Myeloproliferative Neoplasms: an In-depth Case-control Study

MPNs are a group of haematopoietic malignancies resulting from a transformed haematopietic progenitor cell. They are characterised by overproduction of mature functional blood cells. MPNs were historically termed myeloproliferative disorders and have undergone numerous amendments in classification. The 2008 World Health Organisation (WHO) classifies a number of related disorders as MPNs including polycythemia vera (PV), essential thrombocythemia (ET) and primary/idiopathic myelofibrosis (PMF)2. Reported incidence rates vary from 0.02 to 2.8 per 100,000 persons3-5 but are not well characterised with half of patients asymptomatic at diagnosis6. In Northern Ireland (NI) the prevalence of ET is reported to be 18 per 100,000 persons (by correspondence, Mary McMullin). PV and ET are indolent neoplasms with better prognosis (88% and 92% 3-year survival, respectively) compared to PMF which has a 63% 3-year survival in the United States5.

The main study objectives are to:

• Develop a telephone-based questionnaire to investigate potential risk factors associated with MPNs.
• Assess the reliability and validity of the study questionnaire.
• Determine which control group to investigate by investigating:
• Response rates.
• Representativeness of control groups.
• Assess recruitment procedures.
• Determine if offering reimbursement of expenses (£10 per person) encourages control participation and improved representativeness to the general population.

A case-control approach was chosen as it is the most cost-efficient method of obtaining information on the aetiology of rare diseases such as MPNs. A prospective cohort study would require participation from millions of individuals followed over a long period of time. A hospital-based approach was chosen over a population-based approach to ensure the study was more feasible and could be rolled out to a number of centres in the UK.

The disorders under investigation are:

• Polycythaemia vera
• Essential Thrombocythaemia
• Primary myelofibrosis

Initially there will be one site in Belfast, Northern Ireland and one in Southampton, England. Cases will be identified and recruited through Prof Mary Frances McMullin at the Belfast City Hospital (BCH), Belfast Health and Social Care Trust, Northern Ireland and Dr Andrew Duncombe, at the University Hospital Southampton NHS Foundation Trust (UHS), Southampton, England.

Following completion of the exploratory study it is planned that other sites across the UK will be invited to participate through the Clinical Research Network.

. CASE ASCERTAINMENT Recruitment will run over a 12 month period. It is estimated that at least 100 incident and prevalent MPN cases will be recruited from the two sites during this time frame; 50 from Belfast, NI and 50 from Southampton, England. The targeted distribution of cases projected as feasible is 40 PV, 40 ET and 20 PMF cases. Fewer PMF cases are expected as the incidence is lower than PV and ET. Both incident and prevalent cases will be recruited in the exploratory case-control study and the distribution of time since diagnosis evaluated. One of the main limitations of including prevalent cases is that risk factors may differ in those who survive for a long period after diagnosis compared to those who die soon after diagnosis. Since PV and ET have low mortality rates it is expected that this will not impact the evaluation of the aetiology of these conditions to a large extent. However for PMF, which has a higher mortality rate than PV or ET, cases will be recruited as close to diagnosis as possible.

Eligible cases will be identified by the two lead clinicians Prof McMullin (Belfast) and Dr Duncombe (Southampton) during attendance at their routine clinics. Prof McMullin and Dr Duncombe will ensure that potential cases meet the study inclusion criteria (section 2.5) and provide detailed information on the categorisation of the case (PV, ET, PMF), which diagnostic criteria have been met and the clinical history of the patient.

Patients will be invited to participate in the study by letter from their lead clinician with an information sheet. Non-respondents will be randomised to either receive a reminder letter or a telephone contact from their health care team.

The choice of control group is an important consideration in any case-control study. Controls recruited from the general population through General Practitioners, population registers, neighbourhood controls or random digit dialling are ideal at identifying a representative sample of the population from which the case population was identified. However, these groups often have low response rates and selection bias is inferably encountered. Other control groups such as friend, family or hospital controls have much higher response rates but the factors associated with friendship, family or being in hospital may mean that these groups are less representative of the general population. For the purposes of this study two control groups will be recruited; GP controls and non-blood relatives (including spouses) or friend controls. The study will investigate response rates, differences in study characteristics between control groups and generalisability to the population using population-based statistics.

2.8.1 GP Controls Within each study site 5 GPs, reflective of the geographic distribution of MPN cases, will be approached to facilitate the identification and recruitment of controls and for performing phlebotomy. Each practice will be offered £200 for participation in the study.

2.8.1.1 Northern Ireland Control Recruitment in Northern Ireland will be conducted with the help of the Northern Ireland Clinical Research Network (Primary Care) (NICRN (PC)) who work in partnership with GP practices to provide support, guidance and facilitate research. The NICRN nurses who work with practices hold a Confidentiality Agreement and honorary contract with the practice. Practices are supplied with a study file, which is kept in a secure location at the practice and contains the study protocol and all relevant study information including the completed Research Governance form and contact details of researchers and NICRN (PC) staff.

With prior arrangement and agreement with the practice partners the research nurse will make an appointment with the practice manager (PM) who will either facilitate practice staff to conduct a computer search to randomly select potential controls or supply the nurse with their own unique user name and password to do so. The search will be conducted using the practice computer system's inbuilt randomisation programme, based on a numbered list of age and gender-matched patients (approximating 2 per MPN case, calculated on the basis of an attrition rate of 40-50%). The GP will be asked to confirm the eligibility of each potential control patient prior to them receiving study information from the practice.

Study information will be mailed to eligible control patients from the practice on practice headed paper and with a letter, signed by the GP, inviting them to participate. In order to track patient replies, each envelope will have a unique code number written on the front. A patient tracker form with the unique code number written beside the relevant name on the list of control patients is kept in the practice study file to track responders and non-responders for follow up purposes. No patient identifiable data will be removed from the practice at any time.

Non-respondents will be randomised to either receive a reminder letter approx two weeks after the first mail shot or to have a telephone call from practice staff using the follow-up telephone transcript.

2.8.1.2 Southampton The research nurse in Southampton will work in partnership with GP practices to provide support, guidance and facilitate research. An honorary contract and confidentiality agreement will be arranged with the practices.. This will include the nurses NMC number and copies of GCP training certificates. Practices are supplied with a study file, which is kept in a secure location at the practice and contains the study protocol and all relevant study information including the completed Research Governance form and contact details of researchers and NICRN (PC) staff.

With prior arrangement and agreement with the practice partners the research nurse will make an appointment with the practice manager (PM) who will either facilitate practice staff to conduct a computer search to randomly select potential controls or supply the nurse with their own unique user name and password to do so. The search will be conducted using the practice computer system's inbuilt randomisation programme, based on a numbered list of age and gender-matched patients (approximating 2 per MPN case, calculated on the basis of an attrition rate of 40-50%). The GP will be asked to confirm the eligibility of each potential control patient prior to them receiving study information from the practice.

Study information will be mailed to eligible control patients from the practice on practice headed paper and with a letter, signed by the GP, inviting them to participate. In order to track patient replies, each envelope will have a unique code number written on the front. A patient tracker form with the unique code number written beside the relevant name on the list of control patients is kept in the practice study file to track responders and non-responders for follow up purposes. No patient identifiable data will be removed from the practice at any time.

Non-respondents will be randomised to either receive a reminder letter approx two weeks after the first mail shot or to have a telephone call from practice staff using the follow-up telephone transcript.

2.8.2 Relative/Friend Controls Recruitment of 100 non-blood relative/friend controls will be undertaken by asking cases to pass on a flyer to up to 2 or 3 non-blood relatives/friends aged 18 years or older. Controls meeting the exclusion, section 2.7, will not be eligible for participation. Controls will not be individually matched to cases but must reside in the study area (Belfast or Southampton). Blood relatives will not be eligible for participation. Partners or friends MPN cases attending outpatient appointments in Belfast City Hospital or UHS will also be approached to participate in the study as study flyers will be available. A reminder to follow-up with recruitment of non-blood relatives/controls will be undertaken at the time of telephone interview with the cases. Those interested in participating in the study can contact the study team for the information sheet, consent form and other documentation by completing the tear off contact details strip on the flyer and posting (in the FREEPOST envelope that will be attached to the flyer) to the study team or by telephone or e-mail.

INFORMED CONSENT All cases and controls will be contacted by letter with a patient information leaflet, a consent form, patient contact form and work/residence history calendar. A pen will be included in the envelope in addition to a MPD Voice charity information sheet and a trolley token (for 50% of cases and controls). The number of subjects who refuse to participate will be documented. Subjects who refuse to participate at telephone contact will be asked whether they are willing to answer a short questionnaire to assess the representativeness of the recruited subjects. Subjects will be given the opportunity to withdraw from the study at any time however the information provided up to that point will be retained and used within the study. Subjects will be asked for verbal consent at time of telephone interview to have the interview recorded both before and after (if agreeable) the recording device is activated.