Portal de Periódicos da CAPES

Twin‐to‐twin transfusion syndrome: time to forget the Quintero staging system?

2007; Wiley; Volume: 30; Issue: 7 Linguagem: Inglês

10.1002/uog.5221

1469-0705

Y. Ville,

Pregnancy and preeclampsia studies

Twin-to-twin transfusion syndrome (TTTS) affects around 10–15% of all monochorionic twin pregnancies, irrespective of the mode of conception. It is one of the most severe perinatal complications and the most time consuming disorder for fetal medicine specialists. It is also a unique example of a fetal disease that has been shown to benefit from fetal surgery in a randomized controlled trial1. Interest in this syndrome continues to increase, with over 700 papers published in the perinatal literature in the last 20 years, including more than one a week for the last 12 months. A consistently dismal outcome in cases left untreated has stimulated rapid and innovative development of therapeutic strategies, but many questions concerning the diagnosis and pathophysiology of the disease remain unanswered2. For the last 10 years, the Quintero staging system currently used to characterize various forms of the syndrome has been as good as it gets3. However, the system reflects an overly ambitious attempt to combine assessment of natural history with the prognosis of the disease within a single five-stage classification: Stage I: polyhydramnios in the recipient twin and severe oligohydramnios in the donor twin but with a visible bladder in the donor twin; Stage II: polyhydramnios in the recipient twin, stuck donor twin, bladder not visible at any point during the preoperative evaluation, diastolic flow present in the umbilical artery and forward flow in the ductus venosus; Stage III: poly/oligohydramnios, bladder in the donor twin not visible, and critically abnormal Doppler studies defined by at least one of the following: (1) absent or reversed end-diastolic velocity (AREDV) in the umbilical artery, (2) reversed flow in the ductus venosus during atrial contraction, and (3) pulsatile umbilical venous flow; patients with critically abnormal Doppler but with a visible bladder in the donor twin are also classified as Stage III either in the donor (III-D) or in the recipient (III-R); Stage IV: presence of ascites or hydrops (fluid collection in two or more cavities) in either fetus; Stage V: demise of either fetus. An attempt has also been made to combine the Quintero staging system with the identification of artery-to-artery anastomoses (AAA) on the chorionic plate. However, the lack of sensitivity and the poor reproducibility of Doppler identification of AAA, especially in cases with severe polyhydramnios in the second or early third trimester, have limited the clinical applicability of this approach4. Although Quintero's staging system has good specificity for recognizing the most severe forms of TTTS in Stages III and IV, it has led to the unproven postulates that: (1) Stage I can be considered a preclinical stage which does not require treatment apart from amniodrainage to alleviate polyhydramnios and (2) this arbitrary staging system reflects the natural progression of the disease. Ten years down the line, fetoscopic laser surgery has proved to be the best first-line treatment for TTTS, but current controversial issues are focusing on (1) the difference between TTTS and its preclinical stages in the presence of apparently isolated amniotic fluid imbalance; (2) the identification of prognostic factors prior to and following treatment, and (3) the choice of primary treatment, if treatment is justified, of the preclinical or mildest forms of the disease. The difference between Quintero Stages I and II is largely artificial and these stages are likely to reflect a continuum or even an overlap within the staging system, because visualization of the donor twin's bladder is often dependent upon the duration of observation. Indeed, despite the initial description of the staging method by Quintero et al.3, the fetal bladder is rarely subjected to continuous observation over at least an hour in order to characterize urine production. Furthermore, a visible bladder in the donor does not exclude fetal anuria, which can be determined by serial bladder volume measurements to assess urine production rate (UPR). This point is made by Yamamoto et al.5 who found that 48% (14/29) of cases classified as Stage I showed no significant urine production and that around 6% of Stage II (2/31) cases produced some urine in a 5–30-min observation period. Determining when a case is Stage I as opposed to it being preclinical (which could be characterized as Stage 0) is critical. A loose definition of polyhydramnios in TTTS as being present when the deepest vertical pool of amniotic fluid is > 8 cm is likely to include cases of simple discordance in addition to the cases of true polyhydramnios. Huber et al.6 demonstrated that up to only 14% of cases with mild polyhydramnios would progress to severe TTTS requiring fetoscopic laser surgery and that 12% would be controlled with one amniodrainage. The remaining cases were untreated and the overall survival was > 90%. The paper by O'Donoghue et al.7 illustrates the weakness of using this definition of polyhydramnios. Although not statistically significantly so, the prognosis seemed to worsen as the deepest vertical pool in the recipient sac increased, and oliguria in the donor, as determined by subjective assessment of the size of its bladder, seemed more likely. In order to avoid cases in which the diagnosis of TTTS could have been disputed, the randomized controlled trial by the Eurofetus group1 considered only a deepest vertical pool of at least 10 cm after 20 weeks, a definition that is more likely to comply with that of severe TTTS. It seems reasonable to propose that classification into Stages I and II should be abandoned and replaced by the identification of a single clinical form of the disease showing truly abnormal and increasing amniotic fluid volumes within the oliguria-polyuria sequence. This seems to be achievable by quantification of UPR in each fetus5. Cases with isolated mild amniotic fluid discrepancies in which the UPR in either or both fetuses remains normal should be considered as preclinical. These could be managed expectantly, with close monitoring by weekly sonography, including Doppler. Under these circumstances, 80% of cases would do well and timely treatment could be implemented should TTTS develop and be diagnosed during the weekly follow-up6. Bensouda et al. provide evidence that, among Stage-I and -II cases, echocardiography as well as thorough analysis of the ductus venosus flow pattern can differentiate between true TTTS and twins that are discordant in size and amniotic fluid volume8. The proposal of launching a randomized trial of TTTS cases currently labeled as Stage I is unlikely to alleviate the confusion surrounding the differentiation between preclinical forms and severe TTTS with large UPR differences. Numbering the stages from I to V has insidiously promoted the idea of a continuum between these stages. However only Stage V can legitimately claim its position as last, although this stage is of little clinical relevance in terms of treatment. Stages I and II are clearly intermingled and not successive. Stage III is also ambiguous, in that it can refer to two completely different conditions. Stage III in the donor twin (Stage III-D) is largely the result of placental insufficiency, with a growth-restricted fetus showing critically abnormal umbilical blood flow. It does not represent a progression from Stage I or Stage II cases and is likely to present from the late first or early second trimester. The ambiguity stems from the fact that Stage III can also describe a situation in which the donor has normal Doppler but the recipient shows absent or reversed a-wave in the ductus venosus (Stage III-R). This is a marker of hypervolemia and/or abnormal myocardial contractility, which can soon be followed by Stage IV, in which hydrops in the recipient is a crude marker of cardiac overload and insufficiency. This ambiguity is illustrated by the study of O'Donoghue et al.7, in which progression remains ill-defined but reflects marked worsening in the myocardial function of the recipient. Primary expectant management is likely to have allowed potentially irreversible damage to develop in these recipients, while fetoscopic laser surgery would have been more likely to stop the progression of the disease and to eventually cure it. It is also important to recognize and acknowledge the possibility of Stage I or II cases progressing to fetal demise of either or both twins in at least 5% of cases without showing any intermediate worsening stages, as shown by O'Donoghue et al.7 and by the Eurofoetus randomized controlled trial1. Most studies of fetal cardiac function in abnormal monochorionic twin pregnancies and especially in TTTS have shown that, although myocardial performance may be suboptimal in some severely growth-restricted donors, it is mainly the recipient that shows progressive deterioration of myocardial function. This can deteriorate up to the stage of overt cardiac failure or towards the development of morbid ventricular hypertrophy and acquired, potentially critical, pulmonary stenosis9, 10. It is also important to recognize that myocardial performance can be altered while fetal Doppler findings are normal, not only in the donor but also in the recipient, thus defining true severe Stage I or II TTTS11. Michelfelder et al. report that in Stage I and Stage II up to 50% of the recipient twins already show marked cardiac dysfunction12. In non-TTTS monochorionic pregnancies complicated by severe selective intrauterine growth restriction, cardiac dysfunction also predominantly affects the appropriate-for-gestational age cotwin, raising important issues for the management of these equally difficult cases13. It is well established that only up to 20% of TTTS cases managed with amnioreduction will be controlled solely by this modality, and that 80% will require serial amnioreduction. The latter carries an increased risk of miscarriage, very preterm delivery, ischemic or hemorrhagic brain lesions or advanced myocardial dysfunction, especially in recipient twins, when compared with cases treated by fetoscopic laser surgery14, 15. The American Journal of Obstetrics and Gynecology has recently announced the difficult delivery of a randomized, multicenter trial of amnioreduction vs. selective fetoscopic laser photocoagulation for the treatment of severe TTTS16. In this study, Crombleholme et al. reported on 40 cases, a figure not provided in the abstract, recruited over a 5-year period by 36 authors from 17 centers. The sole aim was to assess neonatal survival at 30 days. The methodology, which involved primary amnioreduction followed, in cases in which TTTS progressed, by randomization into serial amnioreduction or laser photocoagulation, might well explain the disappointing results obtained with laser surgery in this study. The authors reported a survival rate of 60% (24/40) following amnioreduction and 45% (18/40) following laser, a rate that is clearly worse than the 62–77% reported across the board with fetoscopic laser surgery1, 6, 17. This is an illustration of the difficulty and probably the inadequacy of performing fetoscopic laser surgery after primary amnioreduction. Tented or ruptured membranes as well as blood-stained amniotic fluid are frequent after amnioreduction and secondary laser procedures are even more complex. Amniotic fluid exchange is not always successful and lengthens the procedure, and amniopatch is frequently unsuccessful and may be associated with an increased risk of fetal brain ischemia related to the profound effect of platelet-serotonin release in the amniotic fluid18. The study did, however, confirm that one of the undesirable consequences of primary amnioreduction is to allow myocardial dysfunction in the recipient to progress to a stage that becomes critical for survival16. In the only randomized study that included sufficient cases to address the issue of survival after primary treatment with either laser surgery or amnioreduction, survival was higher (86 vs. 58%, P < 0.01) and morbidity in the survivors was lower following selective laser surgery at all stages (including 37 Stage I and 36 Stage II cases)1. A large series of fetoscopic laser procedures in a single center provided rates of survival of at least one twin of 93% and 83% in Stage I and Stage II fetuses, respectively, the equivalent rates for double survival being 75% and 61%19. It is disappointing that studies which persist in searching for some virtue in serial amnioreduction, such as those of O'Donoghue et al.7 and Crombleholme et al.16, provide only rudimentary outcomes confined to neonatal survival and do not report the morbidity in the survivors. One could argue that this would not add much to our current knowledge on this issue since the higher morbidity in survivors, especially single survivors, following amnioreduction compared with fetoscopic laser surgery, is well established, with generally accepted approximate rates of 20% and 10%, respectively1, 17. It has also been proved that among premature neonates of less than 34 weeks, those born after successful fetoscopic laser surgery share the prognosis of their matched dichorionic controls and that morbidity is significantly higher when fetoscopic laser surgery fails to control the syndrome and is even higher for those cases managed primarily by serial amnioreduction20. Following an almost 10-year service, for which due credit should be given, I believe it is time to retire the Quintero staging system. We should radically change our assessment of TTTS cases in order to achieve a more precise diagnosis of true and severe cases requiring treatment by fetoscopic laser surgery without delay. This optimal management could include objective assessment of the oliguria-polyuria sequence, but most importantly, a thorough examination of cardiac function in the recipient twin, for which fetal Doppler, even venous, is too poor a substitute. Survival and morbidity figures published in both successful and unsuccessful trials clearly advocate primary fetoscopic laser surgery to be offered in all severe TTTS cases and argue against the call for another randomized controlled trial comparing fetoscopic laser surgery with serial amnioreduction at any stage. Amnioreduction should probably only be offered in critical obstetric situations when facilities for fetoscopic laser surgery cannot be accessed or when semi-elective preterm delivery following lung maturation is an option, especially from 26 weeks onwards. Preclinical stages of TTTS should be submitted to heightened weekly surveillance and isolated and stable fluid imbalance should primarily be managed expectantly. Cases that manifest predominantly with severe selective intrauterine growth restriction and critically abnormal Doppler in the small twin represent a complication of monochorionic pregnancies that is even more complex than TTTS, especially because surgical management is likely to be impaired by a richer intertwin anastomotic vascular placental pattern and the lack of severe polyhydramnios. This is probably the only area in which there is an obvious role for a randomized trial to be conducted. The dismal reputation of complicated monochorionic pregnancies should no longer be an excuse for management strategies that are based on unsubstantiated hypotheses and whose effectiveness is unproven. There is logic, and probably even an ethical obligation, in promoting the development of specialized centers that are willing and able to put sufficient resources into the evaluation and management of complicated monochorionic pregnancies. Appropriate and timely referral can only result from adequate screening strategies, beginning with chorionicity determination in the first trimester and including targeted follow-up21, 22.