Anesthesia for Thoracic Surgery:

1
Anesthesia for Thoracic Surgery

• A potpourri of interesting topics
• Ottawa Hospital
• Journal Club
• October 16, 2006

2
Perioperative Risk Factors for Anastomotic
Leakage After Esophagectomy Influence of
Thoracic Epidural AnalgesiaPierre Michelet MD
Xavier-Benoit DJourno, MD Antoine Roch, MD PhD
Laurent Papazian, MD Jacques Ragni, MD Pascal
Thomas, MD, PhD Jean-Pierre Auffray, MD

• Presented By Dr. Jennifer Ozard

3
Preamble

• Carcinoma of the esophagus
• More common in men than women
• Accounts for approx. 1 of all cancers in the
  North American population
• Mets to liver and lungs primarily
• Risk Factors include age (gt55), alcohol abuse,
  chewing tobacco, poor nutrition, exposure to
  toxins (solvents), HPV, obesity, smoking, hx of
  swallowing lye, mets from larynx/pharynx/tonsils/l
  ung/breast/liver/
• kidney/prostate/testicle/bone/skin ca,
  achalasia, GERD, Barretts esophagus
• Periop mortality rate 3-10
• Most common postop complications include resp
  failure and anastomotic leakage

4
Preamble-2

• TEA
• Most satisfactory analgesia for esophagectomy
  patients
• Has been shown to decrease fatal and nonfatal
  complications in multiple studies

5
The Study

• Purpose
• To evaluate the influence of various
  perioperative factors, including the use of TEA,
  on the incidence of anastomotic leakage after
  esophagectomy
• Design
• Retrospective
• Population
• 207 patients undergoing 1-stage esophagectomy
  between 1998-2003 in a teaching hospital

6
Interventions-Patient Population

• Preop Factors
• Gender
• Age
• Height
• Weight
• BMI
• Smoking hx
• DM
• Tumour histology and stage

• Preop Factors
• Preop rads/chemo
• ASA class
• NYHA class
• PaO2/FVC/FEV1 pred
• FEV1/FVC pred

7
Interventions-Periop Factors and Postop
Complications

• Periop Factors
• Duration of surgery (skin incision to closure),
  duration of 1-lung vent, duration of mech vent,
  periop blood loss, occurrence of hemodynamic
  instability (lt80mmHg for 5 min)
• Level of anastomosis
• Postop Factors/Complications
• Need for blood transfusion, inotrope use, lowest
  value of PaO2/FiO2 within 1st 24 hours postop,
  need for re-intubation
• Pneumonia, ARDS, sepsis

8
Outcome of Interest

• Anastomotic Leakage
• Defined as dehiscence diagnosed during the
  postoperative period
• For mech ventd patients, a routine upper
  endoscopy was performed to confirm or rule out an
  anastomotic leakage as possible cause of resp
  failure
• In patients with short stay in ICU, leakage was
  assessed by a water-soluble monomeric, ionic
  x-ray contrast medium study on postop days 8-10
  and if suspicious, confirmed by endoscopy

9
Methods

• Anesthetic Technique
• Induction Midaz, sufenta, muscle relaxant
• Left-sided double lumen tube
• TV 8-10ml/kg in 2-lung vent
• TV 5-6ml/kg in 1-lung vent with PEEP of 5cm H2O
• TEA and goals
• Placed before induction, T6-8, median approach
• Infusion started at end of esogastric anastomosis
  (ropi and sufenta)
• VASlt3 at rest, nerve block T2-9
• PCA if failure to place, refusal by pt or
  accidental removal

10
Methods, contd

• Esophagectomy
• en bloc esophagectomy with two-field
  lymphadenectomy for malignancy
• Right transthoracic approach with an
  esophagogastric anastomosis in the upper thorax
  for lower third and gastroesophageal tumors (Ivor
  Lewis technique) or in the neck for higher lesions

11
Statistics

• SPSS 12.0 stats software
• Univariate analysis
• Categorical data ?2 or Fisher Exact test
• Continuous data Wilcoxan or Kruskal Wallis
• Used to compare occurrence of anastomotic leak
  against pre/peri/postop variables and development
  of postop complications
• Multivariate analysis
• Odds ratio calculated based on multiple logistic
  regression of variables found to be significant
  (or plt0.25) in the univariate analyses (2-sided
  tests)

12
Results-1

• Anastomotic Leak 11 (23/207)
• 30-day mortality Leak 26
• 30-day mortality - Leak 5.4
• Plt0.002
• OR 8.1 (2.2-30.4, plt0.002)
• Mean length ICU stay Leak 19/-16 d
• Mean length ICU stay -Leak 9/-7 d
• Plt0.008
• Mean length hospital stay Leak 43/-27 d
• Mean length hospital stay -Leak 23/-11 d
• Plt0.001

13
Results-2
14
Results-3

• Preop characteristics
• Significant difference in freq. of leak with
  respect to tumour histology and preop radiation
• Periop characteristics
• No significant difference in freq. of leak with
  regard to surgical duration, mechanical
  ventilation time, OLV duration
• Significant difference in freq. of leakage with
• Cervical anastomosis plt0.001
• Hypotensive event plt0.001
• Estimated blood loss plt0.001
• ARDS plt0.001
• Epidural analgesia p0.002

15
Results-4

• TEA
• 93/207 patients
• Duration 4.4 /- 0.8 d
• 4 accidental losses
• Significant difference in length of ICU stay when
  compared to pts with no TEA
• Significant difference in 30-day mortality when
  compared to pts with no TEA

16
Results-5

• Anastomotic Leakage associated with
• Documentation of surgical difficulties (indicated
  by periop blood loss)
• ?Rate of hypotensive events
• ?Incidence of ARDS postop (5 pts with
  post-leakage ARDS)
• ?Incidence of TEA
• Multiple Logistic Regression Analysis

17
Conclusions

• Anastomotic leakage is not rare (11)
• Anastomotic leakage is associated with
  significantly increased
• 30-day mortality
• ICU and hospital length of stays
• TEA may be protective against conduit ischemia
  and prevent leak
• Support for relationship
• TEA promotes intensive physiotherapy that can
  preserve pulm function and prevent hypoxemia
• Improvement of the gastric tube microcirculation
• Surgical difficulty, cervical anastomosis and
  development of ARDS may increase risk of ischemia
  and leak

18
Conclusions-2

• Offers more justification for spending the MAFAT
  on the thoracic epidural for esophagectomies

19
Critical Discussion-1

• Study Design
• Reference Kabon et al who showed that epidural
  initiated before skin incision can improve
  microcirculation and prevent anastomosis
  insufficiency postop--in this study, not
  initiated until after esogastric anastomosis
  completed
• Retrospective study
• Missing data
• Stats
• could have used stronger stats when doing the
  multiple regression analysis

20
Critical Discussion-2

• Confounders
• Epidural placement only achieved in 93/207
  patients!
• ARDS as a chicken and the egg argument