Infection caused by Haemophilus species is usually caused by Haemophilus influenzae, known for its ability to cause respiratory tract infections and native valve endocarditis. Other species such as Haemophilus parainfluenzae are thought to have comparatively lower pathogenicity. Reported to be part of the normal flora of the oropharynx, urethra, and vagina, H. parainfluenzae is not a frequent cause of infection.1
Despite the rarity with which it causes disease, H. parainfluenzae infections are clinically similar to those caused by H. influenzae.1 In reports, H. parainfluenzae has been noted to cause cases of pharyngitis, epiglottitis, otitis media, conjunctivitis, dental abscess, pneumonia, septic arthritis, osteomyelitis, paraspinal and epidural abscesses, peritonitis, hepatobiliary infection, meningitis, brain abscess, and urinary tract and genital infections.1 To our knowledge, there has been only 1 prior reported case of thoracic empyema caused by H. parainfluenzae.2 We report a second such case of H. parainfluenzae thoracic empyema.
The patient is a 44-year-old woman with a medical history significant for rheumatoid arthritis treated with weekly methotrexate. She was in her usual state of health until 4 days before admission when she noted a "swishing" sound from the right side of her chest while bending down. Three days before admission, she noted the onset of flulike symptoms including fevers (102.5°F), chills, nausea, vomiting, diarrhea, and diaphoresis. On the day of admission, she went to her physician's office, where she was noted to have diminished breath sounds from her right lung and was subsequently sent to the emergency department for further evaluation.
In the emergency department, the patient was noted to have sinus tachycardia of 116 beats per minute with a stable blood pressure. She was afebrile with good oxygen saturation on room air. Physical examination confirmed diminished breath sounds from the right thorax. A chest radiograph and a computed tomographic (CT) scan (Fig. 1) demonstrated a large right-sided hydropneumothorax with no midline shift. A chest tube was inserted with an initial drainage of 2200 mL of puslike fluid. A chest CT scan on hospital day 2 demonstrated reexpansion of the right lung with an ovoid bulla measuring 7 × 15 mm with a slightly thickened wall in the superior segment of the right lower lung.
The patient's methotrexate was held to prevent further immunosuppression, and she was started on an antibiotic regimen of ceftriaxone and azithromycin. Within 24 hours, she became febrile (104°F), and her antibiotics were changed to clindamycin and moxifloxacin for improved anaerobic coverage. She then defervesced. Initial pleural fluid analysis revealed yellow cloudy fluid with a pH of 6.98; 188,120 nucleated cells (97% neutrophils); 4290 red blood cells; lactate dehydrogenase of 15,179 μ/L; glucose level of 2 mg/dL; total protein of 6.4 g/dL; triglycerides of 12 mg/dL; and amylase of 49 μ/L. Gram stain revealed numerous white blood cells and gram-negative rods. At this time, moxifloxacin was discontinued, and the patient was started on aztreonam for better gram-negative coverage. On hospital day 7, the pleural fluid culture grew H. parainfluenzae, which was β-lactamase positive and ampicillin resistant but quinolone sensitive.
The chest tube was removed 9 days after admission, and the patient was discharged the following day on oral moxifloxacin to complete a 4-week course of antibiotics. Chest radiograph upon discharge revealed a residual small loculated right hydropneumothorax. A follow-up chest CT scan 5 months after discharge showed interval resolution of the previous complex pleural effusion.
Although there have been 7 prior reported cases of pneumonia due to H. parainfluenzae, it is not a common respiratory pathogen. The 1 prior reported case of thoracic empyema caused by H. parainfluenzae occurred in the context of alcoholism and chronic obstructive lung disease.2 It is probably safe to assume that immunosuppression is a risk factor for H. parainfluenzae. The use of methotrexate probably served as the risk factor for empyema in our patient. Table 1 shows the prevalence of immunosuppression in the 7 previous case reports of H. parainfluenzae pneumonia.2-6 Of these 7 cases, one was complicated by empyema, and one was complicated by a lung abscess. Our case constituted the second such case of pneumonia with thoracic empyema.
There has been much debate regarding whether H. parainfluenzae, once cultured from the respiratory tract, represents a contaminant or a true pathogen, especially in patients with chronic obstructive pulmonary disease (COPD). In an older prospective study, it was concluded that an isolate of H. parainfluenzae did not represent a true pathogen.7 More recent studies examining surrogate end points, such as inflammatory and immune responses to the organism, have shown conflicting results.
For example, a recent prospective study by Sethi et al8 examined the role of bacterial pathogens in the development of acute exacerbations of chronic bronchitis. In this study, H. parainfluenzae was isolated as the sole organism in 27 of the 81 collected sputum samples. Investigators found that levels of tumor necrosis factor α, interleukin 8, and neutrophil elastase in the 27 samples with H. parainfluenzae were similar to the levels in culture-negative sputum samples. However, the sputum samples of patients with H. influenzae and Moraxella catarrhalis did show increased levels of these same inflammatory markers. This suggests that, unlike other respiratory pathogens, H. parainfluenzae does not cause as intense an inflammatory response in patients with acute exacerbations of chronic bronchitis. However, it is important to note that one of the exclusion criteria to this study was the presence of immunosuppression.
Although the Sethi et al8 study demonstrates lack of a vigorous inflammatory response to H. parainfluenzae, other studies have found a significant immune response to the organism. A small case-control study by Mitchell et al9 investigated the immune response to H. parainfluenzae in 13 patients with COPD and 9 healthy controls. Investigators found that patients who were regularly colonized with H. parainfluenzae had higher serum levels of organism-specific immunoglobulin G titers than did the control subjects. The presence of this immune response suggests that H. parainfluenzae may have a pathogenic role in patients with COPD.
Despite the inconclusive evidence supporting the pathogenicity of H. parainfluenzae, our current case of empyema, and the prior cases listed in Table 1, demonstrate some evidence of the emerging virulence of the organism. Nearly all of these cases, including our own, occurred in the setting of immunosuppression. The clinical features of infections due to H. parainfluenzae resemble those infections due to H. influenzae.2,3,10 In a retrospective study in the 1980s by Rhind et al,10 it was shown that the clinical symptoms caused by both organisms are indistinguishable. The most frequently recorded symptoms included cough with or without sputum production, dyspnea, increased sputum purulence, and wheezing. Because of the similar symptoms, investigators concluded that H. parainfluenzae, when isolated from the respiratory tract, should be considered pathogenic in patients with symptoms of a respiratory tract infection. In addition to the comparable clinical features, both organisms can cause bronchopneumonic infiltrates and areas of consolidation on chest radiograph.6 Thus, neither clinical features nor roentgenographic patterns can be used to distinguish H. parainfluenzae pneumonia from H. influenzae or other bacterial pneumonias.
In the 1970s, ampicillin and chloramphenicol were the drugs of choice to treat infections due to H. parainfluenzae.2,3 However, even at that time, some strains were shown to be resistant to ampicillin possibly secondary to β-lactamase production by the organism.11 Recent susceptibility patterns investigated in multiple European surveys show significant ampicillin resistance of H. parainfluenzae in β-lactamase-producing organisms, with resistance rates of up to 22%.12,13 The Italian Epidemiological Study showed a significant increase in β-lactamase production by the organism from 5% in 1997 to 22% in 1999.13 However, in both of these studies, all isolates of H. parainfluenzae were susceptible to ciprofloxacin. This suggests that ampicillin should no longer be the drug of choice for treating H. parainfluenzae. Other agents, such as quinolones, may be more appropriate as first-line therapy for infections due to H. parainfluenzae.
1. Murphy T. Haemophilus
infections. In: Mandell GL, Bennett JE, Dolin R, eds. Mandell, Douglas, and Bennett's Principles and Practice of Infectious Diseases
. Vol 2. 6th
ed. Philadelphia, PA: Churchill Livingstone, Inc; 2005:2667-2668.
2. Cooney TG, Harwood BR, Meisner DJ. Haemophilus parainfluenzae
thoracic empyema. Arch Intern Med
3. Israel RH, Magnussen CR, Greenblatt DW, et al. Haemophilus parainfluenzae
lung abscess. Respiration
4. Pillai A, Mitchell JL, Hill SL, et al. A case of Haemophilus parainfluenzae
5. Hable KA, Logan GB, Washington GA. Three Haemophilus
species: pathogenic activity. Am J Dis Child
6. Oill PA, Chow AK, Guze LB. Adult bacteremic Haemophilus parainfluenzae
infections. Arch Intern Med
7. Smith CB, Kanner RE, Golden CA, et al. Haemophilus influenzae
and Haemophilus parainfluenzae
in chronic obstructive pulmonary disesase. Lancet
8. Sethi S, Muscarella K, Evans N, et al. Airway inflammation and etiology of acute exacerbations of chronic bronchitis. Chest
9. Mitchell J, Hill S. Immune response to Haemophilus parainfluenzae
in patients with chronic obstructive lung disease. Clin Diagn Lab Immunol
10. Rhind GB, Gould GA, Ahmad F, et al. Haemophilus parainfluenzae
and H. influenzae
respiratory infections: comparison of clinical features. BMJ
11. Mayo JB, McCarthy LR. Antimicrobial susceptibility of Haemophilus parainfluenzae. Antimicrob Agents Chemother
12. Orden MB, Martinez-Ruiz R, Millan PR. Haemophilus spp.
antimicrobial susceptibility in Health Area 6 in Madrid, Spain (2000-2004). Rev Esp Quimioter
13. Nicoletti G, Blandino G, Friscia O, et al. The Italian Epidemiological Survey 1997-1999. Antimicrobial susceptibility data of Haemophilus influenzae, Haemophilus parainfluenzae
and Moraxella catarrhalis
in Italy. Int J Antimicrob Agents