The perspectives of survivors of Hodgkin lymphoma on lung cancer screening: A qualitative study

Hodgkin lymphoma survivors (HLS) are at excess risk of lung cancer as a consequence of HL treatment. HLS without a heavy smoking history are currently unable to access lung cancer screening (LCS) programmes aimed at ever smokers, and there is an unmet need to develop a targeted LCS programme. In this study we prospectively explored HLS perspectives on a future LCS programme, including motivating factors and potential barriers to participation, with the aim of identifying ways to optimise uptake in a future programme.


| INTRODUCTION
Hodgkin lymphoma (HL) is a malignancy which predominantly affects young adults and the elderly. 1 With modern treatments, over 90% of those diagnosed at a young age are cured and live to experience the late effects of treatment. 2 Alkylating agents and radiotherapy used to treat HL put survivors at excess risk of developing subsequent malignant neoplasms (SMNs), with smoking having a multiplicative effect on lung cancer risk. [3][4][5] Compared to survivors of other primary cancers, HL survivors have the highest risk for developing SMNs with those diagnosed with HL in adolescence or young adulthood being at higher risk compared to those diagnosed in childhood or during older adulthood. 6 The most common SMNs are breast cancer, lung cancer and colorectal cancer, with 30-year reported cumulative incidences of 16.6%, 7.1% and 2.5%, respectively. 5 Lung cancer and gastrointestinal cancers (of the upper and lower gastrointestinal tracts) are the leading causes of SMN-related mortality in HL survivors. 3,5,7 Despite evidence of an excess risk of breast, lung and bowel cancer, the sole targeted screening programme currently available to HL survivors in England is the NHS breast screening programme for women at very high risk of breast cancer. 8 Through this programme, women who were treated for HL (and non-HL) with radiotherapy to the breast tissue before the age of 30 are invited to undergo early annual breast cancer screening with an MRI scan and mammogram. In the case of bowel cancer screening, HL survivors aged 60-74 can access the national bowel screening programme in the same manner as the rest of the general population. With regard to lung cancer screening (LCS), several large trials conducted in the general population in the past decade have shown that low-dose CT screening of the thorax reduces lung cancer-related mortality in ever smokers by detecting lung cancers at an early stage. 9-12 In England, LCS is being piloted in ever smokers aged 55-74 with eligibility to undergo screening being determined by lung cancer risk calculators. 13 Such pilots are unlikely to benefit HL survivors because the average age at lung cancer diagnosis (45 years) 14 falls below the screening threshold (55 years) and lung cancer risk calculators do not account for risk associated with HL treatments. Consequently, HL survivors without a heavy smoking history do not meet screening criteria for pilots aimed at ever smokers, even if eligible by age, creating a need for a LCS programme targeted specifically towards HL survivors.
Among ever smokers in the general population, research has shown that LCS is acceptable. 15 Uptake rates are variable, ranging from <5% in the United States 16 to 26% in a recent community-based pilot in Manchester, UK. In the UK Lung Screening pilot trial, higher socioeconomic status was associated with a positive response to a screening invitation and subsequent participation. Current smokers were less likely to participate than former smokers and practical barriers were the most common reasons for nonparticipation. 18,19 Qualitative studies have identified fatalistic attitudes towards lung cancer and smoking-related stigma as barriers to participation in LCS. 20,21 The views of HL survivors towards LCS have not previously been explored so it is not known whether the barriers to LCS participation in ever smokers also apply to HL survivors or whether their views differ. By gaining an understanding of the motivating factors and barriers to LCS participation in HL survivors, it may be possible to design a future LCS programme which reduces barriers to participation, thus optimizing uptake rates. Thus, an exploration of the perspectives of HL survivors towards a future LCS programme is warranted and is addressed by this qualitative study.

| METHODS
This study employed a qualitative design, using semistructured telephone interviews with survivors of HL. Ethical approval was granted by the North West Greater Manchester West Ethics Committee (ref: 20/NW/0025).

| Recruitment
HL survivors aged 18-80 who were known to be at excess risk of lung cancer due to previous treatment and who had survived relapse-free for at least 5 years after completing treatment were eligible for inclusion, regardless of smoking status. The upper age limit for eligibility reflects the highest age threshold for eligibility to participate in LCS trials in ever smokers published to date, whilst the lower age limit reflects the fact that lung cancer cases have been detected at all time periods after 5 years since completion of treatment. The wide age range threshold for also eligibility also reflects the bimodal distribution of HL in the young and the elderly. Patients with a diagnosis of lung cancer at any time, or who had participated in a pilot LCS programme were excluded. We identified potential participants from a prospectively maintained database of lymphoma survivors with at least 5 years follow-up (ADAPT). One hundred and ninety-four HL survivors were eligible for inclusion. Of these, 80 were randomly selected, stratified according to attained age, time since treatment, prior treatment and smoking history (see Table 1 for stratification criteria). Potential participants were sent an invitation letter and the participant information sheet by post. The invitation letter was signed by a doctoral student working in their treating team. Before the interview, participants received brief written information on LCS, provided written consent and completed a short questionnaire. The purpose of the questionnaire was to collect sociodemographic data which was not available from the ADAPT database (ethnicity, employment status and education), to confirm smoking history and collect other health-related data such as self-rated health and prior participation in cancer screening opportunities.

| Data collection
The first author conducted telephone interviews lasting approximately 20 minutes between March and April 2020. The interview schedule explored perceptions of lung cancer and LCS, with prompt questions to explore risk perception, the perceived benefits of LCS and potential barriers to attending.

| Reflexivity statement
The interviewer was a clinician and doctoral student working within the participants' treating team. In the majority of cases, the interviewer had not been involved in the participants' care during treatment or follow-up. In questioning participants and answering their questions about lung cancer risk and screening, the interviewer adopted a nonjudgemental and neutral stance. To respond to participants' questions about lung cancer risk and a future LCS programme, the interviewer referred to their knowledge acquired through review of the relevant literature, taking care to neither promote the benefits nor risks of lung cancer, or other cancer screening. Participants were aware that this study was being conducted with the view to eventually offering LCS to at risk survivors.

| Data analysis
Interviews were audio-recorded and transcribed verbatim by an external company. Transcripts were linked to a pseudo-anonymized study ID number. An inductive approach to thematic analysis was used to analyse the transcripts. 22 This began with familiarization with transcripts. Whole transcripts were examined one by one and data pertinent to the research questions were identified and coded by the first author. Throughout this process, new data were applied to an existing code, or a new code was created. A second researcher followed the same process, coding nine randomly selected transcripts.
The two researchers discussed the codes they had independently developed, their relationship to each other and emergent themes, following which the first author finalized the coding framework. The second researcher was involved throughout the development of the thematic analysis. Participants have been sent a summary of the study findings, but were not involved in the analytic process.

| Participant characteristics
Thirty HL survivors took part in the study. Participants included men and women with a median age of 53 years. Most were white British.
Around half had a university education and most were in employment. The majority had received both chemotherapy and radiotherapy in keeping with treatment guidelines and trends over the last 40 years. Around two thirds rated their health as fair to poor. Five reported a history of SMN. Two thirds reported a prior invitation to undergo cancer screening, reflecting the predominance of female participants in the study invited to early breast cancer screening.
Never smokers were a large majority. Participant characteristics are detailed in Table 2.

| Thematic analysis
The quotes presented here are linked to study ID number, patient gender (M/F) and age (study ID number-gender-age). Tables of quotes illustrative of each theme are available as Supporting Information.

| Lung cancer risk perceptions
As most participants were not aware that their prior cancer treatment increased lung cancer risk before participation in this study, there are two subthemes. The first describes participants' risk perceptions unrelated to cancer treatment (beliefs held before participation in this study) and the second relates to the impact of their prior knowledge and experiences on risk perceptions.

Subtheme 1: Lung cancer risk perceptions before study participation
Most participants were not aware of their excess risk of lung cancer due to their cancer treatment before being contacted about the study. Several factors appeared to influence lung cancer risk perception. Participants associated lung cancer with smoking and a lack of smoking history reduced perceived risk.
One participant referenced the association of lung cancer with occupation or secondary exposure to cigarette smoke: 'I'm not a particularly at risk group for lung cancer. I've never smoked, I've never worked in industry or in a smoky environment' (P43, F, 66).
Despite the association of lung cancer with smoking, several former smokers expected or hoped that quitting smoking abolished lung cancer risk: T A B L E 1 Stratification criteria to guide random selection of participants
(P23, F, 60) Similarly, one participant described being diagnosed with ischaemic heart disease at routine follow-up before any serious consequences.
In contrast, a participant who had suffered a heart attack 'out of nowhere' lamented the lack of screening for ischaemic heart disease before the event.
Whilst the main focus of discussion was early diagnosis, several Reassurance and information about one's health.

| Theme 3: Concerns and potential barriers to participation
Whilst a CT scan was perceived as fast, painless and noninvasive test Waiting for a screening result was frequently described as worrying time, although the severity of worry experienced varied between participants. Some said they would be 'mildly worried' and able to 'put it out of my mind', but others said the result would always be on their mind.
Prior experience of waiting for scan results helped some people cope with worry, or worry less. Almost all participants said that they would attend LCS despite their concerns.
Explaining their willingness to undergo screening, many expressed that it was better to know either way, reflecting uncertainty about their health and desire for surveillance and screening: To not know, is a greater fear than knowing, to me. In this study we explored the perspectives of long-term survivors of HL towards LCS. We report high levels of enthusiasm, possibly reflecting views in the general population towards cancer screening and the positive perceptions of screening held by cancer survivors, 23,24 who are known to be more likely to participate in cancer screening than noncancer survivors. 25,26 LCS research to date has focussed on ever smokers. In this study, we found that HL survivors differ from ever smokers in that they perceive early-stage lung cancer as curable, thereby differentiating them from ever smokers who frequently report fatalistic attitudes as a barrier to undergoing LCS. 15 Half the participants had college or university education, which is associated with higher uptake of cancer screening. 37 It is therefore possible that our findings, in terms of enthusiasm for screening and reported barriers, do not reflect the views of all HL survivors. In particular, those with fewer material resources, for example those who do not own a vehicle, and people with a lower than average reading age who might have difficulty accessing written LCS educational materials, are likely to experience greater barriers to participation than reported by our participants. Current smokers, who are at the highest risk of lung cancer, were significantly underrepresented in our sample, which could reflect smoking practices in HL survivors; the rate of current smoking in HL survivors has been reported as 7%. 28 Nevertheless, we cannot report on the perspectives of HL survivors who currently smoke.
The participants may be more enthusiastic about LCS than those who did not respond to our study invitation, which is likely to have introduced a response bias. HL survivors who remain in follow-up may have better knowledge of late effects than those discharged soon after completion of treatment, and our results may therefore not be representative of the national survivor population discharged from long-term follow-up. Finally, the invitation to this study came from a doctor who worked at the participants' treating centre, which could have led to social desirability bias, with fewer reported barriers reported as a result. 38

| CONCLUSION AND DIRECTIONS FOR FUTURE RESEARCH
Our findings suggest that HL survivors would be willing to attend LCS, motivated by perceived benefits of cancer screening tests, and that uptake of a future LCS programme by HL survivors may exceed uptake by ever smokers. There is no established protocol for survivorship care for people treated for HL, with follow-up care for those in remission varying widely throughout the country. Breast cancer screening for HL survivors is coordinated at a national level by Public Health England rather than individual treating centres and it is likely that a future targeted LCS programme would follow a similar structural approach. However, the most pressing challenge before implementation will be the identification of long-term survivors at risk of lung cancer, many of whom will be discharged from follow-up, which will require a coordinated effort by treating centres. Further steps towards delivering LCS include large scale epidemiological cohort analyses to determine an appropriate lung cancer risk threshold to guide eligibility for screening and the development of LCS educational materials to support decision making and reduce barriers to screening uptake. Future studies should address the feasibility of such a programme and explore barriers to participation in a reallife setting, particularly in current smokers and survivors discharged from routine follow-up.