Predição do parto pré-termo em gestantes com colo curto sob utilização de pessário cervical associado à progesterona vaginal
Data
2022-07-04
Autores
França, Marcelo Santucci 
Orientadores
Moron, Antonio Fernandes
Tipo
Tese de doutorado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Objetivo: Identificar os principais fatores relevantes no determinismo do nascimento prematuro (< 34 semanas) em gestantes com indicação de uso do pessário cervical por colo curto, utilizando-se variáveis demográficas, história obstétrica e ultrassonografia transvaginal do colo (n=475), objetivando-se um novo rastreamento do parto pré-termo. Os objetivos secundários são: (i) validação do rastreamento criado, utilizando o Grupo Progesterona (P) (n=461) com alvo no parto pré-termo < 34 semanas e < 28 semanas (partos e óbitos perinatais), além de uma segunda validação, pela detecção da prematuridade apenas pelo comprimento do colo < 15mm no Grupo Pessário+progesterona (PP) (n=475); (ii) comparar o risco de parto pré-termo < 34, < 32, < 28 semanas entre os grupos randomizados de acordo com o comprimento do colo uterino (n=936); (iii) comparar o risco de óbito perinatal entre os grupos randomizados (n=936). Método: Trata-se de análise post hoc de ensaio clínico randomizado, multicêntrico (Estudo P5), que comparou o uso do pessário associado à progesterona micronizada com o uso isolado da progesterona micronizada vaginal (200mg/dia) em gestantes com comprimento do colo ≤ 30 mm (n=936). Os resultados do Estudo P5 foram revisitados em diferentes comprimentos cervicais (≤ 15, ≤ 20, ≤ 22, ≤ 25 e ≤ 30 mm) em diferentes idades gestacionais de nascimento prematuro (< 28, < 32, < 34 semanas) e óbito perinatal (associado a prematuridade) pelo teste do Qui-quadrado (Odds Ratio e IC95%) e pelas curvas de sobrevida de Kaplan-Meier (n= 936). Procurou-se identificar os casos de maior risco de parto pré-termo no Grupo PP, para tanto, foram utilizadas técnicas de Regressão Logística. Foram obtidas as variáveis que melhor discriminam o parto pré-termo < 34 semanas, idealizando-se uma nova forma de rastreamento do parto pré-termo < 34 semanas em gestantes com colo curto (≤ 30 mm) usuárias de pessário + progesterona. A análise de curva ROC, área sob a curva, cálculo de sensibilidade, especificidade, VPP e VPN para todos os subgrupos foi obtida. A validação dos dados obtidos pela Regressão Logística, foi obtida pelo teste de rastreamento do parto prematuro < 34 semanas e < 28 semanas (Partos e óbitos perinatais) no Grupo P (n=461) e pela comparação com a detecção de prematuros < 34 semanas apenas pelo colo < 15 mm no Grupo PP (n=475). Resultados: O uso do pessário + progesterona em gestantes com comprimento do colo ≤ 25mm, reduziu o risco de parto pré-termo < 34 semanas (OR 0,217), < 32 semanas (OR 0,192) e < 28 semanas (OR 0,129). Também foi encontrada redução do risco de óbito perinatal no Grupo PP (OR 0,379). A Regressão Logística selecionou: cor de pele branca (OR 2,532), ausência de curetagem anterior (OR 0,113); antecedente de parto prematuro < 37 semanas (OR 3,647); gestação única (OR 0,135); idade gestacional no momento do ultrassom < 19 semanas (OR 3,373); comprimento longitudinal do colo em linha reta entre 5,2 e 14,2 mm (OR 4,072); comprimento longitudinal do colo em curva > 21 mm (OR 0,216). O cálculo da área sob a curva para Regressão Logística no Grupo PP foi de 0,978; para gestantes com colo < 15 mm no Grupo PP, de 0,311; para o Grupo P (alvo nos prematuros < 34 semanas), de 0,694 e para o Grupo P (alvo nos prematuros e óbitos < 28 semanas) de 0,765. Considerando o desfecho como parto pré-termo, a sensibilidade, a especificidade, o VPP e o VPN foram, respectivamente, para uma taxa de falso positivo de 20%, de 93,75%, 88,76%, 48,39% e 99,21% para a Regressão Logística no Grupo PP; de 50,00%, 83,37%, 25,26% e 93,68% para gestantes com colo < 15 mm no Grupo PP; de 43,06%, 84,32%, 33,70%, e 88,89% para o Grupo P (alvo nos prematuros < 34 semanas); de 54,84%, 82,56%, 18,48% e 96,21% para o Grupo P (alvo nos prematuros e óbitos perinatais < 28 semanas); e de 78,48%, 85,53%, 33,33% e 97,73% para a junção do Grupos PP com alvo no parto pré-termo < 34 semanas, associado ao Grupo P com alvo no parto pré-termo < 28 semanas (partos e óbitos). Conclusões: A Regressão Logística selecionou variáveis (etnia, prematuridade anterior, ausência de curetagem anterior, gestação única, idade gestacional na inclusão e comprimento do colo) como preditoras do nascimento prematuro em gestantes usuárias de pessário + progesterona com comprimento do colo ≤ 30mm e (i) foi validada pelo teste no Grupo Progesterona com alvo nos prematuros < 28 semanas (parto e óbitos perinatais), assim pode-se dizer que essa RL tem potencial a se destinar aos pacientes com colo curto (≤ 30 mm), como um rastreamento para os casos de prematuridade mais graves. (ii) A associação do pessário + progesterona reduz a chance de parto pré-termo < 34, < 32 e < 28 semanas, (iii) bem como o risco de óbito perinatal.
Objective: Identify the principal prognostic factors to determine preterm birth (< 34 weeks) in gestation with pessary by the shorth cervix. It was considered demographic aspects, obstetric history, and transvaginal ultrasound (n= 475) developing a new screening of preterm birth in pessary users. The secondary objectives were: (i) validation of the screening in Progesterone Group (n=461) with target in preterm birth < 34 weeks and < 28 weeks (birth and perinatal death), further a second validation by detection of preterm birth obtained by cervical length < 15 mm in Pessary group ( n= 475); (ii) compare the risk of preterm birth < 34, < 32, < 28 weeks between randomized groups according to cervical length; (iii) and the risk of perinatal death between randomized groups (n=936). Methods: This study is an analysis post hoc of a multicentric randomized controlled clinical trial with 17 sites in Brazil (P5 Study). Asymptomatic pregnant women, with cervical length ≤ 30 mm, were randomized to pessary plus progesterone (200 mg/d) or progesterone only (200 mg/d) until 36th weeks or delivery (n=936). The results of P5 Study were revisited in different cervical length (≤ 15, ≤ 20, ≤ 22, ≤ 25 e ≤ 30 mm) in different gestational age of preterm birth (< 28, < 32, < 34 weeks) and perinatal death by Chi-square test (Odds Ratio and CI95%) and by curves of survival analysis of Kaplan-Meier (n=936). The risk of preterm birth in pessary plus progesterone group was determined by techniques of Logistic regression, the best variables were obtained with target to discriminated preterm birth < 34 weeks, devising a new screening to preterm birth in gestation with short cervix (≤ 30 mm). It was performed the ROC curve, the area under curve, sensitivity, specificity, PPV and PNV for all subgroups. The validation was obtained by application of logistic regression over Progesterone group (n= 461) with target in preterm birth < 34 weeks and < 28 weeks (birth and perinatal death), and by comparison with detection of preterm birth < 34 weeks by cervical length < 15 mm in pessary plus progesterone group (n=475). Results: The use of pessary plus progesterone reduced the risk of preterm birth < 34 weeks (OR 0.217), < 32 weeks (OR 0.192) e < 28 weeks (OR 0.129). It was observed reduction in perinatal death between groups (OR 0.379). The logistic regression selected: ethnics white (OR 2.532); absence of previous curettage (OR 0.113); previous preterm birth < 37 weeks (OR 3.647); singleton gestation (OR 0.135); gestational age at ultrasound < 19 weeks (OR 3.373); cervical length straight measure < 14.2 mm (OR 4.072); cervical length curve measure > 21 mm (OR 0.216). The area under curve was 0.978 for logistic regression; for gestation with cervical length < 15 mm was 0.311. The area under curve for Progesterone group after application of logistic regression with target in preterm birth < 34 weeks and < 28 weeks was, respectively, 0.694 and 0.765. The sensitivity, specificity, PPV and PNV were respectively, 93.75%, 88.76%, 48.39% and 99.21% for logistic regression to PP Group < 34 weeks; 50.00%, 83.37%, 25.26% and 93.68% for pregnant women with cervical length < 15 mm; 43.06%, 84.32%, 33.70%, and 88.89% for Progesterone group with target in preterm birth < 34 weeks; and 54.84%, 82.56%, 18.48% and 96.21% for Progesterone group with target in preterm birth < 28 weeks (births and perinatal death); and 78.48%, 85.53%, 33.33% and 97.73% for addiction of the logistic regression in PP Group with target preterm birth < 34 weeks plus Logistic regression applied to P Group with target preterm birth < 28 weeks (birth and perinatal death), after fixed 20% of false positive rate. Conclusion: Logistic regression selected variables (ethnics, absence of previous curettage, gestational age at inclusion, previous preterm birth, singleton gestation, and cervical length) as predictors of preterm birth < 34 weeks in pessary plus progesterone group. (i) The validation in Progesterone group with target in preterm birth < 28 weeks (births and perinatal deaths) was performed and is possible assume that LR has a potential to be a predictor of preterm birth in extremely and very preterm birth in gestation with short cervix (≤ 30 mm). (ii) The association pessary plus progesterone reduces the risk of preterm birth < 34, < 32, < 28 weeks, (iii) as well perinatal death.
Objective: Identify the principal prognostic factors to determine preterm birth (< 34 weeks) in gestation with pessary by the shorth cervix. It was considered demographic aspects, obstetric history, and transvaginal ultrasound (n= 475) developing a new screening of preterm birth in pessary users. The secondary objectives were: (i) validation of the screening in Progesterone Group (n=461) with target in preterm birth < 34 weeks and < 28 weeks (birth and perinatal death), further a second validation by detection of preterm birth obtained by cervical length < 15 mm in Pessary group ( n= 475); (ii) compare the risk of preterm birth < 34, < 32, < 28 weeks between randomized groups according to cervical length; (iii) and the risk of perinatal death between randomized groups (n=936). Methods: This study is an analysis post hoc of a multicentric randomized controlled clinical trial with 17 sites in Brazil (P5 Study). Asymptomatic pregnant women, with cervical length ≤ 30 mm, were randomized to pessary plus progesterone (200 mg/d) or progesterone only (200 mg/d) until 36th weeks or delivery (n=936). The results of P5 Study were revisited in different cervical length (≤ 15, ≤ 20, ≤ 22, ≤ 25 e ≤ 30 mm) in different gestational age of preterm birth (< 28, < 32, < 34 weeks) and perinatal death by Chi-square test (Odds Ratio and CI95%) and by curves of survival analysis of Kaplan-Meier (n=936). The risk of preterm birth in pessary plus progesterone group was determined by techniques of Logistic regression, the best variables were obtained with target to discriminated preterm birth < 34 weeks, devising a new screening to preterm birth in gestation with short cervix (≤ 30 mm). It was performed the ROC curve, the area under curve, sensitivity, specificity, PPV and PNV for all subgroups. The validation was obtained by application of logistic regression over Progesterone group (n= 461) with target in preterm birth < 34 weeks and < 28 weeks (birth and perinatal death), and by comparison with detection of preterm birth < 34 weeks by cervical length < 15 mm in pessary plus progesterone group (n=475). Results: The use of pessary plus progesterone reduced the risk of preterm birth < 34 weeks (OR 0.217), < 32 weeks (OR 0.192) e < 28 weeks (OR 0.129). It was observed reduction in perinatal death between groups (OR 0.379). The logistic regression selected: ethnics white (OR 2.532); absence of previous curettage (OR 0.113); previous preterm birth < 37 weeks (OR 3.647); singleton gestation (OR 0.135); gestational age at ultrasound < 19 weeks (OR 3.373); cervical length straight measure < 14.2 mm (OR 4.072); cervical length curve measure > 21 mm (OR 0.216). The area under curve was 0.978 for logistic regression; for gestation with cervical length < 15 mm was 0.311. The area under curve for Progesterone group after application of logistic regression with target in preterm birth < 34 weeks and < 28 weeks was, respectively, 0.694 and 0.765. The sensitivity, specificity, PPV and PNV were respectively, 93.75%, 88.76%, 48.39% and 99.21% for logistic regression to PP Group < 34 weeks; 50.00%, 83.37%, 25.26% and 93.68% for pregnant women with cervical length < 15 mm; 43.06%, 84.32%, 33.70%, and 88.89% for Progesterone group with target in preterm birth < 34 weeks; and 54.84%, 82.56%, 18.48% and 96.21% for Progesterone group with target in preterm birth < 28 weeks (births and perinatal death); and 78.48%, 85.53%, 33.33% and 97.73% for addiction of the logistic regression in PP Group with target preterm birth < 34 weeks plus Logistic regression applied to P Group with target preterm birth < 28 weeks (birth and perinatal death), after fixed 20% of false positive rate. Conclusion: Logistic regression selected variables (ethnics, absence of previous curettage, gestational age at inclusion, previous preterm birth, singleton gestation, and cervical length) as predictors of preterm birth < 34 weeks in pessary plus progesterone group. (i) The validation in Progesterone group with target in preterm birth < 28 weeks (births and perinatal deaths) was performed and is possible assume that LR has a potential to be a predictor of preterm birth in extremely and very preterm birth in gestation with short cervix (≤ 30 mm). (ii) The association pessary plus progesterone reduces the risk of preterm birth < 34, < 32, < 28 weeks, (iii) as well perinatal death.
Descrição
Citação
FRANÇA, M S. Predição do parto pré-termo em gestantes com colo curto sob utilização de pessário cervical associado à progesterona vaginal. São Paulo, 2022. 273 f. Tese (Doutorado em Obstetrícia) - Escola Paulista de Medicina (EPM), Universidade Federal de São Paulo (UNIFESP). São Paulo, 2022.