Documentos de Académico
Documentos de Profesional
Documentos de Cultura
PARTE I:
FISIOLOGÍA DEL SISTEMA RESPIRATORIO
Laringe
El revestimiento interno de la laringe es de origen endodérmico, mientras que los músculos y los
cartílagos provienen del mesénquima de los arcos branquiales. Así, como consecuencia de la
proliferación del mesénquima, el orificio faríngeo adquiere una forma de T (Fig 3A). Posteriormente,
el mesénquima de los dos arcos se transforma en los cartílagos tiroides, cricoides y aritenoides,
identificándose la forma característica del orificio faríngeo en el adulto (Fig 3B). En paralelo, el epitelio
laríngeo se vacuoliza y recanaliza, formando los ventrículos laríngeos que se diferenciarán hacia las
cuerdas vocales falsas y verdaderas. Los músculos de la laringe derivados del cuarto y quinto arco
branquial están inervados por ramos del décimo para craneal (NC X), el nervio vago; el laríngeo superior,
los derivados del cuarto arco branquial; y el recurrente, los del sexto arco.
Maduración pulmonar
Hasta el séptimo mes de desarrollo intrauterino, los bronquiolos se dividen continuamente en
conductos cada vez más pequeños (fase canalicular) y su vascularización aumenta constantemente.
Algunas células de los bronquiolos respiratorios se transforman en células delgadas y planas que
hacen posible la respiración. Estas células están en relación estrecha con los capilares sanguíneos y
linfáticos. Los espacios delimitados por ellas se llaman sacos terminales o alvéolos primitivos. En el
séptimo más ya existen suficientes capilares para el intercambio gaseoso, que permitirían la
supervivencia de un infante prematuro.
En los últimos meses de vida intrauterina y durante algunos años después del nacimiento,
aumenta constantemente el número de sacos terminales. Las células de revestimiento se denominan
células epiteliales alveolares tipo I (neumocitos tipo I) (Fig 7). El contacto íntimo entre los neumocitos tipo
I y los capilares generan la barrera alvéolo-capilar. Antes del nacimiento no existen alvéolos maduros
(Fig 8). En el sexto mes aparece otro tipo celular, denominadas células epiteliales alveolares tipo II
(neumocitos tipo II) responsables de la producción de surfactante pulmonar, que disminuye las fuerzas
de tensión superficial que tienden a colapsar los alvéolos (atelectasia). En la actualidad se acepta que
después de la etapa de desarrollo pulmonar, los neumocitos tipo II son las células precursoras de los
neumocitos tipo I.
Figura 8. Tejido pulmonar del infante. Células epiteliales escamosas delgadas (neumocitos
tipo I) y capilares circundantes. (Sadler, 1996)
Antes del nacimiento, los pulmones tienen líquido rico en cloruro, pocas proteínas y algo de
moco proveniente de las glándulas bronquiales, así como de surfactante pulmonar. El volumen de
surfactante va aumentando a medida que se aproxima el nacimiento. Cuando se inicia la respiración,
la mayor parte del líquido que ocupaba los pulmones es reabsorbido por los capilares sanguíneos y
linfáticos, mientras que una pequeña parte es eliminada por los bronquios y tráquea en el parto. El
surfactante permanece como una delgada capa de fosfolípido sobre las membranas de las células
alveolares (Fig 9A).
El crecimiento del pulmón continúa después del nacimiento, aunque se desconoce con
exactitud el período en que la formación de nuevos alvéolos termina (alrededor de los 7-8 años).
Estructura y función
Los pulmones son fundamentales para el intercambio de gases. Su función principal es permitir el
movimiento del oxígeno (O2) desde el aire a la sangre venosa, y sacar el dióxido de carbono (CO2). Los
pulmones tienen otras funciones, como metabolizar algunos compuestos, filtrar materiales no
deseados desde la circulación y actuar como un reservorio de sangre. Sin embargo, su función esencial
es el intercambio gaseoso.
El aparato respiratorio de los niños posee características especiales derivadas, en gran medida,
de la adaptación al canal del parto estrecho y a la posterior velocidad de crecimiento. Durante la vida
fetal, el intercambio gaseoso que determina la eliminación de CO2 y la incorporación de O2 al torrente
sanguíneo ocurre en la placenta, por medio de unos sinusoides que existen entre los espacios
intervellosos, en los que se vierte la sangre bien oxigenada proveniente de la madre, a través de las
arterias espirales. La sangre arterial materna no se mezcla con la sangre venosa del feto. La
transferencia ocurre por gradiente de presiones parciales de los gases de la sangre fetal y materna
(Tabla I)
Barrera alvéolo-capilar
El O2 y CO2 se mueven entre el aire y la sangre por difusión simple, esto es, desde un área de alta
presión parcial a uno de baja de acuerdo a lo descrito por la ley de Fick, que establece que la cantidad
de gas (flujo) que se mueve a través de la barrera es directamente proporcional al área de la barrera e
inversamente proporcional a su grosor. La barrera alveolo-capilar es extremadamente delgada y tiene
un área de entre 50 a 100 m2.
¿Cómo es posible obtener una superficie tan grande para la difusión, dentro de una cavidad
torácica limitada? Esto se obtiene envolviendo los vasos sanguíneos pequeños (capilares) alrededor de
un número grande de pequeños sacos aéreos tan grandes llamados alvéolos. Existen 300 millones de
alvéolos en el pulmón humano, cada uno de aproximadamente 200-300 m de diámetro. Al nacer son
mucho menos numerosos y más pequeños (50 m). Si ellos fuesen una esfera su superficie total sería
de aproximadamente 85 m2, con un volumen de aproximadamente 4 L (una única espera de este
volumen tendría una superficie interna de solo 1/100 m2).
Los gases llegan al lado alveolar de la barrera a través de las vías aéreas, mientras que la sangre
llega al otro lado a través de los vasos sanguíneos.
Surfactante pulmonar
Una de las funciones más importantes de los neumocitos tipo II es la síntesis de y secreción de
surfactante. En el embrión humano, el surfactante aparece alrededor del 60 % de la gestación. El
surfactante pulmonar está compuesto principalmente por una combinación de fosfolípidos y
proteínas, y una pequeña fracción de colesterol. El fosfolípido más abundante es la fosfatidilcolina (FC,
Fig 9B), que posee dos cadenas de ácido palmítico (16 carbones), un ácido graso saturado. Esto le
confiere gran hidrofobicidad, y permite disminuir la tensión superficial en la interfase aire-líquido de
la superficie alveolar (Fig 9A). Además, presenta otros fosfolípidos, como el fosfatidilglicerol,
fosfatidiletanolamina y esfingomielina.
El surfactante pulmonar tiene, también, un 10% de proteínas. Proteínas específicas del
surfactante, denominadas apoproteínas de surfactante SP-A, -B, -C y -D. La SP-A es la más abundante,
interactúa con un receptor específico del neumocito tipo II y, además, posee una cola tipo-colágeno
que le permite asociarse a otras SP-A para formar un trímero (Fig 9C). Seis trímeros, forman la
estructura molecular definitiva presente en el surfactante. La SP-A participa en la formación de la
mielina tubular, estructura tridimensional de surfactante precursora de la capa mono-molecular de
fosfolípidos de la interfase aire-líquido del alvéolo. Además, tiene una función inmune, que facilita la
respuesta a infecciones pulmonares.
La SP-B, de menor peso molecular que la -A, se presenta de forma monomérica o dimérica.
Rica en residuos aminoacídicos sin carga, es extremadamente hidrofóbica, por lo que ayuda a
mantener la capa mono-molecular de fosfolípidos durante los ciclos de ventilatorios. La SP-C es más
pequeña e hidrofóbica. Es la proteína de expresión más temprana en el pulmón fetal, y participa en la
mantención de la capa mono-molecular de fosfolípidos. La SP-D, estructuralmente bastante semejante
a la SP-A. NO participa directamente en la disminución de la tensión superficial; tiene una función
principalmente inmune.
Durante la vida fetal, todos los componentes del surfactante pulmonar son sintetizados por el
neumocito tipo II, a partir de glucosa, colina y ácidos grasos. Los componentes se sintetizan en
distintas partes de la célula y se ensamblan en los cuerpos lamelares. Estos se secretan, mediante
exocitosis, hacia el lumen del alvéolo donde forman la mielina tubular, estructura precursora de la
capa mono-molecular de fosfolípidos de la interfase aire-líquido. Este proceso requiere la presencia de
Ca2+ y de las proteínas SP-A y -B (Fig 14).
Figura 14. Síntesis del surfactante pulmonar. (Boron & Boulpaep, 2004)
Vasos sanguíneos
Los vasos sanguíneos pulmonares forman una serie de tubos ramificados a partir de la arteria
pulmonar hasta los capilares, y luego la vena pulmonar. Inicialmente, las arterias, venas y bronquios
corren juntos, pero hacia la periferia del pulmón, las venas se mueven y pasan entre los lóbulos,
mientras que las arterias y bronquios permanecen juntos hacia el centro del lóbulo. Los capilares
forman una densa red en las paredes del alvéolo (Fig 15).
Flujo sanguíneo
La circulación pulmonar comienza en la arteria pulmonar, que recibe sangre desde el ventrículo
derecho. Esta arteria se divide en ramas, así como la vía aérea, y de hecho, acompaña a la vía aérea
hasta los bronquiolos terminales. Más allá, se genera el lecho capilar pulmonar, una densa red capilar
asociada a la pared alveolar que genera una estructura altamente eficiente para el intercambio
gaseoso. La sangre oxigenada se recolecta desde el lecho capilar hacia las venas pulmonares pequeñas,
que corren entre los lóbulos, para constituir las venas pulmonares, y retornar la sangre hacia la
aurícula izquierda.
Figura 18. Resistencia vascular en la circulación pulmonar. Cambios en la resistencia vascular pulmonar y en el flujo
sanguíneo pulmonar provocados por aumentos en la presión arterial pulmonar, y dependencia de la resistencia vascular
pulmonar y los volúmenes pulmonares. (Boron & Boulpaep, 2004)
Relaciones presión-volumen
Supongamos que removemos los pulmones de un animal de experimentación, los canulamos a través
de la tráquea y lo ponemos en un recipiente como el que muestra la Figura 23. Cuando la presión en el
recipiente es menor que la atmosférica, el pulmón se expande y es posible determinar su volumen con
un espirómetro (ver más adelante).
Aquí, la presión de expansión se genera con una bomba, pero en el ser humano, esta
disminución de presión se genera por la acción de la pared torácica, a través del espacio intrapleural
(IP), un espacio que posee sólo unos pocos mL de fluido.
Las curvas de inflación y deflación son distintas (histeresis). El volumen durante la deflación es
mayor que durante la inflación. Además, el pulmón posee un volumen mínimo de aire previo al
cambio de presión.
La diferencia de presión entre el interior y exterior del pulmón se llama presión transpulmonar,
que es igual a la presión que rodea el pulmón cuando la presión alveolar (PA) es igual a la presión
atmosférica.
Figura 23. Medición de la curva presión-volumen en un
pulmón aislado. El pulmón se mantiene por algunos
segundos a cada presión indicada, mientras se mide su
volumen. (West, 2005)
Complacencia
La pendiente de la gráfica de la Fig 23, es decir el cambio de volumen (V) por unidad de cambio de
presión (P) se conoce como complacencia. Dentro de rangos normales (presión de expansión de entre -
5 y -10 cm H2O) el pulmón es marcadamente distensible, es decir, muy complaciente. En el ser
humano este valor es de aproximadamente 200 mL/cm H2O.
Aumentos en el tejido fibroso provocan una disminución de la complacencia (fibrosis
pulmonar). Además, disminuye por el edema alveolar y cuando el pulmón permanece sin ventilación
por mucho rato. La complacencia puede aumentar en el enfisema pulmonar y como consecuencia del
envejecimiento, debido a las alteraciones del tejido elástico.
Tensión superficial
Otro factor importante en las curvas presión-volumen es la tensión superficial del líquido que delimita
el alvéolo. La tensión aumenta a medida que las fuerzas atractivas entre las moléculas de líquido
adyacentes son mucho más fuertes que las existentes entre el líquido y el gas, provocando que la
superficie de líquido sea la menor posible (Fig 24A). A medida que se contrae la burbuja, genera una
presión que puede medirse predecirse por medio de la ley de Laplace, donde la presión será cuatro
4*T
veces la razón entre la tensión superficial T y el radio r de la esfera: P .
r
Como el aire se mueve en función del gradiente de presión, sería posible observar el colapso de
los alvéolos de menor diámetro hacia los de mayor, debido a su mayor presión (Fig 24B). El
surfactante pulmonar reduce la tensión superficial. Aparentemente, las fuerzas hidrofóbicas de
repulsión se oponen a las fuerzas de atracción entre las superficies de líquido responsables de la
tensión superficial.
El surfactante pulmonar provoca que, al disminuir la tensión superficial del alvéolo, los
pulmones se hacen más complacientes, disminuyendo el esfuerzo ventilatorio. Además, disminuye la
tendencia de las burbujas pequeñas a colapsar en las de radio mayor (Laplace). Por otro lado, ayuda a
mantener los alvéolos secos. Por último, como consecuencia de los cambios en la densidad de
surfactante durante el ciclo ventilatorio, pone freno a los cambios de volumen tanto durante la
inspiración como durante la espiración (Fig 24C).
Figura 24. Efecto de la
tensión superficial sobre la
complacencia pulmonar. A.
tensión superficial en una
esfera de aire delimitada
por agua. B. presión del aire
dentro de esferas de distinto
radio (ley de Laplace). C.
efecto del surfactante
pulmonar sobre la
velocidad de cambio de
volumen alveolar durante el
ciclo ventilatorio. (Boron &
Boulpaep, 2004)
Las características del flujo laminar fueron descritas por el físico francés Poiseuille, quién
P * * r 4
determinó que la velocidad de flujo Q ,donde P es el gradiente de presión; r, radio de la
8 * * l
tubería; , viscosidad del fluido; y l, el largo de la tubería. Dado que la resistencia R es R = P / Q),
8 * * l
donde Q, es el flujo, P, es el gradiente de presión, entonces tenemos que R . Nótese que si el
*r4
radio disminuye a la mitad, la resistencia aumenta 16 veces. Para el flujo turbulento, la relación es más
compleja, ya que considera la razón entre las fuerzas de inercia y viscosidad.
En el árbol bronquial el sistema es muy complejo, muchas ramas de distinto calibre, paredes
irregulares, etc. por lo que se aplican los conceptos del flujo laminar. Así, el principal determinante de
la resistencia de la vía aérea serán los bronquiolos terminales (generaciones 10-16).
Figura 30. Cambios en la resistencia de la vía aérea. A. efecto sobre la velocidad de inflado pulmonar. B. efecto de sobre el
volumen corriente a distintas frecuencias respiratorias. C. efecto de la frecuencia respiratoria sobre el volumen corriente.
(Boron & Boulpaep, 2004)
I.3.3. Generación del gradiente de presión entre la atmósfera y el alvéolo y esfuerzo ventilatorio y
Supongamos que medimos la presión intrapleural (PIP) y alveolar (PA) durante un ciclo ventilatorio.
Justo antes de la inspiración, la PIP es de -5 cm H2O, presión mínima requerida para mantener
distendida la vía aérea (fuerzas de retracción pulmonar y expansión de la caja torácica). Esto se
conoce, a su vez, como presión transmural (PTM: diferencia de presión radial a través de la pared de la
vía aérea en cualquier parte del árbol traqueobronquial), que a través de la pared alveolar se
denomina presión transpulmonar (PTP = PA - PIP).
Durante la inspiración, la PIP disminuye provocando una disminución de la PA, estableciendo
el gradiente de presión para el ingreso de aire (flujo de entrada de aire, inspiración) (Fig 31). Durante la
espiración, ocurre algo similar, la PIP se hace menos negativa, provocando un aumento de la PA, por la
tanto el aire sale.
Los cambios en la PIP durante la inspiración se deben a la acción conjunta del diafragma y de
los músculos intercostales, que al expandir la caja torácica, provocan una disminución de la presión
intrapleural, que es transmitida a los pulmones, expandiéndolos.
Figura 31. Cambios en la PIP -PTP (B) y en la PA
(C), que producen un flujo de entrada o salida de
aire (D) durante la inspiración y espiración,
respectivamente (A. VT, volumen corriente).
Nótese como estos cambios tienden a superar los
componentes estáticos y dinámicos que
determinan el esfuerzo ventilatorio (ver Fig 32).
(Boron & Boulpaep, 2004)
Esfuerzo ventilatorio
Se requiere un trabajo importante para el movimiento de los pulmones y la pared torácica que
provoquen el ingreso de aire a la vía aérea. Para esto, se deben superar tanto los componentes estáticos
como dinámicos que se oponen al paso de aire. El componente estático es tendiente a superar las
fuerzas elásticas pulmonares que mantienen el volumen pulmonar; y el dinámico, tendiente a superar
la resistencia de la vía aérea al paso de aire y la inercia de éste, para generar un flujo de aire (Fig 32).
El trabajo respiratorio total es difícil de medir, pero se puede estimar en paciente ventilados
mecánicamente durante una parálisis muscular (o voluntarios completamente relajados). Por otro lado
se puede calculare determinando el costo en O2 de la respiración. Asumiendo que la eficiencia es %
eficiencia = 100 * [Trabajo útil / energía total gastada (costo de O2)]. Se cree que este valor de
aproximadamente un 5 a 10%. El costo de O2 en una respiración en reposo es muy bajo (menor al 5%
del O2 consumido).
Volúmenes pulmonares
Antes de entender el movimiento de aire a través de los pulmones, es necesario recordar los
volúmenes estáticos de aire en ellos. La mayor parte de estos se pueden medir mediante un
espirómetro (Fig 34). Durante la espiración, la campana sube y el lápiz baja en el papel de registro.
Inicialmente, se puede observar la ventilación normal (volumen corriente, VT). Luego el voluntario
realiza una inspiración máxima y retorna al VT, obteniéndose el volumen de reserva inspiratorio (IRV). El
voluntario realiza una espiración máxima y retorna al VT, obteniéndose el volumen de reserva espiratorio
(ERV). Si el sujeto realiza una espiración máxima a partir de una inspiración máxima, el volumen de
aire exhalado corresponde a la capacidad vital (VC). Sin embargo, una fracción de aire permanece en los
pulmones luego de una espiración máxima, el volumen residual (RV). El volumen de aire que
permanece en los pulmones luego de una espiración corriente se denomina capacidad residual funcional
(FRC).
Ni la FRC ni el RV se pueden determinar con un espirómetro simple. Para determinar esto
valores se diseñó la técnica espirométrica de dilución de Helio (He), un gas que es prácticamente
insoluble en la sangre (Fig 35). El sujeto respira a través de un espirómetro (VS)con una concentración
conocida de He ([He]inicial), después de algunos ciclos ventilatorios, la concentración de He en el
espirómetro y los pulmones será la misma ([He]final) Así en el equilibrio se tiene que:
Heinicial
VL VS * 1 , donde VL corresponde al volumen pulmonar que se quiere medir: FRC, si la
Hefinal
prueba termina luego de una espiración corriente; RV, si termina luego de una espiración forzada.
Figura 34. Volúmenes y capacidades pulmonares.
(A) espirómetro de campana. (B) registro
espirométrico de VT, IRV, ERV y CV. Note que la
capacidad pulmonar total (TLC), FRC y RV no se
pueden medir mediante espirometría. (Boron &
Boulpaep, 2004)
Otro método para medir la FRC y el RV es la pletismografía de cuerpo completo (Fig 36). En
este método, el sujeto se encuentra en el interior de una cabina. Al término de una espiración, se cierra
la válvula que está conectada a la boca del voluntario y se le pide a éste que realice un esfuerzo
ventilatorio. Con la válvula cerrada, la inhalación provoca un aumento en el volumen pulmonar y del
sujeto, desplazando un volumen igual desde la caja pletismográfica (VL), registrado por el
espirómetro. Al término de la espiración con la válvula abierta, la presión en la boca y en los alvéolos
corresponde a la presión barométrica, registrando (P). Con la válvula cerrada, el voluntario hace un
mínimo esfuerzo ventilatorio, el volumen pulmonar aumentará a VL+VL. El aumento del volumen va
P P
acompañado de una caída de la presión (P), a un valor (P-P). Así, VL VL * , donde VL
P
corresponde al volumen pulmonar que se quiere medir: FRC, si la prueba termina luego de una
espiración corriente; RV, si termina luego de una espiración forzada.
Figura 36. Medición pletismográfica de la FRC y el RV.
Cuando el sujeto realiza un esfuerzo ventilatorio en contra
de una vía aérea cerrada, provoca un discreto cambio del
volumen pulmonar (VL), que está asociado a un cambio
en la presión de la vía aérea (P). (Boron & Boulpaep, 2004)
Ventilación alveolar
Supongamos que el volumen de aire espirado en cada ventilación es de 500 mL, y el sujeto
tiene una frecuencia respiratoria (fR) de 15 min-1. La cantidad de aire que sale de los pulmones cada
minuto es 500 * 15 = 7500 mL/min. Esto se conoce como ventilación total (ventilación pulmonar o
volumen minuto, V E ). El volumen de aire inspirado durante un minuto ( V I ) es levemente mayor al
espirado. Sin embargo, para efectos prácticos se consideran iguales.
No todo el aire que ingresa por la boca llega a los alvéolos, donde ocurre el intercambio
gaseoso. De cada 500 mL de aire inhalado, 150 mL permanecen en el espacio muerto anatómico (VD). Así,
el volumen de aire fresco que llega a la zona respiratoria en un minuto corresponde a (500 - 150) + 15
= 5250 mL/min. Esto es llamado la ventilación alveolar ( V A ).
La ventilación total puede medirse fácilmente si uno tiene un sujeto que respira a través de un
sistema de válvulas que separa el aire inspirado del espirado, y juntamos todo el aire espirado en una
bolsa. La V A es más difícil de determinar. Habitualmente se determina mediante la siguiente
ecuación: V A V E VD . La dificultad del método está dada por la determinación del VD. La V A puede
aumentarse (hiperventilación) ya sea por aumentos en el VT (hiperpnea) o en la fR (taquipnea), o por
ambos (hiperpnea + taquipnea). Aumentar el VT resulta más efectivo puesto que reduce la proporción
de VD en cada ciclo.
Otra manera de medir la V A en sujetos normales es a partir del %CO2 medido al término de la
espiración (%CO2). Dado que no hay intercambio gaseoso en el espacio muerto anatómico, no hay CO2
ahí al término de la espiración (el %CO2 atmosférico es despreciable). Así, todo el CO2 espirado
%CO 2
proviene del gas alveolar, entonces: V CO 2 V A * , donde V CO 2 corresponde al volumen de CO2
100
V CO 2 *100
exhalado por minuto. Así, reordenando tenemos: V A . El valor de %CO2/100 corresponde a
%CO 2
la concentración fraccional de CO2 (FCO2). Así la V A pude determinarse por la razón entre el V CO 2 y la
FCO2. La presión parcial de CO2 (PCO2) es proporcional a la FCO2 en el alvéolo, así PCO2 = FCO2 * K,
V CO 2
donde K es una constante. Por lo que V A * K . Dado que en sujetos normales la PCO2 en el gas
PCO 2
alveolar (PACO2) y arterial (PaCO2) son prácticamente idénticas, se puede usar la PaCO2 para
determinar la V A . La relación entre la V A y la PCO2 es muy importante, ya que si la V A disminuye a la
mitad (y la producción de CO2 no cambia), la PACO2 y la PaCO2 aumentarán al doble.
I.5. Relación ventilación/perfusión ( V Q )
Hipoxemia
La PO2 disminuye a medida que el gas se mueve desde la atmósfera hasta la mitocondria, donde es
utilizado. La PO2 del aire seco es un 20.93%. A nivel del mar, la Pb es de 760 mm Hg y en la tráquea, el
aire está saturado de vapor de agua (47 mm Hg), por lo que el la PO2 inspirada es aproximadamente
150 mm Hg. Una vez que llega a los pulmones, la es PAO2 de 100 mm Hg. Esta caída se debe a dos
factores: el intercambio continuo con los capilares y el movimiento ventilatorio ciclo-a-ciclo. La
oscilación en la PAO2 es de 3 mm Hg. La velocidad de intercambio de O2 depende de la tasa de
consumo de O2 por los tejidos, varía poco durante el reposo, por lo que prácticamente la PAO2 está
determinada sólo por la V A . Lo mismo aplica para la PACO2 que es, habitualmente, de 40 mm Hg, con
una variación 2 mm Hg (Fig 38).
Si disminuye la V A (hipoventilación), la PAO2 disminuirá y, de manera análoga, aumentará la
PACO2. La hipoventilación siempre conduce a un aumento de la PaCO2, así ésta estará dada por:
V CO 2
PaCO 2
* K , dónde V CO 2 es la producción de CO2 y K es constante.
VA
Otra causa para la disminución de la PaO2 son las alteraciones en la difusión de O2 (gobernada
por la ley de Fick), fenómeno que se analizará más adelante.
Efectos de alterar la relación V Q en el pulmón.
Al modificar la relación V Q se verá afectado el intercambio gaseoso. En el modelo esquematizado de
la Figura 39, se observa la situación control (con una relación V Q normal), y el efecto de obstruir la
vía aérea para disminuir la ventilación, dejando el flujo sanguíneo sin cambios, ( V Q bajo), lo que
provocará que la PAO2 y la PACO2 sean iguales a la PO2 y PaCO2 al término de los capilares,
respectivamente. Por otro lado, al obstruir el flujo sanguíneo, dejando la ventilación sin cambios ( V Q
alto), la PAO2 aumenta (PAO2 = PbO2) y la PACO2 = 0. Estos cambios pueden observarse en un diagrama
O2-CO2 y sus relación con el V Q (Fig 40A).
Considerando la distribución heterogénea del flujo sanguíneo y de la ventilación pulmonar
(cambios regionales), es de esperar que la relación V Q dependa de la zona del pulmón considerada.
La relación V Q es extremadamente alta en el ápice pulmonar (donde el flujo sanguíneo es mínimo) y
mucho mas baja en las bases (Fig 40B). Así, si combinamos ambas condiciones, veremos los cambios
en los gases arteriales en las distintas zonas del pulmón (Fig 41) como consecuencia de los cambios
regionales en la relación V Q .
Existen mecanismos compensatorio locales tendientes a compensar cambios en la relación
V Q , para asegurar que permanezcan perfundidos los capilares que estén correctamente ventilados y,
por otro lado, disminuir el flujo sanguíneo hacia los alvéolos probamente ventilados. Anteriormente
señalamos la vasocontricción hipóxica como un mecanismo de regulación de la resistencia de la
circulación pulmonar. Este mismo mecanismo puede utilizarse de manera compensatoria a cambios
en la ventilación (Fig 42). De igual manera, la vasodilatación hiperóxica favorece un aumento del flujo
sanguíneo hacia alvéolos mejor ventilados. Por otro lado, existen mecanismos compensatorios a los
cambios en el flujo sanguíneo, la broncodilatación hipercápnica, provocada por un aumento en el flujo
sanguíneo que conlleva a un aumento del CO2 alveolar, que provoca una broncodilatación y aumento
del flujo de aire, por disminución de la resistencia de la vía aérea. Y una broncoconstricción hipocápnica,
inducida por una disminución del CO2 alveolar consecuencia de una disminución del flujo sanguíneo
(Fig 42).
Figura 39. Efecto de alterar la relación
ventilación/perfusión sobre la PaO2 y la PaCO2. Modificado
de West, 2005
Figura 40. Alteraciones del coeficiente V Q . A.
diagrama O2-CO2; B. variaciones regionales. (Boron
& Boulpaep, 2004)
Figura 41. Resultado de diferencias regionales del coeficiente V Q ,
con el intercambio gaseoso. A una V Q alto, resulta una PO2 alta y
una PCO2 baja. (West, 2005)
Figura 42. Mecanismos compensatorios de ajustes locales en la perfusión (vasocontricción hipóxica – vasodilatación
hiperóxica) a cambios en la ventilación y ajustes ventilatorios (broncodilatación hipercápnica – broncoconstricción
hipocápnica) a cambios en la perfusión, para mantener un coeficiente V Q constante.
Por otro lado, el óxido nitroso (N2O), que no interactúa con la Hb, difunde rápidamente (Fig
46A), provocando un aumento de su presión parcial (Fig 46B). Prácticamente en 1/10 del tiempo que
está en contacto el eritrocito con el capilar (Fig 46C). Así, la cantidad de N2O que difunde del alvéolo
al capilar estará limitada por el flujo sanguíneo -la cantidad de sangre- y no de las propiedades
difusivas a través de la barrera, así, el proceso estará limitado por perfusión ( Q ).
Figura 46. Captación de óxido nitroso (N2O) a lo largo de los capilares pulmonares. (Boron & Boulpaep, 2004)
Figura 51. Trasporte de O2 unido a hemoglobina. A. grupo Hemo. B. tetrámero de hemoglobina. C. curva de saturación de la
Hb por O2 y contenido de O2 en la sangre. Note que la mayor parte del O2 se transporta unido a Hb. (Boron & Boulpaep,
2004)
la HbO2. Así, la presencia de Hb en los tejidos periféricos ayuda a remover el CO2 de éstos (Fig 53A),
mientras que la oxigenación en los capilares pulmonares favorece su eliminación hacia los alvéolos
(Fig 53B). Este efecto se denomina efecto Haldane.
Los compuestos carbamínicos se forman por la interacción del CO2 con grupos amino terminal en
las proteínas de la sangre. Muy importante es la globina de la Hb, que interactúa:
Hb·NH2 + CO2 Hb-NH·CCOH, produciendo carbaminohemoglobina, cuya formación modifica la
afinidad de la Hb por el O2 en los capilares periféricos (Fig 53A).
1. Centro Respiratorio Bulbar (generador del patrón ventilatorio): se encuentra en la formación reticular
del bulbo raquídeo, bajo el piso del cuarto ventrículo. Posee dos áreas identificables: un área dorsal
(grupo respiratorio dorsal, DRG), asociado con la inspiración; y un área ventral (grupo respiratorio ventral,
VRG), principalmente para la espiración (Fig 55). Las neuronas del área inspiratoria tienen una
actividad de descarga intrínseca, por lo que son las responsables de la ventilación corriente. En ausencia
de estímulos, las neuronas del DRG descargan espontáneamente potenciales de acción, que resultan
impulsos nerviosos que viajan al diafragma y los músculos intercostales externos (Fig .56)
2. Centro apnéusico: en la región baja (caudal) del puente. Animales con una transección del
troncoencéfalo por sobre esta región muestran un aumento en la duración de la fase inspiratoria
(apneusis), interrumpida por intentos espiratorios transitorios. Aparentemente, los impulsos
provenientes de esta área estimulan las DRG, tendiendo a prolongar la rampa de potenciales de
acción.
3. Centro pneumotáxico: en la región superior (cefálica) del puente. Esta área parece “apagar” l o inhibir
la inspiración, regulando directamente el volumen e, indirectamente, la fR.
I.7.2. Efectores
Los músculos de la respiración involucran la musculatura primaria y secundaria (Tabla IV) que recibe
inervación eferente desde sus respectivos núcleos motores troncoencefálicos o espinales. Estos núcleos
se activan en repuesta a potenciales de acción provenientes de las RRN y de los DRG y VRG (Fig 55).
Tabla IV. Inervación de la musculatura primaria y secundaria de la ventilación. (Boron & Boulpaep, 2004)
I.7.3. Sensores
Quimiorreceptores centrales.
Responden a cambios en la composición química de la sangre y del fluido circundante. Se encuentran
en la superficie ventral del bulbo, en la vecindad del NC IX y NC X (Fig 58A), donde se encuentran
neuronas quimiosensibles que se activan (Fig 58B) o inhiben (Fig 58C) por la acidosis. Se sabe que el
estímulo específico para los quimiorreceptores centrales son los protones. Sin embargo, la berrera
hemato-encefálica es impermeable a los protones, por lo que quien llega al fluido cerebro espinal es el
CO2, que posteriormente regenera protones que estimulan los quimiorreceptores centrales (Fig 58,
esquema). La respuesta inmediata a la estimulación de los quimiorreceptores centrales es una
hiperventilación (Fig 59).
Figura 61. Efecto sobre la ventilación pulmonar de la estimulación de los quimiorreceptores arteriales periféricos por la
hipoxemia. Además, potenciación de respuestas ventilatorias entre los quimiorreceptores centrales y periféricos. La hipoxia
potencia la respuesta a la hipercapnia (panel medio); mientras que la hipercapnia potencia las respuestas ventilatorias a la
hipoxemia (panel derecha). (Modificado de West, 2004)
Receptores pulmonares
Receptores de estiramiento (receptores de estiramiento de adaptación lenta), se cree que están
en la musculatura lisa de la vía aérea. Responden a la distensión pulmonar, con una actividad
sostenida durante la inspiración. Asciende por el vago como fibras mielínicas grandes. Disminuyen la
fR, por aumentar el tiempo de espiración (reflejo de Hearing-Breuer). Este reflejo determina la
frecuencia y profundidad de la ventilación.
Receptores irritativos (receptores de estiramiento de adaptación rápida), se cree que están
entre las células epiteliales de la vía aérea. Se estimulan por gases nocivos, humo del cigarrillo, polvo
inhalado y aire frío. Asciende por el vago como fibras mielínicas. Podrían ser importantes en la
broncoconstricción y ataque asmáticos.
Receptores J (receptores juxta-capilares), son fibras amielínicas C. Se cree que están en las
paredes alveolares, cercanos a los capilares. Responden rápidamente a sustancias químicas de la
circulación pulmonar. Podrían mediar la sensación de disnea (dificultad respiratoria).
Fibras bronquiales C, reciben información de la circulación bronquial y no de la pulmonar.
Responden rápidamente a químicos inyectados en esta circulación. Broncoconstricción y secreción de
mucus.
Otros receptores
Receptores de la vía aérea superior y nariz, ubicados en la nariz, nasofaringe, laringe y
tráquea, responden a estimulación química y mecánica. Son una extensión de los receptores irritativos.
Gatillan respuestas reflejas de estornudo, tos, broncoconstricción, laringo-espasmo (por estimulación
mecánica directa).
Receptores de las articulaciones y músculos, receptores de las extremidades en movimiento
estimulan la ventilación durante las primeras etapas del ejercicio.
Sistema Gamma, músculos intercostales y diafragma poseen husos musculares que detectan la
elongación del músculo. Control reflejo de la contracción.
Barorreceptores arteriales, aumentos de la presión arterial puede causar hipoventilación
refleja o apnea (ausencia de 3 a 4 ciclos ventilatorios). Una disminución de la presión arterial conduce
a hiperventilación.
Dolor y temperatura, la estimulación de fibras aferentes nociceptivas cambia la ventilación. El
dolor provoca una apnea seguida de hiperventilación. Aumentar la temperatura de la piel provoca
hiperventilación.
El intercambio gaseoso entre la sangre y los alvéolos tiende a mantener los niveles normales de
PaO2 y PaCO2 para satisfacer las demandas de O2 y eliminar el CO2 generado. Esta función resulta de
un mecanismo ventilatorio altamente controlado en el que participan el generador del patrón
ventilatorio (controlador central), diversos receptores sensoriales y las eferencias motoras que
provocan los movimientos de aire en la torácica (Fig 62).
Química ácido—base
Dentro de la fisiología la definición más adecuada para los términos ácido y base es: un ácido es una
sustancia que dona un ión hidrógeno (protón, hidrogenión, H+); mientras que una base, es una
sustancia que puede aceptar un protón de otra sustancia. Un ácido fuerte es una sustancia que se
disocia por completo -o casi por completo- en un protón y su base conjugada en una disolución acuosa;
un ácido débil, permanece levemente disociado en solución acuosa. En general, un ácido fuerte tiene
una base conjugada débil, mientras que un ácido débil, una base conjugada fuerte.
Un buffer (amortiguador, tampón) es una mezcla de compuestos (usualmente un ácido débil y
su base conjugada) que en solución acuosa pueden resistir a cambios en la concentración de protones
cuando se le agrega un ácido o base fuerte.
La acidez de una solución está determinada por la actividad (H+) de los protones en solución.
En soluciones extremadamente diluidas, la H+ está determinada directamente con la concentración de
protones ([H+]) en la solución. La concentración de protones en la sangre es extremadamente baja, por
lo que su actividad está determinada sólo por la [H+]. Dado el amplio rango de [H+] en el cuerpo (10-1
M en el ácido gástrico; 10-8 en la secreción pancreática), se utiliza una escala conveniente de pH,
correspondiente al logaritmo negativo de la [H+]. Así, pH = -log[H+].
El pH arterial es aproximadamente 7.40, con un rango normal de entre 7.35 y 7.45. Valores de
pH bajo este rango se considera acidemia; mientras que superiores, alcalemia. La condición generada
por la retención de H+ o por exceso de eliminación de bicarbonato (HCO3-) u otra base, se denomina
acidosis: la pérdida excesiva de H+ o retención de bases se denomina alcalosis.
Estado ácido-base
El transporte de CO2 tiene un efecto muy importante en el estado ácido-base de la sangre y el cuerpo
completo. El pulmón excreta aproximadamente 10000 mEq de H2CO3 por día, mientras que el riñón
sólo 100 mEq de ácidos no volátiles. Así, alterando la ventilación alveolar, y por ende la eliminación
de CO2, es posible ejercer un control importante sobre el estado ácido-base.
El pH resultante de la disolución del CO2 en la sangre y la subsecuente disociación del ácido
carbónico está dado por la ecuación de Henderson-Hasselbalch, donde:
CO2 + H2O H2CO3 H+ + HCO3-
H CO
2 3
KA
H * HCO
3
CO
2
CO 2
Reordenando,
CO 2
pH pK A log
HCO 3
CO 2
Aplicando la ley de Henry, la concentración de CO2 (mmol/L) puede reemplazarse por el producto
entre la presión parcial del CO2 y su solubilidad (PCO2*0.03)
pH pK A log
HCO 3
0.03 * PCO 2
El valor de pKA es 6.1, la [HCO3 normal en la sangre arterial es de 24 mmol/L. Sustituyendo se tiene:
-]
de base, representado por la línea vertical entre las dos líneas de buffer DE y BAC. EN el otro caso,
una reducción de la [HCO3-] desplaza la línea de buffer hacia abajo, mostrando un exceso negativo de
base (déficit de base).
La razón entre [HCO3-] y PCO2 se puede alterar de cuatro maneras, tanto la [HCO3-] como la
PCO2 pueden aumentar o disminuir.
Figura 64. Curvas de buffer del plasma sanguíneo que contiene 5, 10, 15, o
20 g de hemoglobina por 100 mL de sangre. (Levitzki, 2003)
Acidosis Respiratoria
La acidosis respiratoria es causada por un aumento en la PCO2, que provoca una reducción de la
relación [HCO3-]/PCO2 y, así, disminuye el pH. Esto corresponde a un movimiento desde A hasta B
(Fig 63). Siempre que la PCO2 aumenta, el bicarbonato debe aumentar en algún grado, debido a la
ecuación de disociación del ácido carbónico. Esto se refleja en el aumento hacia la izquierda de la
pendiente de la línea de buffer del plasma sanguíneo (Fig 63B). La razón [HCO3-]/PCO2 puede caer
producto de una hipoventilación o alteraciones de la relación ventilación/perfusión, que provocan la
retención de CO2.
Si la acidosis respiratoria persiste, el riñón responde conservando el bicarbonato. Un estímulo
para esta función es el aumento en la PCO2 en las células tubulares renales, que excretan una orina
más ácida mediante la secreción de H+. Los protones se secretan en la forma de H2PO4- o NH4+; los
iones HCO3- se reabsorben. El aumento en la [HCO3-] mueve la razón [HCO3-]/PCO2 a su condición
inicial. Esto corresponde al movimiento desde B a D por la línea PCO2 = 60 mm Hg en la Figura 63B, y
se conoce como acidosis respiratoria compensada. Así,
6.1 log20 7.4
24
pH 6.1 log (normal)
0.03 * 40
6.1 log15.6 7.29
28
pH 6.1 log (acidosis respiratoria)
0.03 * 60
6.1 log18.3 7.36
33
pH 6.1 log (acidosis respiratoria compensada)
0.03 * 40
Alcalosis Respiratoria
La alcalosis respiratoria es causada por una disminución de la PCO2, que aumenta la razón [HCO3-
]/PCO2 y, así, eleva el pH (movimiento de A a C en la Fig 63B). Una disminución en la PCO2 se puede
inducir por hipoventilación, por ejemplo, a altitudes elevadas. La compensación renal produce un
aumento en la excreción de HCO3-, retornando así la razón [HCO3-]/PCO2 hacia los valores normales
(C a F a lo largo de la línea PCO2 = 20 mm Hg). Después de una estadía prolongada a altitudes
elevadas, la compensación renal puede ser casi completa. Aquí existe un déficit de base.
Acidosis Metabólica
La acidosis metabólica involucra el concepto “metabólico”, que quiere decir un cambio inicial en la
[HCO3-], el numerador de la ecuación de Henderson-Hasselbalch. En la acidosis metabólica, la razón
[HCO3-]/PCO2 cae, disminuyendo el pH. El HCO3- puede disminuir debido a una acumulación de
ácidos en la sangre, como ocurre en la diabetes mellitus descontrolada, o después de hipoxia tisular, que
provoca la liberación de ácido láctico. El cambio correspondiente provoca un desplazamiento de A a G
en la Fig 63B.
En este caso, ocurre una compensación ventilatoria debido al aumento de la ventilación que
disminuye la PCO2 y aumenta la razón [HCO3-]/PCO2. El estímulo para la hiperventilación es el
aumento en la concentración de protones, que estimulan los quimiorreceptores arteriales periféricos.
En la Fig 63B, está esquematizado con el movimiento de G a F. Aquí existe un déficit de base.
Alcalosis metabólica
La alcalosis metabólica se debe a un aumento en la concentración de bicarbonato, que aumenta la
razón [HCO3-]/PCO2, y así el pH. Puede provocarse por la ingesta excesiva de álcalis o por la pérdida
de secreción gástrica ácida por el vómito. Aquí, se produce un movimiento de A a E en la Fig 63B.
Mediante la hipoventilación ocurre una compensación respiratoria leve, que aumenta la PCO2. Así, el
punto E se mueve hacia D. Sin embargo, la compensación respiratoria de la alcalosis metabólica es
muy pequeña o, habitualmente, está ausente. Aquí tenemos un aumento del exceso de base.
Referencias
Berne RM, Levy MN, Koeppen BM & Stanton BA (2004) Physiology, Updated Edition: With
Student Consult Online Access. Elsevier (Mosby). St. Louis, MO (5th Ed).
Bolsover SR, Hyams JS, Shephard EA, White HA & Wiedemann CG (2004) Cell Biology: a
short course. John Wiley & Sons, Inc., Hoboken, New Jersey.
Boron WF & Boulpaep E (2004) Medical Physiology, Updated Edition: With Student Consult
Online Access. Saunders. Collingwood, ON. (1st Ed).
Guyton AC & Hall JE (2005) Textbook of Medical Physiology: With Student Consult Online
Access. Saunders. Collingwood, ON. (11th Ed).
Levitzki MG (2003) Pulmonary Physiology. McGraw-Hill Companies. 6th ed
Lorin MI (1979) Mecanismos de defensa del sistema respiratorio. En. Fisiología respiratoria del
feto, Scarpelli EM (Ed) Fisiología respiratoria del feto, recién nacido y niño. Salvat Editores SA.
Barcelona.
Rizzardini M (1999) Pediatría. Publicaciones Técnicas Mediterráneo, Ltda. Santiago.
West JB (2005) Respiratory Physiology. The Essentials. Lippincott Williams & Wilkins.
Philadelphia (7th Ed).
Elmer ress
Review J Clin Med Res • 2011;4(1):7-16
Articles © The authors | Journal compilation © J Clin Med Res and Elmer Press™ | www.jocmr.org 7
Pierrakos et al J Clin Med Res • 2011;4(1):7-16
um in correlation to impaired fluid remove from the alveolar vances in technology have introduced new treatments and
space result in accumulation of protein-rich fluid inside the improved therapeutic strategies. The following paragraphs
alveoli, thereby producing diffuse alveolar damage, with re- discuss recent developments in the therapeutic approach to
lease of pro-inflammatory cytokines, such as Tumor Necro- ARDS.
sis Factor (TNF), IL-1 and IL-6 [5]. Neutrophils are recruited
to the lungs by cytokines, become activated and release toxic Low tidal volume ventilation
mediators, such as reactive oxygen species and proteases [6].
Extensive free radical production overwhelms endogenous The concept of “baby lung” was introduced in 1980s by
anti-oxidants and causes oxidative cell damage [7]. Gattinoni et al and generated interest in the use of low tidal
Inflammation due to neutrophil activation is key in the volume ventilation as therapeutic strategy in ARDS. Several
pathogenesis of ARDS. Fundamental transcription abnor- animal studies showed that ventilation with large tidal vol-
malities, involving NF-kappa B that is required for tran- umes and high inspiratory pressures resulted in development
scription of genes for many pro-inflammatory mediators, of hyaline membranes and inflammatory infiltrates in the
are present in the lungs of ARDS patients [8]. In addition, lungs, and development of respiratory failure [17].
other factors such as endothelin-1, angiotensin-2 and phos- In the late 1990s four randomized controlled trials
pholipase A-2 increase vascular permeability and destroy (RCTs) evaluated the potential benefit of low tidal volume
micro-vascular architecture, enhancing inflammation and ventilation in ARDS [18-21]. Although all four studies had
lung damage. In conclusion, as several different pathways limited power, one study by Amato et al [21] demonstrated
are involved in ARDS development, no single biomarker can that the low tidal volume group had higher survival, higher
predict outcome in ARDS patients [9]. rate of weaning from mechanical ventilation and reduced
Computed Tomography studies in the 1980s helped us rate of barotrauma.
understand the pathophysiologic alterations in the lungs of Because of conflicting results from these studies, the
ARDS patients [10]. In addition, as lung compliance corre- National Heart Lung and Blood Institute ARDS Network
lates with the degree of the normally ventilated tissue, lung conducted a multicenter RCT on 861 ARDS patients [22],
compliance decreases in ARDS because of decreased lung comparing two group of patients ventilated with low vs. high
size, rather than because of lung stiffness, and this hypoth- tidal volumes. In-hospital mortality was significantly lower
esis introduced the concept of ”baby lung” in ARDS [11]. and the number of days without mechanical ventilation was
Pulmonary hypertension (PH) is widely recognized as significantly higher in the low tidal volume group. Although
a characteristic feature of ARDS [12]. PH etiology includes this study has been criticized for the high difference of tidal
parenchymal destruction and airway collapse, hypoxic pul- volume between groups, it demonstrated that high tidal vol-
monary vasoconstriction, presence of other pulmonary vaso- umes should be avoided, and underlined the importance of
constrictors and vascular compression [13]. maintaining low plateau pressures, with 30 cm H2O as an
The initial phase of fluid accumulation is followed by a acceptable cut-off.
proliferation phase characterized by resolution of pulmonary Low tidal volume ventilation is generally well tolerated
edema, proliferation of type II alveolar cells, fibroblasts and and it has not been associated with clinically important ad-
myofibroblasts, and new matrix deposition. This phase starts verse outcomes, except for hypercapnic respiratory acidosis
early (within 72 h) in ARDS, and lasts for more than 7 days. in some patients. In conclusion, hypercapnia and respiratory
Factors influencing the progression to fibro-proliferation vs. acidosis are expected consequences of low tidal volume ven-
resolution and reconstitution of normal pulmonary paren- tilation. However, there is no evidence that hypercapnia is
chymal architecture are poorly understood [14], but patients harmful in ARDS patients, and it may in fact confer some
who develop pulmonary fibrosis exhibit deterioration of pul- protection against ventilator-associated lung injury.
monary compliance, progressive hypoxia and ventilator de-
pendence, and increased mortality (> 57%) [15]. PEEP
Multiple Organ Failure (MOF) is the leading cause
of death in ARDS, but the pathophysiologic link between PEEP is an essential component of mechanical ventilation
ARDS and MOF is not well defined [16]. However, based for patients with ARDS, as it was early observed that PEEP
on existing data it is not clear whether ARDS is the mani- greatly improves oxygenation in ARDS patients. High PEEP
festation of a disease, or it is a disease that causes the MOF levels may open collapsed alveoli and decrease intrapulmo-
syndrome. nary shunt. Additionally, ventilation-induced alveolar injury
is reduced by decreasing alveolar over-distention, because
the volume of each subsequent tidal breath is shared by more
Treatment open alveoli [23]. On the other hand, high PEEP levels may
decrease repetitive alveolar opening and closing during the
Improved understanding of ARDS pathophysiology and ad- respiratory cycle, thereby promoting lung injury [24].
8 Articles © The authors | Journal compilation © J Clin Med Res and Elmer Press™ | www.jocmr.org
Acute Respiratory Distress Syndrome J Clin Med Res • 2011;4(1):7-16
Three RCTs have evaluated modest vs. high levels of of alveoli, more homogeneous ventilation due to decreased
PEEP in patients with ARDS. The NHLB ARDS Network ventilation-perfusion inequalities and reduced ventilator in-
conducted the ALVEOLI trial (Assessment of Low tidal Vol- duced lung injury [36].
ume and Elevated end expiratory pressure to Obviate Lung Four RCTs have investigated the effect of prone posi-
Injury) [25]. This study showed improved PaO2/FiO2 ratio tioning on outcome. The first trial by the Prone-Supine Study
but no benefits with regards to survival or duration of me- group randomized 304 patients with a wide range of severity
chanical ventilation in the high PEEP group. Several years of acute lung injury [37].Patients remained prone for 7 h/
later, the Canadian Critical Care Trials Group performed a day on average, for up to 10 days, but there was no effect on
similar study to determine whether the combination of low survival. Three years later, Guerin et al conducted a similar
tidal volume ventilation with high PEEP could improve mor- multicenter study [38]: patients remained prone for about 8
tality to a greater extent compared to low tidal ventilation h/day, and prone positioning continued until clinical criteria
alone [26]. Results of this study showed reduced need for for improvement were met, but this study also did not show
other rescue therapies such as prone position or NO, but did a reduction in mortality. Two subsequent RCTs attempted to
not show any benefit in survival. correct some shortcomings of the earlier study: they only in-
In conclusion, based on published data, high levels of cluded patients with ARDS, and patients remained prone for
PEEP do not seem to confer any benefit with regards to mor- most of the day (about 20 h). The first RCT by Mancebo et
tality in ARDS. Because ARDS patients are a heterogeneous al was terminated prematurely, after only including 142 pa-
population, the apparent absence of benefit from high levels tients, because of problems with patient recruitment [39]. A
of PEEP could be due to the beneficial effects of high PEEP more recent multicenter RCT by Taccone et al, included 344
in some ARDS patients being offset by detrimental effects in patients [40], and showed significantly increased frequency
other patients [27]. However, data from the RCTs mentioned of adverse events (airway obstruction, hypotension, vom-
above suggest that high PEEP levels improve lungs function iting, accidental extubation) in patients treated with prone
without any adverse effect on mortality [28]. position. Neither of the last two studies showed any surviv-
al benefit using the prone position in patients with severe
Recruitment maneuvers ARDS.
In conclusion, existing data do not support routine use of
A recruitment maneuver is a transient increase of trans-pul- the prone position in ARDS. However, because all published
monary pressure intended to promote reopening of collapsed studies have shown improved oxygenation, prone position-
alveoli [29]. Techniques described for recruitment maneu- ing is an attractive rescue treatment for ARDS patients with
vers include sustained high-pressure inflation and increased severe hypoxemia, even though a survival benefit has never
PEEP, with concurrent reduction of tidal volume [30], but been demonstrated.
it is not clear if any maneuver is superior to others. Several
studies have shown improved gas exchange with recruitment High-frequency ventilation
maneuvers, but no RCT has shown benefit on ARDS mor-
tality [31] and a recent systematic review by Fan et al [32] The idea of high frequency ventilation (HFV) is to provide
showed that hypotension and decreased saturation occur in tidal volumes below that of anatomic dead space at high fre-
12% and 8% of patients respectively during or after such ma- quency (> 60 breaths per minute). Compared to conventional
neuvers. Based on currently available data, although routine mechanical ventilation, mean airway pressure is higher [41].
recruitment maneuvers are not recommended in ARDS, such Two studies, by Hamilton and Chan, showed reduced risk for
maneuvers can dramatically improve oxygenation in certain barotrauma and lung over-distention, after performing high
patients, and should be considered as rescue therapy in pa- frequency ventilation [42, 43]. High frequency ventilation
tients with life-threatening refractory hypoxia [33]. can be applied by different modes, such as high-frequency
percussive ventilation, high-frequency jet ventilation and
Prone position high-frequency oscillatory ventilation (HFOV) [44]. In the
absence of studies showing superiority of one method over
Prone positioning has been used in ARDS for over 30 years. another, HFOV is the HFV method used more often in adult
In 1976 Piehl et al. reported improved oxygenation in ARDS critical care [43].
patients when they were turned to the prone position [34]. In HFOV very small tidal volumes (1 - 4 ml/kg) are de-
Since then, several observational studies on ARDS have livered at high frequency (3 - 15 Hz) by an oscillatory pump
found similar results, and improvement in oxygenation can [45]. However, the use of HFOV as rescue therapy in pa-
sometimes be dramatic [35]. Mechanisms proposed to ex- tients with refractory hypoxia remains controversial. There
plain the observed beneficial effects of prone positioning are two RCTs comparing HFOV with conventional mechani-
include increased chest wall elastance decreased compres- cal ventilation. The first RCT by Derdak et al found a trend
sion of lung tissue in the dependent regions and recruitment for decreasing 30-day mortality [46] even though relatively
Articles © The authors | Journal compilation © J Clin Med Res and Elmer Press™ | www.jocmr.org 9
Pierrakos et al J Clin Med Res • 2011;4(1):7-16
high tidal volume was used in the control group. The second relaxation may also improve chest wall compliance. NMBAs
RCT by Bollen et al was terminated prematurely because may also have anti-inflammatory effects that could decrease
of slow enrollment, but found an opposite trend in mortal- the inflammation associated with ARDS. However, because
ity [47]. Two meta-analyses also had conflicting results. The there is evidence that NMBAs increase the risk of acquired
first one included only 2 RCTs comparing HFOV vs. con- neuromuscular weakness, thereby making weaning from
ventional ventilation, and did not find any mortality reduc- mechanical ventilation more difficult, and may even increase
tion or improvement of oxygenation [48]. However, a more mortality [56].
recent study by Sud et al included 8 RCTs and found reduced
30-day mortality with HFOV compared to conventional ven- Nutrition
tilation [49]. In conclusion, the role of HFOV in ARDS is
not well defined, and deserves further study from well de- Two RCTs [57, 58] and a meta-analysis by Puntes-Arruda
signed RCTs. et al [59] showed that administration of enteral nutrition
containing high concentrations of eicosapentanoic acid and
Supportive treatment of ARDS γ-linoleic acid and ω-3 fatty acids increased oxygenation and
decreased ICU stay and 28-day mortality in ARDS.
Fluid management
Pharmacologic agents
Early data indicate that increased fluid administration is
correlated with worse outcome in ARDS, whereas negative Vasodilators
fluid balance was associated with better outcome [50]. The
ARDS network conducted a RCT with 1000 patients ran- Because diffuse pulmonary vasoconstriction is part of ARDS
domized to receive conservative or liberal fluid administra- pathophysiology, selective vasodilatation of well ventilated
tion. The conservative strategy group showed improved lung areas seems an attractive method to improve gas exchange in
function and shortened duration of mechanical ventilation, ARDS patients.
but there was no difference in non pulmonary organ failures Inhaled nitric oxide (iNO) causes vasodilation of venti-
and 60-day mortality. lated lung units and redistribution of pulmonary blood flow
away from non-ventilated lung areas, without adverse sys-
Transfusions temic hemodynamic effects. Four RCTs evaluated the effects
of iNO and showed transient improvement in oxygenation
Allogenic blood transfusion is associated with detrimental [60-63], but no improvement in mortality. Similarly, one
immuno-modulatory effects that may result in ARDS [51]. meta-analysis found transient oxygenation improvement but
Consequently, conservative transfusion strategies may de- no survival benefit with iNO [64]. Consequently, iNO may
crease the incidence of ARDS. be useful as short-term rescue therapy in patients with severe
hypoxemic respiratory failure.
Sedation Inhaled prostacycline is another selective pulmonary
vasodilator. Importantly, liposomal PGE1 influences neu-
Several studies have shown that newer ventilation strate- trophil function and decreases neutrophil accumulation
gies using low tidal volume and high levels of PEEP do not and lung leak. Although inhaled prostacycline improves
require high doses of sedation [52]. Furthermore, evidence oxygenation, it has not been shown to reduce duration of
suggests that use of sedatives may increase duration of me- mechanical ventilation or mortality in ARDS patients [65].
chanical ventilation and ICU length of stay and may even be Despite the lack of sufficient data supporting the use of
associated with higher mortality [53]. In addition, there is prostacycline as alternative to iNO, prostacycline is increas-
evidence that spontaneously breathing ARDS patients have ingly used as pulmonary vasodilator, because of the high
improved cardiopulmonary function, presumably by recruit- cost of iNO [66].
ing non ventilated lung units [54]. Therefore, based on cur-
rent evidence, avoidance or minimization of sedation and Vasoconstrictors
paralysis is preferred in ARDS.
Vasoconstrictors can improve oxygenation in ARDS patients
Neuromuscular blocking agents by decreasing intrapulmonary shunt. Phenylephrine [67] and
almitrine [68] have been used in small studies, mainly as
Administration of neuromuscular blocking agents (NMBA) adjuncts during administration of NO. B-blockers have also
in addition to sedation can result in improvement in severe been shown to increase arterial oxygenation in patients with
hypoxemia, because paralysis improves patient-ventilator ARDS [69]. However, the role of these agents in ARDS has
synchronism and reduces oxygen consumption [55]. Muscle not been adequately evaluated, and deserves further study.
10 Articles © The authors | Journal compilation © J Clin Med Res and Elmer Press™ | www.jocmr.org
Acute Respiratory Distress Syndrome J Clin Med Res • 2011;4(1):7-16
Anti inflammatory agents [87] have been evaluated, but none of them has been shown
to be effective for treatment of ARDS.
The interaction between nuclear factor-kappa B (NF-kB)
and glucocorticoid receptor alpha (GRa) is a key mechanism Extracorporeal techniques
regulating the progression of systemic and pulmonary in-
flammation in ARDS [70]. The ability of GC-GRa to down- Extracorporeal membrane oxygenation (ECMO) has been
regulate NF-kB activation is critical for the resolution of sys- studied since 1970s as a method for supporting gas exchange
temic and pulmonary inflammation in ARDS [71]. Although in patients failing conventional ventilation [88]. A RCT con-
several studies showed that corticosteroids confer no benefit ducted in 1979 showed that ECMO use had no effect on long-
and may even cause harm [72], corticosteroids are still used term survival of ARDS patients [89]. Nowadays, ECMO is
in clinical practice. used to support oxygenation of patients with severe ARDS,
Timing of corticosteroid administration and duration of thereby allowing use of decreased ventilator settings (tidal
therapy may be important, and should be taken into consid- volume, respiratory rate, FiO2), in an attempt to minimize
eration. A RCT conducted by the ARDS Network, random- ventilator induced lung injury.
ized 180 patients with persistent (> 7 days) ARDS, to receive In a more recent study 180 patients with severe ARDS
methylprednisolone (2 mg/kg/d) or placebo for 21 days, and were randomized to support by ECMO vs. conventional
showed improved oxygenation and more ventilator-free days treatments. This study showed significantly improved sur-
in the methylprednisolone group, but no significant improve- vival (63% vs. 47%, P = 0.03) at 6 months in the ECMO
ment in mortality [73]. Another RCT evaluated early cortico- group [90]. Although no definite conclusions can be drawn
steroid administration, and showed that methylprednisolone from this study [91], the results of this study suggest that
administration (1mg/kg/d) [74] less than 72 after the onset of ECMO can be used as a rescue therapy in cases of very se-
ARDS reduced mortality. However, these results should be vere ARDS.
interpreted with caution, because this study included a large Extracorporeal CO2 removal (ECCO2R) is an alter-
number of patients with septic shock. native device that uses veno-venous circuit for removal of
Conflicting data exist concerning the correlation be- CO2 at much lower extracorporeal flow rates compared to
tween corticosteroids and ICU neuromyopathy. A sub-analy- ECMO. A RCT conducted before the year 2000 used EC-
sis of study survivors did not show any association between CO2R and showed no effect on mortality [92]. In another
randomization to corticosteroids and increased risk of neuro- study ECCO2R was combined with low frequency positive
myopathy [75]. In conclusion, the relationship between cor- pressure ventilation (2 - 3 b/min), and showed improvement
ticosteroids and ICU neuromyopathy is an important issue in lung mechanics [93]. Overall, extracorporeal support
that deserves further study [76]. technologies produce significant temporary improvement in
Several other anti-inflammatory factors like Ibuprofen ARDS patients with severe respiratory dysfunction, but this
[77], ketoconazole [78], neutrophil elastase inhibitors (ONO improvement does not seem to affect outcome. New, well
5046) [79], NF-KB inhibitors [80], recombinant soluble conducted clinical studies are needed to better evaluate the
complement receptor-1 [81], and liposomal prostaglandin E1 role of ECMO and ECCO2R on survival in ARDS.
[82] have been evaluated in ARDS patients without success.
There are substantial evidences that b2-agonists may play a Conflicting data exist about the evolution ARDS mortality
potential role in the treatment of patients with ARDS. B2 ag- over time. A meta-analysis by Phua et al did not find any
onists have been found to have anti-inflammatory effects by mortality reduction in recent years [94], whereas another
direct influence on neutrophil function and by reducing the meta-analysis by Zambon et al. showed reduced mortality
secretion of several pro-inflammatory cytokines. Addition- in recent years [4]. In the past, several studies evaluated pat-
ally, b2-agonists can reduce the endothelial permeability and terns of ARDS mortality over time within the same institu-
stimulate the fluid clearance from the lungs [83]. In a small tion [95-99], and all studies, except for two [98, 99], found
RCT using the thermodilution method (PiCCO), intravenous decreasing mortality in ARDS. The observed discrepancy
salbutamol (15 μg/kg/h) use for seven days reduced extra- between different studies may be due to different investiga-
vascular lung water compared to placebo [84]. The effects tional methods, but we can conclude that ARDS mortality
of inhaled b-agonists have not, to this date, been adequately remains high (41 - 46%). Regardless of improvements in re-
evaluated, but will be investigated in an ongoing study con- cent years, ARDS mortality is higher in older patients and in
ducted by ARDS network (NCT00434993) medical patients [100]. However, the impact, if any, of newer
Several other pharmacological agents, including gluta- treatment strategies on ARDS mortality has not been evalu-
thione lisofylline [85], N-Acetylcysteine[86], and surfactant ated, because most studies are referred to the period before
Articles © The authors | Journal compilation © J Clin Med Res and Elmer Press™ | www.jocmr.org 11
Pierrakos et al J Clin Med Res • 2011;4(1):7-16
the year 2000. GR, Thompson BT, Brower RG, et al. Prognostic and
pathogenetic value of combining clinical and biochemi-
cal indices in patients with acute lung injury. Chest.
Conflicts of Interest 2010;137(2):288-296.
10. Gattinoni L, Caironi P, Pelosi P, Goodman LR. What has
This work was supported solely by Department funds. All computed tomography taught us about the acute respi-
authors state that they have no conflicts of interest to dis- ratory distress syndrome? Am J Respir Crit Care Med.
close. 2001;164(9):1701-1711.
11. Gattinoni L, Pelosi P, Pesenti A, Brazzi L, Vitale G,
Moretto A, Crespi A, et al. CT scan in ARDS: clini-
Abbreviations cal and physiopathological insights. Acta Anaesthesiol
Scand Suppl. 1991;95:87-94; discussion 94-86.
ARDS: Acute Respiratory Distress Syndrome; ECCO2R: 12. Tomashefski JF, Jr. Pulmonary pathology of acute
Extracorporeal CO2 Removal; ECMO: Extracorporeal respiratory distress syndrome. Clin Chest Med.
Membrane Oxygenation; HFV: High Frequency Ventila- 2000;21(3):435-466.
tion; HFOV: High Frequency Oscillatory Ventilation; ICU: 13. Zapol WM, Snider MT. Pulmonary hypertension
Intensive Care Unit; IL: Interleukin; iNO: Inhaled Nitric in severe acute respiratory failure. N Engl J Med.
Oxide;MOF: Multiple Organ Failure; NMBAs: Neuromus- 1977;296(9):476-480.
cular Blocking Agents; PEEP: Positive End Expiratory Pres- 14. Rocco PR, Dos Santos C, Pelosi P. Lung parenchyma
sure; PH: Pulmonary Hypertension; pRBCs: Packed Red remodeling in acute respiratory distress syndrome. Mi-
Blood Cells; RCT: Randomized Controlled Trial; TNF: Tu- nerva Anestesiol. 2009;75(12):730-740.
mor Necrosis Factor; TV: Tidal Volume 15. Martin C, Papazian L, Payan MJ, Saux P, Gouin F. Pul-
monary fibrosis correlates with outcome in adult respira-
tory distress syndrome. A study in mechanically venti-
References lated patients. Chest. 1995;107(1):196-200.
16. Meduri GU, Annane D, Chrousos GP, Marik PE, Sin-
1. Bernard GR. Acute respiratory distress syndrome: a clair SE. Activation and regulation of systemic inflam-
historical perspective. Am J Respir Crit Care Med. mation in ARDS: rationale for prolonged glucocorticoid
2005;172(7):798-806. therapy. Chest. 2009;136(6):1631-1643.
2. Ashbaugh DG, Bigelow DB, Petty TL, Levine BE. Acute 17. Kolobow T, Moretti MP, Fumagalli R, Mascheroni D,
respiratory distress in adults. Lancet. 1967;2(7511):319- Prato P, Chen V, Joris M. Severe impairment in lung
323. function induced by high peak airway pressure during
3. Bernard GR, Artigas A, Brigham KL, Carlet J, Falke K, mechanical ventilation. An experimental study. Am Rev
Hudson L, Lamy M, et al. The American-European Con- Respir Dis. 1987;135(2):312-315.
sensus Conference on ARDS. Definitions, mechanisms, 18. Brower RG, Shanholtz CB, Fessler HE, Shade DM,
relevant outcomes, and clinical trial coordination. Am J White P, Jr., Wiener CM, Teeter JG, et al. Prospective,
Respir Crit Care Med. 1994;149(3 Pt 1):818-824. randomized, controlled clinical trial comparing tradi-
4. Zambon M, Vincent JL. Mortality rates for patients tional versus reduced tidal volume ventilation in acute
with acute lung injury/ARDS have decreased over time. respiratory distress syndrome patients. Crit Care Med.
Chest. 2008;133(5):1120-1127. 1999;27(8):1492-1498.
5. Martin TR. Lung cytokines and ARDS: Roger S. Mitch- 19. Brochard L, Roudot-Thoraval F, Roupie E, Delclaux C,
ell Lecture. Chest. 1999;116(1 Suppl):2S-8S. Chastre J, Fernandez-Mondejar E, Clementi E, et al. Tid-
6. Windsor AC, Mullen PG, Fowler AA, Sugerman HJ. al volume reduction for prevention of ventilator-induced
Role of the neutrophil in adult respiratory distress syn- lung injury in acute respiratory distress syndrome. The
drome. Br J Surg. 1993;80(1):10-17. Multicenter Trail Group on Tidal Volume reduction in
7. Gadek JE, Pacht ER. The interdependence of lung an- ARDS. Am J Respir Crit Care Med. 1998;158(6):1831-
tioxidants and antiprotease defense in ARDS. Chest. 1838.
1996;110(6 Suppl):273S-277S. 20. Stewart TE, Meade MO, Cook DJ, Granton JT, Hodder
8. Moine P, McIntyre R, Schwartz MD, Kaneko D, Shen- RV, Lapinsky SE, Mazer CD, et al. Evaluation of a ven-
kar R, Le Tulzo Y, Moore EE, et al. NF-kappaB regu- tilation strategy to prevent barotrauma in patients at high
latory mechanisms in alveolar macrophages from pa- risk for acute respiratory distress syndrome. Pressure-
tients with acute respiratory distress syndrome. Shock. and Volume-Limited Ventilation Strategy Group. N Engl
2000;13(2):85-91. J Med. 1998;338(6):355-361.
9. Ware LB, Koyama T, Billheimer DD, Wu W, Bernard 21. Amato MB, Barbas CS, Medeiros DM, Magaldi RB,
12 Articles © The authors | Journal compilation © J Clin Med Res and Elmer Press™ | www.jocmr.org
Acute Respiratory Distress Syndrome -&OLQ0HG5HV Pierrakos et al -&OLQ0HG5HV
Schettino GP, Lorenzi-Filho G, Kairalla RA, et al. Ef- 33. Hodgson C, Keating JL, Holland AE, Davies AR, ventilation for acute respiratory distress syndrome in lated patients with severe sepsis and septic shock. Crit
fect of a protective-ventilation strategy on mortality in Smirneos L, Bradley SJ, Tuxen D. Recruitment ma- adults: a randomized, controlled trial. Am J Respir Crit &DUH0HG
WKHDFXWHUHVSLUDWRU\GLVWUHVVV\QGURPH1(QJO-0HG noeuvres for adults with acute lung injury receiving &DUH0HG 3RQWHV$UUXGD$ 'HPLFKHOH 6 6HWK$ 6LQJHU 3 7KH
mechanical ventilation. Cochrane Database Syst Rev. %ROOHQ&:YDQ:HOO*76KHUU\7%HDOH5-6KDK6 XVHRIDQLQÀDPPDWLRQPRGXODWLQJGLHWLQSDWLHQWVZLWK
22. Ventilation with lower tidal volumes as compared with &' Findlay G, Monchi M, et al. High frequency oscillato- acute lung injury or acute respiratory distress syndrome:
traditional tidal volumes for acute lung injury and the 3LHKO0$%URZQ568VHRIH[WUHPHSRVLWLRQFKDQJHV ry ventilation compared with conventional mechanical D PHWDDQDO\VLV RI RXWFRPH GDWD -3(1 - 3DUHQWHU (Q-
acute respiratory distress syndrome. The Acute Re- LQDFXWHUHVSLUDWRU\IDLOXUH&ULW&DUH0HG ventilation in adult respiratory distress syndrome: a ran- WHUDO1XWU
VSLUDWRU\ 'LVWUHVV 6\QGURPH 1HWZRUN 1 (QJO - 0HG GRPL]HGFRQWUROOHGWULDO>,65&71@&ULW&DUH 0LFKDHO -5 %DUWRQ 5* 6DIÀH -5 0RQH 0 0DUNH-
%ODQFK/0DQFHER-3HUH]00DUWLQH]00DV$%HW- 5 witz BA, Hillier K, Elstad MR, et al. Inhaled nitric ox-
23. Gattinoni L, Pelosi P, Crotti S, Valenza F. Effects of bese AJ, Joseph D, et al. Short-term effects of prone posi- :XQVFK + 0DSVWRQH - +LJKIUHTXHQF\ YHQWLODWLRQ ide versus conventional therapy: effect on oxygenation
positive end-expiratory pressure on regional distribu- tion in critically ill patients with acute respiratory distress versus conventional ventilation for treatment of acute LQ$5'6$P - 5HVSLU &ULW &DUH 0HG 3W
tion of tidal volume and recruitment in adult respira- V\QGURPH,QWHQVLYH&DUH0HG lung injury and acute respiratory distress syndrome. Co-
tory distress syndrome. Am J Respir Crit Care Med. *DWWLQRQL / &DUOHVVR ( 7DFFRQH 3 3ROOL ) *XHULQ FKUDQH'DWDEDVH6\VW5HY&' 7D\ORU5:=LPPHUPDQ-/'HOOLQJHU536WUDXEH5&
C, Mancebo J. Prone positioning improves survival 6XG66XG0)ULHGULFK-20HDGH02)HUJXVRQ1' Criner GJ, Davis K, Jr., Kelly KM, et al. Low-dose in-
&DLURQL3&UHVVRQL0&KLXPHOOR'5DQLHUL04XLQWHO in severe ARDS: a pathophysiologic review and in- :XQVFK+$GKLNDUL1.+LJKIUHTXHQF\RVFLOODWLRQLQ haled nitric oxide in patients with acute lung injury: a
M, Russo SG, Cornejo R, et al. Lung opening and clos- dividual patient meta-analysis. Minerva Anestesiol. patients with acute lung injury and acute respiratory dis- UDQGRPL]HGFRQWUROOHGWULDO-$0$
ing during ventilation of acute respiratory distress syn- tress syndrome (ARDS): systematic review and meta-
GURPH$P - 5HVSLU &ULW &DUH 0HG *DWWLQRQL/7RJQRQL*3HVHQWL$7DFFRQH30DVFKH- DQDO\VLV%0-F 'HOOLQJHU53=LPPHUPDQ-/7D\ORU5:6WUDXEH5&
roni D, Labarta V, Malacrida R, et al. Effect of prone 6LPPRQV 56 %HUGLQH ** 6HLGHQIHOG -- 3ULKRGD 7- Hauser DL, Criner GJ, Davis K, Jr., et al. Effects of in-
%URZHU 5* /DQNHQ 31 0DF,QW\UH 1 0DWWKD\ 0$ positioning on the survival of patients with acute respi- Harris GD, Smith JD, Gilbert TJ, et al. Fluid balance and haled nitric oxide in patients with acute respiratory dis-
Morris A, Ancukiewicz M, Schoenfeld D, et al. Higher UDWRU\IDLOXUH1(QJO-0HG the adult respiratory distress syndrome. Am Rev Respir tress syndrome: results of a randomized phase II trial.
versus lower positive end-expiratory pressures in pa- 38. Guerin C, Gaillard S, Lemasson S, Ayzac L, Girard R, 'LV ,QKDOHG1LWULF2[LGHLQ$5'66WXG\*URXS&ULW&DUH
WLHQWV ZLWK WKH DFXWH UHVSLUDWRU\ GLVWUHVV V\QGURPH 1 Beuret P, Palmier B, et al. Effects of systematic prone 6LOOLPDQ&&%RVKNRY/.0HKGL]DGHKNDVKL=(O]L'- 0HG
(QJO-0HG positioning in hypoxemic acute respiratory failure: a Dickey WO, Podlosky L, Clarke G, et al. Transfusion-re- 7URQF\ ( &ROOHW -3 6KDSLUR 6 *XLPRQG -* %ODLU /
0HDGH 02 &RRN '- *X\DWW *+ 6OXWVN\$6$UDEL UDQGRPL]HGFRQWUROOHGWULDO-$0$ lated acute lung injury: epidemiology and a prospective Ducruet T, Francoeur M, et al. Inhaled nitric oxide in
YM, Cooper DJ, Davies AR, et al. Ventilation strategy DQDO\VLV RI HWLRORJLF IDFWRUV %ORRG acute respiratory distress syndrome: a pilot randomized
using low tidal volumes, recruitment maneuvers, and 0DQFHER - )HUQDQGH] 5 %ODQFK / 5LDOS * *RUGR FRQWUROOHGVWXG\$P-5HVSLU&ULW&DUH0HG
high positive end-expiratory pressure for acute lung in- F, Ferrer M, Rodriguez F, et al. A multicenter trial of *LUDUG7'%HUQDUG*50HFKDQLFDOYHQWLODWLRQLQ$5'6 3W
jury and acute respiratory distress syndrome: a random- prolonged prone ventilation in severe acute respira- DVWDWHRIWKHDUWUHYLHZ&KHVW $GKLNDUL 1. %XUQV .( )ULHGULFK -2 *UDQWRQ -7
L]HGFRQWUROOHGWULDO-$0$ tory distress syndrome. Am J Respir Crit Care Med. .UHVV -3 3RKOPDQ$6 2¶&RQQRU 0) +DOO -% 'DLO\ Cook DJ, Meade MO. Effect of nitric oxide on oxygen-
*UDVVR 6 )DQHOOL 9 &DIDUHOOL $ $QDFOHULR 5 $PD- interruption of sedative infusions in critically ill pa- ation and mortality in acute lung injury: systematic re-
bile M, Ancona G, Fiore T. Effects of high versus low 7DFFRQH 3 3HVHQWL $ /DWLQL 5 3ROOL ) 9DJJLQHOOL ) WLHQWVXQGHUJRLQJPHFKDQLFDOYHQWLODWLRQ1(QJO-0HG YLHZDQGPHWDDQDO\VLV%0-
positive end-expiratory pressures in acute respira- Mietto C, Caspani L, et al. Prone positioning in pa- +HDUG62/RQJWLQH.7RWK,3X\DQD-&3RWHQ]D%
tory distress syndrome. Am J Respir Crit Care Med. tients with moderate and severe acute respiratory dis- 3XWHQVHQ&=HFK6:ULJJH+=LQVHUOLQJ-6WXEHU) 6P\UQLRV 1 7KH LQÀXHQFH RI OLSRVRPHHQFDSVXODWHG
tress syndrome: a randomized controlled trial. JAMA. 9RQ 6SLHJHO 7 0XW] 1 /RQJWHUP HIIHFWV RI VSRQWD- prostaglandin E1 on hydrogen peroxide concentrations
*DWWLQRQL / &DLURQL 3 5H¿QLQJ YHQWLODWRU\ WUHDWPHQW neous breathing during ventilatory support in patients in the exhaled breath of patients with the acute respira-
for acute lung injury and acute respiratory distress syn- GRV6DQWRV&&6OXWVN\$62YHUYLHZRIKLJKIUHTXHQF\ with acute lung injury. Am J Respir Crit Care Med. WRU\GLVWUHVVV\QGURPH$QHVWK$QDOJ
GURPH-$0$ ventilation modes, clinical rationale, and gas transport
(VDQ$+HVV'55DRRI6*HRUJH/6HVVOHU&16H- PHFKDQLVPV 5HVSLU &DUH &OLQ 1 $P *DLQQLHU 0 5RFK$ )RUHO -0 7KLULRQ ;$UQDO -0 6LREDO 06 3XOPRQDU\ YDVRGLODWRUV 5HVSLU &DUH
vere hypoxemic respiratory failure: part 1--ventilatory Donati S, Papazian L. Effect of neuromuscular block-
VWUDWHJLHV&KHVW +DPLOWRQ 33 2QD\HPL$ 6P\WK -$ *LOODQ -( &XW] ing agents on gas exchange in patients presenting with 'RHULQJ (% +DQVRQ &: UG 5HLO\ '- 0DUVKDOO &
%RUJHV -% 2NDPRWR 91 0DWRV *) &DUDPH] 03 E, Froese AB, Bryan AC. Comparison of conventional acute respiratory distress syndrome. Crit Care Med. Marshall BE. Improvement in oxygenation by phenyl-
Arantes PR, Barros F, Souza CE, et al. Reversibility and high-frequency ventilation: oxygenation and lung ephrine and nitric oxide in patients with adult respiratory
of lung collapse and hypoxemia in early acute respira- SDWKRORJ\-$SSO3K\VLRO3W 6HVVOHU &1 7UDLQRIIRXU WR PRQLWRU QHXURPXVFXODU GLVWUHVVV\QGURPH$QHVWKHVLRORJ\
tory distress syndrome. Am J Respir Crit Care Med. &KDQ.36WHZDUW7(0HKWD6+LJKIUHTXHQF\RVFLO- EORFNDGH"&KHVW 5RFK$3DSD]LDQ/%UHJHRQ)*DLQQLHU00LFKHOHW3
latory ventilation for adult patients with ARDS. Chest. *DGHN-('H0LFKHOH6-.DUOVWDG0'3DFKW(5'R- 7KLULRQ;6DX[3HWDO+LJKRUORZGRVHVRIDOPLWULQH
31. Gattinoni L, Caironi P, Cressoni M, Chiumello D, Ran- nahoe M, Albertson TE, Van Hoozen C, et al. Effect of bismesylate in ARDS patients responding to inhaled
LHUL904XLQWHO05XVVR6HWDO/XQJUHFUXLWPHQWLQ +HVV'0DVRQ6%UDQVRQ5+LJKIUHTXHQF\YHQWLOD- enteral feeding with eicosapentaenoic acid, gamma-lin- 12DQGUHFHLYLQJQRUHSLQHSKULQH",QWHQVLYH&DUH0HG
SDWLHQWVZLWKWKHDFXWHUHVSLUDWRU\GLVWUHVVV\QGURPH1 WLRQ GHVLJQ DQG HTXLSPHQW LVVXHV 5HVSLU &DUH &OLQ 1 olenic acid, and antioxidants in patients with acute re-
(QJO-0HG $P VSLUDWRU\GLVWUHVVV\QGURPH(QWHUDO1XWULWLRQLQ$5'6 9LQFHQW-//LJQLDQ+*LOOHW-%%HUUH-&RQWX(,Q-
32. Fan E, Wilcox ME, Brower RG, Stewart TE, Mehta S, 5LPHQVEHUJHU 3& ,&8 FRUQHUVWRQH KLJK IUHTXHQF\ 6WXG\*URXS&ULW&DUH0HG crease in PaO2 following intravenous administration
Lapinsky SE, Meade MO, et al. Recruitment maneuvers YHQWLODWLRQLVKHUHWRVWD\&ULW&DUH 3RQWHV$UUXGD$$UDJDR$0$OEXTXHUTXH-'(IIHFWV of propranolol in acutely hypoxemic patients. Chest.
for acute lung injury: a systematic review. Am J Respir 'HUGDN60HKWD66WHZDUW7(6PLWK75RJHUV0%X- of enteral feeding with eicosapentaenoic acid, gamma-
&ULW&DUH0HG chman TG, Carlin B, et al. High-frequency oscillatory linolenic acid, and antioxidants in mechanically venti- YRQ%LVPDUFN3.OHPP.*DUFLD:LVWDGW&):LQRWR
Articles © The authors | Journal compilation © J Clin Med Res and Elmer Press™ | www.jocmr.org Articles © The authors | Journal compilation © J Clin Med Res and Elmer Press™ | www.jocmr.org
Acute Respiratory Distress Syndrome -&OLQ0HG5HV Pierrakos et al -&OLQ0HG5HV
0RUEDFK 6 6FKXW]H 6 .UDXVH 0) 6HOHFWLYH 1)NDS- WUHVVV\QGURPH&ULW&DUH0HG 3KXD - %DGLD -5$GKLNDUL 1. )ULHGULFK -2 )RZOHU Caldwell ES, Steinberg KP. Causes and timing of death
paB inhibition, but not dexamethasone, decreases acute 82. Vincent JL, Brase R, Santman F, Suter PM, McLuckie A, RA, Singh JM, Scales DC, et al. Has mortality from LQSDWLHQWVZLWK$5'6&KHVW
OXQJ LQMXU\ LQ D QHZERUQ SLJOHW DLUZD\ LQÀDPPDWLRQ Dhainaut JF, Park Y, et al. A multi-centre, double-blind, acute respiratory distress syndrome decreased over (DFKHPSDWL65+\GR/-6KRX-%DULH362XWFRPHV
PRGHO3XOP3KDUPDFRO7KHU placebo-controlled study of liposomal prostaglandin E1 time?: A systematic review. Am J Respir Crit Care Med. of acute respiratory distress syndrome (ARDS) in elder-
0HGXUL*80XWKLDK03&DUUDWX3(OWRUN\0&KURX- 7/& & LQ SDWLHQWV ZLWK DFXWH UHVSLUDWRU\ GLVWUHVV O\SDWLHQWV-7UDXPD
VRV *3 1XFOHDU IDFWRUNDSSD% DQG JOXFRFRUWLFRLG UH- V\QGURPH,QWHQVLYH&DUH0HG 0LOEHUJ -$ 'DYLV '5 6WHLQEHUJ .3 +XGVRQ /' 1DYDUUHWH1DYDUUR35RGULJXH]$5H\QROGV1:HVW5
ceptor alpha- mediated mechanisms in the regulation 83. Perkins GD, Gao F, Thickett DR. In vivo and in vitro ef- Improved survival of patients with acute respira- +DEDVKL15LYHUD5&KLX:&HWDO$FXWHUHVSLUDWRU\
RIV\VWHPLFDQGSXOPRQDU\LQÀDPPDWLRQGXULQJVHSVLV fects of salbutamol on alveolar epithelial repair in acute WRU\ GLVWUHVV V\QGURPH $5'6 -$0$ distress syndrome among trauma patients: trends in ICU
and acute respiratory distress syndrome. Evidence for OXQJLQMXU\7KRUD[ mortality, risk factors, complications and resource utili-
LQÀDPPDWLRQLQGXFHGWDUJHWWLVVXHUHVLVWDQFHWRJOXFR- 3HUNLQV *' 0F$XOH\ ') 7KLFNHWW '5 *DR ) 7KH $EHO 6- )LQQH\ 6- %UHWW 6- .HRJK %) 0RUJDQ &- ]DWLRQ,QWHQVLYH&DUH0HG
FRUWLFRLGV 1HXURLPPXQRPRGXODWLRQ beta-agonist lung injury trial (BALTI): a randomized Evans TW. Reduced mortality in association with the 100.Suchyta MR, Clemmer TP, Elliott CG, Orme JF, Jr.,
338. placebo-controlled clinical trial. Am J Respir Crit Care acute respiratory distress syndrome (ARDS). Thorax. Morris AH, Jacobson J, Menlove R. Increased mortal-
%HUQDUG*5/XFH-06SUXQJ&/5LQDOGR-(7DWH50 0HG ity of older patients with acute respiratory distress syn-
Sibbald WJ, Kariman K, et al. High-dose corticosteroids 5DQGRPL]HG SODFHERFRQWUROOHG WULDO RI OLVRI\OOLQH IRU 6WDSOHWRQ5':DQJ%0+XGVRQ/'5XEHQIHOG*' GURPH&KHVW
in patients with the adult respiratory distress syndrome. early treatment of acute lung injury and acute respiratory
1(QJO-0HG GLVWUHVVV\QGURPH&ULW&DUH0HG
6WHLQEHUJ .3 +XGVRQ /' *RRGPDQ 5% +RXJK &/ 0RUDGL 0 0RMWDKHG]DGHK 0 0DQGHJDUL $ 6ROWDQ
/DQNHQ31+\]\57KRPSVRQ%7HWDO(I¿FDF\DQG 6KDUL¿ 06 1DMD¿$ .KDMDYL 05 +DMLEDED\HH 0 HW
safety of corticosteroids for persistent acute respiratory al. The role of glutathione-S-transferase polymorphisms
GLVWUHVVV\QGURPH1(QJO-0HG on clinical outcome of ALI/ARDS patient treated with
1DFHW\OF\VWHLQH5HVSLU0HG
0HGXUL*8*ROGHQ()UHLUH$;7D\ORU(=DPDQ0 6SUDJJ 5* /HZLV -) :DOPUDWK +' -RKDQQLJPDQ
Carson SJ, Gibson M, et al. Methylprednisolone infu- J, Bellingan G, Laterre PF, Witte MC, et al. Effect of
sion in early severe ARDS: results of a randomized con- recombinant surfactant protein C-based surfactant on
WUROOHGWULDO&KHVW WKHDFXWHUHVSLUDWRU\GLVWUHVVV\QGURPH1(QJO-0HG
+RXJK&/6WHLQEHUJ.37D\ORU7KRPSVRQ%5XEHQ-
feld GD, Hudson LD. Intensive care unit-acquired neu- 1HZODQG3(([WUDFRUSRUHDOPHPEUDQHR[\JHQDWLRQLQ
romyopathy and corticosteroids in survivors of persis- the treatment of respiratory failure--a review. Anaesth
WHQW$5'6,QWHQVLYH&DUH0HG ,QWHQVLYH&DUH
0HGXUL *8 0DULN 3( &KURXVRV *3 3DVWRUHV 60 =DSRO :0 6QLGHU 07 +LOO -' )DOODW 5- %DUWOHWW
Arlt W, Beishuizen A, Bokhari F, et al. Steroid treat- RH, Edmunds LH, Morris AH, et al. Extracorpo-
ment in ARDS: a critical appraisal of the ARDS net- real membrane oxygenation in severe acute respira-
work trial and the recent literature. Intensive Care Med. tory failure. A randomized prospective study. JAMA.
%HUQDUG*5:KHHOHU$35XVVHOO-$6FKHLQ56XP- 3HHN*-(OERXUQH'0XJIRUG07LUXYRLSDWL5:LO-
mer WR, Steinberg KP, Fulkerson WJ, et al. The effects son A, Allen E, Clemens F, et al. Randomised controlled
of ibuprofen on the physiology and survival of patients trial and parallel economic evaluation of conventional
ZLWK VHSVLV 7KH ,EXSURIHQ LQ 6HSVLV 6WXG\ *URXS 1 ventilatory support versus extracorporeal membrane ox-
(QJO-0HG ygenation for severe adult respiratory failure (CESAR).
.HWRFRQD]ROH IRU HDUO\ WUHDWPHQW RI DFXWH OXQJ LQMXU\ +HDOWK7HFKQRO$VVHVV
and acute respiratory distress syndrome: a random- 0RUDQ -/ &KDOZLQ 53 *UDKDP 3/ ([WUDFRUSRUHDO
L]HG FRQWUROOHG WULDO 7KH $5'6 1HWZRUN -$0$ membrane oxygenation (ECMO) reconsidered. Crit
&DUH5HVXVF
.DGRL<+LQRKDUD+.XQLPRWR)6DLWR6*RWR).R- 0RUULV $+ :DOODFH &- 0HQORYH 5/ &OHPPHU 73
VDND7,HWD.3LORWVWXG\RIWKHHIIHFWVRI212LQ 2UPH-)-U:HDYHU/.'HDQ1&HWDO5DQGRPL]HG
patients with acute respiratory distress syndrome. Anes- clinical trial of pressure-controlled inverse ratio venti-
WK$QDOJWDEOHRIFRQWHQWV lation and extracorporeal CO2 removal for adult respi-
80. Lee HS, Lee JM, Kim MS, Kim HY, Hwangbo B, Zo JI. ratory distress syndrome. Am J Respir Crit Care Med.
Low-dose steroid therapy at an early phase of postop- 3W
erative acute respiratory distress syndrome. Ann Thorac *DWWLQRQL/3HVHQWL$0DVFKHURQL'0DUFROLQ5)X-
6XUJ magalli R, Rossi F, Iapichino G, et al. Low-frequency
81. Zimmerman JL, Dellinger RP, Straube RC, Levin JL. positive-pressure ventilation with extracorporeal CO2
Phase I trial of the recombinant soluble complement re- removal in severe acute respiratory failure. JAMA.
ceptor 1 in acute lung injury and acute respiratory dis-
Articles © The authors | Journal compilation © J Clin Med Res and Elmer Press™ | www.jocmr.org Articles © The authors | Journal compilation © J Clin Med Res and Elmer Press™ | www.jocmr.org
Eur Respir J, 1997; 10: 219–225 Copyright ERS Journals Ltd 1997
DOI: 10.1183/09031936.97.10010219 European Respiratory Journal
Printed in UK - all rights reserved ISSN 0903 - 1936
Physiology and pathophysiology of pleural fluid turnover. G. Miserocchi. ©ERS Journals Istituto di Fisiologia Umana, Università
Ltd 1997. degli Studi, Milano, Italy
ABSTRACT: The pleural space contains a tiny amount (≈0.3 mL·kg-1) of hypo-
oncotic fluid (≈1 g·dL-1 protein). Pleural fluid turnover is estimated to be ≈0.15 Correspondence: G. Miserocchi
Istituto di Fisiologia Umana
mL·kg-1·h-1. Pleural fluid is produced at parietal pleural level, mainly in the less Via Mangiagalli 32
dependent regions of the cavity. Reabsorption is accomplished by parietal pleur- 20133 Milano
al lymphatics in the most dependent part of the cavity, on the diaphragmatic sur- Italy
face and in the mediastinal regions. The flow rate in pleural lymphatics can increase
in response to an increase in pleural fluid filtration, acting as a negative feedback Keywords: Fluid dynamics
mechanism to control pleural liquid volume. Such control is very efficient, as a 10 lung interstitium
fold increase in filtration rate would only result in a 15% increase in pleural liq- lung microvessels
uid volume. When filtration exceeds maximum pleural lymphatic flow, pleural effu- lymphatics
micropuncture
sion occurs: as an estimate, in man, maximum pleural lymph flow could attain 30
mL·h-1, equivalent to ≈700 mL·day-1 (≈40% of overall lymph flow). Received: November 2 1995
Under physiological conditions, the lung interstitium and the pleural space behave Accepted after revision July 31 1996
as functionally independent compartments, due to the low water and solute per-
meability of the visceral pleura. Pleural fluid circulates in the pleural cavity and
intrapleural fluid dynamics may be represented by a porous flow model. Lubrication
between lung and chest wall is assured by oligolamellar surfactant molecules strat-
ified on mesothelial cells of the opposing pleurae. These molecules carry a charge
of similar sign and, therefore, repulse each other, assuring a graphite-like lubri-
cation.
Eur Respir J., 1997; 10: 219–225.
A review on pleural space is certainly most welcome, the visceral pleura [1] would imply that pleural liquid
for the simple reason that on opening a textbook of phy- protein concentration would keep increasing with age,
siology the concepts concerning pleural fluid turnover but again there are no indications that this occurs. The
date back to 1927 [1]. The same comment, in fact, applies existence of partial restriction to the movement of large
in general to microvascular water and solute exchange. solutes, compared to that of water molecules, posed sci-
This appears particularly misleading, considering the entists the major problem of explaining how interstitial
major advances achieved in the field over the last 70 volume and protein concentration are kept fairly steady.
yrs. Around the turn of the 19th century, STARLING and Ideologically, this led to re-evaluation of lymphatics as
TUBBY [2] interpreted microvascular fluid and solute the major route for interstitial fluid drainage. In fact,
exchange as resulting from the balance between hydraulic since lymphatics do not sieve proteins, they leave the
and colloidosmotic pressures. This concept is still valid interstitial protein concentration unaltered the latter dep-
today, but the complete formulation of transmicrovas- ending only upon the sieving properties of the filtering
cular exchange has become much more complex because membrane. Accordingly, the description of interstitial
water crosses biological membranes more easily than fluid homeostasis under steady state conditions is now
large solutes (namely, plasma proteins) do [3]. In fact, dif- being explained as a balance between capillary filtra-
ferent equations describe water and solute fluxes [4]. tion and lymphatic absorption. Major validation of this
As a result, the general model of transcapillary fluid model has come from the accumulating experimental
exchange still taught to students, based on fluid filtra- data over the past 30 yrs concerning interstitial tissues
tion at the arteriolar end of the capillary bed and reab- [4] and serous spaces [5]. Interestingly, since the pio-
sorption at the venular end, appears rather simplistic. neering, work of STARLING and TUBBY [2] the pleural
Due to the different permeability to water and solutes, space has been used as a useful experimental model to
one could predict on a mathematical basis that such a study the interaction between microvascular filtration
model would lead to a progressive increase in intersti- and lymphatic drainage.
tial protein concentration over time [3, 4], a condition This article presents an integrated view of pleural fluid
that cannot be confirmed experimentally. turnover, based on data gathered from experimental stud-
Similarly, the old hypothesis claiming that pleural ies on animals over the last 15 yrs. The situation in man
fluid filters at parietal level and is reabsorbed through is, unfortunately, still poorly defined; yet, where possible,
220 G . MISEROCCHI
extrapolations will be made in an attempt to relate the surface to 8,000·cm-2 on the diaphragm and their diam-
present state of knowledge to human pleural patho- eter averages 1 µm (range <1–~40 µm) in size. A sim-
physiology. ilar anatomical disposition was seen for the peritoneal
cavity. Lacunae in man have not been definitely demon-
strated, although their existence is probable. Furthermore,
The pleural compartments they were found in mammals as large as sheep [13].
Mesothelial cells are only about 4 µm thick [11], and
The pleural space, like other serous cavities of the connect to each other by tight junctions on the luminal
body, may be considered an enlarged tissue space. In side and by desmosomes on the basal portion of the inter-
fact, unlike a common interstitial space, it presents a cellular junction [11].
higher ratio of free fluid to solid tissue volume. Solid Microvilli 1–3 µm long are seen on mesothelial cells,
volume may be represented by cells being present in varying in density from 2 to 30 per µm2, and trap high
the pleural fluid and microvilli of mesothelial cells. The concentrations of glycoproteins and hyaluronic acid [11,
pleural space actually contains a tiny amount of fluid, 14]. The thickness of the pleurae is quite variable among
(≈0.3 mL·kg-1 body mass, with a low protein concen- species: in animals with thin pleurae (dogs, rabbits and
tration (≈1 g·dL-1, [6, 7]); the latter appears a peculiar cats [15]), the thickness of the submesothelial intersti-
feature considering that, in physiological conditions, the tium is equal for parietal and visceral pleura, averaging
pressure of the pleural fluid is subatmospheric. In fact, 20 µm; however, in animals with thick pleurae (sheep,
when a subatmospheric pressure is found in other tis- pig, horses and humans), it is about five times thinner
sues, like the lung interstitium for example, protein con- in the parietal compared to the visceral pleura, where it
centration increases and, furthermore, it is well-known can attain about 100 µm [13, 14, 16].
that an opposite condition, namely tissue hyperhydra- It might be useful to recall briefly the concepts con-
tion, causes a decrease in interstitial protein concentra- cerning fluid and plasma protein flux across biological
tion. membranes. All acceptable description of water flux
Figure 1 is a simplified schema of pleuropulmonary (normally indicated as Jv) between two compartments
compartments. Considering the anatomical arrangements, labelled 1 and 2 is given by the revised Starling law:
it appears that five compartments are involved: the pari-
etal systemic microcirculation; the parietal interstitial Jv = Kf [(PH1- PH2) - σ (π1 −π2)]
space; the pleural cavity itself; the lung interstitium; and
the visceral microcirculation (either systemic from bron-
chial artery or pulmonary). The membranes separating where Kf is the filtration coefficient, PH and π are the
such compartments are: the capillary endothelium (on hydraulic and colloidosmotic pressures, and σ is the
parietal and visceral side); and the parietal and the vis- solute reflection coefficient of the membrane. The coef-
ceral mesothelia. The lymphatics provide drainage of ficient, σ, is a number varying from 0 to 1. For σ=1,
the interstitial spaces but also of the pleural cavity, as the radius of the solute (we consider plasma proteins)
they open directly on the parietal pleura (lymphatic sto- is larger than that of the pores of the membrane and,
mata). Stomata, in rabbits and sheep [8–11], are fre- therefore, no solute flux can occur through the mem-
quently grouped in clusters and connect to an extensive brane. For σ=0, the radius of the pores of the membrane
network of submesothelial lacunae [11, 12]. They range is large enough to allow solute to cross the membrane.
in density from 100 stomata·cm-2 on the intercostal For 0 <σ < 1, there is partial restriction to solute move-
ment. The description of solute flux is rather difficult,
as it occurs partly via the water flux and partly down a
Parietal Visceral diffusion concentration gradient [3].
pleura pleura
Basal lamina
Pleural
space Pulmonary The old hypothesis on pleural fluid turnover
Extrapleural s.c. interstitium
parietal In 1927, based on the Starling hypothesis of fluid
p.c.
p.c. exchange, NEERGARD [1] proposed the hypothesis that
interstitium
pleural fluid turnover is entirely dependent upon the dif-
Parietal Alveolus ference between hydraulic and colloidosmotic pressure.
lymphatic Although provocative and relatively unchallenged for
over half a century, this model appears today simplistic
and untenable, as it neglects the existence of interstital
Stoma Pulmonary lymphatic spaces, the permselectivity to water and solutes, and the
existence of pleural lymphatics. NEERGARD [1] reasoned
that pleural fluid filters at parietal level because hydraulic
Microvilli
pressure in systemic microcirculation exceeds colloi-
dosmotic pressure; conversely, fluid is reabsorbed at vis-
ceral level because in the pulmonary microcirculation
Unidirectional
valve the opposite is true. This hypothesis was developed by
AGOSTONI et al. [17], who found that the difference
Fig. 1. – Schema of the morphofunctional design of pleural space; between hydraulic and colloidosmotic pressure in the
s.c.: systemic capillary; p.c.: pulmonary capillary. pulmonary capillaries was large enough to account for
PLEURAL FLUID TURNOVER 221
a) Parietal Visceral a relatively small pressure gradient (fig. 2b). The aver-
pleura pleura age filtration rate decreases with increasing animal size
Systematic Pulmonary and ranges ~0.1–0.02 mL·kg-1·h-1, from rabbits to dogs.
capillary capillary It is important to recall that, in order to derive indica-
tions on pleural fluid turnover, one must carry out expe-
Pleural riments in conditions close to the physiological one that
liquid is characterized by two main features: namely, a small
pleural liquid volume and subatmospheric pleural liquid
pressures, such as those occurring during spontaneous
breathing. In experimental models implying large pleu-
ral effusions and mechanical ventilation [24, 25], the
database is difficult to interpret as it does not reflect a
steady state situation.
b) Parietal Visceral Comparative studies (with animals ranging from a few
pleura pleura gram to 50 kg body mass) reveal that pleural fluid turn-
Systemic capillary Pulmonary capillary over (normalized to body mass) decreases with increas-
Jv=0.153 ing size [5]. In fact, it was found that with increasing
Parietal mass: 1) pleural liquid pressure (considered at the level
interstitium Lung of the right atrium) becomes more negative; and 2) pleur-
Jv=0.194 Pleural Jv=0.05 interstitium
liquid al fluid volume (normalized to body mass) and protein
Jv=0.041 concentration decrease. Thus, it appears that, with increas-
Lung ing body mass, the tendency develops for a greater "dehy-
lymphatics dration" of the pleural space. At the extreme of this scale,
anatomical records indicate that in animals as large as
Lymphatics elephants, the pleural cavity is obliterated.
From the biophysical standpoint, the parietal mesothe-
[Jv]=mL·kg-1·h-1 lium can be modelled as a membrane with few but large
Fig. 2. – a) Neergard's hypothesis (1927) [1]. The old hypothesis pores: this is reflected by a low σ value (≈0.3) but also
concerning pleural fluid turnover, which neglects the existence of by a low solute permeability coefficient [7]. This fea-
interstitial spaces and pleural lymphatics. b) the present view of pleur- ture makes it possible to sieve proteins efficiently, so
al fluid turnover, based on the available experimental database gath- that the protein concentration of pleural fluid is rather
ered with minimally invasive techniques in the rabbit. The pleural
space and the pulmonary interstitium are, under normal conditions, low (about 1 g·dL-1), smaller than the protein concen-
two functionally separate compartments, fluid filters from the micro- tration in the extrapleural parietal interstitium (about 2.5
circulation and is drained by the lymphatics. Subatmospheric hydraulic g·dL-1 [7]). Extrapolation to humans is difficult. Further-
pressures are set by the action of the lymphatic pump. The figure more, in humans, there is the suggestion that some fil-
shows water flux under steady state conditions. Solute (plasma pro-
tein) flux has been omitted. Although figures will vary in man, the tration may occur from the visceral pleura, as its blood
overall situation is unlikely to be qualitatively different based on supply stems from the systemic circulation where func-
anotomofunctional considerations. tional hydrostatic pressure is higher compared to pul-
the subatmospheric pressure of the pleural liquid. This monary circulation; this fact however, may be of no
old hypothesis (fig. 2a) is no longer valid today based relevance as the key variable in this respect is pulmonary
on the present state of knowledge. In fact, a thorough interstitial pressure. The latter was directly measured in
analysis of fluid (and solute) flux has required the devel- rabbits, by using the micropuncture technique through
opment of sophisticated techniques to measure hydraulic the pleural window approach, and was found to be rather
and colloidosmotic pressures in the various pleural com- subatmospheric (≈-10 cmH2O [26]). The same technique
partments [18, 19], to estimate water and solute per- made it possible to describe the pulmonary microvas-
meability coefficients [7] of the membranes separating cular pressure profile [27]. Based on these findings, it
the compartments, and, finally, to measure pleural lym- appears that fluid normally filters from pulmonary micro-
phatic flow in conditions very close to the physiologi- vessels to the lung interstitium. If pulmonary intersti-
cal ones [20–23]. The whole mosaic of knowledge is tial and microvascular pressures in humans were similar
fairly complete in rabbits, where most of the techniques to those measured in rabbits, no gradient should be pre-
were developed by preserving the integrity of the pleural sent to cause fluid filtration through the visceral pleura.
space, and the available information has made it possi-
ble to develop a model of pleural fluid turnover that can
be extended to animals as large as sheep and is com- Pleural fluid drainage
patible with pathophysiological observations in man.
Absorption flows through the visceral pleura are nor-
mally negligible, so that, under normal conditions, the
Pleural fluid filtration pleural space and the pulmonary interstitium are two
functionally separate compartments. This should also be
From experimental determination of variables and true in man, given the great thickness of the visceral
coefficients shown in the Starling equation [5, 7], it pleura, implying a low water and solute permeability.
appears that pleural fluid is filtered at parietal pleural Most pleural fluid drainage (≈75%), estimated from clear-
level from systemic microvessels to the extrapleural ance of labelled proteins in rabbits and dogs, occurs thr-
interstitium and from there into the pleural space down ough parietal pleural lymphatics [22, 23]. Controversial
222 G . MISEROCCHI
data on pleural lymph flow were gathered in sheep, nevertheless, although numbers may vary somewhat rel-
where the same group of investigators found on differ- ative to other species, the situation is unlikely to be
ent occasions that pleural lymphatics would drain either qualitatively different based on anatomofunctional con-
<1% [25] or ≈70% of pleural fluid [24]. siderations and on data gathered in large mammals.
Pleural lymphatics are able to generate a subatmos-
pheric pressure [21] of about -10 cmH2O. Furthermore,
they can increase the flow rate by about 20 fold in res- Control and pathophysiology
ponse to an increase in pleural liquid volume following
an increase in pleural fluid filtration [20]. Lymphatic Another implication of the original hypothesis of
activity is pulsatile in nature, due partly to myogenic NEERGARD [1] was that pleural liquid volume and pro-
rhythmic contraction of the smooth muscles of the tein concentration should vary on changing the Starling
lymphatic walls (intrinsic activity), and partly to tissue balance of pressure across the pleurae. However, both
pressure oscillations related to respiratory movements experimental evidence and medical practice indicate that
(extrinsic mechanism). The two mechanisms account for pleural fluid volume and composition are highly stable
about 40 and 60%, respectively, of total pleural lym- and, in fact, pleural effusions develop when dramatic
phatic flow under physiological conditions, the stroke changes in fluid and solute homeostasis occur. Therefore,
volume for each initial lymphatic stoma being of the order this suggests that some sort of mechanism ought to exist
of 1×10-6 µL·stoma [22]. Some features of the drainage to guarantee a tight control on pleural fluid volume and
are peculiar, as a greater lymphatic drainage occurs in protein concentration. The same happens, of course, in
the lowermost parts of the cavity [28], over the diaphrag- other tissues, although the features of the control vary.
matic surface and in the mediastinal regions [18]. Since In the lung parenchyma, for example, a very tight
filtration and absorption sites are different, this implies control exists to guarantee a minimum amount of inter-
that pleural fluid circulates within the pleural space (see stitial fluid [26, 29] which is crucial to assure gas dif-
below under "Intrapleural fluid dynamics"). fusion. The comparison between pleural space and lung
Figure 2a presents the old hypothesis of pleural fluid interstitium offers, in fact, an important clue to the under-
turnover, whilst the present model based on experimental standing of how a condition of minimum interstitial vol-
results is shown in figure 2b. Fluid filters from the pari- ume is achieved. The pulmonary interstitium has a very
etal microcirculation into the pleural space and is drained low mechanical compliance; accordingly, when facing
by lymphatics: the new feature is that flows, rather than a condition of increased filtration, this leads to a marked
pressure gradients, are presented in this figure in order increase in pulmonary interstitial pressure [29]. This rep-
to describe the pleural fluid turnover rate. The figure resents the so-called "tissue safety factor" against the
also shows that liquid filters from the pulmonary capil- development of pulmonary oedema, as it opposes, based
laries into the lung interstitium, from where it is removed on the Starling balance of pressures, a further filtration.
by the lymphatics. The latter point also represents a new However, due to its high compliance, the pleural space
piece of knowledge [26], based strongly on interstitial has no "tissue safety factor"; therefore, the only mech-
pressure measurements performed with lungs physiolo- anism assuring a control on minimum pleural liquid vol-
gically expanded in the pleural space. The important ume is represented by lymphatic drainage. Since pleural
concept to grasp here is that the lymphatics represent a lymphatics can increase the flow in response to an
drainage mechanism, able to generate a subatmospheric increase in pleural liquid volume, they represent a neg-
pressure (like a vacuum cleaner). From a mechanistic ative feedback mechanism controlling pleural liquid vol-
standpoint, the lymphatics set the pressure in the compart- ume, as they tend to offset the induced perturbation
ment to be drained; therefore, they establish an impor- Such a system seems to be highly efficient. In fact,
tant variable appearing in the Starling pressure balance relatively wide variations in pleural filtration rate result
equation, namely interstitial hydraulic pressure.
As mentioned previously, similar concepts for fluid
turnover are now also being shared for other interstitia Power
Lymphatic
and serous cavities [4], in fact, such a system allows a pump
close control on interstitial volume and protein concen- Filtration
tration. It is worth recalling here that when lungs are
physiologically expanded in the chest, alveolar pressure
is equal to atmospheric and pleural pressure is subat- Parietal
mospheric, so that, from the mechanical standpoint, the pleura
Buoy
lung interstitium is subject to a tensile stress. Note that,
if pulmonary interstital pressure were to be measured
in isolated lungs passively inflated at positive alveolar
pressure, a compressive stress would be applied to the
lung tissue and this may raise interstitial pressure above
atmospheric; clearly, such data bear little relationship Visceral
to the physiological situation. Finally, as can be appre- pleura
ciated from the available database, it is clear that no re-
absorption of pleural fluid can occur into the pulmonary Fig. 3. – Simple schema to depict the features of lymphatic control.
Lymphatics act as a pump controlled by a rheostat that can increase
capillaries, as hypothesized by NEERGARD [1] in 1927. flow rate in response to an increase in pleural liquid volume. When
In man no database is available to attempt an analysis the level of the buoy rises, the power to the pump increases and, cor-
of fluid turnover in the pleuropulmonary compartments; respondingly, the lymph flow increases also.
PLEURAL FLUID TURNOVER 223
in minor deviations from steady state pleural liquid vol- maximum lymph flow, transudate would then form. Exu-
ume [30], the latter being maintained at a minimum va- date would mostly form when protein permeability of
lue. Figure 3 shows a simple functional schema of the the systemic capillaries is increased; a comparable in-
pleural space and of pleural lymphatics. The visceral crease in mesothelial protein permeability would only
pleura is shown as a high resistance pathway as opposed cause a modest increase in pleural liquid protein concen-
to the parietal pleura and to the parietal lymphatics. tration because interstitial protein concentration is low
Parietal lymphatics act as a pump controlled by a rheo- already. Again, pleural effusion of exudate type would
stat; when the level of the buoy rises (as a consequence occur when filtration rate exceeds maximum lymph flow.
of an increased filtration), the power to the pump increa- It must be remembered that on biophysical grounds an
ses and, correspondingly, lymph flow increases also. increase in solute, compared to water permeability, ref-
To give an idea of such a regulatory mechanism, one lects a more severe lesion of the membrane. Finally, it
may consider that for a 10 fold increase in filtration rate, must be noted that the extrapleural interstitium exerts a
the steady state pleural liquid volume would only increase potent buffer action against the increase in pleural fil-
by 15–20% [30], undetectable by common X-rays. The tration rate; indeed, due to its low compliance, an increased
same regulatory capacity is retained even if the maxi- capillary filtration leads to a marked increase in inter-
mum lymph flow rate is decreased 10 fold. It must be stitial pressure, thus opposing further capillary filtration.
emphasized, however, that the increase in lymph flow
to attain its maximum (about 20 fold) is smaller than
the potential increase in Starling dependent filtration Pleural space and lung pathophysiology
flows. In fact, an increase in filtration rate by two orders
of magnitude may result from pathological conditions Some controversies should be reported concerning the
(e.g. inflammation, right heart failure); clearly, an incre- permeability of the visceral pleura. Results from stud-
ase in pleural filtration rate beyond the maximum pleu- ies on stripped specimens of the visceral pleura [33]
ral lymph flow results in pleural effusion. Thus, although suggest that permeability of the visceral pleura to water
lymph flow may increase substantially, lymphatics can- and solutes is fairly high. Contrary to these conclusions
not cope with a massive increase in filtration rate. It are the results from experiments where the integrity of
appears, therefore, more appropriate to regard lympha- the pleural space and spontaneous respiration were pre-
tics as a system mostly effective in controlling pleural served; in these conditions, it was found that albumin
liquid volume close to steady state conditions. transport from pleural space to pulmonary interstitium
These conclusions may be valid in humans too, although accounted for less than 20% (an overestimate) of total
some reservation is due because of the extrapolation. albumin removal from the pleural space with close to
Let us consider a basal pleural lymph flow rate equal normal pleural liquid volume [23]. Thus, the finding of
to that measured in dogs (0.02 mL·kg-1·h-1: for a 70 kg a high permeability of the visceral pleura may suggest
man this would correspond to 1.4 mL-1 or about 34 that experimental damage was induced. In line with the
mL·day-1, about 2% of the overall daily lymphatic flow hypothesis of a highly permeable visceral pleura, other
(1 mL·kg-1·h-1 or 1,680 mL·day-1)). Maximum pleural data [34] suggest that the visceral pleura may represent
lymph flow would amount to ≈30 mL·h-1, ≈700 mL·day-1 a pathway for fluid removal during pulmonary oedema;
(about 40% of overall lymph flow), a remarkable increase, these data are based on the observation that protein rich
and yet, still not enough to counteract the formation of liquid leaks through the visceral pleura of isolated, infla-
large pleural effusions. Beyond maximum lymph flow ted, perfused lungs made oedematous in an artificial
saturation, fluid exchanges depend upon hydraulic and pleural space. Again, the finding, of a protein-rich liquid
osmotic pressure gradients, whilst below saturation level, suggests a marked alteration in membrane permability
lymph flow represents the main drainage pathway. It both of capillaries and visceral pleura, because normal
seems appropriate to recall here the long overlooked hy- pleural fluid is hypo-oncotic.
pothesis proposed by Starling 100 yrs ago concerning Extrapolation of such findings to human pathology
the fluid reabsorption from experimental pleural effu- may be difficult, as protein-rich pleural effusion is a
sion, namely, that after hydraulic and osmotic equili- rather uncommon finding in pulmonary oedema. In fact,
brium is accomplished, "...the absorption of fluid from it was also found that during lung interstitial oedema in
the pleural cavity is extremely slow, so that it might rabbits (that have a thin visceral pleura), pulmonary
perhaps be affected by the lymphatics alone" [31]. interstitial pressure rose well above atmospheric [29],
Based on the fact that pleural fluid is filtered and creating a gradient for liquid filtration from the lung
mostly reabsorbed via lymphatics at parietal level, an parenchyma into the pleural space. However, this did
attempt was made to model pleural fluid turnover [32], not result (at least up to 3 h of observation) in any appre-
as occurring between three compartments in series (sys- ciable increase in pleural liquid volume, indicating that
temic capillaries, extrapleural parietal interstitium and the permeability of the visceral pleura is physiological-
pleural cavity), separated by two resistances in series ly fairly low. Human clinical experience reveals that
(capillary endothelium and parietal mesothelium), and transudative pleural effusion may develop several hours
two draining pathways (extrapleural and pleural lym- after pulmonary oedema. To explain an increase in micro-
phatics). As already mentioned, the visceral pleura does vascular filtration through the visceral pleura, two hypo-
not appreciably account for pleural fluid egress under theses may be put forward: 1) on biophysical grounds,
normal conditions. The model is useful to discuss some pulmonary vascular congestion implies an increase in
important determinants of pleural effusion. A simul- surface area for microvascular exchanges (this may
taneous increase of capillary and mesothelial water perm- involve not only the pulmonary circulation but also the
eability leads to hypo-oncotic fluid; if filtration exceeds bronchial circulation that supplies the visceral pleura);
224 G . MISEROCCHI
20. Miserocchi G, Nearini D. Contribution of Starling and the transition to interstitial edema. J Appl Physiol 1993;
lymphatic flows to pleural liquid exchange in anes- 74: 1171–1177.
thetized rabbits. J Appl Physiol 1986; 61: 325–330. 30. Miserocchi G, Venturoli D, Negrini D, Del Fabbro M.
21. Miserocchi G, Negrini D, Mukenge S, Turconi P, Del Model of pleural fluid turnover. J Appl Physiol 1993;
Fabbro M. Liquid drainage through the peritoneal diaph- 75: 1798–1806.
ragmatic surface. J Appl Physiol 1989; 66: 1579–1585. 31. Leathes JB, Starlin EH. The absorption of salt solution
22. Negrini D, Ballard ST, Benoit JN. Contribution of lym- from the pleural cavities. J Physiol 1895; 18: 106–116.
phatic myogenic activity and of respiratory movements 32. Venturoli D. Modello di scambio di liquido e soluti in
to pleural lymph flow. J Appl Physiol 1994; 76: 2267– un sistema a tre compartimenti: applicazione al turnover
2274. di liquido e soluti nel cavo pleurico. PhD Dissertation,
23. Negrini D, Pistolesi M, Miniati M, Bellina CR, Giuntilli University of Milano, 1995.
C, Miserocchi G. Regional protein absorption rates from 33. Paine DK, Kinasewitz GT, Gonzales E. Comparative
the pleural cavity in dogs. J Appl Physiol 1985; 58: permeability of canine visceral and parietal pleura. J
2062–2067. Appl Physiol 1988; 65: 2558–2564.
24. Broaddus, VC, Wiener-Kronish JP, Berthiaume Y, Staub 34. Broaddus VC, Wiener-Kronish JP, Staub NC. Clearance
NC. Removal of pleural liquid and protein by lymphatics of lung edema into the pleural space of volume-loaded,
in awake sheep. J Appl Physiol 1988; 56: 384–390. anesthetized sheep. J Appl Physiol 1990; 68: 2623–2630.
25. Wiener Kronish JP, Albertine KH, Roos PJ, Staub NC. 35. Miserocchi G, Venturoli D, Negrini D, Gilardi MC,
Protein egress and entry rates in pleural fluid and plas- Bellina RC. Intrapleural fluid movements described by
ma in sheep. J Appl Physiol: Respirat Environ Exercise a porous flow model. J Appl Physiol 1992; 73: 2511–
Physiol 1984; 56: 459–463. 2516.
26. Miserocchi G, Negrini D, Gonano C. Direct measure- 36. Butler JP, Huang J, Loring SH, Lai-Fook SJ, Wang PM,
ments of interstitial pulmonary pressure in in situ lung Wilson TA. Model of a pump that drives circulation of
with intact pleural space. J Appl Physiol 1990; 69: pleural fluid. J Appl Physiol 1995; 78: 23–29.
2168–2174. 37. Grotberg JB, Glucksberg MR. A buoyancy-driven squeeze
27. Negrini D, Gonano C, Miserocchi G. Microvascular film model of intrapleural fluid dynamics: basic con-
pressure profile in intact in situ lung. J Appl Physiol cepts. J Appl Physiol 1994; 77: 1555–1561.
1992; 72: 332–339. 38. Negrini D, Miserocchi G. Size-related differences in
28. Miserocchi G, Negrini D, Pistolesi M, et al. Intrapleural parietal extrapleural and pleural liquid pressure distrib-
liquid flow down a gravity-dependent hydraulic pres- ution. J Appl Physiol 1989; 67: 1967–1972.
sure gradient. J Appl Physiol 1988; 64: 577–584. 39. Hills BA. Graphite-like lubrication of mesothelium by
29. Miserocchi G, Negrini D, Del Fabbro M, Venturoli D. oligolamellar pleural surfactant. J Appl Physiol 1992;
Pulmonary interstitial pressure in intact in situ lung: 73: 1034–1039.