  |
|
|
||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
22-September-2006, Vol 119 No 1242
|
||||||||||||||||||||||||||||||||||
|
Multidisciplinary treatment of colorectal cancer in
New Zealand: survival rates from 1997–2002
John Keating, David Yong, Glenda Cutler, James
Johnston
Colorectal cancer (CRC) is second only to lung cancer as a
cause of years of life lost due to cancer in Australia and New Zealand.1,2
Despite screening initiatives for high-risk patients, improvements in the
surgical management of rectal cancer, the introduction of novel active
chemotherapy agents, and advances in the treatment of liver metastases, survival
statistics have been relatively slow to improve.3
National incidence and survival figures provide an overall
picture of the outcome of patients with CRC, however large differences in
survival have been found between countries, hospitals, and individual surgeons
treating CRC.4–6 Against this background, a detailed database was set up
to analyse the outcome of patients seen and treated with bowel cancer in a
university teaching hospital. The database was designed to capture information
on the patient, the pathology, and the treatment variables that might impact on
ultimate survival.
MethodsAll patients seen with a new CRC from January 1997 to
January 2003 by a single colorectal surgical team were entered into the
database. The great majority were treated in a public university hospital with
the remainder treated in adjacent private hospitals. 100 data fields were
assigned to each patient and the collected data included basic demographics,
detailed information of initial treatment and pathological variables,
perioperative blood transfusion, complications, and ultimate survival.
Tumours were staged according to the Australian
Clinicopathological Staging (ACPS) system. Comprehensive data were also
collected on sites of, time to, and treatment of metastatic disease. Data was
entered into an EPI6 program and analysed in a EPI 2004 program (CDC, Atlanta,
Georgia). The current status of all patients is known. Patients treated by other
surgeons and referred with recurrent disease were not included in the
analysis.
There was a close liaison between the surgeon and
medical and radiation oncologists over the period of the study in the care of
patients with colorectal cancer. An inclusive definition of pelvic recurrence
was adopted.
Local pelvic recurrence after rectal cancer surgery was
defined as any documented clinical, radiological, or histological evidence of
recurrence in the pelvis at any time after surgery.
Curative surgery was defined as the absence of
macroscopic evidence of disease at the completion of the primary operation.
Perioperative blood transfusion was defined as any
transfusion in the 2 weeks prior to surgery or during the index admission.
Patient follow-up consisted of 3-monthly
carcinoembryonic antigen (CEA) estimation and 6-monthly clinical review for 5
years—at which point patients were referred back to their primary
physician.
Colonoscopy was performed at the time of resection, at
1 year in patients with multiple adenomatous polyps in addition to their cancer,
and thereafter 3 yearly. Imaging of the abdomen, pelvis, and chest was only
performed on the basis of a rising CEA or on clinical suspicion of recurrence
and not as a routine. Patients with metastatic liver disease suitable for
resection were referred to the hepatobiliary team.
All patients with mid and low rectal cancers had a
total mesorectal excision (TME). A colonic pouch was constructed in all patients
having a TME from early 1998. Adjuvant preoperative radiotherapy or
chemoradiation was given to extraperitoneal rectal cancer patients if the
clinical examination and preoperative imaging suggested that the tumour was
infiltrating the mesorectum or if it was thought that downstaging would
facilitate a restorative resection.
In patients receiving preoperative chemoradiation,
postoperative chemotherapy was completed regardless of lymph node status.
Adjuvant chemotherapy was offered to node-positive colon and rectal cancer
patients deemed fit for treatment. On completion of adjuvant chemotherapy,
patients had an abdominal and pelvic CT scan and subsequent follow-up was
undertaken by the surgical team.
Operative mortality was defined as any death within 30
days of surgery. Urgent and emergency cases were defined as cases operated on
out of hours or on a non-elective list.
Patients with multiple cancers, either synchronous or
metachronous, had only a single entry in the database and analysis of prognostic
factors was carried out for the most pathologically advanced tumour.
Results244 patients were seen with a new primary colorectal cancer
(CRC) in the 6-year period. Twenty patients had more than one primary cancer at
presentation (9%) and 1 developed a second primary over the period of follow-up.
The median age at operation was 68 years.
The 5-year survival rate of all new patients seen with a
rectal or colon cancer was 58% and 56.5% respectively. The corresponding
cancer-specific survival rates were 65% and 71% for rectal and colon cancer
respectively.
The majority (87%) of patients were treated in the one
public university hospital. Patients treated in the private system were younger
than public patients by an average of 7 years (p=0.004). Thirty-nine of 233
(17%) operations were performed as emergent cases. No patients had a
laparoscopic resection of CRC in this series.
One of 197 patients (0.5%) having an elective resection died
after a myocardial infarction following an elective right hemicolectomy; and 3
of 39 patients (8% ) died following emergency resection, 2 from pulmonary emboli
and 1 from an aspiration pneumonia.
Twenty-two percent of patients having surgery had
postoperative morbidity. The anastomotic leak rate was 1.5% (2/130) for patients
having a colon cancer resection and 11.5% (6/52) for patients with rectal cancer
undergoing restorative resection. All six rectal leaks occurred in patients with
a coloanal anastomosis with a proximal defunctioning ileostomy. There were no
deaths from sepsis.
The stage distribution is shown in Table 1 and is
characteristic of a predominantly unscreened population with only 16% cancers
being at stage A. The stage at presentation was significantly different between
patients with colon and rectal cancer (p<0.0015).
Table 1. Stage distribution (%) at presentation or
resection of 244 primary colorectal cancers
CPR=Complete pathological
remission.
Rectal cancers were more likely to be resected at stage A
than colon cancers (24% vs 11%), however the percentage of early stage disease
(A+B) was higher for colon cancers 57% vs 48%. There was no difference in
survival between patients with colon and rectal cancers (Figure 1).
Screen-detected cancers were the most likely to be detected
at stage A (6/8, 75%) followed by presentation with rectal bleeding (18/76,
24%), whereas emergency cases presenting with obstruction or perforation had the
lowest percentage of A cases (2/28, 7%).
The crude survival by ACPS stage at 5 years was A 77%, B
71%, C 63%, and D 0%, with the cancer specific survival being A 100%, B 92%, C
65%, and D 0% (Figure 2).
Figure 1. Survival of patients with colorectal cancer
by tumour location (C=Colon, R=Rectum)
 Figure 2. Cancer-specific survival by the Australian
Clinicopathological Staging (ACPS) system
 Stage
A:
the cancer is confined to the bowel wall; Stage
B:
the cancer has spread to the outer surface of the bowel wall; Stage
C:
cancer is found in lymph nodes in the area of the bowel; Stage
D:
cancer is found at distant sites—for example, in the liver or lungs;
N:
no residual cancer (complete pathological remission).
On univariate analysis stage, vascular invasion, perineural
invasion, emergency operation, perioperative blood transfusion, and tumour grade
all significantly influenced survival (Table 2), however in a multivariate model
only stage remained an independent predictor of survival.
Table 2. Prognostic factors on univariate analysis for
death from colorectal cancer in addition to stage
P=Poor;
M=Moderate.
Sixty-three patients had complete resection of a lymph
node-positive colon or rectal cancer, of whom 32 (53%) received adjuvant
chemotherapy. Patients receiving adjuvant chemotherapy were significantly
younger with a median age of 60 years as opposed to non-recipients with a median
age of 73 years (p<0.0001).
Of 56 patients with metastatic disease, 43% were treated
with systemic chemotherapy, 25 % had no active treatment, and 16% had
metastectomy as their initial treatment. One patient each had cryotherapy and
selective internal radiation therapy.
Twelve patients (21%) have undergone surgery for metastatic
or recurrent disease; of these, eight are alive and disease-free after a median
of 31 months following resection (Table 3). In addition, 7 of these patients
were found to have surgically resectable disease on imaging (undertaken to
investigate a rising CEA as part of their routine follow-up). Repeat
metastasectomy was performed in 2 of these 8 patients for re-recurrence again on
the basis of a rising CEA, with a third currently awaiting a repeat liver
resection.
Of 103 patients seen with a new rectal cancer, 95 were
resected. Of these 95 patients, 50 had an anterior resection and 16 an
abdominoperineal excision (APE) of the rectum; 6 patients had a local excision
of whom 1 (the only patient with a T2 lesion) recurred and was treated by a
salvage APE and remains well 2 years later.
One-third of patients in this series with a rectal cancer
had neoadjuvant treatment; chemoradiation in 20, long course radiation in 7, and
short course in 3 patients respectively. Two out of 20 patients (10%) having
chemoradiation had a complete pathological remission (CPR). One patient had
postoperative radiation.
Local recurrence in the pelvis was apparent in one patient
after curative resection and four patients having palliative resection giving a
total local recurrence rate of 6% at a median follow-up of 32 months.
Table 3. Details of surgical resection of
metastatic/recurrent disease in 11 patients
The follow-up (F/U) time is
in months from the resection of metastatic disease; *Recurrence after local
excision of rectal cancer treated with an abdominoperineal excision
(APE).
DiscussionAlthough the age-specific incidence of CRC has plateaued
(and looks set to fall) the burden of the disease is likely to continue to
increase due to the effects of an ageing population and a growth in population
size.7 In spite of the continued evolution of most aspects of colorectal cancer
management, regional and national mortality rates have been slow to improve and
improvement has been more marked in women than in men.3,7 Indeed, New Zealand
survival rates for a number of malignancies including CRC have been compared
unfavourably with Australian outcomes.8
A colorectal cancer database with comprehensive fields was
set up in January 1997 to collect data on treatment and outcome of all new
patients seen with CRC to test the hypothesis that “optimised
management” of CRC could improve on nationally derived figures on
survival. The 5-year survival of all new patients in this series with a rectal
or colon cancer was 58% and 56.5% respectively with a cancer-specific survival
of 65% and 71% respectively. This represents a modest improvement on nationally
derived figures covering the same period.2
On univariate analysis, the clinical variables and urgency
of operation and blood transfusion—as well as the pathological variables,
tumour stage, tumour grade, vascular and perineural invasion—predicted a
worse outcome. In a multivariate model however, only stage remained as a
significant predictor of survival. The failure of these same variables to show
an independent effect on prognosis in a multivariate model is likely explained
by the relatively smaller sample size in subset analysis as in larger series
these factors persist as being of independent prognostic value.9
The difference in the stage distribution between colon and
rectal cancer cases is of interest. Significantly more rectal cancers were
resected at stage A than colon cancers, largely at the expense of fewer B cases,
whereas the percentage of advanced cases (stages C and D) was similar in the two
groups. This suggests that an effect similar to the ‘lead time bias’
seen in screening programmes is at play in patients presenting with rectal
bleeding.
Operative mortality was low in this series, a factor that is
important in ultimate survival.10 However, the rate of postoperative morbidity
at 22% remains a significant problem in keeping with other series.11
Rectal cancer is a tumour now widely recognised to be best
managed by surgeons with an interest and expertise in its treatment.12 The total
local recurrence rate for patients who had a resection of rectal cancer in this
series, after a median follow-up of 32 months, was 6%. This compares favourably
with contemporary reports which often only report on rates of pelvic recurrence
following “curative” resection and ignore the recurrence rate in the
25% of patients who have a palliative resection.13–15 A third of patients
in this series had preoperative radiotherapy, however our current practice is to
refer all patients with an extraperitoneal rectal cancer for preoperative
radiotherapy prior to resection.16
The use of adjuvant chemotherapy in this study was
disappointing, with only 53% of patients with a lymph node-positive cancer
receiving treatment. Age, and therefore presumably comorbidity, was a major
factor in determining who received treatment.
Age in itself is not a barrier to completing adjuvant
chemotherapy for CRC, and this underutilisation of adjuvant chemotherapy has
been addressed by leaving the decision-making for adjuvant therapy with the
medical oncologists.17 If it is assumed that the absolute survival benefit from
adjuvant therapy in patients with a lymph node-positive CRC is currently 15% and
given that only just over half of the 26% of eligible patients received
chemotherapy, then the absolute potential gain from adjuvant chemotherapy in
this series of patients was approximately 2% at 5 years.
The utility of follow-up after curative resection of CRC is
a topic of ongoing interest.18,19 Regular follow-up serves the interests of the
surgeon for audit purposes and is generally liked by patients, however the
rationale for follow-up is principally to detect and treat early recurrent
disease. Such a strategy can only be effective if pre-symptomatic recurrent
disease is detected and resected in a timely manner.
Three-monthly CEA and 6-monthly clinical review revealed 13
patients in this series with surgically resectable disease of whom 12 underwent
surgery (1 patient with resectable liver disease was deemed unfit for liver
surgery). After a median follow-up of 31 months, 8 of these 11 patients who
received surgery remain disease-free, of whom 2 have had a second resection both
again detected on rising CEA. These results have encouraged us to continue with
this protocol. Regular liver imaging may have lead to a further small
incremental gain but was not a practical proposition at our institution.18
The results of this series leaves no room for complacency.
Perioperative morbidity remains a common problem for patients having a
colorectal cancer resection. Adjuvant chemotherapy has been underutilised.
A low operative mortality, a low rate of local recurrence in
rectal cancer, and careful follow-up all make small but incremental gains in
survival, however the stage at diagnosis is the overriding predictor of ultimate
survival. Further gains in survival will require a shift to an earlier stage at
diagnosis and more effective systemic therapy.
Author information:
John P Keating, Senior Lecturer in Surgery, Wellington School of Medicine and
Health Sciences; David Yong, Registrar in Surgery, Wellington Hospital; Glenda
Cutler, Database Manager, Wellington School of Medicine and Health Sciences;
James Johnston, Medical Student, Wellington School of Medicine; Wellington
Acknowledgements: I
gratefully acknowledge the expertise of the medical and radiation oncologists of
the Wellington Blood and Cancer Centre as well as Peter Johnston
(hepatobiliary surgeon) for his care of these patients.
Correspondence: Mr J
P Keating, Senior Lecturer in Surgery, Wellington School of Medicine and Health
Sciences, PO Box 7343, Wellington South. Fax: (04) 389 5318; email: keating40@xtra.co.nz
References:
|
|
||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|