BACKGROUND

In patients with squamous cell carcinoma of the anus (SCCA), lymph node positivity (LNP) indicates poor prognosis for survival and is central to radiotherapy planning. Over the past three decades, LNP proportion has increased, mainly reflecting enhanced detection with newer imaging modalities; a process known as nodal stage migration. If accompanied by constant T stage distributions, prognosis for both lymph node-positive and lymph node-negative groups may improve without any increase in overall survival for individual patients; a paradox termed the Will Rogers phenomenon. Here, we aim to systematically evaluate the impact of nodal stage migration on survival in SCCA and address a novel hypothesis that this phenomenon results in reduced prognostic discrimination.

METHODS

We did a systematic review and meta-regression to quantify changes in LNP over time and the impact of this change on survival and prognostic discrimination. We searched MEDLINE, Embase, and the Cochrane Library to identify randomised trials and observational studies in patients with SCCA published between Jan 1, 1970, and Oct 11, 2016. Studies were eligible if patients received chemoradiotherapy or radiotherapy as the main treatment, reported LNP proportions (all studies), and reported overall survival (not necessarily present in all studies). We excluded studies with fewer than 50 patients. We extracted study-level data with a standardised, piloted form. The primary outcome measure was 5-year overall survival. To investigate scenarios in which reduced prognostic discrimination might occur, we simulated varying true LNP proportions and true overall survival, and compared these with expected observed outcomes for varying levels of misclassification of true nodal state.

FINDINGS

We identified 62 studies reporting LNP proportions, which included 10 569 patients. From these, we included 45 studies (6302 patients) with whole cohort 5-year overall survival, 11 studies with 5-year survival stratified by nodal status, and 20 studies with hazard ratios in our analyses of temporal changes. In 62 studies, the LNP proportions increased from a mean estimate of 15·3% (95% CI 10·5-20·1) in 1980 to 37·1% (34·0-41·3) in 2012 (p<0·0001). In 11 studies with prognostic data, increasing LNP was associated with improved overall survival in both lymph node-positive and lymph node-negative categories, whereas the proportions with combined tumour stage T3 and T4 remained constant. In 20 studies, across a range of LNP proportions from 15% to 40%, the hazard ratios for overall survival of lymph node-positive versus lymph node-negative patients decreased significantly from 2·5 (95% CI 1·8-3·3) at 15% LNP to 1·3 (1·2-1·9; p=0·014) at 40% LNP. The simulated scenarios reproduced this effect if the true LNP proportions were 20% or 25%, but not if the true LNP proportions were 30% or greater.

INTERPRETATION

We describe a consequence of staging misclassification in anal cancer that we have termed reduced prognostic discrimination. We used this new observation to infer that the LNP proportions of more than 30% seen in modern clinical series (11 out of 15 studies with a median year since 2007) are higher than the true LNP proportion. The introduction of new staging technologies in oncology might misclassify true disease stage, spuriously informing disease management and ultimately increasing the risk of overtreatment.

FUNDING

Bowel Disease Research Foundation.