Exploring the use of metformin in pregnant women with PCOS: new evidence, new wisdom

메트포민의 유산, 조산 예방 효과에 대한 내용이며 가운데 논문 내용은 http://www.monews.co.kr/news/articleView.html?idxno=202098 을 참고 하였습니다. (PCOS 여성은 제2형 당뇨병 또는 IGT 발병 위험이 증가되어 있습니다). 기존의 PCOS 여성의 후향적 데이터(Endocr Pract. 2007;13(4):373.)는 메트포민이 당뇨병과 IGT으로 전환되는 것을 막거나 지연할 수 있음을 제시하였습니다. 그러나 기존 데이터를 고려하더라도, clinical trial data가 아니어서 정상 혈당 조절이 되는 여성에서 메트포민을 권고하지는 않았습니다. 이번 연구에서는 메트포민이 임신성 당뇨병 발생 예방 효과가 없는 것으로 나타났습니다.

​

다낭성 난소 증후군 (PCOS)을 가진 여성, 특히 안드로겐 과다가 있는 여성은 pre-eclampsia, 임신성 당뇨병 및 조기 분만을 포함하여 임신 합병증의 위험이 증가합니다. 이러한 합병증을 뒷받침하는 메커니즘은 잘 알려져 있지 않지만, 안드로겐 과잉과 인슐린 내성이 관련되어 있다고 여겨져 PCOS가 있는 임산부에서 metformin의 사용을 off-label로 사용됩니다(가능한 증거의 대부분이 무작위되지 않았고, 후향적인 것을 고려할 때 만연된 불확실성이 있습니다)

Women with polycystic ovary syndrome (PCOS), particularly those with androgen excess, are at increased risk for pregnancy complications, including pre-eclampsia, gestational diabetes, and preterm delivery. The mechanisms underpinning these complications are poorly understood, but androgen excess and insulin resistance are believed to be involved, prompting the off-label use of metformin in pregnant women with PCOS—a widespread practice of uncertain relevance, given the mostly non-randomised, retrospective nature of the available evidence.

​

Between 2000 and 2004, Eszter Vanky and colleagues did a randomised, pilot study of metformin treatment versus placebo in 40 pregnant women with PCOS in Norway, with changes in measures of androgen excess as the primary outcome. Between 2005 and 2009, the same group did a larger, multicentre, randomised, placebo-controlled trial (PregMet; n=273), with a primary outcome of gestational complications. In PregMet alone, metformin treatment did not significantly reduce gestational complications; however, a pooled individual participant data analysis of both studies led to the hypothesis that gestational metformin treatment in women with PCOS might reduce late miscarriage and preterm delivery.

​

2012년부터 2017년까지 노르웨이, 스웨덴, 아이슬란드의 14개 병원에서 18~45세 PCOS 환자 487명이 모집됐습니다. 이들은 메트포르민 복용군(메트포르민군, 244명)과 위약군(243명)에 1:1로 무작위 분류됐습니다. 메트포르민군은 임신 첫 주에 메트포르민 500mg 1일 2회 복용 후 용량을 증량해 출산까지 메트포르민 1000mg 1일 2회 투약했습니다. 1차 종료점은 임신 13주와 22주 6일 동안 후기 유산을 겪었거나 임신 23주와 36주 6일 사이에 조산한 경우로 정의했습니다. 2차 종료점으로 임신성 당뇨병, 자간전증, 임신성 고혈압, 신생아의 중환자실 입원 등을 평가했습니다. 치료의향분석(intention-to-treat) 분석 결과, 1차 종료점 발생률은 메트포르민군 5%(12명) 위약군 10%(23명)로 1차 종료점 발생 위험은 메트포르민군에서 50% 낮았습니다(OR 0.50; 95% CI 0.22~1.08; P=0.08). 다만 통계적인 유의성은 없어 메트포르민으로 후기 유산 또는 조산을 예방할 수 있다는 경향만 확인됐습니다. 그러나 PregMet2 연구와 앞서 진행된 예비연구, PregMet 연구 데이터를 통합분석한 결과, 메트포르민군에서 후기 유산 또는 조산 발생 위험이 57% 낮은 것으로 파악됐습니다(OR 0.43; 95% CI 0.23~0.79; P=0.004). 구체적인 후기 유산 또는 조산 발생률은 메트포르민군 5%(397명 중 18명), 위약군 10%(399명 중 40명)로 두 군간 2배가량 발생률 차이가 벌어졌습니다. 2차 종료점 분석에서는 메트포르민의 임신성 당뇨병 예방 효과가 나타나지 않았습니다. 임신 기간에 메트포르민을 복용하면 임신성 당뇨병 예방 효과가 있다는 보고가 있었으나(J Coll Physicians Surg Pak 2015;25(4):237-241), 이번 연구에서는 임신성 당뇨병 발생률은 메트포르민군 25%(60명), 위약군 24%(57명)로 비슷했습니다(OR 1.09; 95% CI 0.69~1.66; P=0.75). 아울러 심각한 이상반응은 산모와 신생아 모두 확인되지 않았습니다.

In The Lancet Diabetes & Endocrinology , Tone Løvvik and colleagues report the results of a larger randomised trial (PregMet2; n=487), done between 2012 and 2017 in Norway, Sweden, and Iceland. The composite incidence of late miscarriage and preterm birth (primary outcome) occurred in 23 (10%) of 240 women in the placebo group and 12 (5%) of 238 women in the metformin group (odds ratio [OR] 0·50, 95% CI 0·22–1·08; p=0·08). There were no significant differences between the groups for secondary endpoints, including gestational diabetes. However, in a post-hoc pooled analysis of individual participant data from of all three placebo-controlled studies (n=800), a favourable effect of metformin on the incidence of late miscarriage and preterm birth was seen (40 [10%] of 399 women in the placebo group vs 18 [5%] of 391 women in the metformin group; OR 0·43, 95% CI 0·23–0·79; p=0·004), yielding a number needed to treat of 18·4. This finding is a remarkable achievement, crowning almost two decades of relatively solitary exploration by this group of investigators.

​

Strengths of the PregMet2 study include the participation of 14 academic centres across three countries, the use of study medication (placebo or metformin) from the same manufacturer, the extensive masking approach used, the low number of patient withdrawals, and the overall commendable mean adherence to study medication (>80%). Study limitations include slow recruitment, which prolonged the study duration and led to a sample size smaller than had been planned. Inclusion of only hyperandrogenic PCOS phenotypes would have further undermined recruitment, although such an approach might have accelerated the attainment of statistical significance for the primary outcome.

​

The PregMet2 findings provide an obvious wealth of knowledge, but also a potential pearl of wisdom somewhat hidden in the study appendix. As part of their post-hoc individual participant data analysis of the three trials, the investigators looked at various subgroups, including the many women already receiving metformin at inclusion (n=222; 28% of the study population) versus those who were not—the result of this analysis might provide a key to unlock further progress. These data suggest that gestational metformin treatment reduced the incidence of late miscarriage and preterm delivery when given in the absence of pre-study metformin treatment (OR 0·33, 95% CI 0·15–0·68; p=0·001), but not in the presence of pre-study metformin treatment (1·00, 0·29–3·46; p=1·00). Admittedly, the absence or presence of such pre-study treatment was not randomly assigned, and this analysis is post-hoc, so this finding should be treated with caution. However, the apparent dependence of efficacy on pre-study metformin treatment, along with the finding that exclusive pre-study treatment seemed as effective as gestational treatment, suggests the possibility that metformin treatment before pregnancy might be an alternative to treatment during pregnancy. An exclusive preconception strategy would also obviate the need for the conceptus to be exposed to metformin, thereby removing any concerns about the potential lifelong sequelae of such early exposure.

​

PCOS is a post-pubertal central obesity syndrome, and preconception monotherapy with metformin is unlikely to lower the incidence of gestational diabetes in women with PCOS. During the years before conception, the risk of gestational diabetes can already be estimated by factors such as birthweight, age at menarche, overweight and obesity, liver adiposity, circulating adiponectin, and measures of insulin resistance. A new niche in preventive medicine should be dedicated specifically to improving the metabolic health of women who anticipate becoming pregnant. The focus of attention during the preconception year (or years) might include not only a normalisation of body composition with lifestyle intervention, but also the use of old and safe medications (under contraceptive protection) that lead to a preferential loss of excess hepato-visceral fat, and to improvements in insulin sensitivity, adiponectin concentration, and post-treatment ovulation rate. Since the intake of an oral oestrogen-progestogen contraceptive might not have a favourable effect on post-treatment ovulation rates and risk factors for gestational diabetes, other contraceptive measures are preferable.

​

When Roald Amundsen and his team departed in 1910 from Norway with the North Pole as an apparent destination, they were the first to reach somewhere equally important: the South Pole. A century later, the PregMet2 investigators launched—also from Norway—their consecutive explorations of metformin treatment in pregnant women with PCOS. Their original endpoints included androgen excess, pre-eclampsia, and gestational diabetes, but they reached an equally important target: a reduction in late miscarriages and preterm births. Hopefully the evidence and wisdom in their report will inspire further progress.

​

참고문헌 :

1. UpToDate 2019.04.21

2. Lancet Diabetes & Endocrinology, The, 2019-04-01, Volume 7, Issue 4, Pages 242-243

3. http://www.monews.co.kr/news/articleView.html?idxno=202098