Late complications affecting Hodgkin lymphoma (HL) survivors are well described in paediatric and adult-based publications. This study determined the late morbidity and mortality risk for 442 teenage and young adult (TYAs) 5-year HL survivors, diagnosed at 15-24 years of age between 1970 and 1999, identified from the British Columbia Cancer Registry. Treatment details were abstracted from charts. Survivors and a matched comparison cohort were linked to provincial administrative health datasets until December 2006 and regression analysis was performed, providing risk ratios regarding mortality, secondary malignancy and morbidity causing hospitalisation. Sixty (13·6%) survivors experienced late mortality with excess deaths from secondary cancer [standardised mortality ratio (SMR) 18·6; 95% confidence interval (CI) 11-29·4] and non-malignant disease (SMR 3·6; 95% CI 2·2-5·5). Excess secondary cancers (standardised incidence ratio 7·8; 95% CI 5·6-10·5) were associated with radiotherapy [Hazard ratio (HR) 2·7; 95% CI 1-7·7] and female gender (HR 1·8; 95% CI 1-3·4). Of 281 survivors treated between 1981 and 1999, 143 (51%) had morbidity resulting in hospitalisation (relative risk 1·45; 95% CI 1·22-1·73). Hospitalisation significantly increased with combined modality therapy, chemotherapy alone and recent treatment era. TYA HL survivors have excess risk of mortality and secondary malignancy continuing 30 years from diagnosis. Radiotherapy is associated with secondary malignancy and current response-adapted protocols attempt to minimise exposure, but late morbidity causing hospitalisation remains significant.

The role of the microenvironment in high-grade lymphoma is not well defined. In this report, we employ immunohistochemistry to characterise programmed death-1 (PD-1/CD279) and FoxP3 expression in 70 cases of diffuse large B-cell lymphoma (DLBCL). PD-1 is a surface marker characteristic of follicular helper T-cells whilst FoxP3 is characteristic of Tregs. We demonstrate variable infiltration with CD4(+) T-cells (<10 to >50 % of all lymph node cells) and PD-1(hi) cells (0.1 to 1.5 % of all cells). CD4(+) T-cells can be distributed in clusters or more diffusely and PD-1(hi) cells, but not FoxP3(+) cells, are found in rosettes around lymphoma cells. Cases with high CD4(+) T-cell numbers tended to have higher numbers of both PD-1(hi) and FoxP3(+) cells. Cases with total CD4(+) T-cell, PD-1(hi) and FoxP3(+) numbers above the median associate with better clinical outcome. Overall, we demonstrate that infiltration by CD4(+) T-cells, including both FoxP3(+) and PD-1(hi) subsets, correlates with prognosis in DLBCL. In distinction to previous reported series, patients (91 %) were treated with rituximab-containing regimens, suggesting that the effects of CD4+ T-cell infiltration are maintained in the rituximab era. This work suggests that determinants of total CD4(+) T-cell infiltration, either molecular characteristics of the lymphoma or the patients' immune system, and not individual T-cell subsets, correlate with clinical outcome.