Chemotherapy and radiotherapy use in patients with lung cancer in Australia, Canada, the UK and Norway 2012-2017: an ICBP population-based study

Abstract

Background: International variation in lung cancer survival may be partly explained by variation in stage-specific treatment use, but relevant comparative evidence is sparse. As part of the International Cancer Benchmarking Partnership, we examined use of chemotherapy and radiotherapy in population-based cancer registry data.

Methods: Linked population-based data sources were used to describe use and time to first treatment for either chemotherapy or radiotherapy in patients with lung cancer diagnosed in study periods during 2012-2017 in 16 jurisdictions of Australia, Canada, the UK and Norway.

Results: There was large variation in the proportions of patients with lung cancer receiving chemotherapy (ranging from 23% in Northern Ireland to 45% in Norway) and radiotherapy (ranging from 32% in England to 48% in New South Wales and 50% in Newfoundland and Labrador). Across jurisdictions, chemotherapy use decreased steeply with increasing age, regardless of stage at diagnosis. For radiotherapy use, in stage 1-3 cancer three patterns were observed: (a) steep decrease with increasing age (UK jurisdictions, Saskatchewan-Manitoba); (b) a relatively flat pattern (Norway, Alberta, British Columbia, Atlantic Canada, New South Wales) and (c) increasing use with increasing age (Ontario).Time to radiotherapy initiation was longer in the UK jurisdictions than elsewhere; time to chemotherapy was longer in the UK and Canadian jurisdictions except Ontario.

Discussion: Use of chemotherapy and radiotherapy in patients with lung cancer varied substantially between jurisdictions during the mid-2010s within age-stage strata. Reasons for these variations are unclear. Differences in non-surgical treatment use are plausibly associated with international variation in lung cancer survival.

Keywords: chemotherapy; epidemiology; lung cancer (non-small cell); lung cancer (small-cell); radiotherapy.

Figure 1. Proportion of patients with lung cancer in each jurisdiction who received chemotherapy (top) or radiotherapy (bottom) treatment, overall, for non-advanced stage, for advanced stage and for those with no recorded stage. Grey diamonds (bottom of panels) show the pooled estimate from a random-effects meta-analysis, with the widest point being the central estimate, the edges of the diamond representing 95% CIs, and the wider grey lines representing 95% prediction intervals. Same colour is used for estimates of jurisdictions of the same country. D, distant; L, localised; R, regional. See also online supplemental appendix 2 table 1.

Figure 2. (A–B) Proportion of patients treated with chemotherapy (A, top) and radiotherapy (B, bottom) by age, for all stages (left), non-advanced stage (middle) and advanced stage (right). Coloured lines show jurisdictional results, with jurisdictions within the same country having the same colour lines. Dashed vertical lines show 95% CIs, noting that they often align closely to central estimates, making them hard to decipher. Grey lines show results for all other jurisdictions for ease of comparisons. Proportions based on counts of between 1 and 9 are not shown. Atl. Canada, Atlantic Canada; BC, British Columbia; L, localised; NI, Northern Ireland; NSW, New South Wales; R, regional; RT, radiotherapy; SK-MB, Saskatchewan-Manitoba. See also online supplemental appendix 2 table 2.

Figure 3. Proportion of patients treated with chemotherapy (top) and radiotherapy (bottom), by sex. Squares or circles show the point estimate, and horizontal lines show 95% CIs. L, localised; R, regional; SK-MB, Saskatchewan-Manitoba. See also online supplemental appendix 2 table 3.

Figure 4. Cumulative percentage of patients in each country who had received chemotherapy (top) or radiotherapy (bottom) treatment, by elapsed time since diagnosis, for all stages, stages 1–3 (or localised (L)/regional (R)) or stage 4 (or distant). Stage-specific results for Australia are based on New South Wales data only. See also online supplemental appendix 2 tables 4 and 5.

Figure 5. (A—left, B—right) Scatter plot of jurisdictional use of surgery versus chemotherapy and surgery versus radiotherapy for a subset of jurisdictions. Coloured circles show jurisdiction-specific use, all patients with lung cancer, with jurisdictions within the same country having the same colour circle. Data related to surgery use do not always align with study period for chemotherapy or radiotherapy use, or population basis (see ‘Methods’ section). Pearson’s r for surgery versus chemotherapy=0.75 (95% CI to 0.15, 0.97); Pearson’s r for surgery versus radiotherapy=0.60 (95% CI to 0.42, 0.95).

Figure 6. Scatter plot of jurisdictional use of radiotherapy versus chemotherapy, for all stages, stages 1–3 (or localised (L)/regional (R)), or stage 4 (or distant). Coloured circles show jurisdiction-specific use, with jurisdictions within the same country having the same colour circle.

Figure 7. Scatter plot of jurisdictional 5-year net survival against use of chemotherapy, radiotherapy and surgery. Data for 5-year net survival from Arnold et al; data on chemotherapy and radiotherapy use from the study analysis samples; for data on surgery use, please see ‘Methods’ section. Same colour is used for estimates of jurisdictions of the same country.