Abstract
Tuberculosis is an ancient human scourge that continues to be an important public health problem worldwide. As impaired immunity is closely associated with the pathogenesis of TB, an increase in the incidence of the disease among patients with cancer is anticipated. A total of 20 cases of haematological cancers co-existing with TB in a tertiary hospital in an African setting between January 2005 and September 2008 were analysed. Clinico-pathologic profiles, types of TB infection and outcome were noted. The incidence of TB in patients with haematological malignancies was 3.4%. Lymphoma and leukaemia were the most common malignancies associated with TB. With the high incidence of such cases we recommend the exclusion of an active disease prior to the initiation of cancer therapy. Chemoprophylaxis might be indicated in TB endemic areas.
Introduction
TB has remained a major public health problem worldwide with several reports documenting an increasing incidence in cancer patients. 1,2 It has been recognized that cancers and cytotoxic chemotherapy increase the risk for TB. This has been observed especially during treatment for leukaemias, Hodgkin's lymphoma and other malignant diseases. 3,4 The increasing incidence of TB has been attributed to improved cancer survival which may have increased the number of people at risk including immunocompromised patients. However, there is a dearth of studies examining the incidence and clinical characteristics of TB in adults with haematological cancers in Nigeria. The aim of this study is to evaluate the incidence and clinico-pathologic profile of TB in adult patients with haematological cancer.
Patients and methods
All case files of haematological cancer patients managed at the University of Benin Teaching Hospital, Benin City, Nigeria, between January 2005 and September 2008 were studied. Haematological cancer coexisting with TB infection formed the study group. Only cases involving patients with documented outcomes were included. Patients with any risk factor for the reactivation of TB, except corticosteroids included in anticancer regimens, were excluded. The pretherapy, clinico-pathological and demographic features, including the type of associated TB infection, were analysed.
The haematological malignancies were diagnosed based on typical clinical features, standard cytomorphological and histological findings. A diagnosis of TB was based on bacteriological, pathological and clinical findings. Chest X-rays and abdominal ultrasound were occasionally used. The overall survival was calculated with reference to the date of diagnosis and the date of death/last documented clinic visit. Data analysis was done using the Instat statistical package.
Results
Twenty consecutive patients with haematological cancers associated with TB were studied. This constituted 3.4% (20/581) of the total haematological cancers seen. There were 13 men (65.0%) with a mean age of 40.6 ± 15.9 and seven women (35.0%) with a mean age of 50.3 ± 10.8. The male-to-female ratio was 1.9:1. The overall median age was 50 years (range, 18–60 years) with a mean age of 44.0 ± 14.8.
Lymphomas and leukaemias were the most common haematological cancers coexisting with TB infection (Table 1). The baseline characteristics of patients shown in Table 2 revealed that the performance status as assessed by the Eastern Cooperative Oncology Group showed that all the patients (100%) were within the worst scale. 2–4 Cervical and inguinal lymph nodes were the most common sites involved.
Types of haematological malignancies occurring in association with tuberculosis
Baseline characteristics of haematological cancer patients associated with tuberculosis
A TB diagnosis was made on positive findings of acid-alcohol-fast bacilli (AAFB) in sputum smears in 14 (70%) patients, lymph node biopsy and sputum analysis in 10 (50%) patients and a demonstration of AAFB in pleural aspirate in two (10%) patients. The type of TB infection associated with the cancers were pulmonary TB (n = 15), miliary TB (n = 4) and spinal TB (n = 1). Of the 15 patients with pulmonary TB: six were diagnosed as having TB before a diagnosis of the cancer; four were concomitantly diagnosed as TB with osteosclerotic myeloma and Richter's syndrome. Importantly, 40% of TB infection was diagnosed during the initial admission necessitated by haematological problems, while the remaining cases were diagnosed within the first eight to 10 months of initiation of chemotherapy (50%) or cessation of therapy (10%), suggesting the reactivation of latent mycobacterium during a period of impaired immunity.
The patients were treated with standard chemotherapy and six months of triple therapy for TB. For relapses a different combination therapy was used. The median duration of follow-up after completion of therapy was 15 months. The interval between the completion of a specific cancer therapy and the development of a TB infection was 10 months. Five patients had complete remission and two had partial remission. Seven patients died in less than one year and six were lost to follow-up.
Discussion
An increasing number of hospital case pulmonary TB in Nigeria in recent years has been reported.
5
In patients with underlying malignancies associated with TB, three main issues are of concern: A diagnostic dilemma between TB versus malignancy versus the coexistence of both; An increase in the predisposition for TB due to impaired immunity (underlying malignancy and cytotoxic therapy); Atypical presentation and/or behaviour of TB resulting in a delayed diagnosis and poor outcome with standard anti-TB treatment.
3
The incidence of 3.4% in our studied patients is higher than the incidence range of 0.72% to 2.6% given on available data concerning TB patients undergoing anticancer chemotherapy.
6
Factors attributed to the high incidence are: poverty and high prevalence of TB in the country; impairment of host defenses; poor nutritional and debility. Our study demonstrated a reactivation of TB in some patients with a past history of TB. Patients had abnormal radiographs at diagnosis of the cancers such as granuloma, healed scar with higher bacillary load and spread of infection from the site of healed scar. This is in contrast to the findings of Nair et al.
3
where no reactivation of TB in non-Hodgkin's lymphoma patients was reported.
The features of atypical extrapulmonary/disseminated diseases is similar to the study conducted in allogeneic stem-cell transplant recipients of TB with malignancies. 7 However, Kim et al. 8 suggested that malignancy has no effect on either the presentation or the response to anti-TB treatment. The poor outcome in this study was due to poverty, ignorance, illiteracy, cultural setting, self medication which promotes late presentation and the non-implementation of adequate treatment among other strong limiting factors.
The coexistence of TB with leukaemia has been reported. 9 Acute myeloid leukaemia patients were more likely to develop TB compared to acute lymphoblastic leukaemia (ALL) patients, despite the wider use of steroids and radiotherapy. 9 In this study, a firm statement as to which type of leukaemia is more prone to TB infection could not be made due to the small number of cases involved. However, TB is not a commonly described infection in acute leukaemia and a high index of suspicion is needed, especially in areas endemic for TB.
In conclusion, our study found high incidence of TB infection of 3.4% in haematological cancer patients in an endemic area. Lymphomas and leukaemias were the most common malignancies associated with TB. We suggest that efforts be made to exclude any active disease prior to the initiation of cancer therapy, particularly in TB-endemic areas.
