The studies included as part of this review included descriptive studies, observational studies, case reports and a meta-analysis (Fig. 3a). The cases identified were grouped into four categories namely those where ABPA was 1. previously misdiagnosed as TB, 2. ABPA in patients with previous history of TB with or without PTLD, 3. concomitant ABPA-TB and 4. allergic sensitization to Aspergillus in patients with tuberculosis (vide Fig. 3b). A brief summary of all the studies and case reports included in this systematic review is provided in Supplementary appendix 1.
Overall, the most common pattern for the 397 patients included in the final analysis was ABPA misdiagnosed as TB (51%) followed by allergic sensitization to Aspergillus in patients with past or current tuberculosis (23%) and ABPA in previous tuberculosis (22%). Concomitant ABPA and TB was reported in 4% of cases (Fig. 3b). A detailed analysis of radiological features of patients with one of the patterns ABPA-tuberculosis alone was difficult, since most of the studies which evaluated the association have reported radiological features for the overall population screened. Of a total of 43 patients in groups 2,3,4 for whom radiological characteristics were reported, 12 (28%) had central bronchiectasis, mucoid-impaction/high-attenuation mucus/glove and finger shadows were seen in 12 (28%), consolidation/mass/nodule was seen in seven (16%), fungal ball in seven (16%) and cavitary lesions in 16 (37%). Of the specific subset of patients who developed ABPA after previous tuberculosis or with concomitant tuberculosis, treatment response was reported only for 12, ~ 67% of whom received steroids and antifungals. All 12 (100%) showed adequate clinical response with either steroids alone or with antifungals. As with radiological features, both treatment options and response rates were mostly reported only for the overall population included in the studies and the criteria for response used were variable.
ABPA misdiagnosed as tuberculosis
A total of 20 original articles (which dealt with the different manifestations of ABPA, including eight original studies, a meta-analysis which included 11 studies (18) and 11 case reports) reported the prevalence of ABPA in patients misdiagnosed previously as TB. The overall compiled data from these studies is shown in Table 1. As shown, out of a total of 625 patients with ABPA reported in these studies, 203 (32.5%) were previously misdiagnosed and treated as TB. While the criteria used for diagnosis of ABPA varied across the different studies, most of the cases of TB were presumptively treated likely on clinic-radiological basis and a definitive diagnostic criteria for diagnosis of TB is not mentioned for most of the included studies. Of the 19 studies and 11 case reports which studied this association, seven studies used ISHAM criteria for diagnosis of ABPA, Greenberger and Rosenberg-Patterson criteria were used in five and four studies respectively and three studies used either a clinical diagnosis of ABPA or did not report the exact criteria used for diagnosis. Only three case reports specified the criteria used for diagnosis of ABPA (two used Rosenberg Patterson while one used ISHAM criteria). None of these studies mentioned the criteria used for diagnosis of tuberculosis, which in most cases were likely to be clinic-radiological.
Table 1
ABPA misdiagnosed as tuberculosis
Total number of original articles = 9* (19 studies) |
Case reports = 11 |
Total number of patients with ABPA = 625 |
Number misdiagnosed as tuberculosis (%) = 203/625 ( 32.5%) |
| Observational studies (614 patients) | Case reports (11 patients) |
1. Criteria for ABPA diagnosis used in included studies | (n = 19 studies **) | (n = 11 case reports) |
ISHAM: | 7 | 1 |
Rosenberg-Patterson: | 4 | 2 |
Greenberger | 5 | 0 |
NA/Clinical/Others: | 3 | 8 |
2. Criteria for TB diagnosis in included studies | (n = 19 studies **) | (n = 11 case reports) |
Clinico-radiological | 0 | 4 |
AFB positive | 0 | 0 |
NAAT | 0 | 1 |
NA | 19 | 6 |
*1 was a meta-analysis which included 11 studies |
** Includes the studies reported in the meta-analysis |
ABPA in patients with previous tuberculosis
A total of six original articles and ten case reports reporting ABPA in patients with previous TB were retrieved. Out of the total of 175 patients with ABPA, 88 (50.3%) had previous history of TB as shown in Table 2. 36.1% of patients with previous TB subsequently developed ABPA. Out of the 16 articles, a single article used the ISHAM criteria, two articles each used Greenberger and Rosenberg Patterson criteria for the diagnosis of ABPA and eleven articles (two studies and nine case reports) did not report the exact criteria used for diagnosis of ABPA. Only two articles (both case reports) reported the exact criteria used for diagnosis of tuberculosis in these patients as shown in Table 2. Radiological findings were reported individually for 25 patients with previous TB who developed ABPA, out of which ten (40%) had central bronchiectasis, while cavitary lesions were present in 16 (64%), mucoid-impaction/HAM/Glove and finger shadows in nine (36%), consolidation/mass/nodule in five (20%), and fungal ball in six (24%). Six out of the total ten patients for whom treatment was reported had received steroids, while seven received antifungals. Clinical outcomes were good for all these patients as shown in Table 2.
Table 2
ABPA diagnosed in patients with previous tuberculosis
Total number of observational studies = 6 |
Case reports = 10 |
Total number of patients with ABPA = 175 |
Total number of patients with previous TB = 244 |
Patients with ABPA and previous tuberculosis = 88 |
Previous tuberculosis in patients with ABPA (%) = 50.3% (88/175) ABPA in patients with previous tuberculosis = 36.1% (88/244) |
1. Criteria for ABPA diagnosis | Observational studies (n = 6 studies) | Case reports (n = 10 case reports) |
ISHAM: | 1 | 0 |
Rosenberg-Patterson: | 1 | 1 |
Greenberger | 2 | 0 |
NA/Clinical : | 2 | 9 |
2. Criteria for TB diagnosis | Observational studies (n = 6 studies) | Case reports (n = 10 case reports) |
Clinico-radiological | 0 | 1 |
AFB positive | 0 | 1 |
NAAT | 0 | 0 |
NA | 6 | 8 |
3. Radiological features | Observational studies (n = 15 patients) | Case reports (n = 10 patients) |
Central bronchiectasis | 4 | 6 |
Cavitatory lesion | 11 | 5 |
Mucoid-impaction/HAM/Glove and finger shadows | 3 | 6 |
Fungal ball | 6 | - |
Fleeting opacities/Patchy infiltrates | | 1 |
Consolidation/Mass/Nodule | 0 | 5 |
4. Treatment | | N = 10 patients |
| Treatment specific to ABPA-Previous TB subgroup not reported | Systemic steroids + Antifungal = 5 Only antifungal = 2 Only steroid = 1 Omalizumab = 1 Thoracotomy = 1 |
5. Response | | N = 10 patients |
| Outcomes specific to ABPA-Previous TB subgroup not reported | All showed clinical improvement |
Co-existent ABPA-TB
A total of six articles (three studies and three case reports) were retrieved which described co-existence of ABPA and active tuberculosis. Out of a total of 40 ABPA patients who were screened, 15 had concomitant ABPA and active tuberculosis as shown in Table 3. Of the six articles, ABPA was diagnosed using Greenberger, Rosenberg-Patterson and ISHAM criteria in one each while three either did not report the criteria used for diagnosis or used a clinical criteria for diagnosis of ABPA; tuberculosis was diagnosed microbiologically in three articles (all case reports) while the criteria was not mentioned in three articles. The radiographic features were only reported for three patients while treatment options used and outcomes were only reported for two patients as shown in Table 3.
Table 3
Total number of observational studies = 3 |
Case reports = 3 |
Total number of patients with ABPA = 40 |
Number with co-existent tuberculosis (%) = 15/40 (37.5%) |
| Observational studies (n = 3 studies) | Case reports (n = 3 case reports) |
1. Criteria for ABPA diagnosis | | |
ISHAM: | 0 | 1 |
Rosenberg-Patterson: | 1 | 0 |
Greenberger | 1 | 0 |
NA/Clinical : | 1 | 2 |
2. Criteria for TB diagnosis | Observational studies (n = 3 studies) | Case reports (n = 3 case reports) |
Clinico-radiological | 0 | 0 |
AFB positive | 0 | 2 |
NAAT | 0 | 1 |
NA | 3 | 0 |
3. Radiological features | Observational studies (n = 37 patients) | Case reports (n = 3 patients) |
Central bronchiectasis | Not reported | 1 |
Mucoid-impaction/HAM/Glove and finger shadows | Not reported | 1 |
Pleuropulmonary fibrosis | Not reported | 0 |
Fleeting opacities/Patchy infiltrates | Not reported | 0 |
Consolidation/Mass/Nodule | Not reported | 2 |
Miliary pattern with centrilobular nodules | Not reported | 0 |
4. Treatment | Observational studies (n = 37 patients) | Case reports (n = 3 patients)* |
Not reported | ATT + Steroids + Antifungal = 1 ATT + Steroids = 1 Not reported = 1 |
5. Response | Observational studies (n = 37 patients) | Case reports (n = 3 patients)* |
Not reported | Both showed clinical improvement |
*Reported only for two patients |
Aspergillus sensitization in tuberculosis patients (concurrent or post-tubercular lung disease)
A total of four articles (including one which also reported the incidence of ABPA in patients with previous tuberculosis) evaluated the prevalence of Aspergillus sensitization in patients with either concurrent tuberculosis or post-tubercular structural lung disease. Of the 265 patients with past history of tuberculosis with or without structural lung disease, reported from three studies, 82 (31%) showed evidence of Aspergillus sensitization. The criteria used for defining Aspergillus sensitization are shown in Table 4. Only one study out of the three that assessed Aspergillus sensitization in patients with previous tuberculosis clearly reported the criteria used for diagnosis of tuberculosis, which was either microbiological or clinic-radiological with response to antitubercular therapy. Of the 16 patients with previous tuberculosis who developed Aspergillus sensitization for whom radiological features were described, central bronchiectasis, mucoid-impaction/HAM/Glove and finger shadows and fungal ball were each reported in one patient.
Table 4
Aspergillus sensitization in tuberculosis
Total number of observational studies = 4 |
Case reports = 0 |
Total number of patients with previous tuberculosis = 265 |
Aspergillus sensitization = 82/265 (31%) |
Total number of patients with active tuberculosis = 101 |
Aspergillus sensitization = 9 (10%)* |
1. Criteria for Aspergillus sensitization | Studies on patients with previous Tuberculosis (n = 3 studies) | Studies on patients with concurrent TB (n = 1 study) |
Type I reaction to A. fumigatus antigen or an A. fumigatus-specific serum IgE level .0.35 kUA/l | 1 | - |
A fumigatus-specific IgE > 0.35 kUA/L and IgG > 27 mgA/L | 1 | - |
A fumigatus-specific IgE > 0.35 kUA/L | - | 1 |
NA/Clinical : | 1 | 0 |
2. Criteria for TB diagnosis | Studies on patients with previous Tuberculosis (n = 3 studies) | Studies on patients with concurrent TB (n = 1 study) |
Microbiological/ Clinico-radiological with ATT response | 1 | - |
Microbiological/ Clinico-radiogical | - | 1 (93%/7%) |
NA | 2 | 0 |
3. Radiological features | Previous Tuberculosis (n = 16)** | Concurrent TB (n = 9 patients) |
Central bronchiectasis | 1 | Not reported |
Cavitatory lesion | - | Not reported |
Mucoid-impaction/HAM/Glove and finger shadows | 1 | Not reported |
Fungal ball | 1 | Not reported |
Fleeting opacities/Patchy infiltrates | - | Not reported |
Consolidation/Mass/Nodule | - | Not reported |
Centrilobular nodules | - | Not reported |
*9/93 (Only 93 had Aspergillus specific IgE evaluated) |
** Reported only for 16 with Aspergillus sensitization |
A single study reported the prevalence of Aspergillus sensitization in HIV-infected patients with active pulmonary tuberculosis diagnosed either microbiologically (93%) or clinico-radiologically (7%). This study showed a prevalence of 10% for Aspergillus sensitization, defined as Aspergillus fumigatus specific IgE > 0.35 kUA/L at the end of treatment duration of 24 months. The radiological features specific to the subgroup of patients with Aspergillus sensitization were not reported as shown in Table 4.