Popularity of less frequent follow up for breast cancer in randomised study: initial findings from the hotline study

Introduction

Most patients who relapse after treatment for primary breast cancer do so within the first five years of follow up.1 As curable intramammary relapses are more sensitively detected by mammography than by clinical examination,2 regular mammography after lumpectomy has now become accepted practice in most centres. Asymptomatic detection of disseminated disease is also achievable by using routine radiological or serologic testing, or both, but objective benefit from this approach has not been shown. Indeed, the only undisputed effect of intensive monitoring is that of diminished relapse free survival.3 It is therefore not surprising that many retrospective studies have cast doubts on the effectiveness of routine non-mammographic investigations or clinical follow up, or both, for breast cancer.4 5 6 7 Despite these doubts routine follow up continues to be the norm, with many physicians assuming that patients are reassured by frequent contact with their oncologist.8 This view is not universal, however, as evidenced by wide disparities in follow up practice on both sides of the Atlantic9 10 11; disparities that have been attributed to defensive perceptions by physicians unsure as to the efficacy of treatment and monitoring.12 13 For these reasons many authorities have lately acknowledged the need for a prospective randomised study to determine the benefits of the tradition of routine follow up.14 15 16 17

In this pilot study we compared the experience of a cohort who received conventional follow up with that of a similar cohort in whom review, in the absence of symptoms, was restricted to the time of mammography.

Patients_and_methods

Participants–All eligible patients seen at the Chelsea and Westminster combined breast clinic over a 24 month period were offered randomisation. Eligibility criteria were a history of breast cancer proved by biopsy; lack of known recurrence of cancer; current lack of symptoms consistent with recurrence; no active management apart from adjuvant tamoxifen; home telephone; fluency in English; and informed consent. Details of disease stage for patients who declined participation were recorded anonymously to provide an internal control for recruitment bias. Primary treatment and staging details were obtained from information in medical records, the Charing Cross CTS computer system, correspondence with patients, and archives. Median follow up for the present analysis was 16 months.

Follow up and mammography scheduling–Table 1 presents the schedule for visits. No restrictions were placed on attendance at other hospital or local clinics. Patients older than 70 years and for whom over five years had passed since diagnosis were not routinely scheduled for mammography but continued outpatient review according to the same schedule shown in table 1. All patients were advised to examine their breasts each month and to request an immediate appointment if they felt a lump or developed any other symptom.

Telephone access and questionnaires–All patients were instructed to telephone immediately if they developed symptoms or had other concerns; in these circumstances an appointment to be seen within five working days was made. Records were kept concerning the content of all telephone calls from patients in the study. Questionnaires based on the Medical Research Council quality of life questionnaire were used and included additional questions relating to frequency of follow up in hospital and general practice, satisfaction with clinic visits, and the time and expense entailed in clinic attendance; these were sent to patients annually, beginning six months after study entry.

Protocol violation and termination–Documentation of progression of disease or second malignancy automatically terminated participation in the study (10 patients) as did any decision after randomisation by any patient to pursue an alternative schedule of appointments or mammograms to that specified in the protocol (three patients).

Results

Patterns of patient recruitment–Age distribution, duration since diagnosis, and primary disease stage did not differ substantially between the cohorts (table 2). Fifteen eligible patients (7%) declined participation; this group was distinguished by a trend towards a higher stage of primary disease, younger age, and more recent diagnosis; other parameters were not appreciably different (table 3). We consider that the minor degree of selection bias suggested here does not confound the broad conclusions of this study.

Discussion

The present study is limited by its size and duration and does not justify publication of immature data concerning recurrence and quality of life. A study of survival outcomes would require several thousand patients and many years of follow up to establish the relative incidence of uncontrolled local recurrence, particularly late stage recurrence within the breast. The early phase of such a study would also need to focus on the possibility of more local or distant recurrences, or both, being detected in the “conventional” group, thus implying a shorter lead time (and, perhaps, more effective local treatment). Interim analysis of the present results indicates that no such trend has yet emerged; indeed, of the recurrences documented to date, all have been detected at interval (as opposed to scheduled) visits. We acknowledge, however, that these unpublished observations fall short of establishing the safety of less frequent follow up.

What does seem clear from the present study is that this cohort of patients with breast cancer was highly supportive of the option to pursue less frequent follow up and seemed willing to assume responsibility for seeking medical attention in the event of symptoms. These findings are consistent with those of other groups.22 Moreover, the data reveal surprisingly little hesitancy from subgroups one might have expected to be reluctant–that is, the young, the recently diagnosed, and those with more aggressive primary disease–notwithstanding that this profile was somewhat over-represented among the patients who did not participate.

The potential benefits of less intensive monitoring are clear. Clinicians are able to rationalise their practices and thus improve the efficiency of patient care; anxiety and inconvenience for patients without symptoms is minimised; outpatient clinic waiting time is reduced, increasing evaluation time for sick patients; nursing and reception staff are able to provide higher quality outpatient service by having fewer arrivals and room changes to deal with; the efficiency of general practitioner referral to hospital specialists improves; and healthcare purchasers reap the rewards of greater cost-effectiveness. Given this potentially attractive harvest, financial support for this pilot study was been surprisingly hard to find (table 6). None the less, we submit that these preliminary findings establish both the feasibility and desirability of a larger and more comprehensive national initiative to establish the value of routine follow up for cancer and other diseases.