Cervical Ultrasound

Cervical Ultrasound

  

• Griff Jones, Darryl Maxwell (Division of Perinatology, Ottawa General Hospital and Fetal Medicine Unit Guy’s and St.Thomas’ Hospital, London)
  
  
  Cervical Ultrasound in Pregnancy
  
  
  Prog. Obstet & Gynaec, 14, p.80
  
        
  
  
  The uterine cervix undergoes considerable physiological, biochemical and anatomical changes during the transition between the antenatal and intrapartum periods. Digital examination consistently underestimated length by more than 13mm. In contrast, ultrasonographic measurements correlated well with those obtained using a ruler on the postoperative specimen.
  
          
  
  
  Transabdominal ultrasound
  
  The average cervical length to be between 35-40mm. Up to 6mm dilatation at the internal os was normal with no significant cervical change seen from the 10th to the 36th week of gestation.
  
       
  
  
  115 women with risk factors for cervical incompetence were studied. Patient management was largely based on clinical (as opposed to ultrasound) findings. Of the women with evidence of moderate cervical weakness on ultrasound, one-third were found to have clinically normal cervices. With expectant management, 75% delivered before 34 weeks’ gestation. Most of the remaining two-thirds, underwent cerclage with 25% delivering before 34 weeks.
  
         
  
  
  Riley et al report a 50% risk of preterm birth in women co-incidentally found to have cervical shortening and membrane funneling on transabdominal scanning.
  
      
  
  
  Transvaginal ultrasound:
  
  Brown et al suggest that adequate cervical images were more likely to be obtained by transvaginal scanning than by transabdominal U.S.G. The cervix could be visualized transvaginally 99% of the time. When transabdominal and transvaginal cervical length measurements from the same woman were compared, transabdominal measurements were on average 5mm longer, presumably secondary to bladder filling. Of the 30 women, 29 demonstrated an increase in cervical length with increasing bladder volume. Confino et al also reported cases in which quantified bladder filling or manual pressure exerted via the transducer led to significant reductions in internal os dilatation in patients with suspected incompetence.
  
         
  
  
  Digital examination only assesses that portion of the cervix below the vaginal wall. Goldberg et al found the mean cervical lengths in 43 pregnant women measured by transvaginal ultrasound (39mm) to be over twice as long as those measured digitally (19mm) by experienced examiners. Similarly, Sonek et al found that digital measurement of cervical length showed a poor correlation with vaginal ultrasonography.
  
        
  
  
  Safety and acceptability:
  
       
  
  
  Heath et al questioned 100 consecutive women about the acceptability of a transvaginal scan at 23 weeks gestation: 94% experienced no, or minimal, discomfort only and 98% reported no or only mild embarrassment. Half of the women found it to be less
  uncomfortable than a speculum examination.
  
         
  
  
  Normal values for cervical length
  
  There appears to be no clinically significant difference between the mean cervical lengths of primiparous and
  multiperous women.
  
         
  
  
  It is possible that ethnic differences in cervical length exist. The work of National Institute of Child Health and Human Development Maternal-Fetal Medicine Unit Network has shown a short cervix (£ 25mm) to be significantly more common in high risk women.
  
         
  
  
  DIAGNOSTIC ROLE
  
  Preterm Labour
  
    
  
  
  Murakawa et al studied 32 women presenting with threatened preterm labour. None of the 15 women with cervical lengths greater than 30 mm delivered preterm. If the cervix was less than 30mm, 65% delivered preterm and all early births were identified. All women with a cervical length below 20mm delivered preterm.
  
       
  
  
  Gomez et al developed the term cervical index to include information about endocervical length and funnel length in a single figure. Funneling was present in 58% of women and associated with a nearly 3-fold increase in the risk of preterm birth. A complex statistical analysis suggested that a cervical index ³0.52, cervical length < 18mm, funnel length > 9mm and funnel width > 6mm were all significantly associated with preterm birth. All the patients who delivered preterm had a funnel present.
  
        
  
  
  Cervical anatomy
  
  The ability to diagnose (or exclude) placenta praevia is one of the main benefits ascribed to antenatal ultrasound. Provided that the placental edge is at least 20mm from the internal os, a vaginal birth appears safe.
  
          
  
  
  Cervical incompetence
  
  Cervical conization may iatrogenically predispose to cervical incompetence.
  
      
  
  
  Management Role
  
  Assessment of risk of preterm birth
  
  Cervical length
  
        
  
  
  3000 women of mixed-risk were screened at both 24 and 28 weeks of gestation. The mean cervical length was approximately 35mm and the overall incidence of spontaneous preterm birth £ 35 weeks’ gestation was 4.3%. There was a clear inverse relationship between cervical length and risk of preterm birth. A short cervix (£ 25mm) was associated with an 8% probability of spontaneous preterm delivery £ 35 weeks gestation in low-risk parous women. In high-risk women, this probability climbed to 31% with the same cervical measurement.
  
           
  
  
  The second study was carried out at a single inner city centre in the UK. Results from cervical scans at 23 weeks’ gestation were blinded unless the length was less than 16mm. A cervical length of £ 15mm at 23 weeks’ gestation carried a risk of spontaneous delivery £ 32 weeks of 50% and correctly identified 58% of these births.
  
          
  
  
  The third study was from a well-defined geographical area in Helsinki, Findland with a 99% white population. The mean cervical length was just over 40mm and a short cervix was defined as one £ 29mm, corresponding to the 3rd centile.
  
          
  
  
  All three studies confirm the ability of transvaginal cervical ultrasound to reliably stratify women by risk of spontaneous preterm birth. Even more importantly, the background or pre-test risk of spontaneous preterm birth must be taken into account when interpreting cervical sonograms.
  
          
  
  
  Membrane funneling
  
  If the funnel length was greater than the length of closed cervix below it, 75% of patients delivered before 37 weeks’ gestation. A funnel width exceeding 15mm was another risk factor. Heath et al found that all women with a cervical length £ 15mm exhibited funneling compared to only one-third of those with a cervical length of 16-25mm.
  
    
  
  
  Dynamic changes.
  
  Sonek subsequently reported the same phenomenon in women not believed to have cervical incompetence, using transvaginal scanning. He stated that the funneling could be accentuated or brought on by gentle manual pressure on the uterine fundus and advocated this as an ‘internal os stress test’.
  
    
  
  
  At present, transfundal pressure cannot be standardised. More recently, reports have appeared advocating the use of a postural challenge, with the cervix being scanned first while the mother is supine and then when upright.
  
         
  
  
  Cervical cerclage:
  
  The final report of the MRC/RCOG Multicentre Randomised Trial of Cervical Cerclage suggested benefit in only 4% of cases with prior clinical uncertainty. Detailed inspection revealed that cerclage improved outcome only after three or more previous very early deliveries. The mean interval between cerclage removal and delivery was 2.3 weeks.
  
       
  
  
  Cerclage placement
  
  Both Andersen et al and Quinn reported all the McDonald sutures in their series to be in the middle third of the cervix. The optimal placement of a suture was obtained in the single patient that was treated with a Shirodkar-type cerclage, involving preliminary dissection of the anterior vaginal wall.
  
        
  
  
  Post cerclage follow up
  
  Between 25-30% of patients developed funneling above the cerclage.
  
           
  
  
  Twin pregnancies
  
  Goldenberg et al scanned 147 twin pregnancies at 24 and 28 weeks’ gestation. A cervical length of less than 25mm was twice as common in twin pregnancies as in singletons and became more common as gestation advanced, occurring in 18% of women at 24 weeks and 33% at 28 weeks.
  
      
  
  
  Cervical ultrasound and routine antenatal care
  
  Although Zalar suggested that knowledge of transvaginal cervical ultrasound measurements can lead to a reduction in spontaneous low birth-weight deliveries, it cannot yet be concluded that cervical ultrasound has a place in routine antenatal care in low-risk pregnancies.
  
      
  
  
  Preterm prelabour rupture of the membranes
  
  The authors were unable to demonstrate a significant relationship between cervical length and the number of days to spontaneous labour.
  
       
  
  
  TECHNIQUES FOR CERVICAL SCANNING
  
  Transabdominal
  
  Cervical images are best obtained transabdominally with a full bladder. Unfortunately, this is associated with artificial lengthening of the cervix and potential closure of a dilated internal os. Therefore, it will lead to false reassurance in some cases. The external os can also be difficult to identify transabdominally. In the studies of Varma et al, an inflated Foley catheter balloon was placed against it to overcome this problem.
  
       
  
  
  Transvaginal
  
  This remains the gold standard for cervical imaging. Sonek et al have advocated the use of a probe with a 240° scanning field, this would appear to be unnecessary. Pressure on the cervix can falsely increase the measured length and obscure funneling at the internal os. Failing to appreciate that not all cervical canals follow a straight line can lead to underestimates of length. At least 3 measurements should be taken and the shortest (not the average) used. Most authors advise scanning over approximately 5 min to detect dynamic changes. The presence of a funnel should be noted and funnel width and length recorded. Some authors have reserved the term ‘funnel’ for membrane protrusions greater than 5mm down the endocervical canal, referring to anything less as ‘nippling’.
  
          
  
  
  Transperineal scanning
  
  A glove-covered 3.5MHz or 5 MHz sector or curvilinear transducer is applied to the perineum to visualize the cervix. It is reportedly easily tolerated by patients, but is best performed before digital or speculum examination which, by introducing air into the vagina, produce artefact. The technique has been particularly useful in cases of prolapsing membranes when cerclage is not being undertaken. The contrast provided by the funnel or membranes usually allows excellent visualization and serial monitoring. Transperineal ultrasonography may prove to be an acceptable mass screening technique.
  
      
  
  
  CONCLUSIONS:
  
       
  
  
  As proposed by Iams et al, ultrasound imaging suggests cervical competence to be a continuous variable. Infection has attracted considerable interest as one of the most important aetiologies behind preterm labour and delivery. A weak or short cervix could offer less resistance to ascending infection. Half have positive amniotic fluid cultures. 
  
      
  
  
  Iams et al have recently reported that a cervical length of less than 25mm at 24 weeks’ gestation is strongly associated with subsequent perinatal infection. Bacteria can ascend the genital tract attached to motile sperm. Antibiotics may play just as important a role as cervical cerclage in preventing prematurity. As prostaglandins are implicated in the process of cervical ripening, non-steroidal anti-inflammatories such as indomethacin may have role for ultrasonographic cervical change, at least before 28 weeks’ gestation. Cervical ultrasound undoubtedly allows a far better assessment of risk than clinical examination.