POW Journal Club

1
POW Journal Club

• 6th March 2006

2
(No Transcript)
3
Study population

• Patients prospectively entered into a database
  which was updated annually
• 459 patients
• breast conserving surgery
• within a 25 year period from 1973 to 1998
• clear margins
• at least 5 years of follow-up

4
Definition of margins

• Clear
• No invasive or non invasive disease within a mm
  of the inked margin
• Close
• Disease within a mm
• Positive
• Disease at the inked margins

5
RESULTS

• 28 (6) of 459 patients had recurrences
• Multivariate analysis demonstrated that
  recurrence is significantly related to
• Omission of radiotherapy
• Young age
• No of re-excisions necessary to obtain clear
  margins

6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
Authors explanation for the results

• The necessity for re-excision of non-negative
  margins indicates a higher risk of local
  recurrence because it reflects a more extensive
  disease with a higher risk of local recurrence

10
More on margins

• Before 1996
• No markings of the tumour margins with sutures
• Complete re-excisions of all walls of the
  lumpectomy cavity
• After 1996
• Margins marked with sutures
• Directed re-excisions of the involved walls

11
Adjuvant therapy

• Adjuvant radiotherapy given to most patients
  (90)
• 45 Gy to entire breast
• 15 Gy boost to lumpectomy bed
• Chemotherapy given to 30
• Premenopausal women with node positive disease or
  tumours gt 2cm
• Post menopausal women with oestrogen receptor
  negative tumours gt 2cm
• Cyclophosphamide and 5-FU with doxorubicin or
  methotrexate

12
Tumours

• Clinically detected 55 (251)
• Mammographic abnormalities 45 (208)
• Infiltrating ductal 82
• DCIS 10
• Infiltrating lobular 8

13
First excised specimen

• Margins negative 46 (210)
• Close 10 (47)
• Positive 15 (68)
• Unknown 29 (135)
• 229 patients underwent 439 re-excisions

14
Criticisms of paper

• Unknown margins in 29 - why margins unknown
  not stated
• Numerous Numerical Inconsistencies
• No of negative margins initially stated to be 210
• No of patients who had no re-excisions 209
• No of re-excisions
• 229 patients had re-excisions
• Total no of positive and unknown margins 68135
  203
• Total no of positive, close and unknown margins
  6847135 250
• Indications for re-excisions not stated
• ? Close / Positive / Unknown margins

15
Insufficient information

• Pathology protocol not stated unchanged over 25
  years?
• Variables such as grade, type of tumour LVI not
  included in multivariate analysis
• May well be confounding

16
Discussion

• Appropriateness of the statistical methods
• This paper really describes an audit
• It does not state its aim clearly
• Seems to highlight an interesting finding from
  analysis of the database namely that
  recurrences seem statistically related to
  re-excisions
• Whilst there may be statistical significance, is
  there clinical significance to the findings?

17

• Will the results change our practice today?
• Appropriateness of the margins for DCIS
• Will we aim for bigger margins initially? Is more
  breast better?
• Will we recommend mastectomy instead of 2
  re-excisions?