Mednotes

L9 Public Health Screening for lung cancer

Properties2
tagspublichealth, lecture
aliasesL9 Public Health Screening for lung cancer

19 min read

Overview

This lecture covers general principles of screening, then builds the case for a national lung cancer (LC) screening programme in Aotearoa New Zealand through an equity lens: the scale of Māori/non-Māori disparities in lung cancer incidence, mortality and survival; international trial evidence that low-dose CT (LDCT) screening reduces mortality; and the Te Oranga Pūkahukahu research programme, which re-examined cost-effectiveness with equity weighting and is running a clinical trial comparing primary-care-led versus centralised invitation to risk assessment and LDCT scanning.

Principles of Screening

  • Purposes of screening:
    • Early identification of risk of an illness/condition
    • Early detection of a condition/illness
    • For cancer specifically: early detection and management of pre-cancerous lesions, and earlier detection and management of cancers
  • Screening can be delivered as an organised national programme, organised within a practice, or opportunistically
  • Rationale: screening saves work later on and screening saves lives

Criteria for assessing (organised) screening programmes

  1. The condition is a suitable candidate for screening
  2. There is a suitable test
  3. There is an effective and accessible treatment or intervention for the condition identified through early detection
  4. There is high-quality evidence, ideally from randomised controlled trials, that the programme reduces mortality or morbidity
  5. Potential benefit should outweigh potential physical and psychological harm (from the test, diagnostic procedures and treatment)
  6. The health care system can support all elements of the screening pathway, including diagnosis, follow-up and programme evaluation
  7. Social and ethical issues are considered
  8. Cost-benefit issues are considered

(National Advisory Committee on Health and Disability (National Health Committee), 2003)

Why Lung Cancer Screening Is Needed: the Equity Case

  • Incidence (1996-2017): Māori females >4x higher than non-Māori and increasing; Māori males >3x higher but reducing
  • Mortality: lung cancer is the leading cancer cause of death for both Māori and non-Māori; Māori mortality is >3x higher than non-Māori
  • Lung cancer is the largest single cancer contributor to the life-expectancy gap between Māori and non-Māori. For males, of a 7.2-year total life-expectancy gap, 1.4 years is attributable to cancer and ~0.75 years to lung cancer specifically. [flag: the source text runs a second, female-specific set of figures into this same line without a clear break (“~0.75 y6.7 y, cancer 2.1 y, lung cancer 1.1 y”); it likely reads “Females - total 6.7 y, cancer 2.1 y, lung cancer 1.1 y” but the female label is missing from the extracted text, so these female figures are reported here with that caveat rather than asserted as certain]
  • Lung cancer mortality is far higher than any other major cancer shown (breast, colorectal, stomach, pancreas, prostate, leukaemia, eye/brain CNS, non-Hodgkin lymphoma, melanoma), and shows by far the largest Māori/non-Māori gap of the cancers compared - Māori markedly higher, non-Māori much lower. [flag: chart carried no numeric data labels, so only the relative pattern is recorded, not exact rates]
  • Lung cancer is among the cancers with the lowest 5-year survival (<50%), sitting between pancreas (lowest) and liver, though survival improved somewhat between 1998/99 and 2016/17. [flag: exact percentage values were not labelled on the chart; positions are visual estimates]
  • Geographic variation in 5-year survival (cancers diagnosed 2010-2016), by region and ethnicity (nMnP = non-Māori non-Pacific):
    • Northern: nMnP >18%, Māori 12%, Pacific 16%
    • Te Manawa Taki: nMnP ~11%, Māori ~8%, Pacific 14%
    • Central: nMnP ~11%, Māori ~6%, Pacific ~15%
    • Te Waipounamu: nMnP ~12%, Māori ~11%, Pacific ~17%
    • In every region, Māori 5-year survival is the lowest of the three groups
  • One slide presented two personal photographs (a young boy in a sailor’s uniform; an older man) with no caption or text. [flag: no caption confirms it, but the two photos appear to be the same person as a boy and as an older man, likely used as a personal/community story illustrating the lecture - recorded as visible content only, not confirmed]

The Case for Screening and Evidence of Benefit

  • To improve Māori lung cancer outcomes, action is needed across the whole pathway: preventing uptake (of smoking), smoking cessation, screening of asymptomatic people, early detection of symptomatic disease, improved diagnostics/treatment, improved mutation testing, and drug therapies
  • High-quality international trials (NLST; NELSON) have demonstrated LDCT screening reduces lung cancer mortality by 20-26% (36% in women); meta-analysis confirms this effect
  • Stage shift with targeted screening (Manchester Lung Health Check pilot vs normal Greater Manchester figures): screening shifted diagnoses toward earlier stages - roughly 65% vs ~19% at stage 1, similar low rates (~10-13%) at stage 2, ~7% vs ~24% at stage 3, and ~13% vs ~48% at stage 4. [flag: values were read from bar heights with no data labels, so are approximate]
  • For Aotearoa NZ, lung cancer screening:
    • Cannot be inequitable, unlike other current cancer screening programmes
    • Needs to sit within a national organised screening programme, because opportunistic programmes (e.g. CVDRA, DAR) have failed to deliver for Māori
    • Offers a focus and a lever for improvements and investment across the whole care pathway

How Lung Cancer Screening Differs From Other Cancer Screening

Lung cancer screening is a two-stage, risk-based process rather than screening the whole eligible population:

  1. Identifying people potentially at high risk - invitation to risk assessment
  2. Risk assessment of those above a risk threshold - invitation to LDCT scan

Because only a risk-selected subset is scanned, the balance of benefits and harms depends on the risk-prediction process and the threshold chosen, unlike programmes where the entire eligible population is screened. This balance needs to be understood separately for Māori and non-Māori, and a health-literacy lens together with shared decision-making (SDM) are critical.

Screening pathway: Health Promotion Initiatives -> Invitation -> Screening Procedure (branching to a negative or positive result) -> Diagnosis -> Treatment, with a Recall feedback loop from the Screening Procedure back to Invitation for subsequent rounds.

Key features of the programme:

  • Identification of potentially eligible people
  • Risk assessment
  • Shared decision-making
  • LDCT scan
  • Nodule identification, leading to either nodule management or immediate investigation depending on risk
  • Incidental findings
  • Subsequent screens, at either 1- or 2-yearly intervals
  • Adjacent opportunities: smoking cessation, COPD assessment and management

Te Oranga Pūkahukahu Research Programme

The programme is led by Prof Sue Crengle with a Māori steering group, Te Hā Kotahi (Māori consumer group) and a Technical Research Group. Its structure includes a Māori Caucus and Māori data governance groups (spanning the whole programme), Te Hā Kotahi and a Manawatu Advisory Group feeding into a Steering Group, a Technical Advisory Group and international collaborators feeding into a Core Research Team, whose current work comprises the Invitation Trial, a COPD Study and a Risk Prediction Study, plus proposed further projects.

Timeline (2020-2029): cost-effectiveness analysis (2020) -> surveys (2020-21) -> focus groups (2021) -> Te Hā Kotahi (ongoing 2021-2029) -> Invitation Trial, 514 scans (2021-2024) -> COPD Study (2022-2024) -> Risk Prediction Study, 1,200 scans (2023-2025) -> Second Round Scans, 1,714 scans (2024-2027) -> AHRR study (2024-2028) -> Māori Provider study, 400 scans (2025-2028) -> AI project (2026-2028).

Cost-Effectiveness Re-Analysis

  • A 2018 NZ-based cost-effectiveness analysis concluded CT screening for lung cancer was unlikely to be cost-effective for any population group
  • The research group was concerned equity was not sufficiently weighted in that analysis, and had questions about some of its assumptions, so conducted a re-analysis with different base assumptions, updated parameters from more recent evidence, and specific equity analyses
  • The re-analysis found a national LC screening programme is likely to be cost-effective for all groups, improving total population health and reducing inequities for Māori
  • Māori are likely to see greater health gains: ICERs ranged from NZ39,100 per HALY for non-Māori males; HALYs per capita were approximately twice as great for Māori females compared with non-Māori females, and 25% greater for Māori males compared with non-Māori males
  • Absolute inequities in lung cancer reduce with CT screening, but relative inequities in early-stage treatment are little affected - meaning current pathway inequities limit the benefit screening alone can achieve, so inequities along the whole health pathway also need addressing now

Survey of Attitudes and Beliefs

  • Completed in Auckland, Waitematā and Northland DHBs (Auckland/Waitematā: 306 main participants, 103 whānau members; Northland: 82 main participants, 30 whānau members), and ongoing in MidCentral DHB
  • 91% (+8% maybe) said they would attend a screening programme if told they were at “high risk” and offered a free lung CT scan
  • Biggest potential barriers to belief/understanding: 38% agreed lung cancer usually can’t be cured once symptomatic; 32% worried about health workers judging them for having smoked; 25% agreed treatment is usually worse than the cancer itself
  • Factors most likely to increase attendance: desire to find any lung cancer early (76%, no differences by age, gender, education or smoking status); reassurance from a clear scan (63% overall, stronger in the 60-69 age group (72%) than the 50-59 group (50%))
  • Factors most likely to reduce attendance: other responsibilities (16%), cost (15%), worry about judgement over smoking history (15%), transport/timing of appointments (14%), worry about radiation exposure (14%)

Invitation Trial: Primary Care vs Central Hub

Design: a randomised trial comparing primary-care-led versus centralised invitation approaches, in Māori aged 55-74 not classified as “never smoker,” including risk assessment and SDM, aiming for 514 CT scans. Primary research question: does inviting people to lung screening through primary care achieve higher Māori participation than a central hub-based invitation method?

Secondary outcomes: proportion of Māori potentially eligible for screening; proportion of potentially eligible people who agree to risk assessment; proportion of those risk-assessed who take up screening; screening outcome data (positive scans, cancers, follow-up scans, investigations, incidental findings, harms); burden on participants and on GP practices in the primary-care arm; experience and acceptability of study processes. Add-on studies: ctDNA and biomarkers, and a COPD study.

Process after randomisation: letter with information -> up to 5 contact attempts -> explain study -> risk assessment -> shared decision-making -> CT scan (with bloods, spirometry and symptom assessment) -> results -> any required follow-up.

Trial flow (CONSORT-style):

  • 48 practices recruited, 2,639 potentially eligible subjects identified, then pairwise randomised (24 pairs)
  • Primary care (PC) arm: 24 practices, 1,179 patients, 28% excluded
  • Central hub (CH) arm: 24 practices, 1,460 patients, 23% excluded
  • Cohort to be contacted: PC n=847, CH n=1,120
  • Not contacted after 5+ attempts: PC 23% (n=193), CH <1% (n=4)
  • Uncontactable: PC 22%, CH 22%
  • Contacted: PC 466, CH 873
  • Declined: PC 12%, CH 21%; Ineligible: PC 10%, CH 12%
  • Extra patients receiving an LDCT scan after PC practices censored: stated as 117 in the slide text. [flag: the diagram box for this figure printed “n=113,” a possible discrepancy with the “117” stated in the slide text; both figures are recorded here]

Preliminary primary outcome (proportion eligible for contact who completed a CT scan, N=847 PC, N=1,120 CH): 24% in the central hub arm completed a scan versus 14% in the primary care arm; weighted risk difference 8.7% (95% CI 5.4-11.9%), p<0.0001; crude RR 1.65. For comparison, the Manchester Lung Health Check T0 pilot (Crosbie et al., Thorax 2018) sent letters to 9,926 estimated ever-smokers: 29% booked an appointment, 26% attended, 25% were risk-assessed, and 14% had an LDCT performed.

Other preliminary outcomes (central hub vs primary care): contacted - 78% vs 55% (weighted difference 20.6%, 95% CI 16.7-24.2%, p<0.0001, crude RR 1.42); risk-assessed - 45% vs 34% (p<0.0001; weighted difference 9.2%, 95% CI 4.6-13.9%, p<0.0001, crude RR 1.31); risk ≥2% - 26% vs 15% (p<0.0001; weighted difference 9.8%, 95% CI 6.1-13.6%, p<0.0001, crude RR 1.68). The central hub arm outperformed the primary care arm at every stage measured.

Risk assessment outcomes (including the 117 extra patients): of 1,009 people risk-assessed, 55% were eligible for a CT scan (547 participants); 92% of those consented to a scan (505/547). Outcome breakdown: proceeded with CT 505, did not attend/changed mind 18, CT declined 16, SDM declined 6, other 2.

CT results - nodules: 15 cancers detected to date, a detection rate of ~3%, consistent with the anticipated rate. Findings by risk category: PC1 (no significant nodules, risk <1.5%) 64% (n=322); PC2 (low risk, 1.5% to <6%) 25% (n=124), 12-month follow-up scan; PC3 (moderate risk, 6% to <30%) 10% (n=49), 3-month follow-up scan; PC4 (high risk, >30%) 1% (n=5), immediate follow-up; PC5 (suspicious: mass/lesion, mediastinal or hilar lymphadenopathy irrespective of size) 1% (n=5), immediate follow-up.

CT results - incidental findings: IF4 (major, potentially life-threatening, direct hospital referral) 0%; IF3 (indicative of cancer elsewhere, e.g. breast, or other significant findings e.g. dilated aortic aneurysm, requiring urgent referral) 1% (n=7); IF2 (other non-cancer findings requiring non-urgent referral, e.g. fibrotic interstitial lung disease, TB, bronchiectasis) 21% (n=106); IF1B (non-cancer findings manageable in primary care, e.g. COPD, ectatic aorta, coronary calcification) 49% (n=245); IF1A (findings not usually associated with a beneficial intervention, e.g. bronchial wall thickening) 29% (n=147).

Clinical outcomes (as of August 2024): 12 participants diagnosed with lung cancer (2.4%): 7 had surgery, 2 had chemotherapy/radiation/immunotherapy, 2 had radiotherapy, and one died prior to treatment (had lung cancer plus bilateral breast cancer with metastatic disease, and died after a fall). A further 3 cases were diagnosed since, bringing the rate to ~3%. Of the 12 cases, 11 were treated with curative intent (92%). Current 5-year survival for lung cancer overall is <19%, but ~9 of the 11 curative-intent cases (80%) are expected to survive 5 years. 2 people had a recurrence as of August 2024.

Participant experience (interim reports, 2022): participants described being pleased to receive an invitation and wanting the programme rolled out nationally like breast and bowel screening; feeling the programme addressed long-standing inequities in lung cancer/cancer for Māori; valuing early detection (“early detection could save my life”; “I would prefer to know so it could be treated”); feeling settled knowing the study could help themselves and others; valuing the shared decision-making discussion (“without it I was unable to make a decision”) and friendly, informative staff who made themselves available for further questions. One participant’s response was mixed: “50/50 pleased and relieved, but a little disappointed [a] fuller review was not done to totally negate concerns.”

Self-test

  1. List the three ways screening can be delivered, and give the two overall justifications the lecture gives for screening.
  2. List the 8 National Health Committee criteria for assessing whether an organised screening programme should go ahead.
  3. Describe the pattern of lung cancer incidence and mortality inequity between Māori and non-Māori described in the lecture.
  4. Explain why lung cancer screening cannot follow the same “screen the whole eligible population” model as other NZ cancer screening programmes, and describe the two-stage process it uses instead.
  5. Describe the screening pathway from health promotion through to treatment, including its feedback loop.
  6. What mortality reduction have international trials (NLST, NELSON) shown for LDCT lung cancer screening, and what evidence type supports this?
  7. Describe the “stage shift” finding from the Manchester Lung Health Check pilot.
  8. Why did the research group re-analyse the 2018 NZ cost-effectiveness assessment of lung cancer screening, and what did the re-analysis conclude, including its equity findings?
  9. In the invitation-trial survey, what proportion of respondents said they would attend screening if told they were high risk, and what were the top three barriers to belief identified?
  10. Distinguish the primary care and central hub invitation arms of the Invitation Trial in terms of the proportion of eligible people who completed a CT scan, including the statistical comparison given.
  11. Describe the risk-assessment outcomes for the 1,009 people risk-assessed in the Invitation Trial: how many were eligible for a CT scan, and how many of those consented?
  12. Distinguish the PC1-PC5 nodule risk categories used to report the Invitation Trial’s CT findings, with the proportion found in each.
  13. Distinguish the IF1A/IF1B/IF2/IF3/IF4 incidental finding categories, including one example of each and its urgency.
  14. Summarise the clinical outcomes for the 12 lung cancers diagnosed by August 2024, including treatment intent and expected survival.
  15. A Māori patient aged 60, a long-term ex-smoker, is invited to a risk assessment through a central-hub lung cancer screening trial rather than through their GP. Using the trial’s findings, predict how this is likely to affect their probability of completing a CT scan compared with a primary-care invitation, and explain why.

Answers

Graph View

Backlinks