| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters |
University of Auckland
Auckland, New Zealand
National Women's Hospital
Auckland, New Zealand
University of Auckland
Auckland, New Zealand
We disagree with Drs. Spring and Barkan [1] that the therapeutic effect of a hysterosalpingography (HSG) is purely a myth. Evidence from meta-analysis of randomized controlled trials confirms that oil-soluble contrast media significantly increase pregnancy and live-birth rates versus no intervention [2]. No such data exist for water-soluble contrast media versus no intervention. We were impressed by the quality and size of the randomized controlled trial undertaken by Spring's group [3], but it remains unclear why the results of that trial—in which no significant difference was found in live-birth rate following oil-soluble contrast media versus water-soluble contrast media—differed from those of a previous meta-analysis of five randomized controlled trials, in which a significant benefit of oil-soluble contrast media over water-soluble contrast media was found [2]. There is now substantial evidence-based doubt over the relative fertility efficacy of oil-soluble contrast media and water-soluble contrast media.
We showed a significant overall benefit of HSG using oil-soluble contrast media (iodized oil [Lipiodol, Andre Guerbet]) versus no intervention in our randomized controlled trial involving 158 women [4]. The fertility benefit among 96 women with unexplained infertility was marginal (a statistically significant benefit could be detected only by metaanalysis of our trial with a similar trial performed in the United Kingdom [2]). However, the benefit of Lipiodol for women with endometriosis, albeit with a small sample size of 62 women, was highly statistically significant; the relative "risk" for pregnancy was 4.4 with a 95% confidence interval of from 1.6 to 12.2 (p = 0.001). Furthermore, we could find no evidence of increased sexual frequency or increased focus of sexual activity to the fertile phase of the cycle in women randomized to Lipiodol versus those randomized to no intervention. Given that the therapeutic effect in the endometriosis population differed from that in women without diagnosed endometriosis, we speculated that Lipiodol could be correcting an immunodysfunction in women with endometriosis.
In 2002, we started investigating a hypothesis that Yun and Lee [5] have now outlined: Oil-soluble contrast media exert a fertility-enhancing effect through endometrial immunomodulation [5]. Indeed, we previously proposed that such a mechanism could correct an endometrial implantation dysfunction, thereby leading to improved endometrial receptivity [3]. We also have found that Lipiodol alters uterine dendritic cell populations in mice [6]. Such a change in dendritic cell populations may lead to the shift in the T-helper cell balance suggested by Yun and Lee. Alternatively, changes in dendritic cell populations may alter the ability of the endometrial immune system to take up and present antigens from the histoincompatible fetoplacental unit, resulting in a uterine environment that is more receptive to implantation.
The finding that three women with known bilateral patent fallopian tubes became pregnant immediately after a Lipiodol procedure in which patency of neither fallopian tube could be shown (from a total of 58 pregnancies after 173 consecutive procedures performed from February 2000 to April 2004) further supports an endometrial bathing effect rather than a tubal flushing or peritoneal bathing mechanism.
Whatever the mechanism of effect, the compelling evidence of fertility enhancement after Lipiodol administration has led to it being offered on a wider scale in New Zealand to couples who fulfill strict selection criteria that include previous confirmation of normal patent fallopian tubes. As a simple, low-cost, minimally invasive intervention that carries no increased risk of multiple pregnancy and a low risk of complications, Lipiodol treatment is, for many, a more appealing choice than established fertility treatments such as in vitro fertilization.
References
Kaiser Permanente Oakland Medical Center
Oakland, CA
94611
Kaiser Permanente Santa Rosa
Medical Offices
Santa Rosa, CA
95403
We are grateful for the rigorous Cochrane Library database analysis [1] and recommend it and its predecessors to interested readers.
Three separate questions are at the heart of the discussion of oil- and water-based contrast agents used in hysterosalpingography (HSG). First, is there a therapeutic value of HSG procedures that results in enhanced rates of live births? Second, is there an added therapeutic value of using oil-based over water-based contrast agents in HSG? Furthermore, if either or both of these questions can be answered in the affirmative, what is the underlying mechanism?
Question 1: Is there therapeutic value of HSG procedures that results in enhanced rates of live births?
We are less confident of a therapeutic effect of HSG than Drs. Johnson, Hadden, and Chamley and Yun and Lee [2] seem to be. In the previous Letter to the Editor, Johnson et al. refer to two therapeutic studies that we presume to be Nugent et al. [3] and Johnson et al. [4]. The results of both of these studies reached statistical significance only weakly when considered alone. The first study [3] evaluated a subgroup of women with unexplained infertility. It found an increase in the live-birth rate after HSG using iodized oil (Lipiodol, Andre Guerbet) versus no HSG that reached statistical significance, albeit marginally and with wide confidence intervals (CIs) resulting from the study's small sample size. An editorial commentary suggested that the statistical measures attained in that study were "skating on thin ice" [5]. The author of the commentary observed that if one patient in the control group had become pregnant, the p value would have shifted from 0.0445 to 0.17, and the findings would have lost statistical significance.
The second study [4], one conducted by Johnson et al. and termed "The FLUSH Trial," found that the association of Lipiodol with live births failed to reach significance for those with unexplained infertility (p = 0.13, not significant) but did reach statistical significance for those with endometriosis not involving the fallopian tubes or ovaries (p = 0.0007). The pattern of statistical significance levels for conception was similar—that is, the association was not statistically significant for those with unexplained infertility but reached statistical significance for those with endometriosis. Although the control group did not undergo HSG, Nugent et al. [3] attributed the increased live-birth rate to oil-soluble contrast media rather than to the performance of HSG using Lipiodol per se. We will discuss whether the results reached in that and other prior studies support such a specific conclusion.
The FLUSH Trial found that endometriosis, in the absence of tubal or ovarian involvement, had no negative effect on pregnancy rates. This finding is consistent with the findings of other trials [6, 7]. Why the FLUSH Trial showed such a striking difference in live-birth rates between women who underwent HSG using oil-soluble contrast media and those who did not undergo HSG among the members of their sample with endometriosis is unclear. The diagnostic strategies followed before HSG may have differed between those with and those without endometriosis. All the patients in the endometriosis group had prior tubal patency confirmed at the time with laparoscopy. We do not know the diagnostic strategy used in the nonendometriosis group. Their clinicians may have had fewer opportunities to identify other coexistent unrecognized pelvic disorders, including endometriosis. (Given the stated preferences of the authors, all HSG procedures may have used oil-soluble contrast media. Interestingly, all of the study participants had failed to become pregnant after screening HSG, which is one reason they were eligible to participate in the FLUSH Trial.) Please note that in both of these trials the intervention compared with no treatment was HSG with oil-soluble contrast media, not oil-soluble contrast media per se.
Question 2: Is there an added value of oil-based over water-based contrast agents in HSG?
Rasmussen and colleagues [8] found that HSG performed using oil-soluble contrast media was followed by higher live-birth rates than was HSG using water-soluble contrast media. Our study [9] aimed to generate findings similar to those of Rasmussen et al. If our findings had been similar, we had intended to encourage the use of oil-based agents in our practice setting to maximize live-birth pregnancy outcomes among our patients. This would be a simple and relatively inexpensive intervention to address a common clinical complaint. Our study was not designed to compare outcomes with and without contrast media. The explanation for the differences between Rasmussen and colleagues' results and ours remains obscure to us.
In contrast to the two flushing trials referenced earlier [3, 4], both our study [9] and the study by Rasmussen et al. [8] compared outcomes among women imaged using different contrast media (i.e., water-soluble contrast media vs oil-soluble contrast media). Neither our study nor that of Rasmussen and colleagues attempted to validate or refute a hypothesis comparing contrast agents with control subjects who did not undergo HSG. To our knowledge, no one has conducted a randomized controlled trial to examine either the question of water-soluble contrast media versus no intervention or of all types of contrast media versus no intervention (i.e., there has been no study of oil-soluble contrast media versus water-soluble contrast media versus not treatment, to our knowledge) in trying to show a benefit. We are encouraged that, except for Rasmussen et al., the other randomized controlled trials included in the Cochrane analysis [10-12] each had CIs that were also consistent with no preferential effect by either agent. Only the Rasmussen et al. study has shown confidence limits that seem to support a clear preference for oil-soluble contrast media. None of the other studies have CIs that support a clear preference.
We agree that these questions of whether HSG per se carries a treatment benefit and of whether that treatment benefit varies among contrast agents are important. Their answers carry important implications for the investigation and management of infertility, a common disorder. Their answers also carry financial implications for couples seeking to have children because fertility enhancement by tubal flushing alone is less expensive than other interventions such as in vitro fertilization. We encourage further research to examine these questions.
Question 3: If HSG enhances fertility and live-birth rates for selected couples, how does HSG do it?
Yun and Lee [2] and Johnson and colleagues [4] believe that the use of oil-soluble contrast media results in an increased rate of live births. Assuming this to be true and seeking an explanation, these authors postulate immunomodulation to explain a therapeutic effect, and Johnson et al. added another hypothesis: A chain of oil-soluble contrast media-mediated events enhances uterine receptivity.
These hypotheses are interesting but highly conjectural. Other possible explanations abound. One obvious suggestion, implied in the title "FLUSH Trial" [4], is that tubal flushing washes out obstructions, debris, adhesions, and other material. This has been stated as a mechanism at work in fallopian tube recanalization procedures [13]. One might then need to hypothesize that the effect they have suggested exists for only women with endometriosis. Such a group might be more likely to have intratubal debris. However, this too is a hypothesis.
Finally, as suggested by the Cochrane Analysis [1], it is likely that the couples in the Rasmussen et al. [8] study and those in our study differ. We suspect this is true and wish we could identify which factors in our population differ from those of the population studied by Rasmussen and colleagues (Scandinavians) and those for whom Johnson et al. [4] recommend oil-soluble contrast media therapy. Important factors may include ethnicity and the participating women's ages.
On the basis of the data of Johnson and colleagues [4], we are willing to entertain the hypothesis that HSG with either oil-soluble contrast media or water-soluble contrast media may have a therapeutic effect in selected women with endometriosis not involving the fallopian tubes or ovaries. We think that the study most needed is one that compares the outcomes obtained after HSG is performed using oil-soluble contrast media, HSG performed using water-soluble contrast media, and no HSG in a group of comparable women receiving care for infertility of similar characteristics. That study should be stratified by the presence or absence of endometriosis.
References
Stanford University
Palo Alto, CA 94310
We appreciate the correspondence from Johnson et al. responding to the article by Drs. Spring and Barkan [1] regarding the fertility enhancement seen with iodized oil (Lipiodol, Andre Guerbet) hysterosalpingography. The results of the FLUSH Trial support the view that oil-soluble contrast media may promote fertility [2].
Johnson et al. [2] found the benefit of Lipiodol to be greater in women with endometriosis than in those without endometriosis. Infertility exists in some patients with endometriosis despite proven tubal patency [2]. We share the view that Lipiodol may enhance fertility in these women through endometrial immunomodulation. This finding suggests a new consideration in our understanding of endometriosis: Its pathogenesis and its link to infertility may involve dysfunction of helper T cells (TH1). Although more confirmatory data are needed, peritoneal T cells in endometriosis patients appear to be primarily TH1 in nature [2]. A shift from TH1 to TH2 induced by contrast media may explain the excess fertility benefit of Lipiodol in women with endometriosis [2].
Notably, emerging data point to an immunomodulatory role of seminal fluid [2, 3]. Constituents of seminal fluid such as norepinephrine, aldosterone, transforming growth factor-B1 (TGF-B1), and prostaglandin E2 (PGE2) have been shown to dampen maternal TH1 immunity and promote TH2 immunity [4, 5]. This phenomenon may reflect an evolutionary adaptation to ensure that the immune environment in the female reproductive tract remodels sufficiently toward TH2 bias, a state that favors the survival of the male gametes and the fetal allograft [4]. Live-birth rates for assisted reproduction techniques are markedly improved if the females are exposed to seminal fluid around the time of embryo or gamete transfer [4]. Thus, oil-based contrast agents may buttress natural immunomodulatory processes that facilitate fertility. Whether a declining exposure to seminal fluid and its immunomodulatory effect contributes to the modern proliferation of various gynecologic dysfunctions remains to be investigated.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
