[Computed tomographic appearance of pulmonary mucormycosis]

Has the mortality from pulmonary mucormycosis changed over time? A systematic review and meta-analysis

OBJECTIVES

Pulmonary mucormycosis (PM) is increasingly being reported in immunocompromised patients and has a high mortality. Our aim was to assess the mortality of PM and its trend over time. We also evaluated the role of combined medical-surgical therapy in PM.

METHODS

We performed a systematic review of Pubmed, Embase, and Cochrane central databases. Studies were eligible if they described at least five confirmed cases of PM and reported mortality. We also assessed the effect of combined medical-surgical therapy versus medical treatment alone on PM mortality. We used a random-effects model to estimate the pooled mortality of PM and compared it across three time periods. The factors influencing mortality were assessed using meta-regression. We evaluated the risk difference (RD) of death in the following: subjects undergoing combined medical-surgical therapy versus medical therapy alone, subjects with isolated PM versus disseminated disease, and PM in diabetes mellitus (DM) versus non-DM as a risk factor.

RESULTS

We included 79 studies (1544 subjects). The pooled mortality of PM was 57.1% (95% confidence interval [CI] 51.7-62.6%). Mortality improved significantly over time (72.1% versus 58.3% versus 49.8% for studies before 2000, 2000-2009, and 2010-2020, respectively, p 0.00001). This improved survival was confirmed in meta-regression after adjusting for the study design, the country's income level, and the sample size. Combined medical-surgical therapy was associated with a significantly lower RD (95%CI) of death: -0.32 (-0.49 to -0.16). The disseminated disease had a higher risk of death than isolated PM, but DM was not associated with a higher risk of death than other risk factors.

CONCLUSIONS

While PM is still associated with high mortality, we noted improved survival over time. Combined medical-surgical therapy improved survival compared to medical treatment alone.

[Pneumonia in immunosuppressed patients]

CLINICAL/METHODICAL ISSUE

Pulmonary infections are a common complication in immunosuppressed patients with a frequently fatal prognosis despite modern prophylactic therapy. An early and correct diagnosis is important for initiation of the appropriate therapy.

STANDARD RADIOLOGICAL METHODS

Chest radiography is the preferred initial imaging examination but is not accurate enough for the detection of pulmonary infections in immunosuppressed patients.

METHODICAL INNOVATIONS

Pneumonia is caused by a broad spectrum of pathogens in immunocompromised patients. In addition to imaging, the clinical history and epidemiology also play an important role in the diagnostics.

PERFORMANCE

Using epidemiological and anamnestic information, computed tomography (CT) shows a significantly better sensitivity and specificity particularly for the diagnosis of atypical forms of pneumonia. Due to the exact imaging of the different infiltration patterns CT provides an increased sensitivity with respect to the etiological classification of pulmonary infections.

PRACTICAL RECOMMENDATIONS

This article reviews in particular the radiological findings of commonly occurring pulmonary infections in immunosuppressed patients.

Comparison of computed tomographic findings in pulmonary mucormycosis and invasive pulmonary aspergillosis

Because there are no available molecular markers for pulmonary mucormycosis (PM), which has low culture sensitivity, early diagnosis and treatment rely heavily on imaging modes such as computed tomography (CT). However, there are limited data comparing CT findings for PM with those for invasive pulmonary aspergillosis (IPA). Adult patients who met the modified criteria for proven and probable PM (over an 11-year period) and IPA (over a 6-year period, owing to the availability of the galactomannan assay) according to the modified European Organization for Research and Treatment of Cancer/Mycosis Study Group definitions were retrospectively enrolled. IPA cases were selected at a 1 : 4 (PM/IPA) ratio. Thoracic CT scans were reviewed by two experienced radiologists blinded to the patients' demographics and clinical outcomes. A total of 24 patients with PM, including 20 (83%) with proven PM and four (17%) with probable PM, and 96 patients with IPA, including 12 (13%) with proven IPA and 84 (87%) with probable IPA, were eventually analysed. The reverse halo sign was more common in patients with PM (54%) than in those with IPA (6%, p < 0.001), whereas some airway-invasive features, such as clusters of centrilobular nodules, peribronchial consolidations, and bronchial wall thickening, were more common in patients with IPA (IPA 52% vs. PM 29%, p 0.04; IPA 49% vs. PM 21%, p 0.01; IPA 34% vs. PM 4%, p 0.003, respectively). The reverse halo sign was more common, and airway-invasive features were less common, in patients with PM than in those with IPA. These findings may help physicians to initiate Zygomycetes-active antifungal treatment earlier.

Predictors of pulmonary zygomycosis versus invasive pulmonary aspergillosis in patients with cancer

BACKGROUND

Pulmonary zygomycosis (PZ), an emerging mycosis among patients with cancer, has a clinical manifestation similar to that of invasive pulmonary aspergillosis (IPA). Most cases of PZ in such patients develop as breakthrough infections if treatment with antifungal agents effective against Aspergillus species is administered. However, clinical criteria to differentiate PZ from IPA are lacking.

METHODS

We retrospectively reviewed the clinical characteristics and computed tomography (CT) findings for 16 patients with cancer and PZ and for 29 contemporaneous patients with cancer and IPA at the time of infection onset (2002-2004). Patients with mixed infections were excluded. Parameters predictive of PZ by univariate analysis were included in a logistic regression model.

RESULTS

Almost all patients with PZ (15 of 16) and IPA (28 of 29) had underlying hematological malignancies and typical risk factors for invasive mold infections. In logistic regression analysis of clinical characteristics, concomitant sinusitis (odds ratio [OR], 25.7; 95% confidence interval [CI], 1.47-448.15; P = .026) and voriconazole prophylaxis (OR, 7.76; 95% CI, 1.32-45.53; P = .023) were significantly associated with PZ. The presence of multiple (> or = 10) nodules (OR, 19.8; 95% CI, 1.94-202.29; P = .012) and pleural effusion (OR, 5.07; 95% CI, 1.06-24.23; P = .042) at the time that the patient underwent the initial CT were both independent predictors of PZ in the logistic regression analysis of radiological parameters. No difference occurred in the frequency of other CT findings suggestive of pulmonary mold infections (e.g., masses, cavities, halo sign, or air-crescent sign) between the 2 patient groups.

CONCLUSIONS

PZ in immunocompromised patients with cancer could potentially be distinguished from IPA on the basis of clinical and radiological parameters; prospective validation is needed.