We read with great interest the article by Myles and colleagues on
the relationship between type of anaesthesia and cancer recurrence in
patients undergoing major abdominal surgery.[1] Annually millions of
people require tumour surgery; therefore it is important to question some
of the results and conclusions. A wealth of basic science data supports
the hypothesis that the surgical stress response increases the likelihood
of cancer dissemination and formation of metastases in patients undergoing
cancer surgery.[2] Anaesthetic management of the cancer patient,
therefore, could potentially influence long-term outcome. However, the
results by Myles et al do not support this hypothesis, as the use of
epidural block in abdominal cancer surgery was not associated with
improved cancer-free survival in their published trial. In our opinion the
study suffers from several problems we will discuss below:
The study population is a secondary analysis of subjects enrolled in
the MASTER trial.[3] The major criticism of the MASTER trial was the high
rate of protocol violations. Only 50% of patients randomized to the
epidural group actually received epidural analgesia throughout the entire
study period; as this secondary analysis is based on intention to treat
principle, this might have relevant implications for the outcome.[1,4]
Additionally, the authors state that their study is likely to be the only
randomized data on this subject for years to come. However, this is not a
randomized prospective study for the outcome of cancer survival; it is a
retrospective follow-up study of a previous RCT for an outcome not
intended at point of randomization. The use of the terminology 'This
randomized trial' in the conclusion of the paper is therefore incorrect.
Certainly, a lot of information about the perioperative anaesthetic
course is given to the reader; unfortunately important oncological data
are missing. For example, a lot of patients in both groups were not graded
either to the Duke or TNM system (Group "epidural" n=230: Dukes
characteristics reported for n=58 patients, TNM characteristics reported
for n=125 patients; Group "no epidural" n=216: Dukes characteristics
reported for n=54 patients, TNM characteristics reported for n=106.) Does
this missing data imply that patients without information provided were
patients potentially with metastatic diseases at time of surgery?
Additionally, there is no information given to the reader about the
quality of oncological resection to obtain R0 results (no evidence of
microscopic cancer in the margins). How many patients got a chemotherapy
or radiotherapy after cancer surgery and how many patients refused this?
No information is given about stringent postoperative follow-up, which has
been shown to help the early identification of potentially radically
treatable oligometastatic disease. Those resections of lung or liver
metastases after colorectal cancer substantially improve 5-year and 10-
year survival rate.[5] In the Result section the reader cannot find any
information about treatment with adjuvant chemotherapeutics that further
improves survival in colon and rectal cancers.[5]
The power analysis is questionable for three reasons:
1.The authors have performed a power analysis on a dichotomous outcome:
percentage of five-year survival. However, they defined cancer-free
survival as the main outcome measure in the abstract and reported and
discussed survival times in the manuscript that were analysed by the log
rank test. This would need a power analysis, in which the alternative
hypothesis is expressed as a hazard ratio. This power-analysis, using a
realistic hazard ratio might result in a significant bigger sample size as
compared to the power analysis for a one third treatment effect in the chi
-square test.
2.The authors assume a five-year survival of approximately 50%. They
report survival rates of 42 % and 44 % respectively. Thus, the study may
lack external validity and the authors cannot be sure that the conclusions
drawn about cause-effect-relationships observed in this study can be
generalized and do apply to patients undergoing abdominal surgery for
cancer.
3. In their power analyses Myles et al state that most of these patients
had primary colorectal cancer, for which five year survival is
approximately 50% and 12 year survival is less than 20%. If epidural block
reduced recurrence-free survival by 25%, then 250 patients per group would
provide 89% power with a two sided alpha of 0.05.[1] Why would epidural
block reduce, not increase survival rate? According to results of basic
science studies the cancer free survival rate or time should be increased
with the use of epidural block. [2] On the other hand, if the author meant
increase then the hypothesis seems too optimistic. For example in the
QUASAR study of 3239 patients with mostly stage II colorectal cancer, who
were randomly assigned to fluorouracil and folinic acid or to observation,
there was a significant improvement in overall survival (risk reduction of
18% and 3.6% overall survival gain).[6] If this hypothesis of Myles et al
would be confirmed, than the use of epidural block is more effective than
chemotherapy.
In summary, in the Introduction the authors present important
arguments in favour of an effect of epidural analgesia on the long-term
prognosis after cancer surgery. These arguments gave reason to carry out
the follow-up of the MASTER-trial, which due to its retrospective nature
presents a variety of shortcomings. The authors have therefore to observe
caution in the evaluation of their study and the discussion of their
findings. It is difficult to qualify the study as a randomized trial and
dangerous to conclude that "use of epidural block in abdominal surgery for
cancer is not associated with improved cancer-free survival." In view of
the good arguments to conduct this study and its many methodological
problems such a conclusion might put cancer patients at harm if the
alternative hypothesis is still true and epidural anaesthesia improves
cancer-free survival. It would have been more appropriate to conclude that
better designed studies are needed to evaluate the use of epidural block
in abdominal cancer surgery. Even though their study does not support the
hypothesis that epidural anaesthesia improves recurrence-free survival, it
has too many flaws to refute this hypothesis.
In our opinion this paper has too many "Questions marks". Therefore,
we still have to wait for the results of the randomized prospective
studies, which are focused primarily on the cancer outcome for patients
undergoing tumour surgery with different types of anaesthesia.
Sincerely,
Antje Gottschalk, Clinical Research Fellow, Department of
Anaesthesiology and Intensive Care Medicine, University Hospital Muenster,
Muenster, Germany
Stephan A Schug, Professor of Anaesthesiology and Chair of
Anaesthesiology, Pharmacology and Anaesthesiology Unit, School of Medicine
and Pharmacology, Director of Pain Medicine, Royal Perth Hospital, Perth,
Australia
Hugo K Van Aken, Professor of Anaesthesiology and Head of the
Department of Anesthesiology and Intensive Care Medicine, Department of
Anaesthesiology and Intensive Care Medicine, University Hospital Muenster,
Muenster, Germany
Gerhard Brodner, Professor of Anaesthesiology and Head of the
Department of Anesthesiology, Intensive Care Medicine and Pain Therapy,
Fachklinik Hornheide, Muenster, Germany
References
1. Myles PS, Peyton P, Silbert B, Hunt J, Rigg JR, Sessler DI:
Perioperative epidural analgesia for major abdominal surgery for cancer
and recurrence-free survival: randomised trial. BMJ 2011; 342: d1491
2. Gottschalk A, Sharma S, Ford J, Durieux ME, Tiouririne M: Review
article: the role of the perioperative period in recurrence after cancer
surgery. Anesth Analg 2010; 110: 1636-43
3. Rigg JR, Jamrozik K, Myles PS, Silbert BS, Peyton PJ, Parsons RW,
Collins KS: Epidural anaesthesia and analgesia and outcome of major
surgery: a randomised trial. Lancet 2002; 359: 1276-82
4. Van Aken H, Gogarten W, Brussel T, Brodner G: Epidural anaesthesia
and analgesia in mayor surgery. Lancet 2002; 360: 568; author reply 569
6. Gray R, Barnwell J, McConkey C, Hills RK, Williams NS, Kerr DJ:
Adjuvant chemotherapy versus observation in patients with colorectal
cancer: a randomised study. Lancet 2007; 370: 2020-9
Corresponding Author:
Antje Gottschalk, MD
Department of Anaesthesiology and Intensive Care Medicine, University of
Muenster
Albert-Schweitzer-Strasse 33
48149 Muenster, Germany
Email: antjegottschalk@gmx.net
Competing interests:
No competing interests
27 April 2011
Antje Gottschalk
Clinical Research Fellow
Stephan A Schug, Hugo K Van Aken, Gerhard Brodner
Department of Anaesthesiology and Intensive Care Medicine, University Hospital Muenster, Muenster, G
Rapid Response:
Too many "Question marks"
We read with great interest the article by Myles and colleagues on
the relationship between type of anaesthesia and cancer recurrence in
patients undergoing major abdominal surgery.[1] Annually millions of
people require tumour surgery; therefore it is important to question some
of the results and conclusions. A wealth of basic science data supports
the hypothesis that the surgical stress response increases the likelihood
of cancer dissemination and formation of metastases in patients undergoing
cancer surgery.[2] Anaesthetic management of the cancer patient,
therefore, could potentially influence long-term outcome. However, the
results by Myles et al do not support this hypothesis, as the use of
epidural block in abdominal cancer surgery was not associated with
improved cancer-free survival in their published trial. In our opinion the
study suffers from several problems we will discuss below:
The study population is a secondary analysis of subjects enrolled in
the MASTER trial.[3] The major criticism of the MASTER trial was the high
rate of protocol violations. Only 50% of patients randomized to the
epidural group actually received epidural analgesia throughout the entire
study period; as this secondary analysis is based on intention to treat
principle, this might have relevant implications for the outcome.[1,4]
Additionally, the authors state that their study is likely to be the only
randomized data on this subject for years to come. However, this is not a
randomized prospective study for the outcome of cancer survival; it is a
retrospective follow-up study of a previous RCT for an outcome not
intended at point of randomization. The use of the terminology 'This
randomized trial' in the conclusion of the paper is therefore incorrect.
Certainly, a lot of information about the perioperative anaesthetic
course is given to the reader; unfortunately important oncological data
are missing. For example, a lot of patients in both groups were not graded
either to the Duke or TNM system (Group "epidural" n=230: Dukes
characteristics reported for n=58 patients, TNM characteristics reported
for n=125 patients; Group "no epidural" n=216: Dukes characteristics
reported for n=54 patients, TNM characteristics reported for n=106.) Does
this missing data imply that patients without information provided were
patients potentially with metastatic diseases at time of surgery?
Additionally, there is no information given to the reader about the
quality of oncological resection to obtain R0 results (no evidence of
microscopic cancer in the margins). How many patients got a chemotherapy
or radiotherapy after cancer surgery and how many patients refused this?
No information is given about stringent postoperative follow-up, which has
been shown to help the early identification of potentially radically
treatable oligometastatic disease. Those resections of lung or liver
metastases after colorectal cancer substantially improve 5-year and 10-
year survival rate.[5] In the Result section the reader cannot find any
information about treatment with adjuvant chemotherapeutics that further
improves survival in colon and rectal cancers.[5]
The power analysis is questionable for three reasons:
1.The authors have performed a power analysis on a dichotomous outcome:
percentage of five-year survival. However, they defined cancer-free
survival as the main outcome measure in the abstract and reported and
discussed survival times in the manuscript that were analysed by the log
rank test. This would need a power analysis, in which the alternative
hypothesis is expressed as a hazard ratio. This power-analysis, using a
realistic hazard ratio might result in a significant bigger sample size as
compared to the power analysis for a one third treatment effect in the chi
-square test.
2.The authors assume a five-year survival of approximately 50%. They
report survival rates of 42 % and 44 % respectively. Thus, the study may
lack external validity and the authors cannot be sure that the conclusions
drawn about cause-effect-relationships observed in this study can be
generalized and do apply to patients undergoing abdominal surgery for
cancer.
3. In their power analyses Myles et al state that most of these patients
had primary colorectal cancer, for which five year survival is
approximately 50% and 12 year survival is less than 20%. If epidural block
reduced recurrence-free survival by 25%, then 250 patients per group would
provide 89% power with a two sided alpha of 0.05.[1] Why would epidural
block reduce, not increase survival rate? According to results of basic
science studies the cancer free survival rate or time should be increased
with the use of epidural block. [2] On the other hand, if the author meant
increase then the hypothesis seems too optimistic. For example in the
QUASAR study of 3239 patients with mostly stage II colorectal cancer, who
were randomly assigned to fluorouracil and folinic acid or to observation,
there was a significant improvement in overall survival (risk reduction of
18% and 3.6% overall survival gain).[6] If this hypothesis of Myles et al
would be confirmed, than the use of epidural block is more effective than
chemotherapy.
In summary, in the Introduction the authors present important
arguments in favour of an effect of epidural analgesia on the long-term
prognosis after cancer surgery. These arguments gave reason to carry out
the follow-up of the MASTER-trial, which due to its retrospective nature
presents a variety of shortcomings. The authors have therefore to observe
caution in the evaluation of their study and the discussion of their
findings. It is difficult to qualify the study as a randomized trial and
dangerous to conclude that "use of epidural block in abdominal surgery for
cancer is not associated with improved cancer-free survival." In view of
the good arguments to conduct this study and its many methodological
problems such a conclusion might put cancer patients at harm if the
alternative hypothesis is still true and epidural anaesthesia improves
cancer-free survival. It would have been more appropriate to conclude that
better designed studies are needed to evaluate the use of epidural block
in abdominal cancer surgery. Even though their study does not support the
hypothesis that epidural anaesthesia improves recurrence-free survival, it
has too many flaws to refute this hypothesis.
In our opinion this paper has too many "Questions marks". Therefore,
we still have to wait for the results of the randomized prospective
studies, which are focused primarily on the cancer outcome for patients
undergoing tumour surgery with different types of anaesthesia.
Sincerely,
Antje Gottschalk, Clinical Research Fellow, Department of
Anaesthesiology and Intensive Care Medicine, University Hospital Muenster,
Muenster, Germany
Stephan A Schug, Professor of Anaesthesiology and Chair of
Anaesthesiology, Pharmacology and Anaesthesiology Unit, School of Medicine
and Pharmacology, Director of Pain Medicine, Royal Perth Hospital, Perth,
Australia
Hugo K Van Aken, Professor of Anaesthesiology and Head of the
Department of Anesthesiology and Intensive Care Medicine, Department of
Anaesthesiology and Intensive Care Medicine, University Hospital Muenster,
Muenster, Germany
Gerhard Brodner, Professor of Anaesthesiology and Head of the
Department of Anesthesiology, Intensive Care Medicine and Pain Therapy,
Fachklinik Hornheide, Muenster, Germany
References
1. Myles PS, Peyton P, Silbert B, Hunt J, Rigg JR, Sessler DI:
Perioperative epidural analgesia for major abdominal surgery for cancer
and recurrence-free survival: randomised trial. BMJ 2011; 342: d1491
2. Gottschalk A, Sharma S, Ford J, Durieux ME, Tiouririne M: Review
article: the role of the perioperative period in recurrence after cancer
surgery. Anesth Analg 2010; 110: 1636-43
3. Rigg JR, Jamrozik K, Myles PS, Silbert BS, Peyton PJ, Parsons RW,
Collins KS: Epidural anaesthesia and analgesia and outcome of major
surgery: a randomised trial. Lancet 2002; 359: 1276-82
4. Van Aken H, Gogarten W, Brussel T, Brodner G: Epidural anaesthesia
and analgesia in mayor surgery. Lancet 2002; 360: 568; author reply 569
5. Cunningham D, Atkin W, Lenz HJ, Lynch HT, Minsky B, Nordlinger B,
Starling N: Colorectal cancer. Lancet 2010; 375: 1030-47
6. Gray R, Barnwell J, McConkey C, Hills RK, Williams NS, Kerr DJ:
Adjuvant chemotherapy versus observation in patients with colorectal
cancer: a randomised study. Lancet 2007; 370: 2020-9
Corresponding Author:
Antje Gottschalk, MD
Department of Anaesthesiology and Intensive Care Medicine, University of
Muenster
Albert-Schweitzer-Strasse 33
48149 Muenster, Germany
Email: antjegottschalk@gmx.net
Competing interests: No competing interests