Active tuberculosis disease has been classified as either primary or postprimary tuberculosis (TB)1, 2). Primary TB is common in the pediatric age group, and it is caused by initial Mycobacterium tuberculosis (M. tuberculosis) infection. The radiological characteristics are focal lung infiltration or homogeneous consolidation that usually affects the middle and lower lobes and lymphadenopathy is also present1). Postprimary TB is common in adults; this is mainly located in the apical and posterior segments of the upper lobes and it is characterized by cavitary lung lesion2).

While the radiographic features of childhood TB have apparently not changed, there has been an increased number of reports of atypical TB in adults3-10). One of these atypical radiological findings in adult is tuberculous pneumonia that resembles childhood primary TB. The radiographic characteristics of tuberculous pneumonia are homogeneous, segmental or lobar consolidation6, 7) and it occurs in any site of the lung6, 10).

There have been many reports concerning the pathogenesis of postprimary TB in adults. Stead et al. proposed that postprimary TB was usually caused by reactivation of dormant M. tuberculosis rather than by new exogenous reinfection11). Yet the recent reports that have employed mycobacterial genotyping techniques have suggested that exogenous reinfection was also a significant cause of postprimary TB in adults, and especially in an area with a high incidence of tuberculosis12-15). But it is still unknown if tuberculosis pneumonia in adults is caused by primary infection, endogenous reactivation or exogenous reinfection. Comparing the characteristics of tuberculous pneumonia with that of typical postprimary tuberculosis might be useful for understanding the pathogenesis of tuberculous pneumonia. Despite conducting a search of the related articles, we could not find any literature that compared tuberculous pneumonia with cavitary TB.

In this study we analyzed the clinical and radiological characteristics of tuberculous pneumonia and cavitary TB, and we tried to determine the factors associated with the two characteristic radiological patterns.

Primary TB is caused by an airborne infection of M. tuberculosis and its location reflects the pulmonary airflow. This can occur in any site of the lung, yet it is more frequently in the mid or lower lung field due to these areas greater ventilation. From the primary focus, tuberculous bacilli are spread via the lymphatics or blood stream. Dormant states mainly occur in such areas such as the apicoposterial segment of the upper lobe or the apical segment of the lower lobe, where lymph production and drainage are deficient and high oxygen tension is present2). Stead et al. proposed that postprimary TB could occur at these sites by reactivation of dormant disease states11). Lung cavitation is the hallmark of postprimary TB and this appears in about half of the patients17).

Tuberculous pneumonia, similar to childhood primary tuberculosis18), is the unusual radiographic finding in adults and this occurs in any site of the lung6, 10) Tuberculosis occurring in the lower lung fields in adult is a frequent characteristics of tuberculous pneumonia19), which is homogeneous segmental or lobar consolidation6, 7). Fever was more common and the duration of symptom until the hospital visit was shorter for tuberculous pneumonia than for cavitary TB. Because tuberculous pneumonia may frequently be indistinguishable from bacterial pneumonia, the diagnosis at a hospital is generally more delayed6, 10). The white blood cell count in the peripheral blood is frequently normal in both groups, and this is one of the clues to distinguish tuberculosis pneumonia from bacterial pneumonia19).

The patients with tuberculous pneumonia in our study had a tendency to have a lower yield for a positive sputum AFB smear and culture for M. tuberculosis than did the patients with cavitary TB, and bronchocopic examination is useful for these cases to make the diagnosis of tuberculosis20). In our study the positive culture rate for M. tuberculosis in a specimen was lower than that of other reports3, 6, 10). Three cases with positive AFB smears didnt have culture for M. tuberculosis ordered. In our hospital, culture for M. tuberculosis is done at another laboratory, so the result of culture might be influenced by many factors, including storage, transport and processing.

Many authors have reported that lower lung field TB and tuberculous pneumonia more commonly occur in specific groups of patients such as those with diabetes mellitus (DM), cancer or infection with human immunodeficiency virus (HIV)3, 6, 19, 21). But in our study, the prevalence of comorbidity such as cancer, DM, alcoholism and long-term steroid use was not different between the two groups. The prevalence of DM, cancer and long-term steroid use has been variable in the reports on tuberculous pneumonia6, 10) and these studies did not compare tuberculous pneumonia with cavitary TB3, 6, 10, 19, 21).

In our study the patients with tuberculous pneumonia were older and had lower levels of serum albumin and hemoglobin than did those patients with cavitary TB. In many reports, lung cavitation was less frequent in the elderly patients22-24) and lower lung field TB and tuberculous pneumonia are more common in older patients than in the younger patients22, 25). Because elderly people have more impaired T-lymphocyte function than younger patients26), the decrease in their immunologic status associated with aging might be related to the development of tuberculous pneumonia. Also, the findings of lower serum albumin and hemoglobin levels in patients with tuberculous pneumonia might reflect malnutrition or more severe catabolic states in older patients than in those patients with cavitary TB.

Recent reports concerned with the use of mycobacterial genotyping techniques have suggested that exogenous reinfection appears to be a significant cause of postprimary TB in adults, and the incidence of exogenous reinfection is variable according to the patients age and the incidence of TB in a country12-15). The incidence of TB in our country was 64.5 cases per 100.000 persons in 2004 (http://tbnet.go.kr) and exogenous reinfection might be a substantial cause of postprimary TB in Korean adults. In our study, the main lesion in tuberculous pneumonia patients occurred in any site of the lung, which is similar to other studies6, 10). This finding suggests that tuberculous pneumonia might be caused by exogenous reinfection via airflow implantation of M tuberculosis rather than by reactivation of dormant bacilli. Further, the elderly might be more susceptible to exogenous reinfection due to impaired host immunity.

In addition to exogenous reinfection, endobronchial stenosis in adults might be involved in the pathogenesis of tuberculosis pneumonia, as compared to the enlarged lymph nodes in the pathogenesis of childhood TB. Lymphadenopathy in tuberculous pneumonia patients was uncommon in our study and also in other studies7, 10) and it was different from that of childhood primary TB, in which lymphadenopathy was observed in almost all such patients27). Goussard et al. speculated that the pathogenesis of tuberculous pneumonia in children was enlarged tuberculous lymphadenitis that ruptured into the bronchus; the caseous material was aspirated into the affected lobe and the subsequent exudative hypersensitivity reaction to the aspirated tuberculoprotein caused tuberculous pneumonia28). Tuberculous pneumonia, including that in the lower lung in adults, is frequently associated with endobronchial lesions7, 10) and it is more common in female patients25, 29). In our study, seven non-responders to antibiotic therapy and they had negative sputum AFB smears were examined by bronchofibroscopy, and endobronchial lesion was discovered in six of them.

Therefore we can speculate that tuberculous pneumonia might develop by reinfection with aging, and implantation of M. tuberculosis occurred via the airflow and this caused a random distribution of lesion. In addition to exogenous reinfection, endobronchial narrowing hinders the drainage of infected materials and so contributes to the development of consolidative hypersensitivity reaction.

Our retrospective study had the following limitations. Although there were some limitations in interpreting the tuberculin skin test in our country because of BCG vaccination30), the enrolled patients in our study were not examined by tuberculin skin test and the scar from the BCG vaccination was not recorded. However, the prevalence of a previous TB scar on radiographs, which is visible in one third of healed primary tuberculosis patients2), was not different between the two groups. Second, the exact prevalence of combined endobronchial lesion could not be estimated because bronchofibroscopy was not performed in all of the patients of both groups.

A large scale prospective study using DNA genotyping methods and bronchofibroscopic examination is needed to elucidate the pathogenesis of tuberculosis pneumonia and cavitary TB.