathol 1991;138:213–21. 10. Abramov Y, Barak V, Nisman B, Schenker JG. Vascular endothelial growth factor plasma levels correlate to the clinical picture in severe ovarian hyperstimulation syndrome. Fertil Steril 1997;67:261–5. 11. Agrawal R, Tan SL, Wild S, Sladkevicius P, Engmann L, Payne N, et al. Serum vascular endothelial growth factor concentrations in in vitro fertilization cycles predict the risk of ovarian hyperstimulation syndrome. Fertil Steril 1999;71:287–93. 12. Agrawal R, Sladkevicius P, Engmann L, Conway GS, Payne NN, Bekis J, et al. Serum vascular endothelial growth factor and ovarian stromal blood flow are increased in women with polycystic ovaries. Hum Reprod 1998;13:651–5. 13. Friedman CL, Seifer DB, Kennard EA, Arbogast L, Alak B, Danforth DR. Elevated level of follicular fluid vascular endothelial growth factor is a marker of diminished pregnancy potential. Fertil Steril 2000;70: 836 –9. 14. Chen CD, Chen HF, Lu HF, Ho HN, Yang YS. Value of serum and follicular fluid cytokine profile in the prediction of moderate to severe ovarian hyperstimulation syndrome. Hum Reprod 2000;15:1037–42. 15. Polson DW, Adams J, Wadsworth J, Franks S. Polycystic ovaries: a common finding in normal women. Lancet 1988;1:870 –2. 16. Golan A, Ron-El R, Herman A, Soffer Y, Weinraub Z, Caspi E. Ovarian hyperstimulation syndrome: an update review. Obstet Gynecol Surv 1989;44:430 –41. 17. Artini PG, Fasciani A, Monti M, Luisi S, D’Ambrogio G, Gennaziani AR. Changes in serum vascular endothelial growth factor levels and the risk of ovarian hyperstimulation syndrome in women enrolled in an in vitro fertilization programme. Fertil Steril 1998;70:560 –4. 18. Ludwig M, Jelkman W, Bauer O, Diedrich K. Prediction of severe ovarian hyperstimulation syndrome by free serum vascular endothelial growth factor concentration on the day of human chorionic gonadotrophin administration. Hum Reprod 1999;14:2437–41. 19. Krasnow JS, Berga SL, Guzick DS, Zeleznik AJ, Yeo KT. Vascular permeability factor and vascular endothelial growth factor in ovarian hyperstimulation syndrome: a preliminary report. Fertil Steril 1996;65: 552–5. 20. Wolfgang J. Pitfalls in the measurement of circulating vascular endothelial growth factor. Clin Chem 2001;47:617–23.
Vol. 78, No. 6, December 2002
CONTROVERSY: VASCULAR ENDOTHELIAL GROWTH FACTOR
Copyright ©2002 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A.
Serum vascular endothelial growth factor levels are poorly predictive of subsequent ovarian hyperstimulation syndrome in highly responsive women undergoing assisted conception Raj Mathur, M.D.,a Grant Hayman, M.D.,b Amolak Bansal, M.D.,b and Julian Jenkins, D.M.a Bristol University Centre for Reproductive Medicine, Bristol, and St. Helier’s Hospital, Carshalton, United Kingdom
Objective: To determine whether serum vascular endothelial growth factor (VEGF) levels can distinguish highly responsive women who subsequently develop ovarian hyperstimulation syndrome (OHSS) from women with a similar ovarian response who do not. Design: Prospective controlled study. Setting: University IVF unit. Patient(s): Women undergoing IVF who met predetermined risk criteria for OHSS. Patients developing OHSS were compared with patients who did not develop OHSS. Intervention(s): Long-protocol pituitary down-regulation followed by FSH stimulation by a standard protocol without coasting. A maximum of three embryos was transferred. Vaginal progesterone was used for luteal support. Main Outcome Measure(s): Occurrence of OHSS; serum VEGF concentrations on the day of embryo transfer (ET) and at 5 and 10 days after ET. Result(s): Serum VEGF levels at any time point did not differ significantly between 9 OHSS cases and 9 controls. Vascular endothelial growth factor levels in samples collected from cases before the onset of OHSS were higher than levels in time-matched samples from controls (medians, 177.6 [range, 64.02–549.1] pg/mL vs. 100.7 [range, 37.59 –246] pg/mL, respectively), but the difference failed to reach statistical significance (P⫽.0518), and there was considerable overlap between cases and controls. Conclusion(s): Serum VEGF levels in the luteal phase do not distinguish between high-risk women who subsequently develop OHSS and women with a similar risk profile who do not develop OHSS. (Fertil Steril威 2002;78:1154 – 8. ©2002 by American Society for Reproductive Medicine.) Key Words: OHSS, VEGF, IVF, prediction, complications Received May 29, 2002; revised and accepted July 5, 2002. Reprint requests: Raj Mathur, M.D., Reproductive Medicine Unit, Rosie Hospital, Cambridge, CB2 2SW, United Kingdom (FAX: 00-44-1223-215327; E-mail: rajmathur@ doctors.org.uk). a Bristol University Centre for Reproductive Medicine. b Department of Immunology, St. Helier’s Hospital. 0015-0282/02/$22.00 PII S0015-0282(02)04243-7
1154
As subfertility treatment moves into the 21st century, OHSS remains a significant iatrogenic complication of ovarian stimulatory treatments, causing considerable morbidity and even mortality (1). Attempts to reduce the incidence of OHSS have foundered on the lack of certainty regarding its pathogenesis and the difficulties involved in the clinical prediction of significant OHSS (2). Evidence suggests that the pathogenesis of OHSS may involve the immune system and a complex network of interacting cytokines and
growth factors (3). Of these, VEGF has been implicated as having a major role in the development of OHSS. Follicular and peritoneal fluid from patients who responded to gonadotropin stimulation with ⱖ25 follicles was found to increase permeability across endothelial cell monolayers, whereas follicular fluid from natural cycles had no effect (4). Follicular fluid was shown to contain VEGF by Western blot (5). The vascular permeability– enhancing effect of ascitic fluid from patients of severe OHSS was largely blocked by antiserum directed against VEGF (6).
Cultured granulosa cells are known to produce VEGF, and this production is increased in a time- and dose-dependent fashion by human chorionic gonadotropin (7). In theory, the actions of VEGF on vascular endothelium could account for the pathophysiological changes seen in OHSS. These actions include increased capillary permeability and neovascularization (8) and up-regulation of procoagulant factors (9). Serum VEGF concentrations during the course of OHSS have been reported to reflect the clinical severity of the disease (10). It has been suggested (11) that measurement of serum VEGF concentrations may have a role in the clinical prediction of OHSS. This study was performed to investigate whether serum VEGF concentrations before the onset of OHSS can be used to distinguish high-risk women who develop OHSS from women who do not develop OHSS despite a similar response to ovarian stimulation. VEGF concentrations in women undergoing assisted conception may vary with ovarian response, patient age, and the presence or absence of polycystic ovaries (12, 13). Hence, unless efforts are made to control for these factors, they may act as potential confounders when VEGF levels are compared between populations of women with and without OHSS. To avoid this, this study recruited women who satisfied prospectively established criteria putting them at high risk of OHSS. The aim was to achieve a control group that was similar to the OHSS group in all relevant respects except the occurrence of OHSS. A previous report in the literature examined VEGF levels in highly responsive women on the days of hCG administration, oocyte retrieval (OR), and embryo transfer (ET) (14). However, clinically severe OHSS often does not present till several days after ET (2), and VEGF levels remote from onset may not reflect the likelihood of developing late-onset OHSS. Hence, we aimed to investigate VEGF levels at later points in the luteal phase as well.
MATERIALS AND METHODS This was a prospective observational study carried out on patients undergoing in vitro fertilization at the Bristol University Centre for Reproductive Medicine between 1 January 1998 and 31 March 1999. Pituitary down-regulation and ovarian stimulation were carried out by a standardized regime that has been described elsewhere (2). Coasting was not carried out, and all patients had embryos transferred in the treatment cycle. A maximum of three embryos were transferred. In all cases, luteal support was by progesterone suppositories (400 mg twice per day) for 14 days, starting on the day of ET. Prior approval for the study was obtained from the institutional ethics committee. Individual patients gave informed, written consent for their participation in the study. Patients were approached for recruitment if they were thought to be at high risk for the development of OHSS on the basis of any of the following criteria: FERTILITY & STERILITY威
1. History of moderate or severe OHSS 2. Polycystic ovaries (PCO), diagnosed by ultrasonographic criteria (15) 3. Presence of ⱖ30 follicles in total, or ⱖ10 follicles of ⱖ13 mm in diameter on the final ultrasound scan before administering hCG for follicular maturation 4. Peak estradiol level of ⱖ10,000 pmol/L
Patients were excluded from the study if they suffered from significant systemic disease, (e.g., diabetes, thyroid disease) or were receiving corticosteroid or antiinflammatory medication. Serum samples were collected from subjects at the following times: at the time of OR, immediately after ET, between 9 AM and 12 PM 5 days after ET, and between 9 AM and 12 PM 10 days after ET. Ten mililiters of blood was collected and immediately centrifuged at 600 RPM for 6 minutes. Serum was pipetted off and stored in sterile, nonpyrogenic polypropylene microtubes (Sarstedt Limited, Leicester, United Kingdom) at ⫺80°C. Patients were assessed to determine the presence and severity of OHSS on the day of OR, the day of ET, 5 days after ET, and 10 days after ET. In addition, patients reporting symptoms of OHSS at other times were seen within 24 hours for assessment, in most instances by the same person (R.M.). OHSS severity was graded according to the criteria of Golan et al. (16). Further management including hospital admission was arranged according to the clinical picture in the individual case. One patient was admitted with OHSS in another hospital before she could be assessed; however, sufficient information was available from the admitting unit to make an accurate diagnosis of severity. Serum VEGF levels in samples from subjects who developed OHSS (cases) were compared with levels in samples from subjects who did not develop OHSS (controls). One control was matched for each case from the subjects recruited. Controls were patients treated closest in time to the OHSS case for whom all the same serum samples were available as the case. Not all planned samples were available for all cases of OHSS: day-of-OR samples were available for eight cases, day-of-ET samples for five cases, and ET ⫹5 and ET ⫹10 samples, for seven cases each. The same samples were assayed for each control as the corresponding case. Serum samples from patients and controls were assayed in a single batch for VEGF 165 using an ELISA based on the sandwich method of R&D Systems (Abingdon, Oxon, United Kingdom). No cross-reactivity has been found between VEGF and other recombinant cytokines and growth factors using this assay. The immobilized anti-human VEGF antibody was monoclonal, and the second antibody, conjugated with horseradish peroxidase, was polyclonal. All samples were run in duplicate and standards were included in the assay. Serial dilutions were tested to confirm assay linearity. 1155
TABLE 1 Characteristics of patients developing OHSS. No. 1 2 3 4 5 6 7 8 9
Age (y)
PCO
FSH (no. of Ampoules)a
Follicles (n)
Peak E2 (pmol/L)
Oocytes (n)
Outcome (no. of Sacs)
Onset (d after OR)
Severity
34 33 28 29 39 34 34 36 36
N N Y N Y N N N Y
22 22 26 16.5 52 24 22 42 24
24 24 55 24 43 35 39 15 32
NM NM 13,440 NM 33,712 NM 13,000 14,563 9194
13 23 14 14 21 19 26 7 17
NP P (1) NP NP NP NP P (3) P (1) P (1)
3 4 3 5 2 3 11 10 10
Moderate Moderate Moderate Moderate Severe Moderate Moderate Severe Moderate
Note: Y ⫽ yes; N ⫽ no; NM ⫽ not measured; NP ⫽ not pregnant; P ⫽ pregnant. Total ampoules, each containing 75 IU of FSH.
a
Mathur. Serum VEGF is poorly predictive of OHSS. Fertil Steril 2002.
The intra-assay and interassay coefficients of variation were 9% and 12%, respectively. The assay range was from 31.25– 2000 pg/mL. Absorbance was read at 405 nm using a DPC plate reader (Diagnostics Products Corporation, Los Angeles, CA). A standard curve was generated by plotting the optical density of the standards against the concentration of the standards. The VEGF concentration of the samples was determined by comparing their optical densities to the standard curve. The statistical software employed for calculations and graphs was Arcus Biomedical (version 6.80, Medical Computing, Cambridge, United Kingdom). Patient characteristics and VEGF concentrations were compared between OHSS cases and controls using a nonparametric method (MannWhitney U test).
RESULTS Fifty-four women were recruited, of whom 9 (16.6%) developed OHSS. The patient details, treatment cycle characteristics, and the times of onset and severity of OHSS are shown in Table 1. There was no significant difference between women who developed OHSS and controls in terms of patient age, body mass index, infertility diagnosis, incidence of PCO, duration, and type of infertility (primary vs. secondary; Table 2). OHSS cases and controls did not differ significantly in regard to treatment cycle characteristics and outcomes (Table 3 ). Peak estradiol concentrations were measured in five cases of OHSS and in three controls and did not differ significantly between the two groups (13,440 pmol/L [range, 9194–33,712] vs. 7289 pmol/L [range, 2869–16,268]), respectively.
Serum VEGF Concentrations in Cases and Controls Serum VEGF concentrations on the days of OR and ET, 5 days after ET, and 10 days after ET did not differ significantly between OHSS cases and controls (Table 4 ). 1156 Mathur et al.
Serum VEGF is poorly predictive of OHSS
Serum VEGF Concentrations Before Onset of OHSS Fifteen serum samples were available before the onset of OHSS from the nine cases. Vascular endothelial growth factor concentrations in these overlapped considerably with concentrations in an equal number of samples from controls matched for time of collection (Fig. 1 ). Samples from OHSS cases showed higher VEGF concentrations before the onset of OHSS than did matched control samples (median, 177.6 [range, 64.02–549.1] pg/mL vs. 100.7 [range, 37.59 –246] pg/mL, respectively), but the difference failed to reach statistical significance (P⫽.0518 by Mann-Whitney U test).
DISCUSSION The present study set out to investigate whether lutealphase serum VEGF concentrations can distinguish between
TABLE 2 Patient characteristics of OHSS cases and controls. Characteristic Age in y, median Infertility diagnosis, n (%)a Sperm dysfunction Tubal Endometriosis Ovulatory Unexplained PCO, n (%) Primary infertility, n (%) Duration of infertility in y, median (range) Body mass index, median (range)
OHSS cases (n ⫽ 9)
Controls (n ⫽ 9)
33 (28–39)
33 (29–40)
2 (21.2) 1 (11.1) 2 (22.2) 3 (33.3) 1 (11.1) 3 (33.3) 5 (55.5) 3 (1–7) 23 (19–25)
3 (33.3) 3 (33.3) 1 (11.1) 1 (11.1) 2 (22.2) 1 (11.1) 5 (55.5) 4 (3–8) 22 (21–31)
Note: None of the comparisons were statistically significant. a Some patients had more than one infertility diagnosis, hence, total may exceed 100%. Mathur. Serum VEGF is poorly predictive of OHSS. Fertil Steril 2002.
Vol. 78, No. 6, December 2002
TABLE 3
FIGURE 1
Treatment cycle characteristics of OHSS cases and controls. Characteristics
OHSS cases (n ⫽ 9)
Controls (n ⫽ 9)
No. of ampoules of FSH No. of stimulation days No. of follicles ⱖ12 mm Follicles ⬎12 mm Oocytes Embryos transferred Pregnancies, n (%) Multiple pregnancies, n (%) Implantation rate (%)
23 (16.5–52) 11 (11–14) 10 (2–21) 17 (11–35) 15 (7–26) 3 (2–3) 4 (44.4) 1 (11.1) 22.7
26 (22–46) 13 (11–17) 10 (2–12) 16 (12–33) 15 (7–31) 3 (2–3) 5 (55.5) 2 (22.2) 31.8
Vascular endothelial growth factor concentrations in serum samples collected from OHSS cases before the onset of OHSS and in time-matched samples from controls.
Note: All data are median (range) unless otherwise specified. No comparisons were statistically significant. Mathur. Serum VEGF is poorly predictive of OHSS. Fertil Steril 2002.
women who develop OHSS and controls, once the confounding factors of patient age, ovarian response, and treatment cycle outcome are excluded. The study aimed to compare women with similar characteristics and similar magnitudes of ovarian response to stimulation who had similar treatment cycle outcomes, differing only in that some of them developed OHSS whereas others did not. As is apparent from Tables 2 and 3, this aim was achieved; the case and control groups were similar in respect to infertility characteristics, ovarian response, and cycle outcome. Other investigators have compared women with “excessive” ovarian responses with those with more modest responses (11, 17). In such a comparison, it may be difficult to decide whether any detected difference in VEGF activity relates to the magnitude of ovarian response or to the occurrence of OHSS. Agrawal et al. (11) observed higher VEGF concentrations throughout the IVF cycle in 20 women who developed OHSS, compared with 87 women who did not develop OHSS. However, the OHSS group differed significantly from the control group in their ovarian response, with significantly higher peak E2 concentrations and significantly
TABLE 4 Serum VEGF (pg/mL) at different points of the treatment cycle in OHSS cases and controls. Day of cycle OR ET ET ⫹ 5 ET ⫹ 10
OHSS cases
Controls
126.7 (64.02–549.1) 191 (92.75–449.2) 128.3 (72.87–376.8) 135.8 (107.8–265.6)
115.65 (37.59–252.8) 91.55 (64.02–124.3) 120.75 (64.02–206.9) 178.4 (49.33–291.7)
Note: Values are expressed as median (range. No comparisons were statistically significant. Mathur. Serum VEGF is poorly predictive of OHSS. Fertil Steril 2002.
FERTILITY & STERILITY威
Mathur. Serum VEGF is poorly predictive of OHSS. Fertil Steril 2002.
greater number of follicles and oocytes retrieved in the OHSS group. In addition, the prevalence of PCO was significantly higher in the OHSS group (85% vs. 29.8%). The observed difference in serum VEGF concentrations may relate to these differences, rather than to the occurrence of OHSS per se. Serum VEGF levels are higher in women with PCO than in women with ultrasonically normal ovaries (12). Artini et al. (17) observed higher plasma VEGF concentrations on the day of OR in 15 women who were considered to be at high risk of developing OHSS compared with women whose ovarian responses were more modest. Another study (18) found significantly higher free-serum VEGF concentrations on the days of hCG administration and ET, but not the day of OR, in 10 women who developed OHSS compared with 15 controls. It is possible that at least part of the difference observed in serum VEGF concentrations between the two groups in that study may be accounted for by the significantly higher prevalence of PCO in the OHSS group compared with in the control group (30% vs. 6.7%). In contrast, the findings of the present study are in agreement with other studies comparing women with similar ovarian responses. A pilot study (19) observed no difference in serum VEGF concentrations between four women who 1157
developed OHSS and four controls with a similar magnitude of ovarian response, until the actual onset of OHSS. No significant difference was found in serum VEGF concentrations between 12 women developing OHSS and 12 high-risk controls on the days of hCG administration, OR, or ET (14). Similar findings were noted in the present study, which also examined serum VEGF concentrations at later points in the luteal phase of the cycle (5 and 10 days after ET). This was considered important because the onset of clinically significant OHSS is often delayed till at least a week after ET (2). Serum VEGF concentrations have been suggested as having a possible role in the clinical prediction of OHSS (11). The present study found no significant difference in VEGF concentrations between samples collected from cases before the onset of OHSS and samples collected from controls at the same time in the treatment cycle. Although concentrations were higher in women who subsequently developed OHSS, the difference did not reach statistical significance. From the overlap between both sets of values (Fig. 1) it may be surmised that although it is possible that a bigger sample may show a statistically significant difference, the clinical utility of this in predicting the risk of OHSS is likely to be limited. A rise in total VEGF concentration between the day of hCG administration and OR was found in one study to have a positive predictive value of 40% in determining the risk of OHSS (11). Because the present study did not measure VEGF concentrations before OR, it cannot provide data regarding the usefulness of this parameter. However, a rise in free or total VEGF between the days of hCG and OR or ET was not found in another study in either the OHSS group or controls (18). It is possible that simple measurement of VEGF in serum is not an adequate reflection of the true VEGF status of the individual. Vascular endothelial growth factor release by platelets at clotting may give falsely high results, whereas soluble VEGF receptors and ␣2-macroglobulin may neutralize some circulating VEGF, making it difficult to interpret the significance of VEGF concentrations on their own (20). Awareness of these limitations strengthens further our conclusion of a lack of value, in present-day clinical practice, for serum VEGF measurement in predicting the risk of OHSS in highly responsive women. However, it cannot be ruled out that a more sophisticated assessment of VEGF status may in future turn out to be useful for prediction of OHSS. A limitation of the current study is the relatively small number of patients, which is in keeping with other studies of this nature on OHSS, a relatively uncommon condition that is difficult to predict. In addition, not all serum samples were available for all cases, as in some instances, patients were unable to attend at the correct time for the blood sample to be taken. This highlights the problems of performing clinical research in this area.
1158 Mathur et al.
Serum VEGF is poorly predictive of OHSS
In conclusion, this study did not show significant differences in serum VEGF levels between high-risk women who developed OHSS and controls with similar risk profiles. Serum VEGF concentrations before the onset of OHSS are unlikely to be clinically useful in distinguishing highly responsive women who subsequently develop OHSS from women with a similar ovarian response who do not develop OHSS. References 1. Semba S, Moriya T, Youssef EM, Sasano H. An autopsy case of ovarian hyperstimulation syndrome with massive pulmonary edema and pleural effusion. Pathol Int 2000;50:549 –52. 2. Mathur RS, Akande AV, Keay SD, Hunt LP, Jenkins JM. Distinction between early and late ovarian hyperstimulation syndrome. Fertil Steril 2000;73:901–12. 3. Elchalal U, Schenker JG. The pathophysiology of ovarian hyperstimulation syndrome—views and ideas. Hum Reprod 1997;12:1129 –37. 4. Goldsman MP, Pedram A, Dominguez CE, Ciuffardi I, Levin E, Asch RH. Increased capillary permeability induced by human follicular fluid: a hypothesis for an ovarian origin of the hyperstimulation syndrome. Fertil Steril 1995;63:268 –72. 5. Amato P, Zwain I, Yen SSC. Identification of vascular endothelial growth factor in human follicular fluid: a possible role in ovarian hyperstimulation syndrome. Proceedings of the 52nd Annual Meeting of the American Society of Reproductive Medicine; Birmingham, Alabama. Boston:1996. 6. McClure N, Healy DL, Rogers PA, Sullivan J, Beaton L, Haning RV, Jr, et al. Vascular endothelial growth factor as capillary permeability agent in ovarian hyperstimulation syndrome. Lancet 1994;344:235–6. 7. Neulen J, Yan Z, Raczek S, Weindel K, Keck C, Weich HA, et al. Human chorionic gonadotropin-dependent expression of vascular endothelial growth factor/vascular permeability factor in human granulosa cells: importance in ovarian hyperstimulation syndrome. J Clin Endocrinol Metab 1995;80:1967–71. 8. Ferrara N, Houck K, Jakeman L, Leung DW. Molecular and biological properties of the vascular endothelial growth factor family of proteins. Endocr Rev 1992;13:18 –32. 9. Brock TA, Dvorak H, Senger DR. Tumour secreted vascular permeability factor increases cytosolic Ca2⫹ and von Willebrand factor release in human endothelial cells. Am J Pathol 1991;138:213–21. 10. Abramov Y, Barak V, Nisman B, Schenker JG. Vascular endothelial growth factor plasma levels correlate to the clinical picture in severe ovarian hyperstimulation syndrome. Fertil Steril 1997;67:261–5. 11. Agrawal R, Tan SL, Wild S, Sladkevicius P, Engmann L, Payne N, et al. Serum vascular endothelial growth factor concentrations in in vitro fertilization cycles predict the risk of ovarian hyperstimulation syndrome. Fertil Steril 1999;71:287–93. 12. Agrawal R, Sladkevicius P, Engmann L, Conway GS, Payne NN, Bekis J, et al. Serum vascular endothelial growth factor and ovarian stromal blood flow are increased in women with polycystic ovaries. Hum Reprod 1998;13:651–5. 13. Friedman CL, Seifer DB, Kennard EA, Arbogast L, Alak B, Danforth DR. Elevated level of follicular fluid vascular endothelial growth factor is a marker of diminished pregnancy potential. Fertil Steril 2000;70: 836 –9. 14. Chen CD, Chen HF, Lu HF, Ho HN, Yang YS. Value of serum and follicular fluid cytokine profile in the prediction of moderate to severe ovarian hyperstimulation syndrome. Hum Reprod 2000;15:1037–42. 15. Polson DW, Adams J, Wadsworth J, Franks S. Polycystic ovaries: a common finding in normal women. Lancet 1988;1:870 –2. 16. Golan A, Ron-El R, Herman A, Soffer Y, Weinraub Z, Caspi E. Ovarian hyperstimulation syndrome: an update review. Obstet Gynecol Surv 1989;44:430 –41. 17. Artini PG, Fasciani A, Monti M, Luisi S, D’Ambrogio G, Gennaziani AR. Changes in serum vascular endothelial growth factor levels and the risk of ovarian hyperstimulation syndrome in women enrolled in an in vitro fertilization programme. Fertil Steril 1998;70:560 –4. 18. Ludwig M, Jelkman W, Bauer O, Diedrich K. Prediction of severe ovarian hyperstimulation syndrome by free serum vascular endothelial growth factor concentration on the day of human chorionic gonadotrophin administration. Hum Reprod 1999;14:2437–41. 19. Krasnow JS, Berga SL, Guzick DS, Zeleznik AJ, Yeo KT. Vascular permeability factor and vascular endothelial growth factor in ovarian hyperstimulation syndrome: a preliminary report. Fertil Steril 1996;65: 552–5. 20. Wolfgang J. Pitfalls in the measurement of circulating vascular endothelial growth factor. Clin Chem 2001;47:617–23.
Vol. 78, No. 6, December 2002