Elsevier

European Journal of Cancer

Volume 36, Issue 3, February 2000, Pages 384-389
European Journal of Cancer

Long-term survival in Hodgkin's disease patients: a comparison of relative survival in patients in trials and those recorded in population-based cancer registries

https://doi.org/10.1016/S0959-8049(99)00267-1Get rights and content

Abstract

The prognosis of Hodgkin's disease (HD) has improved during the last 30 years. This study was planned to analyse long-term survival of HD patients and to compare survival rates estimated from clinical trials and population-based data. Individual data were analysed on 2755 adult HD patients entering randomised clinical trials of the British National Lymphoma Investigation (BNLI) between 1970 and 1987, and 5064 patients with HD incident 1978–1984 recorded in the UK population-based cancer registries participating in the EUROCARE study. Relative survival of Hodgkin's disease patients allowing for mortality expected from general population rates was analysed by a proportional hazards regression model including covariates. Although relative mortality decreased with longer follow-up, it was still significantly positive at 9–10 years after diagnosis in both the clinical trials and the population-based data sets. Relative mortality was worse for late stage than for early stage patients even at 10–15 years after first treatment (BNLI data). Whereas 10-year relative survival was identical in trials and population-based patients at ages under 45 years (>69%), it was much higher in BNLI older patients than in the population-based patients. In the older age group (65–74 years) the BNLI patients had 39% relative survival whilst for the population-based patients it was only 27%. Generalisation of clinical trials results to the general population must be done with caution, especially for older patients.

Introduction

The prognosis of Hodgkin's disease (HD) has greatly improved during the last 30 years, and as a consequence it has become important to determine the extent of long-term survival, and whether after several years survival, mortality approximates that of the general population. Crude survival analysis 1, 2, as frequently used in the analysis of trials, gives only partial information because, especially as the patients grow older, some deaths would be expected if the patients simply suffered general population mortality rates (i.e. if they ceased to suffer adverse mortality as a consequence of their past Hodgkin's disease). Adjustment for competing causes of death is then necessary if one wants to analyse the late effects of treatment on survival.

Survival is most often estimated from clinical series, but such series are selected to meet entry criteria for the trials and also may be unrepresentative of general population survival because they may receive better care within the trials. At a public health level, it would be interesting to know how survival rates obtained from clinical trial series compare with survival rates from population-based data, and hence the extent to which trial results may imply prognosis for patients generally. Although studies comparing the prognosis of patients included in clinical trials with those not included have been published for several sites [3], there appear to have been no studies for Hodgkin's disease.

To address these questions, we analysed survival in patients participating in the randomised clinical trials of the British National Lymphoma Investigation (BNLI) 4, 5 between 1970 and 1987, and compared their long-term relative survival 6, 7 with that in population-based UK data [8].

Section snippets

BNLI data

The BNLI is a large British clinical collaborative group which since 1970 has conducted clinical trials of treatment of lymphoma. Approximately two-thirds of patients at the collaborating centres are entered into the trials, a greater proportion at younger than at older adult ages. The present analyses were conducted on the records of 2755 HD patients aged 15–79 years, all of the subjects included in BNLI trials between 1970 and 1987. Demographic, clinical, and biological variables were

Distribution of cases by age

Fig. 1 shows the distribution by age of the BNLI cases (median=32 years; 15–79) and population-based cases (median=41 years; 15–79). The median values of age in the four age groups (15–24; 25–44; 45–64; 65–79) were respectively 20, 32, 54 and 69 years for BNLI patients and 20, 33, 56 and 71 years for population-based cases.

Long-term survival and net annual mortality

To follow current practice in the analysis of Hodgkin's disease data [15], results were presented distinguishing BNLI early stages (CS I and CS II) and advanced stages (CS

Discussion

An excess of mortality compared with that in the general population was still observed up to 15 years after diagnosis for advanced stage HD, and a smaller excess was also observed for early stage HD. HD incidence is higher in populations with elevated social status, who have a higher life expectancy than the general population. The use of general population mortality tables for comparison will, therefore, have tended to lead to overestimation of the relative survival rates of the patients. A

Acknowledgements

The authors acknowledge a referee for helpful comment for comparison of the two datasets.

References (23)

  • F. Berrino et al.

    Survival of cancer patients in Europe. The Eurocare study

    (1995)
  • Cited by (59)

    • Impact of Clinical Trial Participation on Survival of Patients with Metastatic Non-Small Cell Lung Cancer

      2021, Clinical Lung Cancer
      Citation Excerpt :

      It is unclear whether access to experimental agents or contextual changes in care translate into improved survival among trial participants as compared to nonparticipants receiving usual care. The association between clinical trial participation and survival has been studied in a variety of tumor types7–11, including small cell lung cancer12–15 and non-small cell lung cancer (NSCLC)16–18. There has not been a clear consensus, with results split between either no difference in survival or improved survival associated with trial participation7.

    • Ten-year survival outcomes for patients with early stage classical Hodgkin lymphoma: An analysis from Kentucky Cancer Registry

      2020, Hematology/ Oncology and Stem Cell Therapy
      Citation Excerpt :

      Importantly, the advantage of our cohort study is that it reflects a real-life experience outside a clinical trial setting. Studies have shown discrepancies in outcome in patients treated for cHL in a clinical trial [17], or even in a comprehensive cancer center [18] compared with the community. Our data provide a longer follow-up for our patients than results reported in the H10 trial, which could have led to an increased incidence of secondary malignancy, leading to a lack of earlier benefits seen with the combined approach in the unfavorable group [19].

    • A retrospective multicenter analysis of elderly Hodgkin lymphoma: Outcomes and prognostic factors in the modern era

      2012, Blood
      Citation Excerpt :

      Survival rates for elderly Hodgkin lymphoma (eHL), typically defined as more than equal to 60 years of age, are disproportionately inferior compared with younger patients.1 Studies in the 1970s to the 1990s established 5-year progression-free survival (PFS) or freedom from treatment failure rates for advanced-stage eHL of 30% to 45% with 5-year overall survival (OS) rates of 40% to 55%.2-8 There are limited data examining eHL outcomes in the modern era.

    • Endocrine Management of the Cancer Survivor

      2011, Williams Textbook of Endocrinology, Twelfth Edition
    View all citing articles on Scopus
    View full text