From the 4th World Congress on Controversies in Obstetrics, Gynaecology and Infertility  

Head to Head: Nifedipine versus Atosiban

• View the interview (requires free Windows Media Player)

Hans van der Slikke, MD, PhD: “It’s April, 2003 and we are in Berlin at the Controversies in Obstetrics conference and next to me is Nick Fisk from London. Welcome Nick.”

Nick Fisk, MD: “Thank you.”

Hans van der Slikke, MD, PhD: “You did a wonderful presentation this afternoon about tocolytics, and you discussed all tocolytics, or 'meant-to-be' tocolytics, and after all there are now two reasonable options: nifedipine and atosiban. Could you discuss your preference for one or both?”

Nick Fisk, MD: “In Europe, where atosiban is licensed and nifedipine is not, the choice comes out to one of the two. I think it is now completely unacceptable to use betamimetic drugs anymore. They are clearly accepted as having a high side effect profile mainly in the mother but also in the baby. The newer drugs don’t have these problems. Atosiban is a drug that’s specific for uterine muscle and very little of it crosses the placenta so it has very few side effects. 

Nifedipine is a drug that affects all vasculature, in fact that’s why we use it in obstetrics for pre-eclampsia, etc. And so it still has some side effects: flushing, headache, tachycardia, but they’re not usually a major problem at least in well people. The other side of the equation in side effects is always the baby. There are no side effects with atosiban in the baby the only side effect of the atosiban in the mother is nausea in about 12%. 

In the foetus there is a concern with calcium antagonists that it might impair uterine perfusion and thus the amount of blood flow through to the baby and that this could have an effect on the baby’s pH and oxygenation. In the late 80s there were a whole lot of sheep and goat and monkey studies showing that nifedipine and or nicardipine, a similar drug, dropped the foetal pH. And this is very controversial. Some people argue that it’s due to the use of much higher doses than we use in humans, and that’s partially true. Or that it’s perhaps due to the vehicle, the alcohol in which the drug was put. And that’s probably partially true also, but there are some studies with low dose without a vehicle effect still suggesting that there is an influence on foetal pH. So there’s just some concern there that’s not yet resolved. I mean we can’t say that human studies don't appear to show any worrying effects. But the human studies were done with fairly crude tools, like doppler and foetal heart rate monitoring, etc. 

I guess this really boils down to one issue: we have one drug that has been thoroughly investigated to high standards as required to get drugs licensed in Europe versus, it’s really unfortunately so often the case in obstetrics in our field largely ignored by the pharmaceutical industry, that for the other we have small investigator-led studies that aren’t done with nearly the rigour that we’d expect if they were done by the pharmaceutical industry going for licensing. And that’s why we have these concerns about nifedipine.”

Hans van der Slikke, MD, PhD: “So there are contra-indications for the mothers who are hypertensive themselves, do you give them nifedipine as well?”

Nick Fisk, MD: “There’s an interesting paradox. I don’t use nifedipine for pre-term labour, because I prefer to use the licensed drug, which is going to be safe. But those who use nifedipine tend to be very cautious in mothers who are hypertensive. But it’s a paradox that they actually on the other hand use it as a treatment for hypertension in the mothers who are hypertensive, but not for pre-term labour. 

There are two other big things in terms of this balance of nifedipine versus atosiban. One is cost, which just directly follows whether one’s licensed or not. In Europe, euros and dollars are almost the same at about $360 for a 24-hour course of atosiban and nifedipine probably about $36, so a nine to tenfold difference. You can save on atosiban by only using it for a shorter period of time. But it is considerably more expensive because it’s a licensed drug reflecting all the work that goes into getting that status. 

The other thing is, and which should be the big question with tocolytics, how well they reduce the chance of the baby dying and reduce the chance of the baby getting complications in the new-born nursery. This has really been a very difficult one for the whole tocolytic field because we don’t have good proof from proper, blinded randomised placebo controlled trials that tocolytics have big effects. We know they definitely delay labour for one, two, and even seven days. The problem with all the studies that have been done is that people haven’t capitalised on that time interval. In the sort of two-day reprieve that we say you gain from tocolysis you should transfer the patient to a major unit. All the studies are done in major units so you aren’t going to show that. You should also give them steroids but many of these studies done up to 10 years ago didn’t give steroids at that time. And finally it may work at really early gestations, like 24, 25 weeks when survival increases by a couple of percent per day. But all these studies are done at 31, 32 weeks when you are unlikely to show that. 

There is a trend for betamimetics I think. What we call a Baysian trend: if you believe it works the data are consistent with it working. But is there a gold standard proof at p = 0.05 that they reduce perinatal morbidity and mortality? - Absolutely not. Atosiban appears to work as well as betamimetics in that regard. With nifedipine, the studies rather surprisingly show that they reduce respiratory distress syndrome and necrotising enterocolitis and inter-ventricular haemorrhage through a slight reduction in the number of babies delivering within two days. They deliver a few days later with nifedipine than with other drugs. And if that’s so, that’s terrific – because that’s a tocolytic that actually makes a big difference. 

It’s always very surprising where tocolytics versus placebo haven’t been proven to have an effect and someone comes and compares a new tocolytic to an old tocolytic and shows an effect. That should be much more difficult. We do need to remember with the calcium story that they have never, ever been tested against placebo. So, if you want to be a purist you can argue there are not a tocolytic but yet they seem better than current tocolytics without being a tocolytic. It’s all very bizarre. I’m not sure what we should conclude about the perinatal advantages of reduced morbidity with nifedipine. These studies weren’t brilliantly done. They were all small. They were not blinded, so the investigators knew which drug women were on. They often used very high doses of betamimetics, which upped their side effects. They analysed the outcome variables un-blinded. They had been put to a meta-analysis, which attempted to overcome some of the shortages, and the effects are significant, so it’s not very high quality data but the data there are suggest they do, or at least they may, reduce perinatal morbidity. So it’s all a bit of a mess in terms of the nifedipine story.”

Hans van der Slikke, MD, PhD: “First of all, talking about blindness, it’s almost impossible because as soon as you give ritodrine you know it is ritodrine because of the side effects.”

Nick Fisk, MD: “We certainly found that during the atosiban study, which is where we gave everyone atosiban and a dummy betamimetic, or dummy atosiban and a betamimetic as it had a big influence on how much you increased or reduced the dose. In particular not blinding a study makes a big difference to the assessment of the outcome variable. And some things like whether a baby had necrotising enterocolitis or not or has respiratory distress syndrome or not, if they’re assessed in a non-blind fashion, whether the investigators mean to or not, there is an incipient bias that can creep in.”

Hans van der Slikke, MD, PhD: “So, do you think there will be a chance to have, in the future, good enough studies to prove your point?”

Nick Fisk, MD: “I can’t see anyone doing a study of atosiban versus nifedipine easily. You would need huge numbers, because they’re both pretty good. So it’s going to be even more difficult; you’ll need bigger numbers than done in the other studies, and I certainly wouldn’t expect the pharmaceutical industry to do that: to compare a licensed with an unlicensed drug. 

I think the hope for the future comes in the development of even better tocolytics. Really specific drugs. They could be newer oxytocin antagonists, they could be interleukin inhibitors, and they could be hormonal support, various things. But any new drug then would need to be tested definitively against placebo in a huge study. And if you could show that the new drug was better than placebo that would be the first agent for which that’s proven to benefit the neonate and that would get around the very high standards that the FDA has set. In the USA today you don’t have any drug you can use that is licensed for pre-term labour.”

Hans van der Slikke, MD, PhD: “That’s right. In the United States they use a large magnesium sulphate.”

Nick Fisk, MD: “Now magnesium sulphate is arguably not a tocolytic…” 

Hans van der Slikke, MD, PhD: “…which doesn’t work.”

Nick Fisk, MD: “…because it doesn’t prevent pre-term labour, and it does increase the chance of the baby dying almost threefold and maternal side effects nine-fold. You won’t find anyone in Europe, and to my knowledge outside the US, using magnesium. It’s just not a tocolytic".

Hans van der Slikke, MD, PhD: “No. So we’re still waiting for the good solution and we are happy to have in Europe these possibilities to try to prevent pre-term labour.”

Nick Fisk, MD: “Yes.”

Hans van der Slikke, MD, PhD: “Thank you very much.