¿Cómo tratar los derrames pleurales malignos?

Tratar un derrame pleural maligno (DPM) cuando el pulmón no se logra expandir puede ser un problema muy frustrante. El pulmón atrapado por un derrame pleural maligno es una situación que se encuentra con relativa frecuencia. Se puede optar por realizar toracocentesis seriadas, pero este no es un procedimiento exento de complicaciones y se necesita una solución más definitiva. Algunos dirán de colocar un tubo de tórax, pero si el pulmón está atrapado, por más que se intente realizar pleurodesis a través del mismo, este tubo nunca logrará retirase. Una solución muy práctica y sencilla es colocar un pequeño catéter tunelizado en la cavidad pleural. Créanme que resuelve un problema que en muchas ocasiones parece no tener fin. Este catéter le permite al paciente y su familia, drenarse líquido pleural en la casa cada vez que haga falta. Vayamos a un ejemplo, tenés un paciente con un DPM y estás pensando entre realizar una videotoracoscopía con pleurodesis o colocar un catéter pleural. La videotoracoscopía con pleurodesis sólo funcionará si el pulmón expande lo suficiente para contactar la pared torácica, de lo contrario nunca va a resolver el problema. Si el pulmón no expande, el catéter pleural es la solución. Lo que podés hacer es realizar la videotoracoscopía y ver si el pulmón expande o no; si expande hacés la pleurodesis, ahora si no expande, en ese mismo momento le colocás un catéter pleural. De manera alternativa y si ya sabés que el derrame pleural es maligno y no necesitás la videotoracoscopía para realizar biopsias pleurales, podrías colocar el catéter pleural directamente, sin la necesitad de la videotoracoscopía ni la anestesia general. En la siguiente tabla enumero algunas de las ventajas y desventajas que se me ocurren de cada uno de los métodos.

A nosotros nos gusta realizar una videotoracoscopía en los pacientes con sospecha de DPM. Drenamos todo el líquido y hacemos biopsias pleurales para confirmar la sospecha. Una vez hecho esto le pedimos al anestesiólogo que nos insufle el pulmón y haga algunas maniobras de Valsalva y vemos que tanto expande el pulmón. Si expande lo suficiente para contactar la parrilla costal, colocamos 5 gr de talco estéril en la cavidad pleural; si no expande colocamos un catéter pleural que lo tunelizamos antes de exteriorizarlo.

Como catéter usamos un tipo Tenkhoff de 15 Fr. Lamentablemente, no tenemos disponible en mi país el original, que se llama PleurX. Al paciente lo educamos para que drene día por medio líquido de su pleural y le damos pautas específicas a partir de cuando puede espaciar el drenaje cada 3, 4, 5 días o una vez por semana. En el video se muestra esta situación de un paciente con un derrame pleural que se demostró maligno y que el pulmón no terminaba de expandir al insuflarlo, por esto decidimos colocar una cráter pleural. Hay que recordar que en estos pacientes con DPM el objetivo más importante es paliar la disnea por y en este aspecto el catéter pleural es óptimo.

¿Alguien tiene alguna experiencia con su uso?

 

Digg This
Reddit This
Stumble Now!
Buzz This
Vote on DZone
Share on Facebook
Bookmark this on Delicious
Kick It on DotNetKicks.com
Shout it
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Opacidades pulmonares en vidrio esmerilado y centrales, ¿cómo tratarlas?

¿Qué harían con un caso como este? Un hombre de 65 años, ex fumador de 20 cigarrillos por día por 30 años, que se le realizó una TC de tórax por otro motivo y en la misma se ve una imagen nodular en vidrio esmerilado en el centro del lóbulo superior izquierdo de 1 cm de diámetro. El médico que lo vio por primera vez le pidió un PET-TC que mostraba que el nódulo hipercaptaba con SUV de 5. No hay otros focos hipermetabólicos. Tiene una broncoscopía normal y su función pulmonar muestra un VEF1 de 1.35 litros (50%) y una DLCO de 65%. No tiene otros problemas significativos de salud. El paciente no está muy preocupado por este hallazgo y querría saber tu opinión acerca de qué hacer. ¿Qué opciones hay?Una opción sería realizar una punción con aguja fina, pero debería realizarla un radiólogo muy experimentado en este tipo de punciones debido a que el nódulo es pequeño y central. Además, va a tener que atravesar un volumen significativo de pulmón y el riesgo de neumotórax puede ser significativo. Yo personalmente no tengo ninguna experiencia con navigational bronchoscopy, estaría bueno escuchar alguna opinión si alguien tiene experiencia. Otras opciones que se me ocurren son la cirugía o la vigilancia. Yo no sería muy partícipe de la vigilancia, pero entiendo que conversando con el paciente los riesgos y beneficios alguien podría recomendar el control un unos pocos meses. La cirugía requeriría una lobectomía, el nódulo es central y está muy cerca de una rama de la arteria pulmonar. ¿Cuál piensan que es la mejor opción? ¿Piensan que puede ser resecado con una resección menor?

Digg This
Reddit This
Stumble Now!
Buzz This
Vote on DZone
Share on Facebook
Bookmark this on Delicious
Kick It on DotNetKicks.com
Shout it
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Small Central Ground Glass Opacity: how do you proceed?

What do you do with a case like this? It’s a 65 year, smoking history of 20 cigarettes a day for 30 years, that got a CT scan for other reason and a GGO spot was found in the left upper lobe. The nodule was sized in 1 cm and the doctor that saw the patient ordered a PET-CT that showed SUV of 5. Nothing lighted up in the mediastinum or elsewhere. Bronchoscopy was normal. FEV1 is 1.35 liters (50%) and DLCO 65%. He hasn’t any other significant health issue. The patient is not very anxious about this finding and he wants to know your suggestion. What are the options?

I guess you might try to stick a needle on it, but your radiologist should be very skilful to target this tiny spot in the middle of the lung, especially in a patient with emphysema. The risk of pneumothorax is significant.

What other options? Well, I don’t have any experience with navigational bronchoscopy, but if any has, it’ll be great to hear any input. The two other options I can think of are surgery or just wait. I’m not very keen on waiting in a case like this, but I accept somebody may have this as a suggestion. Surgery will take a lobectomy, as the spot is in the middle of the upper lobe and very close to a PA branch.

What do you think is the best option? Do you think you can take it out with a lesser resection than lobectomy?

 

Digg This
Reddit This
Stumble Now!
Buzz This
Vote on DZone
Share on Facebook
Bookmark this on Delicious
Kick It on DotNetKicks.com
Shout it
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

How to deal with malignant pleural effusions

Treating a malignant pleural effusion with a lung that does not expand can be a frustrating problem. Repeat thoracocentesis is not complication free and as there is not apposition between both pleural layers, pleurodesis is not an option. Placing a flexible pleural catheter is a great option in these cases. They allow the patient to drain himself fluid out of the chest and they can last for long periods of time. Let’s say you have a patient with a recurrent MPE; you can either do a VATS pleurodesis or place a pleural catheter. VATS pleurodesis will only work if the lung expands enough to touch the ribs; if it does not you will need to place a pleural catheter. In the following table the advantages and disadvantages of each method are listed.

What we usually do in a patient with a suspicion of recurrent pleural effusion is VATS. We drain the fluid and do some pleural biopsies to confirm the diagnosis. Once this is done, we ask the anaesthesiologist to inflate the lung and do some Valsalva manoeuvre and check whether the lung expands enough to reach the rib cage or not. If it does expand, we insufflate 5 grams of sterile talc, place a chest tube and leave it in place for 3 to 4 days or until the output is less than 300 cc a day. If the lung does not expand we place a pleural catheter. We used a Tenkhoff-type catheter that we introduce through a second tiny port and exteriorize the catheter through a subcutaneous tunnel. Unfortunately, we do not have available in my country the catheter that is specifically design for pleural use. We instruct the patient and his or her family to drain fluid with a 50 cc syringe every other day at the beginning and give instructions on when they should switch to once every three days, four days, five days and finally once a week.  In the video we show how it looks like doing VATS with a lung that does not expand and how a pleural catheter is placed at that moment.

Digg This
Reddit This
Stumble Now!
Buzz This
Vote on DZone
Share on Facebook
Bookmark this on Delicious
Kick It on DotNetKicks.com
Shout it
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Are we following the guidelines?

New piece of evidence suggest that guidelines are often times not followed. The EJCTS published last month an article about thoracic and general surgeon’s adherence to mediastinal lymph sampling during lung resection for lung cancer. This article describes how many LN stations were sampled and compares the number to the gold standard –in this case they considered the European guidelines as the gold standard-.
As an example, the guide recommends to dissect all mediastinal LN station when doing lung resection for cancer. In this paper only 4% of 216 patients had this kind of mediastinal exploration.
Also, the guidelines said that when dealing with a peripheral T1 lesion, you might be OK exploring only 3 LN stations. Station 7 should always be dissected, regardless of the affected lobe. The study shows that less than 50% of patients had station 7 dissected and even a lower percentage had at least 3 mediastinal LN stations explored.
This is not the only paper that address this issue, there are many paper from the US as well that underlines the suboptimal staging of the mediastinal in lung cancer patients.
For much time, mediastinal lymph node dissection has been considered a measure of quality in lung cancer treatment; however, data is showing that this measure is seldom times fulfilled. Are we looking into the right set of data? Is MLND a real measure of quality or should we look at something else? This is one more report that shows how far away are the guidelines from everyday practice.
Should MLND be considered a measure of quality in lung cancer treatment? I believe it should, but it be great to hear your ideas.

Digg This
Reddit This
Stumble Now!
Buzz This
Vote on DZone
Share on Facebook
Bookmark this on Delicious
Kick It on DotNetKicks.com
Shout it
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

How do you manage postoperative chylothorax?

Chylothorax is one of the complications I considered the most fearful. Patients having this PO complication deteriorate so quickly that when you start considering reoperation might be too late. It’s not the same a chylothorax after esophageal surgery than after lung surgery. Prior publications have shown that chylothorax after lung surgery is more likely to seal by itself. However, we’ll refer here only to the basics and some ideas that are common to both PO.

You usually suspect chylothorax when after POD#1 or 2 the chest tube output is more than you expected. Let’s say you did a lobectomy + MLND and CT output is 1200 cc on POD#1. That should give a suspicion clue. Esophageal cases can drain that much on POD#1, 2 or even 3, especially when the patient had preoperative chemorads. But, it’s not until the patient is fed that you realize that chyle is pouring out. CT output becomes milky and if any doubt exists you can test triglycerides in the fluid. Have in mind that if the patient is fasting, the fluid won’t be milky and only a high output will be the clue to the diagnosis. If the fluid has a value of triglycerides of more than 110 mg/dl, chylothorax is almost certain. If the value is below 50 mg/dl the diagnosis will be unlikely. In between are the grays. Chylomicrons can be measure and if they are present they confirm the diagnosis as well.

Once the diagnosis is made, a prompt solution should be looked for. If the output is less than 500 cc/day a little bit of time can be taken to think what to do. However, if the output is more than 1000 cc/day a surgical repair will be very likely. In any event, NPO should be the rule. In my experience no diet ever worked. It sounds neat to think about the medium chain TGL going through the portal vein and not thought the lymphatics, but at least in my experience that has never worked. Just put the patient NPO and start total parenteral nutrition right away. If the output is significant (>800 ml/d) and you don’t replace looses fast, you’ll start having trouble soon –low urinary output, problems with electrolytes, hypotension, renal failure and a very fats general deterioration of the general status of the patient-. There’s no guarantee that doing an adequate reposition of fluid and electrolytes won’t lead you to the same complications. This is why you shouldn’t wait too much to take back the patient to the OR if the output doesn’t slow down. Most would say that you can wait 7 days, but I rather just wait 3 to 4 days at the most.

In summary, surgery involves to mass ligate the thoracic duct. I achieve this more efficiently doing a right thoracotomy, but right VATS can be tried. Don’t even bother to look for the duct, just mass ligate with two silks all the tissue between the aorta’s adventitia and the esophagus, including the inferior hemiazigos vein. We can address this is another post, but the question is how many days do you think is worthwhile to wait until you take a patient back to the OR for PO chylothorax when drainage of more than 700 cc/day doesn’t slow down?

Digg This
Reddit This
Stumble Now!
Buzz This
Vote on DZone
Share on Facebook
Bookmark this on Delicious
Kick It on DotNetKicks.com
Shout it
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)