Stanford School of Medicine
Former President, ASRM
Standford School of Medicine
In this section we’ll review the treatment options for endometriosis, specifically various combinations of surgery, medications, and assisted reproductive technologies (like intrauterine insemination and in-vitro fertilization). Below is a chart that summarizes our findings:
Ultimately, we have the following observations:
In our analysis, we categorize a treatment’s rate of success by the type of patient that is being treated. We characterize patients based upon their ASRM staging, which rates the severity of endometriosis on a 1 - 4 (mild to severe) scale. A few notes on the ASRM staging approach:
When it comes to improving fertility, this is how we’d characterize each treatments effectiveness, within the staging rubrik.
During diagnostic surgery, the surgeon visually inspects and/or biopsies surfaces of the pelvic organs. This is different than treatment with therapeutic surgery, where the surgeon will try to cut out or destroy any endometriosis lesions they see and then repair organs which have been damaged by endometriosis.
It’s important to note that the data for therapeutic surgery differs depending on the goals of surgery: the success rate for treating pain is different than the success rates for treating infertility. As such, it’s important to evaluate options for surgery depending on whether treating pain or infertility is the primary goal.
If you have endometriosis-related pain, particularly if treatment with medications has not been effective, therapeutic surgery improves pain and quality of life in the majority of patients.
Unfortunately 30 to 60% of patients will have recurrences within a year of surgical treatment unless they have additional treatment. These rates can be reduced through post-operative treatment with drugs, which we’ll discuss below.
For patients whose primary symptom of endometriosis is infertility, surgery can also be helpful, but it can be extremely inefficient depending on the patient type.
For patients with minimal or mild endometriosis, there are two randomized controlled trials where patients were assigned to receive either therapeutic surgery or diagnostic surgery only before trying to conceive. Again:
The results are conflicting. In one study, performing therapeutic surgery did not result in additional pregnancies. In the other study, performing therapeutic surgery did result in a small increase in the number of pregnancies.
In a study combining the two clinical trials, investigators calculated the “Number Needed to Treat” (NNT) which reveals how many surgeries must be performed to derive one extra pregnancy. Here are the findings:
Patients With Endometriosis: For patients who do in fact have endometriosis, the NNT is 12, which means for every 12 treatment surgeries performed on patients who have endometriosis, only 1 additional pregnancy was achieved (ASRM 2012, Jacobson 2010). If surgical treatment was risk-free, that might be tolerable, but as we’ll show in the next section, surgery has negative attributes too.
Patients Who May Have Endometriosis: Roughly half of patients suspected of having endometriosis truly have it. Thus, performing surgery on those who have symptoms of endometriosis yields an NNT of 24 or 24 surgeries per one additional birth.
Patients Who May Have Endometriosis Based On Infertility: Roughly one third of women with endometriosis suffer from fertility issues. Thus, performing surgery on women who just present with infertility yields an NNT of 40. Said differently, 40 women with infertility would need to be treated surgically to have one additional child.
For women with the symptoms of endometriosis, or infertility, but who have not been surgically-diagnosed, while surgery may increase the likelihood of getting pregnant, it is incredibly inefficient and there may be better options in light of the physical and financial burden it presents.
For women with moderate to severe endometriosis, several studies demonstrate that 44 - 63% of women conceive naturally within 2 - 3 years of endometriosis surgery.
Amongst this patient population, this success from surgery rates favorably versus not having any treatment (0 - 25% pregnancy rates), or fertility interventions like IUI (10% success rates after four months). However, it looks roughly comparable to undergoing a single cycle of IVF, which is costly and invasive but delivers similar outcomes to surgery in a matter of months, not years.
The Endometrial Fertility Index (EFI) can be a useful tool to determine if a woman will conceive naturally after endometriosis surgery.
The inputs of the EFI come in two batches. On the left-hand panel, the patient provides basic quantifiable information on their age, years they have tried to conceive and if they’ve conceived in the past.
In the middle panel, doctors provide information (often gathered during surgery as a diagnosis) that measure the extent to which endometriosis has spread and the impact it has had on the anatomy.
However, even without having information to fill in the middle panel, the EFI can still be useful in determining the best course of treatment.
Thereafter, a patient and doctor can calculate the patient’s odds of conceiving naturally three years after therapeutic surgery. It is important to note that the EFI specifically pertains to people who would not need assisted reproduction for other reasons, meaning they don’t have something like blocked tubes or severely low sperm count, as the estimated success rates would not apply to these patients.
For instance, if a young couple has a score of 9, they likely have a 75% chance of conceiving in the next three years and might be encouraged to undergo surgery. If an older couple has a score of a 4, they have a 30% chance of conceiving after 3 years and may be better suited to advancing to IVF.
Some scores can be estimated without diagnostic surgery to help determine if treatment surgery is the most appropriate option. For instance:
Her “history total,” on the left-hand panel, is 0, meaning even if her “Surgical total,” on the middle panel, is as high as possible (5), her maximum score possible is 5, which corresponds to a 10 - 45% pregnancy rate over the next 3 years.
Thus the probable outcome is that she will not conceive after surgery. Meantime, the likelihood IVF will work for her would plummet in that time, as she would have gone from age 40 to 43.
The implication is that if this woman did surgery but later needed to resort to IVF (again, the probable outcome), she would be staring at the possibility of needing to do 10 - 15 IVF cycles to have a live birth. Whereas, if she started with IVF she would likely need on average to 2 - 3 cycles to have a live birth.
This patients “history total” (left-hand panel) score is 5, meaning that her least possible total score is a 5. That correlates to a minimum projected 45% pregnancy rate over the next three years. And those projections could reach 75% depending on her findings during surgery (middle panel). This may be more acceptable to a younger patient, particularly if they cannot afford IVF.
More likely than not this patient will conceive after surgery, but if not and she must resort to IVF, she is still within an age range where IVF is highly likely to succeed within a few cycles.
It’s unclear whether having therapeutic surgery performed before starting IVF leads to better success rates compared with immediately starting IVF without therapeutic surgery. Most studies are similar to the one shown below and reveal there is no benefit.
In one study where surgery was being done to treat ovarian endometriosis, patients subjected to surgery required more stimulation during IVF to get fewer eggs and yet ultimately had similar outcomes to women who bypassed surgery and proceeded directly to IVF. We can extract from this that surgery does not improve egg quality or likely improve odds of success with IVF.
In general it’s unclear if surgery between IVF retrieval and transfer improves transfer success rates.
But one scenario where we know surgery is highly likely to improve IVF outcomes is in the case of a hydrosalpinx, which can be associated with endometriosis.
A hydrosalpinx occurs when scarring traps fluid in the fallopian tube. In the fluid there can be debris and inflammatory material which can travel from the fallopian tube to the uterus, lowering the likelihood that an embryo can implant in the uterus and a pregnancy can proceed.
As a result, removing a hydrosalpinx (effectively, removing the tube) can double success rates compared to leaving it in place.
So, if you have a hydrosalpinx, surgery after IVF egg retrieval and before embryo transfer is warranted, and otherwise it’s unclear. We’ll soon address how taking certain medications between IVF retrieval and transfer may improve outcomes for endometriosis patients.
While in many situations surgery can improve fertility in patients with endometriosis, it has risks. In this section we will cover the most significant risks.
When women who have endometriosis on their ovary (known as an endometrioma) surgically treated, they experience a drop in AMH (a proxy for egg count). That’s because even the best surgeons will remove or damage some normal ovarian tissue with removal of endometriosis from the ovary.
While troubling, AMH itself is not a predictor of a woman’s ability to conceive naturally and so a degraded egg supply after surgery may not hurt a woman’s ability to have a baby.
But for patients with an already low ovarian reserve and/or those who will need IVF, such decline could hurt their chances to conceive. Also, studies have shown patients who had surgery on both ovaries went through menopause 4 - 5 years earlier.
If you are planning surgery to remove endometriosis from the ovaries and are starting with low ovarian reserve, it is worth considering freezing eggs or embryos before surgery.
Occasionally removal and repair of parts of the bowel, bladder or ureter (the tube that drains urine from the kidney to the bladder) are required to treat endometriosis. Not all gynecological surgeons are trained to perform these more complicated procedures.
While some report injury and complication rates as low as 0.09%, others report much higher complication rates from endometriosis surgery.
Therapeutic surgery to treat endometriosis is hard and potentially dangerous. A variety of doctors consider this their domain and they have an array of qualifications, namely:
In pelvic surgery, experience clearly matters and there are two studies that bear this out. In one French study, increasing surgeon experience coincided with half the rates of complications, though complications were rare across both groups.
In another study, surgeons who had completed over 30 rectovaginal endometriosis surgeries recorded lower “incomplete surgery rates” (incomplete surgeries drive higher recurrence rates), shorter surgery times, and lower patient blood loss.
While complications from surgery are rare (single digits), they can be serious, inflicting injury to the bladder or bowel.
You must ask your surgeon if issues arise during surgery whether your doctor can treat it themselves or has adequate support from those who can nearby. If you’re being treated at a hospital, oftentimes assistance is close by.
If you are planning surgery for an endometrioma on your ovary, it is important to have discussed your ovarian reserve with your surgeon and whether fertility preservation is indicated before your surgery.
Additionally, you should discuss with your surgeon if they plan to “excise” the cyst wall or “drain it.” Excising the cyst wall decreases the recurrence risk by half but may have a higher risk of damaging normal ovarian tissue.
Finally, infection and pain can be common byproducts of endometriosis surgery. When a surgeon can remove or destroy endometriosis through minimally invasive surgery (using several small incisions instead of one large incision), the recovery times are shorter with less postoperative pain. Thus, it’s helpful if your doctor is comfortable with such an approach.
Again, if you’re considering therapeutic surgery, we recommend consulting the EFI (above) to chart the possible range of your scores based on your history to see if surgery makes sense.
Many women with infertility associated with endometriosis, even those who have surgery, still require fertility treatment to conceive. The most common treatments considered are
In this section we will review some of the benefits and drawbacks of both, starting with IVF.
IVF presents the treatment option with the fastest “time to pregnancy” for patients with endometriosis.
During IVF, patients’ ovaries are stimulated with stronger medications and doses than typically used with other fertility treatments, like IUI (which we’ll cover shortly). Just prior to ovulation, the patient undergoes a procedure in which the eggs are surgically removed from the ovaries (usually under anesthesia). In an embryology laboratory, the eggs are then fertilized with sperm, and grown to a stage where the embryo can be transferred back into the uterus or frozen to be transferred at a later time.
Efficacy: IVF is more successful than the alternatives in driving faster pregnancies in endometriosis patients of all stages.
As you can see by the data below (which is not broken down by stage type, unfortunately), successive IVF treatments in endometriosis patients can be extremely effective.
First, IVF bypasses the fallopian tubes. This is important because endometriosis can elicit damage and inflammation in the fallopian tubes. When that happens the risk of an ectopic pregnancy (where the embryo implants in the fallopian tube wall) climbs. Ectopic pregnancies can be fatal.
Next, oftentimes multiple embryos are created with IVF and they can be frozen to help conceive future children, which may be important depending on the age, ovarian reserve and desired family size of the patient when they start treatment. In approximately 20% of IVF cycles, enough embryos are created that eventually lead to multiple, separate pregnancies and deliveries.
Finally, during IVF embryos can be transferred one at a time, thereby dropping the risk of twins to 1% of pregnancies (low compared to IUI, where 8 - 30% of pregnancies are twins). The reason this matters is because twin pregnancies are inherently more dangerous to both mother and offspring.
IVF Costs More: Unlike endometriosis surgery, IVF is seldom covered by insurance — less than 20% of patients in the US are covered. As a result, patients are more likely to pay out of pocket for IVF versus surgery. In comparison to IUI (also seldom covered by insurance), IVF patients pay 5 to 40x more per cycle ($20K vs $500 - $4K).
IVF Is More Invasive: While IVF involves a minor surgery (egg retrieval) it pales in comparison to endometriosis surgery with regard to risk of complication and the amount of recovery time required (days versus weeks). In comparison to IUI, though, IVF creates real burden. Unlike most types of IUI, IVF requires frequent monitoring and a surgical procedure that demands anesthesia.
Before and during IVF, endometriosis patients have a number of decisions to make and we’ll cover which have an impact on success rates and which don’t. Below is a quick summary of what we’ll cover next.
As you’ll recall, estrogen causes endometriosis to grow, and estrogen is produced by the ovary. There is evidence to support having a woman take “ovarian suppression” medications (like lupron) for 3 - 6 months before an IVF cycle begins. But many providers do not feel the potential benefits outweigh the delay in treatment this creates nor the side effects this imposes (like hot flashes with lupron). Later, we’ll cover “ovarian suppression” medication and it’s strengths (e.g. reducing pain) and limitations (e.g. must be used with IVF to improve fertility).
Once the IVF cycle has begun, there is no conclusive data that a particular stimulation protocol (which drugs you take) is better for patients with endometriosis. One retrospective study suggested that a “long” protocol may be better than an antagonist protocol in patients with mild endometriosis, yielding a 43% live birth rate vs. a 27% live birth rate. The long protocol used a low dose of lupron starting at the end of the cycle prior to stimulation to prevent ovulation, and the antagonist protocol used GnRH antagonist to prevent ovulation.
However, this difference was not “statistically significant” (meaning there was more than 5% chance this observation was due to random chance and not the different medications). Also, no differences were observed in patients with moderate and severe endometriosis.
During IVF eggs can be fertilized with “conventional insemination” (placing eggs and sperm together in a petri dish) or by ICSI, which involves injecting a single sperm directly into the egg.
Patients with endometriosis have lower fertilization rates than patients without endometriosis. The best data we have shows endometriosis patients are better off having their eggs fertilized with ICSI. As you can see below, this leads to higher fertilization rates, more embryos created, and a higher a live birth rate.
During IVF, embryos can be either transferred immediately after they are created (3 - 6 days after eggs are retrieved) or frozen and transferred at a later date.
As you’ll recall, estrogen feeds endometriosis, and the drugs given to stimulate the ovaries in an IVF cycle induce extremely high estrogen levels. A “fresh” transfer can take place immediately after estrogen levels have spiked, or embryos can be frozen and transferred the next cycle or later, so that the transfer doesn’t immediately follow such a dramatic estrogen spike. This gives the body time to reset before focusing on getting the uterus ready to accept an embryo, which is likely preferable.
A recent study suggests women with endometriosis benefit by waiting for at least a month to transfer their embryos rather than having them transferred immediately. This translates into lower miscarriage rates and higher ongoing pregnancy rates:
You may recall we addressed “ovarian suppression” earlier, but that was in the lead-up to an IVF cycle. Here we’re talking about taking “ovarian suppression” drugs after an IVF egg retrieval and before an embryo transfer.
Some believe if you are going to suppress a woman’s ovaries, before the transfer is the best time to do so. Proponents of this theory believe in three tenets:
One recent study showed impressive embryo implantation rates in women with endometriosis when ovarian suppression, after retrieval and before transfer was applied. In fact, results were on par with non-endometriosis patients, which is a group that typically has much higher success rates.
IUI, or intrauterine insemination, is an alternative to IVF. At its core, an IUI is when the most viable sperm is separated from a semen sample during a process called “a wash,” and then placed into the uterus on the day of ovulation through a small straw in a generally-painless procedure that requires no anesthesia. This is often done in combination with some form of ovarian stimulation, so that more than one egg develops to increase the odds that an embryo will be created when sperm is placed into the uterus.
Compared with natural conception, IUI looks expensive (~$500 - $4,000 per cycle) and like an imposition. But compared with IVF ($20,000), IUI looks cheap and less burdensome, with fewer monitoring appointments and no surgery.
Compared with both natural conception and IVF, IUI carries a higher risk of “multiple gestations per pregnancy,” meaning twins, triplets, or higher, depending on whether you use oral medications (1 - 13% per pregnancy) for ovulation induction or injectable hormones (20 - 30% risk). Multiple gestation pregnancies are significantly riskier to both mother and baby. For comparison, in natural conceptions and IVF (with single embryo transfer), only 1% of pregnancies are multiples.
IUI with ovulation induction in Stage 1 and 2 endometriosis patients drives roughly 10% pregnancy rates per cycle, which rates favorably against the alternative of trying to conceive without treatment in head-to-head studies.
However, a single IUI treatment cycle clearly looks inferior to rates of success compared with one treatment surgery or one IVF cycle.
For endometriosis in Stage 3 or 4, there are only a few head-to-head studies comparing IUI plus ovarian stimulation versus no treatment (timed intercourse). One study showed that IUI plus ovarian stimulation after surgery led to no improvement, and potentially worse pregnancy rates, compared to trying to conceive naturally following therapeutic surgery.
In this patient population, these monthly success rates certainly look worse than doing therapeutic surgery or IVF.
Either way, after undergoing four IUI cycles (or four months of timed intercourse) the odds of conception with each passing month are close to zero, and continuing on this path can be a crucial misuse of time. Advancing to IVF is often the best alternative at this point.
Aromatase inhibitors (like letrozole) are one type of medication used to stimulate the ovaries in conjunction with an IUI. They reduce circulating estrogen levels which is why some doctors prefer to use them in patients with endometriosis. However, one randomized controlled trial did not observe any difference in success for endometriosis patients using letrozole versus clomid for ovarian stimulation.
We alluded to the use of medications and “ovarian suppression” in the IVF section and we’ll take a closer look here. As you recall, endometriosis needs estrogen to grow and wreak havoc on organs in the pelvis. Estrogen is primarily produced in the ovaries and so the drug strategies that turn off estrogen production also prevent ovulation and thus pregnancy. In our minds the 3 key takeaways are:
These medications prevent the pituitary gland in the brain from stimulating the ovaries. This induces a temporary menopause and starves endometriosis lesions of the estrogen it needs do inflict damage.
Pain Reduction: In a randomized controlled trial, 93% of patients who were given lupron had significantly improved dysmenorrhea (painful periods) compared to 5% given a placebo. However, in at least half of patients, pain recurs after stopping treatment.
Fertility: These drugs can be used in combination with fertility treatments. While studies have shown there is no fertility benefit to treatments with medications alone (like lupron), some studies suggest an increased IVF success rate after ovarian suppression with lupron for three to six months. For this reason, the European Society for Human Reproduction and Embryology (ESHRE) says this approach can be effectively used before IVF.
As we discussed previously, there is debate whether ovarian suppression is best used in conjunction with IVF before the egg retrieval or between the egg retrieval and transfer.
Because ovarian suppression with GNRH agonists like lupron induces a temporary menopause, side effects like hot flashes and vaginal dryness are common. Due to the detrimental effect on bone density, treatment (and therefore improvement) is only recommended for a maximum of 6 months.
Many patients do well without ovarian suppression, prompting many providers to not accept the tradeoff of side effects and delays associated with ovarian suppression.
Other similar strategies that provide progesterone to counteract the body’s natural ability to create estrogen include treatments with:
We should note that these are birth control methods, and are effective for treating pain symptoms but clearly are not good for infertility.
There is a need for better data to evaluate whether lifestyle modifications in endometriosis patients are capable of alleviating pain, or improving fertility. However, there is some data to support a few interventions:
- Surgery makes more sense for patients with pain affecting their quality of life than for those with infertility without other symptoms from their endometriosis.
- Patients with lower ovarian reserve who are planning surgery — particularly if endometriosis could involve their ovaries — should consider fertility preservation before their operation.
- If it is likely someone will need IVF with or without surgery to conceive (based on other infertility diagnoses, EFI, etc.), it is important to truly justify surgery — surgery is not likely to improve their IVF outcome.