Abstract

Background: Pre-operative chemoradiotherapy prior to surgery has become
a mainstay of treatment for rectal cancers with threatened surgical margins. Many trials
using different chemo/ radiotherapy regimes, have examined various predictive and
prognostic factors [1-4]. We aimed to assess the impact of pre-treatment haemoglobin
(Hb) and length of the rectal tumour (clinically and radiologically) on 3-year disease
free survival and local recurrence in our patient population.
Methods: We have prospectively collected data on all consecutive patients with
histologically confirmed adenocarcinoma of the rectum treated with pre-operative
radiotherapy or chemoradiotherapy at the Mount Vernon Cancer Centre (MVCC) between March 1994 and December 2008. MRI staging has been routine since 1997.
This included data on patient demographics, presenting symptoms, (anaemia, rectal
bleeding), tumour stage (TNM Classification), differentiation, treatment regime (5FU,
capecitabine or two cytotoxic drugs), toxicity and outcomes (3 year disease free
survival). Haemoglobin level of 10g/dL or above was considered normal. Tumour
length was measured clinically using rectal examination, CT scan and by MRI (for
patients after 1997). Differences between populations were assessed using the chisquared
test, and survival analyses were performed using the Kaplan-Meier method
and the logrank test. In addition, a survival analysis was performed implementing
a Cox Regression with Backwards Stepwise (Wald) Regression re-evaluating the
potential prognostic factors found to be significant in univariate analysis. All analyses
were carried out using SPSS 14.0 (SPSS, USA).
Results: Three hundred and eighty four patients were eligible for our analysis. Median
age of diagnosis was 66 years; male:female ratio was 65:35. 191 were treated with
a concurrent treatment regimen consisting of 5FU/ folinic acid, 134 with Capecitabine, 18
with short course radiotherapy and 41 with combinations of 2 cytotoxics. Of the 384
patients, 20 had presenting pre-treatment Hb below 10 g/dL. On univariate analysis, Hb
was related to tumor length (p=0.02, eta square=0.03) and tumor T stage (p<0.01, eta
square=0.04) but not the lymph node status (p=0.568, eta square=0.01). Eta square (used
as an indication of the effect size) for the final model is 0.08 which reflects a moderate
strength. After undertaking a Cox-Regression Survival Analysis, neither Hb nor tumour
length showed any relation with 3 year disease free survival whereas treatment regime
(0.933 general exp. and p=0.01) and T Stage with patients being at stage 4 being at greater
risk (p=0.027 and exp. 0.604 vs p=0.405 and exp. 0.706 for those at stage 3) showed
significant impact. As for local recurrence, again Hb (p=0.150) and the tumour length
(p=0.676) showed no relation, however, T Stage was the most significant determinant as
of whether local recurrence occurred or not (p=0.04 with exp 2.132).
Conclusions: Contrary to previous studies [2, 3], our series does not show any relation
between Hb and/or tumour length on 3 year disease free survival and local recurrence.
Tumour stage remains the most significant determinant in predicting whether local recurrence occurred or not.