Recurrent pregnancy loss in general practice: two losses, investigations, and what actually helps

Chris Arthur

Private obstetrician and gynaecologist, Kundara Place, with public practice as well. Interests he named: fertility, hysteroscopic surgery, pregnancy, fibroids — everything else in gynaecology land, except prolapse. He was covering the night because the scheduled speaker could not.

The Australian definition changed: three consecutive is now two

The reason the topic was on the night, he said, is that something in this area has moved. RANZCOG — the Australian definition of recurrent pregnancy loss — was published as changed last year.

The old bar was three consecutive losses. The new bar, as he taught it: two losses before 20 weeks, of an intrauterine pregnancy. Ectopic pregnancy does not count. It is not the same worldwide; other jurisdictions still use different cut-offs.

It does not have to be consecutive. Two at any stage is now enough to enter the cohort.

Why did the guideline group move? There does not seem to be much difference in the numbers overall. Three consecutive, or two at any stage, captures roughly the same group of couples, and the differences in their characteristics are very, very low, if present at all. Staying on three consecutive would potentially underserve people who already look like the recurrent-loss population.

That cohort is about one to four percent of couples. At least 95 percent — he also said 98 percent — never hit this. So it is not walking through the GP door every day. The point of the hour was that there is also not that much you have to remember once they do.

Background risk, age, and aneuploidy

If you treat everyone as one homogeneous group, the risk of miscarriage in a recognised pregnancy sits around the room’s guesses: about 15 to 20 percent. That number is not fixed across reproductive life.

• Someone in their early 20s is closer to 10 percent.
• Someone in their mid-40s is looking at about a 60 percent chance of miscarriage.

Once you have had one miscarriage, the next pregnancy’s risk goes up by about 10 percent, and it does that for every subsequent loss. After three consecutive miscarriages he put the next-pregnancy miscarriage risk at about 45 percent — which still means something like an 80 percent chance the next pregnancy is successful. Couples hear the 45 and not the 80. Both matter.

Take out advanced maternal age — mothers further from 21, closer to the end of their reproductive life — and about 70 percent of pregnancy losses are associated with aneuploidy (a chromosome number that is not 46). If someone has three in a row, it is actually less likely that each of those was “just aneuploidy.” Recurrent loss is a clue that something else may be running underneath, not that every miscarriage is a random chromosome event.

Risk factors worth asking about

The room offered the usual list. He agreed, then widened it.

• Antiphospholipid syndrome (the clotting story everyone reaches for).
• Age — younger than 20 and older than 35.
• Thyroid disease (he came back to this in the investigation block).
• Anything that increases physiologic or psychological stress: drinking, smoking, increased weight, night-shift work, air pollution, pesticides, other toxins.
• The male partner over 40, and DNA-fragmentation concerns (he named the idea; he did not walk a protocol).

He is an obstetrician-gynaecologist, and he was blunt about his bias: pregnancy is a fantastic stress test. Recurrent miscarriage sits on a continuum with later cardiovascular disease and venous thromboembolism — the magnitude of that association is small, but it is a reason to take a complete history when a new woman joins the book. The comparison he always comes back to: severe preeclampsia is a much greater cardiovascular risk factor than heavy smoking, and it is not sitting in every CVD calculator you open. Recurrent loss is not as loud as severe preeclampsia. It is still a prompt to think about the whole woman, not only the next pregnancy.

Four investigations — and what they cannot tell you

For anyone who meets the recurrent-loss criteria, he recommends a short list. He had four on the slide and joked that he had only put three on the handout. As spoken, the work-up to do is:

Those four exist because there is something you could potentially do if you find them. The pick-up rate is very low. The vast majority of couples will be unexplained.

Cavity problems you can sometimes do something about

A plain ultrasound is useful-ish. The cavity is a closed space. Fill it, and you see what a 2-D slice can hide.

• Asherman syndrome (intrauterine adhesions), especially after an over-zealous curette. Clues on a plain scan: an indistinct endometrial–myometrial junction. On a filled study: bands crossing the cavity. Dividing scars and using oestrogen afterwards gives a chance. After dense, repeated adhesions he still quoted maybe less than a 50 percent chance of a functional cavity — which is better than zero, and needs honest counselling.
• Polyps can look like a slightly thick endometrium until the cavity is opened and the polyp is hanging in the space. Cutting them out maybe helps.
• Submucosal fibroids change the shape of the cavity; the tissue over the fibroid also responds to progesterone a bit differently, so a slice of the cavity is less receptive. Cutting them out may or may not change the next pregnancy. He was not selling surgery as a guarantee.

The counselling line he wanted GPs to hear: give some hope, and temper it. Most of this cohort will not have a fixable cavity problem.

True obstetric APS: aspirin and heparin

This is the one antibody result that changes drugs in a way he was willing to call genuine, not TikTok.

Typical obstetric antiphospholipid syndrome, as he listed it, needs the antibody and one of:

• a pregnancy loss at more than 10 weeks with a previously normal scan, or
• three consecutive losses at less than 10 weeks, or
• a birth at less than 34 weeks with placental insufficiency — very abnormal CTG, growth restriction, severe preeclampsia, the placenta was not working.

Those women have a better chance of a successful next pregnancy with aspirin and heparin, and a better chance of avoiding terrible blood-pressure problems with hydroxychloroquine. That is a real syndrome. It is not a reason to start heparin on a weakly positive antibody and a single early loss.

Thyroid and parental karyotype

Thyroid

His grading, as spoken:

• Overt hypothyroidism — yes, treat.
• Subclinical hypothyroidism — treat, maybe.
• Euthyroid with anti-TPO antibodies — maybe.

Most of the thyroid story in miscarriage, and in later neurodevelopmental worry, is entangled with iodine deficiency. Replacing iodine in a deficient area is what moves the needle for future pregnancies; thyroxine replacement is maybe, potentially, possibly.

That being said, if TSH is 19 he is definitely replacing; 10, probably; 5, sometimes; a bit under 5 he is still trying to get TSH down toward normal — “as long as she has three” was the target he named. He did not walk a full thyroid-in-pregnancy protocol. He named a treat-it instinct once TSH is clearly high, and a softer instinct in the grey zone.

Karyotype, IVF and PGT

If there is a parental balanced translocation, ordinary screening can miss it, and it can keep causing miscarriage. IVF with pre-implantation genetic testing lets you use a euploid embryo. He was not shouting this from the rooftops for every recurrent-loss couple. The group where testing (and thinking about euploid transfer) earns its keep is mothers closer to the end of reproductive life, because any embryo you put back is more likely not to implant in the first place.

Lifestyle, again, in one breath: lose weight, go easy on caffeine. He did not claim those reverse recurrent loss. He still says them.

Progesterone: who, when, and how he actually offers it

This was the longest argument of the night, because it is the drug GPs are already writing.

A trial run out of the Mater Hospital in Brisbane a couple of years ago, in recurrent miscarriage, was stopped early for futility. For recurrent miscarriage as a stand-alone indication, progesterone did not earn its keep in that analysis.

There is evidence in spotting or bleeding in early pregnancy — particularly if there have already been three losses. That is the PROMISE and PRISM story. On the forest plot he recalled, the confidence interval for the three-miscarriage group sat around 1.27 to 1.54 (he was recalling the numbers in the room; check the papers before you quote them as gospel). Once you fold in two losses and one loss, the interval moves closer to 1 and some guidelines have extended the offer to anyone with a previous miscarriage who is now bleeding. When you look at the true subgroup analyses, the signal that actually holds is the three previous losses group.

Does he still prescribe it? Yes — not as a default, and not frequently, but he will not stand on the subgroup table while someone is in tears.

His practical ladder in the room:

• Three previous losses plus bleeding this pregnancy — recommend. This is the group he will actually start.
• Fewer than three — offer / suggest. This is a thing that exists, that is sometimes used, that sometimes might make a difference. Not “I think you should be taking this.”
• Someone who has already started it, or who has arrived convinced after a friend or TikTok — he will educate, stay positive, and not pick a fight. He would definitely not say no, this is a bad thing.

IVF is a different planet. In an HRT cycle there is no corpus luteum keeping progesterone going, so they have to replace it. In practice they usually supplement even in a modified natural cycle.

Harms, cost, and the fancy sugar pill

Are we doing harm? He said yes, but the magnitude depends on how you look. There is a possible hypospadias signal — not massive. Most trials and reviews that declare progesterone harmless were designed to find benefit, not harm, so the “it doesn’t hurt” evidence is low quality. The flip side from IVF: super-therapeutic doses, and they are not seeing a nasty signal.

Cost: a lot of people tick the PBS restricted box. That box has been broadened, but a few years ago it was narrower and still being ticked — shifting a couple of hundred dollars from the patient onto the taxpayer for a treatment with thin utility in that person. It also teaches the room that a friend’s story should dictate therapy rather than trials.

The reason he still writes it more often than the forest plot alone would justify: stress itself increases miscarriage risk. If the person in front of you is terrified, has spotted this morning, and needs to feel they are doing something, a fancy sugar pill that also happens to be progesterone can give them a sense of control. He named older evidence that extra early-pregnancy scans made people feel better with no biological reason the extra pictures should change outcome. Same psychology. Different prescription.

Early scans, serum progesterone, and language

Extra weekly scans in early pregnancy have no biological reason to improve live-birth rates. People felt better when they had them. The catch: the gap between scans can be very stressful, and if you scan too early — thinking you are at eight weeks, actually at five, nothing much to see — you have just given them ten to fourteen days of extra fear before a scan that was always going to look like a five-week pregnancy. First cardiac activity is expected around seven weeks. That is when you can usually say, yes, I can see a heartbeat.

Do not test serum progesterone and then treat the number. Progesterone is high later because a developing pregnancy is telling the corpus luteum to keep producing it. If the pregnancy is failing, progesterone should be low. The low number is the consequence, not the cause. Luteal-phase deficiency exists as a concept; it is hard to diagnose cleanly, and supplementation in that frame is another maybe.

Language, if you remember one thing from the hour: offer, suggest, recommend, strongly recommend. Three losses: he is in recommend. Fewer than three: he is in offer/suggest. Meet the person where they are. Stay honest. Do not pretend a forest plot is a hug, and do not pretend a hug is a forest plot.

Take-home messages for clinic

1. Australian RPL is now two intrauterine losses before 20 weeks, not three consecutive. Ectopic does not count. Consecutive is not required. About 1–4% of couples.
2. Age dominates the raw numbers. About 10% in the early 20s, about 15–20% overall, about 60% in the mid-40s. Each prior loss adds about 10%. After three consecutive losses, next-pregnancy miscarriage is about 45% — which is still about 80% success.
3. Most sporadic losses are aneuploidy. Three in a row is less likely to be “just chromosomes” each time.
4. Four tests once they meet criteria: anticardiolipin (beta-2 GP1 optional), a filled-cavity study, thyroid, parental karyotype. Pick-up is low. Do not use them as a prediction panel in a newly pregnant couple.
5. True obstetric APS (the clinical syndrome, not a lonely antibody) is the drug-changing one: aspirin plus heparin, and hydroxychloroquine for the blood-pressure story.
6. Thyroid: treat overt hypo; grade the rest. Very high TSH he replaces. Target he named was heading down toward 3.
7. Progesterone: strongest for three previous losses plus bleeding this pregnancy. Mater recurrent-miscarriage trial stopped for futility. He still offers it more gently when fear is the presenting complaint. Do not treat a low serum progesterone as the cause of a failing pregnancy.
8. Scan timing: cardiac activity around seven weeks. Early extra scans reassure some people and frighten others. Time the first look so there is something to see.

Dr Kotha · Gold Coast · recurrent-pregnancy-loss.drkotha.com