Pneumococcal Pneumonia in

Patients Requiring Hospitalization:

Effects of Bacteremia and HIV
Seropositivity on Radiographic
Appearance
Rosita M. Shah 1 OBJECTIVE. Our purpose was to establish the incidence of lobar versus bronchopneu-
Sanjay Gupta 2 monia patterns in patients hospitalized for pneumococcal pneumonia and to determine
Elin Angeid-Backman 3 whether bacteremia or HIV status affects the radiographic appearance.
Judith O’Donnell 4 MATERIALS AND METHODS. Eighty-one patients with community-acquired pneumo-
coccal pneumonia and positive findings on admission chest radiographs were selected from a
group of 105 patients with positive findings for Streptococcus pneumoniae from sputum or blood
cultures. Patients less than 16 years old and those with nosocomial pneumonia, aspiration pneu-
American Journal of Roentgenology 2000.175:1533-1536.

monia, or coexistent lung disease were excluded. The dominant pattern was classified as lobar
or lobular bronchopneumonia by two radiologists who were unaware of results from blood cul-
tures and HIV testing.
RESULTS. Forty-three females and 38 males, with a mean age of 48 years (age range, 16–92
years), required admission for pneumococcal pneumonia. Fifty-nine (73%) of the 81 patients
were bacteremic and 20 (25%) of the 81 patients were HIV-positive, including 14 patients (17%)
who were both bacteremic and HIV-positive. Among all the patients, focal lobar consolidation
was the most common pattern, observed in 48%. Multifocal lobar consolidation was the next
most frequent pattern, occurring in 33%. Multifocal and focal bronchopneumonia patterns were
seen in 16% and 2% of the patients, respectively. Overall, multifocal consolidation occurred in
49%. The dominant radiographic pattern and incidence of multicentric disease were not affected
by HIV seropositivity (p = 0.61) or bacteremia (p = 0.17).
CONCLUSION. Lobar consolidation, involving single or multiple lobes, is the most com-
mon radiographic pattern of community-acquired pneumococcal pneumonia in patients requiring
hospitalization. The pattern of consolidation is not influenced by bacteremia or HIV status.

Received April 22, 1999; accepted after revision
May 12, 2000.
Presented at the annual meeting of the American
Roentgen Ray Society, San Francisco, April–May 1998.
1
Department of Radiology, Thomas Jefferson University
Hospital, 111 S. 11th St., Ste. 3390 Gibbon, Philadelphia, PA
19107. Address correspondence to R. M. Shah.
2
Department of Radiology, Yale University School of
Medicine, 333 Cedar St., New Haven, CT 06520.
3
Department of Radiology, Medical College of Georgia,
1120 15th St., Augusta, GA 30912.
4 teaching hospital between January 1994 and March
Department of Infectious Disease, Medical College of
Pennsylvania, 3300 Henry Ave., Philadelphia, PA 19129.
AJR 2000;175:1533–1536
0361–803X/00/1756–1533
© American Roentgen Ray Society
A
mong immune-competent and
immune-suppressed hosts, Strep-
tococcus pneumoniae is readily
recognized as a frequent cause of commu-
nity-acquired pneumonia. In hospitalized pa-
tients with advanced age or associated risk
factors, mortality is said to exceed 50% [1, 2].
In addition to localized lobar consolidation,
studies addressing the radiographic appearance
of pneumococcal pneumonia in hospitalized or
immunosuppressed patients frequently report
atypical appearances, including bronchopneu-
monic or interstitial patterns [3–5].
We have observed a number of cases of
multifocal pneumonia with striking lobar pat-
terns that were later diagnosed as pneumococ-
cal pneumonia in which HIV test results were
positive. It was unclear to us if this variant pat-
tern of pneumonia reflected HIV seropositivity
or if it was a manifestation of disease severity
leading to hospitalization. Although the previ-
ously mentioned literature has described the
appearance of pneumococcal pneumonia in
HIV-positive and hospitalized patients, we are
not aware of a study directly comparing HIV-
positive and HIV-negative patients who have
pneumococcal pneumonia [4–6].
Our purpose was to determine if the dominant
chest radiographic pattern of pneumococcal
pneumonia in hospitalized patients was influ-
enced by HIV status or general disease severity.
We chose bacteremia as an indicator of severity.
Materials and Methods
Our study population consisted of all consecutive
patients with community-acquired pneumococcal
pneumonia requiring admission to a Philadelphia
1997. There were 105 patients with positive findings
for S. pneumoniae from sputum or blood cultures, or
both, in whom admission radiographs revealed posi-
tive findings. Patients less than 16 years old and
those with suspected nosocomial pneumonia, aspira-

AJR:175, December 2000 1533

 tion pneumonia, or known coexistent lung disease Fig. 1.—50-year-old HIV-negative
Shah et al. woman with bacteremia. Chest ra-
were excluded, producing a study population of 81
patients with community-acquired pneumococcal diograph reveals lobar pattern of
Streptococcus pneumoniae. Note
pneumonia. All patients were symptomatic and re- homogeneous, confluent air-space
quired hospital admission. consolidation in right upper lobe.
The dominant pattern seen on admission chest
radiographs was classified as lobar or lobular
bronchopneumonia by the consensus interpretation of
two radiologists. According to classic radiologic de-
scriptions, a lobar pattern of consolidation was consid-
ered present in the setting of homogeneous confluent
air-space density, with or without associated findings
of air bronchograms or the presence of acinar shadows
[7, 8] (Fig. 1). The lobular or bronchopneumonia pat-
tern was considered present when consolidation was
heterogeneous, nonconfluent, and with prominent
peribronchial thickening. [7, 8] (Fig. 2). Regardless of
the size of the consolidation, any heterogeneity was
classified as bronchopneumonic unless limited to the
advancing edge of the infiltrate. The frequency of mul-
tilobar consolidation was also recorded. The reviewers
were unaware of the clinical results of blood cultures
and HIV testing.

Results
American Journal of Roentgenology 2000.175:1533-1536.

Forty-three females and 38 males with a

mean age of 48 years (age range, 16–92 years)
required admission to the hospital for docu-
mented pneumococcal pneumonia. Fifty-nine
(73%) of the 81 patients were bacteremic and 20
(25%) of the 81 patients were HIV-positive, with
14 bacteremic patients also being HIV-positive.
Medical records were available for 17 of the 20
HIV-positive patients. Known risk factors in-
cluded IV drug use in seven patients, dialysis or
transfusion-related complications in two, and
promiscuity in two. The diagnosis of pneumo-
coccal pneumonia represented the first opportu-
nistic infection in 10 of the 17 patients with
available records. CD4 lymphocyte counts were
available in 15 patients and ranged from 1 to 799
cells/µL, with a mean of 309 cells/µL. Patients
without documented positive HIV serology
were classified as HIV-negative. Six patients
with multiple culture sites that showed positive
findings were classified as bacteremic. Because
only the positive culture results were retrievable,
it is unknown for how many patients both blood
and sputum cultures were performed.
By consensus, lobar consolidation was
present in 66 (81%) of the 81 patients. Lobu-
lar or bronchopneumonic consolidation was
present in 15 (19%) of the 81 patients. Initial
disagreement regarding the dominant pattern
occurred in four cases, three of which were
ultimately classified as lobar and one as bron-
chopneumonic. In the three cases of lobar
consolidation, heterogeneity was limited to
the periphery. Fig. 2.—36-year-old HIV-negative
Focal lobar consolidation was the most man with bacteremia. Chest radio-
common pattern, identified in 39 patients graph reveals bronchopneumonic
pattern of Streptococcus pneumo-
(48%), followed by multifocal lobar consoli- niae. Note patchy nonconfluent con-
dation in 27 (33%), multifocal bronchopneu- solidation in left lower lobe.

1534 AJR:175, December 2000

 Radiography of Pneumococcal Pneumonia

monia 13 (16%), and focal bronchopneumonia

in two (2%) (Figs. 1, 3, and 4 and Table 1).
The dominant radiographic pattern and the
incidence of multicentric disease were not af-
fected by HIV seropositivity or bacteremia. Us-
ing the Kruskal-Wallis test for a three-group
comparison, we found no significance (p = 0.61)
between the incidence of focal or multifocal lo-
bar consolidation and bronchopneumonia with
HIV seropositivity. Similarly, no significance
( p = 0.17) was found between focal or multifo-
cal lobar consolidation or bronchopneumonia
patterns and positive results from blood cul-
tures. Power analysis revealed a power of 0.80
to detect differences of 35% between the groups
(α = 0.05, two-tailed).
Pleural effusions were noted in 12 patients,
including 10 HIV-negative and two HIV-posi-
tive patients. Neither cavitation nor abscess
formation was detected on any of the admis-
sion radiographs. Follow-up radiographs were
American Journal of Roentgenology 2000.175:1533-1536.

not assessed.

Discussion Fig. 3.—47-year-old HIV-negative woman with bacteremia. Chest radiograph reveals multifocal lobar consolida-
tion in both lower lobes.
The morphologic classification of pneumo-
nia into lobar, lobular, and interstitial patterns is
well established in the radiography literature [7,
8]. This classification scheme reflects the initial
pathologic distribution of the inflammatory pro-
cess. Contiguous, nonsegmental alveolar filling
is seen histologically in lobar pneumonias,
whereas patchy or nonconfluent and segmental
peribronchial inflammatory changes are ob-
served in bronchopneumonia [7]. Although cor-
relation with radiographic appearance can be
expected frequently, limitations are well known.
Accurate pattern recognition requires imaging
during an early phase of inflammation and nor-
mal underlying lung architecture. Increasing in-
flammatory infiltrates and edema in advanced
bronchopneumonias or interstitial pneumonias
can mimic lobar pneumonia patterns. Alterna-
tively, preexisting diseased lung can contribute
to apparent heterogeneity, potentially altering
the appearance of lobar pneumonia. The litera-
ture suggests basic differentiation between alve-
olar and interstitial pneumonias can be difficult
even under optimal conditions. In a radio-
graphic–pathologic correlative study, accurate
differentiation between bronchopneumonias
and interstitial pneumonias occurred in only
30% of the cases [9]. This difficulty in differen-
tiation can be expected because prominent
peribronchial thickening is observed in both.
Because our results reflect a consensus inter-
pretation of the radiographs, disagreement on Fig. 4.—37-year-old HIV-positive man without bacteremia. Chest radiograph reveals multifocal bronchopneumonia in
the basic pattern could have occurred in bor- parahilar regions of both lungs.

AJR:175, December 2000 1535


TABLE 1 Radiographic Patterns in 81 Patients with Pneumococcal Pneumonia
HIV Negative HIV Positive
Pattern
(n = 61) (n = 20)
Lobar
Focal 31 (51%)
Multifocal 20 (33%)
Lobular
Focal 1 (2%)
Multifocal 9 (15%)
derline cases. We did try to minimize this by
classifying any heterogeneity as bronchopneu-
monia unless it was limited to the margin of an
otherwise homogenous opacity.
Despite significant limitations, the radio-
graphic pattern approach to the diagnosis of
pneumonia remains important. This may espe-
cially be true in the HIV-positive patient pre-
senting with acute symptoms indicative of
pneumonia. Treatment regimens differ signifi-
cantly among pneumonias because of routine
American Journal of Roentgenology 2000.175:1533-1536.
bacterial causes, Pneumocystis carinii, and
other atypical organisms. Improvement in
chest radiographic abnormalities during the
course of treatment and adequate clinical re-
sponse may be sufficient to prevent further di-
agnostic evaluation in most patients.
Familiarity with the common and uncommon
radiographic appearances of pneumococcal
pneumonia in the AIDS population is extremely
important. According to the 1993 revised AIDS
case definition by the Centers for Disease Con-
trol and Prevention [10], recurrent bacterial
pneumonia is considered an AIDS-defining ill-
ness in the setting of positive HIV serology. Of
the bacterial pneumonias in AIDS, S. pneumo-
niae is the most common, occurring five to 18
times more frequently than in the population at
large [11–13]. The literature suggests a declining
incidence of pneumocystis pneumonia, replaced
by a rising incidence of bacterial pneumonia. In a
prospective study of 1130 patients with HIV in-
fection, Hirschtick et al. [11] found bacterial
pneumonia to be the most common cause of pul-
monary infection. This trend likely reflects the
B-cell and neutrophilic dysfunction that are now
recognized to occur in the early and advanced
immunosuppressive stages of AIDS. Defects in
B-cell or humoral immunity are primarily re-
sponsible for the increased rate of infection with
encapsulated bacteria. Furthermore, there is an
apparent tendency toward frequent complica-
tions, including abscess formation, empyemas,
and bronchiectasis [14–16].
In agreement with early literature on the ra-
diographic appearance of pneumococcal pneu-
monia, we were able to show that the focal lobar
pattern of consolidation is the most common
pattern of streptococcal pneumonia in all pa-
tients examined—regardless of HIV status or
1536
Shah et al.
Bacteremia No Bacteremia
(n = 59) (n = 22)
8 (40%) 30 (51%) 10 (45%)
7 (35%) 21 (36%) 6 (27%)
1 (5%) 1 (2%) 0
4 (20%) 7 (12%) 6 (27%)
associated bacteremia. Multifocal lobar type
consolidation, observed in one third of our pa-
tients, was the next most frequent variant pat-
tern, also occurring independently of HIV status
or bacteremic state. Together, multifocal lobar
or bronchopneumonic patterns accounted for
nearly half of all cases and were apparently un-
related to HIV status or bacteremic state. The
high incidence of multifocal consolidation in
our study may be related to the severity of ill-
ness leading to hospitalization in our patient
population. Prior studies have reported a high
incidence of morbidity and mortality in patients
with bacteremia, contributing to our decision to
use bacteremia as an indicator of severity [17–
19]. A limitation of our series is that an outpa-
tient population, presumedly those with less
severe disease, was not studied. We also did not
have results of HIV serology for most patients;
thus, we potentially included some HIV-posi-
tive patients in the group with negative serologic
findings. Because the study was not performed
prospectively and patients were not enrolled
into the study, this was unavoidable.
We did not attempt to include the interstitial
pattern as a descriptor for the radiographic find-
ings of pneumococcal pneumonia. There is sig-
nificant overlap in the imaging features of the
bronchopneumonias and interstitial pneumo-
nias, largely related to a prominent component
of peribronchial thickening occurring in both. It
is unclear whether prior studies reporting inter-
stitial patterns in streptococcal pneumonia reli-
ably distinguished true interstitial abnormality
from bronchopneumonia.
Underlying lung disease can alter the appear-
ance of usual radiographic patterns, often con-
tributing to heterogeneity. For this reason, we
excluded patients with known pulmonary disor-
ders, but we likely retained some patients with
unsuspected disease including those with em-
physema. These patients might have increased
the number of patients classified as having bron-
chopneumonia. Differentiation may have been
facilitated with CT. Although radiographic–CT
correlation would have been optimal, CT is in-
frequently performed in the initial treatment of
pneumonia. Because evaluation of the chest ra-
diograph remains integral to patient workup, our
reliance on descriptions of radiography of the
chest for classification of radiographic abnor-
malities remains important.
Focal or multifocal lobar consolidation is the
most common pattern of pneumococcal pneu-
monia in patients requiring hospitalization re-
gardless of HIV status or bacteremic state.
References
1. Watanakunakorn C, Stroh K, Blend D, et al. Pneumo-
coccal bacteremia in three community teaching hos-
pitals from 1980 to 1989. Chest 1993;103:1152–1156
2. Hook EW III, Horton CA, Schaberg DR. Failure of
intensive care unit support to influence mortality
from pneumococcal bacteremia. JAMA 1983;249:
1055–1057
3. Ziskind MM, George RB, Weill H. Acute local-
ized and diffuse alveolar pneumonias. Semin
Roentgenol 1967;2:46–60
4. Magnenat JL, Nicod LP, Auckenthaler R, et al. Mode
of presentation and diagnosis of bacterial pneumonia
in human immunodeficiency virus-infected patients.
Am Rev Respir Dis 1991;144:917–922
5. Boiselle PM, Tocino I, Hooley RJ, et al. Chest radio-
graph interpretation of Pneumocystis carinii pneumo-
nia, bacterial pneumonia, and pulmonary tuberculosis
in HIV-positive patients: accuracy, distinguishing fea-
tures, and mimics. J Thorac Imaging 1997;12:47–53
6. Janoff EN, Breiman RF, Daley CL, et al. Pneu-
mococcal disease during HIV Infection. Ann In-
tern Med 1992;117:314–324
7. Groskin SA. Heitzman’s the lung: radiologic-
pathologic correlations, 3rd ed. St. Louis: Mosby
1993:194–205
8. Felson B. Chest roentgenology, 1st ed. Philadel-
phia: Saunders, 1973:288–299
9. Tew J, Calenoff L, Berlin BS. Bacterial or non-
bacterial pneumonia: accuracy of radiographic di-
agnosis. Radiology 1977;124:607–612
10. Centers for Disease Control and Prevention. 1993
Revised classification system for HIV infection
and expanded surveillance case definition for
AIDS among adolescents and adults. Morb Mor-
tal Wkly Rep 1992 Dec 18;41(RR-17):1–19
11. Hirschtick RE, Glassroth J, Jordan MC, et al. Bacterial
pneumonia in persons infected with the human immu-
nodeficiency virus. N Engl J Med 1995;333:845–851
12. Miller RF, Foley NM, Kessler D, et al. Community
acquired lobar pneumonia in patients with HIV in-
fection and AIDS. Thorax 1994;49:367–368
13. Polsky B, Gold JWM, Whimbey E, et al. Bacterial
pneumonia in patients with the acquired immunodefi-
ciency syndrome. Ann Intern Med 1986;104:38–41
14. Janoff EN, Breiman RF, Daley CL, et al. Pneu-
mococcal disease during HIV infection. Ann In-
tern Med 1992;117:314–324
15. McGarry TM, Rohman M, Huang CT. Pneumato-
cele formation in adult pneumonia. Chest 1987;
92:717–720
16. McGuinness GM, Naidich DP, Garay S, et al.
AIDS associated bronchiectasis: CT features. J
Comput Assist Tomogr 1993;17:260–266
17. Tilghman RC, Finland M. Clinical significance of
bacteremia in pneumococcic pneumonia. Arch In-
tern Med 1937;59:602–619
18. Hook EW III, Horton CA, Schaberg DR. Failure of
intensive care unit support to influence mortality
from pneumococcal bacteremia. JAMA 1983;
249:1055–1057
19. Marrie TJ. Bacteremic pneumococcal pneumonia: a
continuously evolving disease. J Infect 1992;24:
247–255
AJR:175, December 2000