Cancer immunosurveillance is a major function of the immune system.1 Strong evidence for the role of the immune system in antitumour surveillance mechanisms comes from the observation of the increased risk of malignancies among immunocompromised patients, including patients with primary immunodeficiencies, a heterogeneous group of diseases caused by >300 gene defects affecting natural and acquired immunity. The knowledge of defective genes and pathways2 offers the opportunity to dissect the role of cancer immunosurveillance. An increased prevalence of cancer has been observed in patients affected by common variable immunodeficiency (CVID), which is the most commonly diagnosed primary antibody defect.3,4 In particular, the risk for malignant lymphomas among patients with CVID was found to be 259-times higher in a USA cohort5 and 30-times higher in a British cohort,6 while the risk for gastric cancer was 47-times higher than expected in the British study.5 In the last decade, thanks to immunoglobulin G replacement therapy, CVID patient life expectancy increased due to improvements in surveillance, prevention, and treatment of recurrent and severe infections; however, the cancer mortality rates of these patients have not changed.
In a large CVID Italian cohort of 462 patients followed-up in three primary immunodeficiency care centres in Italy, we assessed cancer prevalence over a 30-year period. We aimed to estimate the prevalence and mortality rate due to haematological and gastrointestinal malignancies, as well as other cancers. Data on cancers in CVID patients were compared to normative data provided by the Italian Registry for Malignancies (AIRTUM).7
We collected data for a cumulative period of 5,326 years across all patients (mean ± standard deviation: 11.7±9.0 years). The prevalence of malignancies was 26.0%. For CVID patients, the risk of developing cancer was 50.0% at 65 years of age, whereas for the general population it was between 33.3% (females) and 50.0% (males) at 85 years of age. In the general population of Italy, breast and prostate cancer were the most frequent cancers diagnosed in females and males, respectively, followed by colorectal cancer. By contrast, in CVID patients, lymphoproliferative malignancies were the most commonly diagnosed cancers in both sexes (10%), followed by gastric cancer (6%). This emphasises the need for specific cancer screening programmes in CVID. The most common haematological malignancies were non-Hodgkin’s B cell lymphomas, often involving extra nodal sites; however, T cell lymphomas were also recorded. While cardiovascular diseases were the primary cause of death in the Italian general population, malignancies were the primary cause of death in CVID patients, accounting for 58% of deaths, followed by infections (23%), chronic lung disease complications (13%), and autoimmunity (7%). The overall survival for patients affected by haematological cancers, gastric cancer, and other malignancies was 67%, 54%, and 88% at 1 year, respectively; 61%, 36%, and 80% at 2 years, respectively; and 61%, 27%, and 29% at 20 years, respectively. The treatment of CVID-associated cancers was similar to the treatment of cancers in other clinical settings. We observed a low rate of infection during chemotherapy and a high incidence of severe malabsorption in patients who underwent gastrectomy.
In conclusion, in this Italian cohort, cancer was the main cause of death in CVID patients. Despite immunodeficiency, patients with CVID with cancer might receive a full therapy regimen due to a risk of infection similar to that observed in the non-CVID cancer population. Cancer prevention strategies should be improved to ameliorate survival.