Tonsillitis, tonsillectomy and Hodgkin’s lymphoma Hanne Vestergaard, Tine Westergaard, Jan Wohlfahrt, Henrik Hjalgrim and Mads Melbye Department of Epidemiology Research, Statens Serum Institut, Copenhagen, Denmark It has been debated whether the reported increased risk of Hodgkin’s lymphoma (HL) after tonsillectomy could be due to some underlying factor rather than the surgery itself. We studied whether not only tonsillectomy but also tonsillitis was associated with HL. This nationwide cohort study included all Danish residents during 1977–2001. Information on a diagnosis of tonsillitis, tonsillectomy, or HL was obtained from national registries. During 124 million person-years we observed 2,988 HL patients of whom 58 were tonsillectomized (most with preceding tonsillitis) and 14 were diagnosed with only tonsillitis at more than 1 year before HL diagnosis. Tonsillectomy was associated with a significantly increased HL risk in persons under 15 years of age as follows: 1–4 years after tonsillectomy, relative risk (RR) 5 3.9 [95% CI: 1.4–11; n 5 4]; >5 years after tonsillectomy, RR 5 3.5 [1.4–8.5; n 5 5]. No young cases of HL occurred among persons diagnosed with only tonsillitis. In contrast, 1–4 years after a hospital diagnosis of tonsillitis without subsequent tonsillectomy we found an increased HL risk in persons aged 15 years or above as follows: 15–34 years of age at HL diagnosis, RR 5 3.5 [1.6–7.7; n 5 6]; 351 years, RR 5 5.9 [2.2–16; n 5 4]. Age at tonsillitis or tonsillectomy did not modify HL risk within the 3 age strata. An increased HL risk was found both after tonsillectomy and after an isolated diagnosis of tonsillitis. These results suggest that tonsillitis is a risk factor for HL and not that, as previously reported, only the surgical removal of tonsils is a risk factor. A prevailing hypothesis for an association between tonsillec- tomy and Hodgkin’s lymphoma (HL) has been that removal of the ‘‘lymphoid tissue barrier’’ would predispose to the de- velopment of HL. 1 However, so far, epidemiological studies have been conflicting, 1,2 as case–control studies have reported either an increased risk of HL following tonsillectomy 1,3–6 or no association. 2,7–18 One nation-wide cohort study found a marginally increased risk for all age groups, but a 4-fold increased HL risk was found for persons tonsillectomized before the age of 12 years. 19 The latter finding is of particular interest because the tonsils’ immunological functions are believed to decline from childhood to adulthood. 20 However, few studies on tonsillectomy and HL have included age at tonsillectomy, 6,7,18,19,21 leading to conflicting results. In many cases, tonsillectomy can be regarded as an indi- cator of severe tonsillitis. Accordingly, in a population-based study we found that 81% of patients aged less than 20 years who had been tonsillectomized at a hospital had also been admitted for chronic tonsil disease within the previous year. 22 However, investigations studying the potential associa- tion between tonsillitis and HL are sparse. 2,8,13 As tonsillitis most often precedes tonsillectomy, it is of particular interest to study their potential effect on HL risk concomitantly as it might be that tonsillitis rather than removal of the tonsils is associated with HL. In a national-wide cohort study we analyzed whether ton- sillectomy or hospital-diagnosed tonsillitis was associated with an increased risk of HL. In addition, we also studied whether age at first registration of tonsillectomy or tonsillitis influenced HL risk. Material and Methods Study population We identified all Danish residents alive anytime between 1977 and 2001 in the Danish Civil Registration System (CRS) as the study population. The CRS was established on April 1, 1968, and since that date all persons living in Denmark have been assigned a personal identification number. The CRS number allows identity secure linkage between registers, as data in Danish registers were recorded by this number. Infor- mation in the CRS is regularly updated; data concerning per- sonal identification number, date of birth, sex, emigration and vital status were extracted from the CRS. Patients with tonsillectomy were identified in the Danish National Patient Registry (NPR) and the Danish Health Secu- rity System, whereas patients with tonsillitis were only ascer- tained from the NPR. The NPR includes registrations of the patients’ personal identification number, dates of hospital admission and discharge, discharge diagnoses and surgical operation codes registered according to the International Classification of Diseases (ICD-8 and ICD-10). Reporting of discharge diagnosis to the NPR has been mandatory since the start of the NPR on January 1, 1977. Since January 1, 1995, outpatients tonsillectomized in hospital-based settings Key words: epidemiology, Hodgkin’s lymphoma, risk factor, tonsillitis, tonsillectomy DOI: 10.1002/ijc.24973 History: Received 14 Aug 2009; Accepted 1 Oct 2009; Online 16 Oct 2009 Correspondence to: Mads Melbye, Department of Epidemiology Research, Statens Serum Institut, Artillerivej 5, DK-2300 Copenhagen, Denmark, Tel: þ45-32-68-31-63, Fax: þ45-32-68-31-65, E-mail: mme@ssi.dk Epidemiology Int. J. Cancer: 127, 633–637 (2010) V C 2009 UICC International Journal of Cancer IJC