Radiological characteristics of pulmonary cryptococcosis in HIV-infected patients
Radiological characteristics of pulmonary cryptococcosis in HIV-infected patients
Zhiliang Hu 0
Qingfang Xiong 0
Yandan Zhong 0
Yongfeng Yang 0
Hongxia Wei 0
Omar Sued, ARGENTINA
0 Department of Infectious Disease, the Second Affiliated Hospital of Medical School of the Southeast University , Nanjing, Jiangsu , China , 2 Department of Infectious Diseases, Shanghai Public Health Clinical Center, Fudan University , Shanghai , China , 3 Department of Pathology, Nanjing Brain Hospital, Nanjing Medical University , Nanjing, Jiangsu , China , 4 Department of radiology, the Second Affiliated Hospital of Medical School of the Southeast University , Nanjing, Jiangsu , China
Data Availability Statement: All relevant data are
within the paper and the Supporting Information
Funding: This work is supported by clinical special
grant BL-2013003 from Department of Science
and Technique, Jiangsu (http://www.jstd.gov.cn/),
and the project of Jiangsu Province Medical Youth
Talent. The funders had no role in study design,
data collection and analysis, decision to publish, or
preparation of the manuscript.
Competing interests: The authors have declared
that no competing interests exist.
60 cases were identified. The median CD4 T-cell counts were 20 cells/μL (range, 0±205
cells/μL). Chest CT scans demonstrated nodular lesions in 93.3% of the studied patients.
Those nodular lesions were usually cavitated and solitary nodule was the most common
form. Pleural effusions and pneumonic infiltrates occurred in 11.6% and 31.7% of the cases
respectively. Those lesions were usually had co-existing nodular lesions. Etiological
analysis suggested that 76.8% of the nodular lesions could have a relationship with PC that
12.5% of the nodular lesions were ªlaboratory-confirmedº cases, 48.2% were ªclinically
confirmedº cases and 16.1% were ªclinically probableº cases. 85.7% of the pleural effusions
could be ªclinically confirmedº cases of PC. At least, 38.5% of the diffuse pneumonic
infiltrates may be clinically attributed to pneumocystis pneumonia.
This study suggested that pulmonary nodules but not diffuse pneumonia are the most
common radiological characteristics of HIV-associated PC. HIV-infected patients with
pulmonary nodules on Chest CT should particularly be screened for cryptococcal infection.
In human immunodeficiency virus (HIV)-negative patients, pulmonary cryptococcosis (PC)
has been extensively studied and nodular masses are the most common radiological
manifestations [1±4]. However, PC in HIV-infected is less well defined. Knowledge of HIV-associated
PC are largely based on studies performed about 2 decades ago which reported that the most
common findings on chest radiograph were diffuse interstitial infiltrates [5±7]. In those
HIVinfected immunocompromised patients, PC was thought to be an important contributor to
fetal respiratory failure [
]. Although computed tomography (CT) has been widely applied in
evaluating pulmonary lesions, few studies on the CT findings of HIV-associated PC are
available. In our previous study, we evaluated 11 HIV-infected patients with pulmonary
abnormalities on chest CT scan and confirmed cryptococcal diseases outside of the lung. Interestingly, 9
patients had solitary pulmonary nodules that could be clinically related to cryptococcal
]. This finding is inconsistent with the dominant perception of HIV-associated PC and
necessitates further investigation. Here, we extend our previous study to expand the
understanding of CT findings of HIV-associated PC.
Patients and methods
HIV-infected patients with crytococccal infection diagnosed by positive cryptococcal antigen
test and/or isolation of Cryptococcus spp., between September 2010 and May 2016 in the
Second Affiliated Hospital of the Southeast University, were retrospectively analyzed. Those with
pulmonary abnormalities were further evaluated for the possibility of pulmonary
cryptococcosis. Confirmed cases of tumors, mycobacterial infections and other fungal infections were
excluded from the analysis. The medical records, such as age, sex, antiretroviral therapy,
clinical features, laboratory tests, chest CT scan imaging results, treatment and outcome, were
collected. This retrospective study was approved by the Ethics Committee of the Second Affiliated
Hospital of the Southeast University. The data were analyzed and presented anonymously.
The chest CT scan findings were mainly classified into three patterns, which were the nodular
lesion with clear boundary, the pneumonic infiltrates with ill-defined margin, and pleural
effusions, as described by Zhang et al with modification[
]. Based on the number of the nodule,
nodular lesions were further subdivided into single nodular (solitary) and multi-nodular.
Cavitation was also recorded. The pneumonic infiltrateswere subdivided into focal, multi-focal and
diffuse pneumonic infiltrates. The special pneumonic infiltrates, defined as diffuse
groundglass opacities on CT scans, were also recorded.
Diagnosis of PC
If pulmonary lesion responded to anti-cryptococcal therapy alone, the case was defined as
ªclinically confirmedº PC. If additional drugs were also used to achieve an improvement, PC
2 / 9
was considered ªclinically probableº. PC was considered ªlaboratory confirmedº if
cryptococcal elements were revealed by lung biopsy or sputum culture from a ªclinically confirmedº PC
patient grew Cryptococcusspp.. PC was considered ªunlikelyº if pulmonary lesions improved
without an anti-cryptococcal treatment. The special lesions of diffuse ground-glass
opacitieswere attributed to pneumocystis pneumonia (PCP) if lesions resolved with anti-PCP therapy
Identification of cases
From September 2010 to May 2016, 80 HIV-infected patients were confirmed to have
cryptococcal infection in the Second Affiliated Hospital of the Southeast University, of whom 66
cases have pulmonary abnormalities on chest CT scan. After excluded 4 cases of pulmonary
tuberculosis and 2 cases of Kaposi's sarcoma, the remaining 60 cases with pulmonary
abnormalities were further evaluated for the possibility of PC(S1 Table). Of these 60 cases, 88.3%
(53/60) were men and 76.7% (46/60) had cryptococcal meningitis (CM). 44 patients with
cryptococcal meningitis were treated with amphotericin B-based combination therapy followed by
fluconazole treatment as recommended by clinical practice guidelines [
]. Of those 44 patients
treated with amphotericin B, 32 patients had blood culture results available that 56.3% (18/32)
of the blood cultures grew Cryptococcus spp. Patients without CM were treated with
fluconazole alone. The median age was 38 years (range, 12-71years). The median CD4 T-cell counts
were 20 cells/μL (range, 0±205 cells/μL). Six patients had initiated anti-retroviral therapy for a
median of 3 months (range, 2 weeks to 2 years), of whom two developed virological failure.
Findings on chest CT scan
Chest CT scans demonstrated nodular lesions in 93.3% (56/60) of the studied patients (Fig 1),
and solitary nodule accounted for the majority of the cases (69.6%, 39/56). These nodular
lesions were usually cavitated with a cavitation rate of 78.6% (44/56). Pleural effusions
occurred in 11.6% (7/60) of the studied patients, and 71.4% (5/7) of the patients with pleural
effusions had co-existing nodular lesions. Pneumonic infiltrates occurred in 31.7% (19/60) of
the patients; of which, 68.4% (13/19) were diffuse pneumonic infiltrates (all diffuse
groundglass opacities), 21.1% (4/19) were focal pneumonic infiltrates and 10.5% (2/19) were
multifocal pneumonic infiltrates. Of these 19 patients with pneumonic infiltrates, 89.5% (17/19) had
co-existing nodular lesions.
Etiologies of pulmonary lesions
76.8% (43/56) of the nodular lesions could be demonstrated a relationship with PC that 12.5%
(7 /56) of the nodular lesions were ªlaboratory-confirmedº cases, 48.2% (27/56) were
ªclinically confirmedº cases and 16.1% (9/56) were ªclinically probableº cases (Table 1). Of the 7
ªlaboratory-confirmedº cases of PC with nodular lesions, 5 were confirmed by lung needle
biopsies that revealed cryptococcal elements (Fig 2, S1 File) and culture of the biopsy
specimens grew Cryptococcus spp.; the other 2 ªlaboratory-confirmedº cases of PC fulfill the criteria
of ªclinically confirmedº cases and culture of the sputum grew Cryptococcusspp. (S1 File). All
the above mentioned 9 ªclinically probableº cases of PC also had pneumonic infiltrates that
anti-cryptococcal drugs and other antibiotics were used to achieve an improvement of
pulmonary lesions(S1 File). Finally, it was difficult to deduce the etiologies of the nodular lesions in
23.2% (13/56) of the patients because follow-up chest CT scans to define clinical responses
were unavailable(S1 File).
3 / 9
Fig 1. Chest CT findings of HIV-infected patients with cryptococcal infection and pulmonary
abnormalities. SN, solitary nodule; MN, multiple nodules; PE, pleural effusions; PI, pneumonic infiltrates.
85.7% (6/7) of the pleural effusions were ªclinically confirmedº cases of PC, and the
remaining one was undefined due to lack of follow-up chest CT scan. For the 19 cases of pneumonic
infiltrates, one with solitary cavitary nodule and multi-focal pneumonia fulfill the criteria of
PC, pulmonary cryptococcosis; SN, solitary nodule; PI, pneumonic in®ltrates; PE, pleural effusions; MN, multiple nodules.
a: Lung needle biopsies were performed in ®ve patients which con®rmed PC. Sputum cultures were available in 33 patients, of whom 2 grew Cryptococcus
spp., 10 grew Candida albicans thought to be related to oral candidiasis, and the remaining 21 showed negative results.
b: Etiologies of pulmonary lesions were unde®ned as clinical data were insuf®cient.
c: All these 4 patients had mixed pulmonary lesions. By thoroughly analyzing the clinical data, nodular lesions were considered ªclinically con®rmedº cases
of PC. The diffuse pneumonic in®ltrates, speci®cally the diffuse ground-glass opacities, were attributed to pneumocystis pneumonia
d: Before admitted to our hospital, this patient had severe respiratory symptoms and chest CT scan showed diffuse ground-glass opacities. After treated
with empirical co-trimoxazole for pneumocystis pneumonia, respiratory symptoms gradually improved. When he presented to our hospital with neurological
symptoms, he had received about 3 weeks of co-trimoxazole and did not complain any respiratory symptoms. Chest CT scan showed residual ground-glass
opacities thought to be related to previous pneumocystis pneumonia.
4 / 9
Fig 2. A representative case of laboratory-confirmed pulmonary cryptococcosis. A 49-years-old man
presented to hospital with Lymphadenopathy for 3 months and fever for 1 month. He had HIV infection with a
CD4 count of 108 cells/μL. Serum cryptococcal antigen titer was 1:40. Serum galactomannan assay was
negative. Chest CT scan demonstrated multiple pulmonary nodules (A, arrow). Lung biopsy showed
encapsulated yeast-like fungal cells consistent with cryptococcal infection(arrows in B, alcian blue stain).
ªclinically confirmedº case and culture of the sputum grew Cryptococcus spp., therefore was
considered ªlaboratory-confirmedº PC. Of the 13 cases of diffuse pneumonic infiltrates, 53.8%
(7/13) were "clinical probable" cases of PC as additional drugs, aside from anti-cryptococcal
drugs, were used to achieve pulmonary improvement (those 7 cases also received anti-PCP
therapy therefore PCP could not be excluded); 38.5% (5/13) were unlikely to related to PC (Fig
3, a representative case of clinically deducing the etiology of pulmonary lesion) but were
perhaps clinically attributed to PCP. Etiology of diffuse pneumonic infiltrates was undefined in 1
case due to lack of follow-up chest CT scan.
In this study, 82.5% (66/80) of the HIV-infected patients with confirmed cryptococcal
infection had pulmonary lesions. On one hand, those lesions could be related to cryptococcal
infection, but on the other hand those lesions may be caused by other pathogens or tumors. Due to
immunodeficiency, mixed infection should not beunderestimated, especially in patients with
multiple forms of pulmonary lesions. The initial identification of possible cases of PC excluded
4 cases of pulmonary tuberculosis and 2 cases of Kaposi's sarcoma. Intriguingly, 56 out of the
60 remaining cases showed pulmonary nodules which usually cavitated and solitary nodule
was the most common form. Based on the available patients' data, at least 76.8% (43/56) of the
nodular lesions could be demonstrated a relationship with PC. Those findings were closely
consistent with our previous observation which suggested that solitary cavitary pulmonary
nodule may be a common CT finding in HIV-associated PC[
]. Nevertheless, pulmonary
nodular lesions with or without cavitation, although reported in other studies, were not recognized
as common manifestations of HIV-associated PC [
6, 7, 10
]. More recently, Lin et al reported
that PC contributed to about 17% of the cavitary pulmonary lesions in HIV-infected patients
The high frequency of diffuse interstitial infiltrates, attribute to HIV-associated PC in other
], was not seen in our study. Diffuse pneumonic infiltrates, specifically diffuse
ground-glass opacities, were only present in 21.7% of (13/60) the patients and almost always
coexisted with pulmonary nodules. Of note, at least some of those diffuse pneumonic infiltrates
may be related to PCP (Table 1, Fig 3). It is therefore better to include anti-cryptococcal therapy
as well as anti-PCP drugs in empirical treatment for patients with confirmed cryptococcal
5 / 9
Fig 3. A representative case of clinical deduction of pulmonary cryptococcosis. A 47-year-old woman
presented to hospital with fever, cough and shortness of breath. She was confirmed of HIV infection with a
CD4 count of 60 cells/μL. The serum cryptococcal antigen titer was 1:16. Lumbar puncture was performed and
there was no evidence of cryptococcal meningitis. A chest CT scan showed diffuse ground-glass opacities
(DGO). Also on the Chest CT, there was a small nodule (A, arrow). She was given empirical treatment with
co-trimoxazole (1440 mg, every eight hours) for pneumocystis pneumonia, and methylprednisolone as an
adjunctive therapy. A chest CT scan after 3 weeks showed nearly complete resolution of the DGO; however,
the nodular lesion progressed(B, arrow). The patient then received secondary prophylaxis for pneumocystis
pneumonia with co-trimoxazole(960 mg/d) and treated with fluconazole (400mg/d). A follow-up CT scan after
seventeen days of fluconazole treatment showed partial resolution of the nodule (C, arrow). Taken together,
the nodule lesion was attributed to pulmonary cryptpcoccosis, while the DGO were related to pneumocystis
infection (for example CM) and diffuse pneumonic infiltrates. Aside from PCP, it is also
necessary to exclude other causes of diffuse pneumonic infiltrates in HIV-infected
immunocompromised persons, such as cytomegalovirus pneumonia and mycobacteriosis[
]. This is
important as diffuse pneumonic infiltrates, unlike nodular lesion, would result in severe
respiratory symptoms, and may quickly progress to respiratory failure if left untreated.
Although pleural effusions were less frequently seen in our study, the presence of pleural
effusions was suggestive of PC in our patients. Of note, pleural effusions could be caused by
multiple common conditions in HIV-infected patients, such as tuberculosis and Kaposi
]. Extensive laboratory evaluation of pleural effusions was required torule out
those comorbidities, especially when pleural effusions did not respond to anti-cryptococcal
As discussed above, cryptococcal infection seems likely to cause localized pulmonary
lesions in HIV-infected patients similar with that in HIV-negative patients [1±4], although
cavitation of the nodules seems to be more common in HIV-infected patients. The imaging
modality of Chest CT scan used in our study is more sensitive to reveal pulmonary lesions
6 / 9
]. As shown in Fig 3, the small nodular lesion was surrounded with diffuse pneumonic
infiltrates which may be very difficult to be revealed by chest radiograph. It is possible that nodular
lesions were underestimated in former studies of HIV-associated PC[5±7]. Moreover, as
shown in our study and others' case report[
], pulmonary nodules without pneumonic
infiltrates or pleural effusions did not often cause respiratory symptoms, while former studies on
HIV-associated PC generally included patients with respiratory symptoms [5±7]. All the above
mentioned factors may lead to the low prevalence of pulmonary nodules in HIV-associated PC
in other studies. Nevertheless, as our study was conducted in Chinese population, different
host-pathogen interactions may exist and contribute to the discrepancy.
There are limitations to our study. Only a minority of the cases of PC were
laboratory-confirmed, and most cases were diagnosed by clinical deduction. Response of pulmonary lesions to
anti-cryptococcal treatment is suggestive of PC, however other fungal infections, such as
aspergillosis, penicilliosismarneffei and histoplasmosis that could cause similar lesions [16±18],
should also be excluded. Nevertheless, the clinically confirmed cases of PC were less likely to be
associated with aspergillosis, penicilliosismarneffei or histoplasmosis, as anti-cryptococcal
treatments used in our study were not optimal therapies for the above mentioned diseases[
penicilliosis was unlikely considered because our patients were not from an endemic area and
did not have skin lesions or blood culture results consistent with penicilliosis. Aside from fungal
infections, pulmonary nodules could also be caused by many other reasons, such as
mycobacteria infection and malignancy [11, 13, 20±22], and those etiologies should be included in the
In conclusion, our study suggested that pulmonary nodules but not diffuse pneumonia are
the most common radiological characteristics of HIV-associated PC. HIV-infected patients
with pulmonary nodules on chest CT should particularly be screened for cryptococcal
S1 Table. Patients' data. CMV, cytomegalovirus; CT, computed tomography; PC, pulmonary
cryptococcosis; PCP, pneumocystis pneumonia. a: The lower limit of detection for CMV was
500 copies/mL. b: All the patients with CD4 cell counts less than 200 cells/μL received
prophylaxis therapies for pneumocystis pneumonia. Those treatments were not shown in the category
of other anti-infective therapies.
S1 File. Representative cases of laboratory-confirmed, clinically confirmed, clinically
probable pulmonary cryptococcosis and undefined pulmonary lesions. Cases 1±6 were
laboratory-confirmed pulmonary cryptococcosis. Cases 7 and 8 were clinically confirmed
pulmonary cryptococcosis. Cases 9 and 10 were clinically probable pulmonary cryptococcosis.
Etiologies of the pulmonary lesions of Cases 11 and 12 were undefined.
We thank Dr. Danielle Minkoulou at Southeast University for assistance with editing the
Conceptualization: CX HW.
7 / 9
Funding acquisition: YY.
Investigation: ZH JC QX YZ YY CX JW.
Writing ± original draft: ZH.
Writing ± review & editing: ZH JC CX.
8 / 9
(LDCT) and chest X-ray (CXR). Lung cancer. 2007; 58(1):50±8. https://doi.org/10.1016/j.lungcan.2007.
05.009 PMID: 17624475
1. Kohno S , Kakeya H , Izumikawa K , Miyazaki T , Yamamoto Y , Yanagihara K , et al. Clinical features of pulmonary cryptococcosis in non-HIV patients in Japan . Journal of infection and chemotherapy: official journal of the Japan Society of Chemotherapy . 2015 ; 21 ( 1 ): 23 ± 30 .
2. Zhang Y , Li N , Zhang Y , Li H , Chen X , Wang S , et al. Clinical analysis of 76 patients pathologically diagnosed with pulmonary cryptococcosis . The European respiratory journal . 2012 ; 40 ( 5 ): 1191 ± 200 . https://doi.org/10.1183/09031936.00168011 PMID: 22408204
3. Ye F , Xie JX , Zeng QS , Chen GQ , Zhong SQ , Zhong NS . Retrospective analysis of 76 immunocompetent patients with primary pulmonary cryptococcosis . Lung . 2012 ; 190 ( 3 ): 339 ± 46 . https://doi.org/10. 1007/s00408-011 -9362-8 PMID: 22246551
4 . Chang WC , Tzao C , Hsu HH , Lee SC , Huang KL , Tung HJ , et al. Pulmonary cryptococcosis: comparison of clinical and radiographic characteristics in immunocompetent and immunocompromised patients . Chest . 2006 ; 129 ( 2 ): 333 ± 40 . https://doi.org/10.1378/chest.129.2.333 PMID: 16478849
5. Visnegarwala F , Graviss EA , Lacke CE , Dural AT , Johnson PC , Atmar RL , et al. Acute respiratory failure associated with cryptococcosis in patients with AIDS: analysis of predictive factors. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America . 1998 ; 27 ( 5 ): 1231 ± 7 .
6. Meyohas MC , Roux P , Bollens D , Chouaid C , Rozenbaum W , Meynard JL , et al. Pulmonary cryptococcosis: localized and disseminated infections in 27 patients with AIDS. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America . 1995 ; 21 ( 3 ): 628 ± 33 .
7. Batungwanayo J , Taelman H , Bogaerts J , Allen S , Lucas S , Kagame A , et al. Pulmonary cryptococcosis associated with HIV-1 infection in Rwanda: a retrospective study of 37 cases . Aids. 1994 ; 8(9):1271±6 . PMID: 7802979
8. Hu Z , Xu C , Wei H , Zhong Y , Bo C , Chi Y , et al. Solitary cavitary pulmonary nodule may be a common CT finding in AIDS-associated pulmonary cryptococcosis . Scandinavian journal of infectious diseases . 2013 ; 45 ( 5 ): 378 ± 89 . Epub 2012/12/19. https://doi.org/10.3109/00365548. 2012 .749422 PMID: 23244589
9. Perfect JR , Dismukes WE , Dromer F , Goldman DL , Graybill JR , Hamill RJ , et al. Clinical practice guidelines for the management of cryptococcal disease: 2010 update by the infectious diseases society of america. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America . 2010 ; 50 ( 3 ): 291 ± 322 .
10. Miller KD , Mican JA , Davey RT . Asymptomatic solitary pulmonary nodules due to Cryptococcus neoformans in patients infected with human immunodeficiency virus. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America . 1996 ; 23 ( 4 ): 810 ± 2 .
11. Lin CY , Sun HY , Chen MY , Hsieh SM , Sheng WH , Lo YC , et al. Aetiology of cavitary lung lesions in patients with HIV infection . HIV medicine . 2009 ; 10 ( 3 ): 191 ±8. https://doi.org/10.1111/j.1468- 1293 . 2008 . 00674 . x PMID : 19207599
12. Steininger C , Puchhammer-Stockl E , Popow-Kraupp T . Cytomegalovirus disease in the era of highly active antiretroviral therapy (HAART) . Journal of clinical virology: the official publication of the Pan American Society for Clinical Virology . 2006 ; 37 ( 1 ):1± 9 .
13. Benito N , Moreno A , Miro JM , Torres A . Pulmonary infections in HIV-infected patients: an update in the 21st century . The European respiratory journal . 2012 ; 39 ( 3 ): 730 ± 45 . https://doi.org/10.1183/09031936. 00200210 PMID: 21885385
14. Restrepo CS , Martinez S , Lemos JA , Carrillo JA , Lemos DF , Ojeda P , et al. Imaging manifestations of Kaposi sarcoma. Radiographics: a review publication of the Radiological Society of North America , Inc. 2006 ; 26 ( 4 ): 1169 ± 85 .
15. Blanchon T , Brechot JM , Grenier PA , Ferretti GR , Lemarie E , Milleron B , et al. Baseline results of the Depiscan study: a French randomized pilot trial of lung cancer screening comparing low dose CT scan
16. Muldoon EG , Sharman A , Page I , Bishop P , Denning DW . Aspergillus nodules; another presentation of Chronic Pulmonary Aspergillosis . BMC pulmonary medicine . 2016 ; 16 ( 1 ): 123 . PubMed Central PMCID : PMC4991006. https://doi.org/10.1186/s12890-016 -0276-3 PMID: 27538521
17. Chitasombat M , Supparatpinyo K. Penicillium marneffei infection in immunocompromised host . Current Fungal Infection Reports . 2013 ; 7 ( 1 ): 44 ± 50 .
18. Ye C , Zhang G , Wang J , Chai Y. Histoplasmosis presenting with solitary pulmonary nodule: two cases mimicking pulmonary metastases . Nigerian journal of clinical practice . 2015 ; 18 ( 2 ): 304 ±6. https://doi. org/10.4103/ 1119 - 3077 .151075 PMID: 25666013
19. Iribarren JA , Rubio R , Aguirrebengoa K , Arribas JR , Baraia-Etxaburu J , Gutierrez F , et al. Prevention and treatment of opportunistic infections and other coinfections in HIV-infected patients: May 2015 . Enfermedades infecciosas y microbiologia clinica . 2016 .
20. La Ferla L , Pinzone MR , Nunnari G , Martellotta F , Lleshi A , Tirelli U , et al. Kaposi' s sarcoma in HIV-positive patients: the state of art in the HAART-era. European review for medical and pharmacological sciences . 2013 ; 17 ( 17 ): 2354 ± 65 . PMID: 24065230
21. Hare SS , Souza CA , Bain G , Seely JM , Frcpc, Gomes MM , et al. The radiological spectrum of pulmonary lymphoproliferative disease . The British journal of radiology . 2012 ; 85 ( 1015 ): 848 ± 64 . PubMed Central PMCID : PMC3474050. https://doi.org/10.1259/bjr/16420165 PMID: 22745203
22. Kisembo HN , Boon SD , Davis JL , Okello R , Worodria W , Cattamanchi A , et al. Chest radiographic findings of pulmonary tuberculosis in severely immunocompromised patients with the human immunodeficiency virus . The British journal of radiology . 2012 ; 85 ( 1014 ) :e130±9 . PubMed Central PMCID : PMC3474111. https://doi.org/10.1259/bjr/70704099 PMID: 21976629