Din aceeași categorie

Oncolog-Hematolog.ro

Tratamentul carcinomului renal metastatic cu terapiile TKI şi mTOR: prezentare de caz

Oncolog-Hematolog.ro

De actualitate în cancerul pulmonar

Oncolog-Hematolog.ro

Tratamentul cancerului gastric în spitalele judeţene, între standardul de aur şi practica clinică

Introduction

The etiology of pericardial effusion is various^(1,3). There are two main categories of factors that can determine the abnormal intrapericardial accumulation of fluid: infectious and non-infectious causes⁽¹⁾. In a patient with known malignancy, beside metastatic tumoral pericardial involvement, there are other factors that can cause pericardial effusion: chemotherapy and radiotherapy, infections, lymphatic tumoral obstructions or tumoral bleedings^(2,4). CT examinations represent the main investigation useful in this pathology, being required generally for the evaluation of the tumoral extent and for a superior evaluation of loculated and hemorrhagic pericardial effusion^(5,6). Establishing the cause of pericardial effusion has implications in therapy and in patient’s outcome, therefore it is important to determine the most relevant CT imaging biomarkers for different types of pericardial pathologies, focused on oncologic patients^(1,2).

Materials and method

We made a retrospective cross-sectional study, between January 2011 and March 2022, for a period of 11 years and 2 months, and we selected 47 patients with different pathologies and associated pericardial effusions detected on CT examinations, 14 patients with tumoral pericarditis, 20 cases with hemopericardium – 17 of them with recent cardiovascular surgery, two cases of infectious pericarditis, and 11 patients with inflammatory pericarditis. The inclusion criteria were patients with pericardial effusion detected at CT evaluations, with a confirmed diagnosis by cytology and histopathologic examinations, patients with tumoral pericardial mass detected at CT and patients with recent cardiac surgery (less than seven days). All CT evaluations were performed on 16, 64 and 256 multidetector CT scanners. The imaging protocol included one non-enhanced phased (NECT) and two phases post-contrast media administration (arterial phase at 25 seconds and delayed phased at 100 seconds). Image analysis included the following elements: type of pericardial effusion based on attenuation values measured on NECT examination (transudate ≤10 UH, exudate >10 UH, hemopericardium >60 UH or density value above that of the paravertebral muscle)^(1,7), the distribution of the pericardial effusion (circumferential, loculated), the presence and type of pericardial thickening (more than 3 mm thickness, nodular/smooth), the presence of tumoral pericardial mass, mediastinal lymphadenopathy, pulmonary lesions (metastases, consolidations) and pleural effusion.

Statistical analysis. Each CT imaging biomarker was assessed from the point of view of diagnostic performance in etiological search by sensitivity, specificity, using receiver operating characteristic curves (ROC). Fisher Exact Test was performed to analyze statistical associations and the Mann-Whitney test was used to test the differences between the groups of patients. The area under the ROC curve (AUC) values were tested against the 0.5 reference value (no determination). A binary logit model with classification table was used to measure the specificity and sensitivity of binary variables and respectively to assess the capacity of factors to contribute to a correct classification of cases. Specific characteristics of pericardial effusion, like the density of pericardial effusion (DPE), tumoral mass and the presence of mediastinal lymphadenopathy, were used as factors in the tested models.

In each type of statistical analysis, a p-value below the 0.05 threshold suggested a statistical significance.

Graphical representation and statistical analysis were performed in Microsoft Excel, IBM SPSS Statistics for Windows, version 21 (IBM Corp., Armonk, N.Y., USA) and MedCalc® Statistical Software version 20.112 (MedCalc Software Ltd, Ostend, Belgium; https://www.medcalc.org; 2022).

Results

We studied 47 patients (35 men, 12 women; age range 18-83 years old, with a mean of 54.8 years old) with pericardial effusions detected at CT examinations. In our study, 20 of 47 patients presented pericardial effusion with density value more than 30 UH, 17 of them with recent cardiovascular surgery and one patient with tumoral pericardial involvement. The median value of pericardial effusion for each entity is presented in Figure 1.
 

A patient with pericardial effusion with a density value higher than 30 HU is more likely to have a postoperative hemopericardium (Case 1) compared with tumoral pericarditis (Figure 2), with a cut-off point of 31 HU (Fi­gure 3). Applying the Mann-Whitney test, we found no statistical difference (p=0.169) between inflammatory pericarditis groups (Case 2) by comparison with tumoral pericarditis (Case 3).
 

The curves drawn in Figure 3 reveal that, in our case, the cut-off point is estimated to be around 30 HU. Lower values of density suggest a higher sensitivity, while values above 30 suggest higher specificity. Thus, the values of DPE below 30 HU suggest a higher probability to identify true positive cases.
 

In the case of the diagnosis of inflammatory pericarditis versus tumoral pericarditis, the density of pericardial effusion has a lower chance (65.9%) to be used as a differential factor.

As shown in Figure 4, from a statistical point of view, the density value of pericardial effusion cannot be considered a significant factor in the differential diagnosis between tumoral and inflammatory pericarditis (p=0.166).
 

The main CT imaging findings in patients diagnosed with tumoral pericarditis (Case 3) are presented in Fi­gure 5.
 

Loculated pericardial effusion was statistically associated only with postoperative hemopericardium (p<0.005) – Case 1. There were no statistically associations between isolated pericardial thickening, the presence of a large pericardial effusion, pulmonary consolidations or metastases, with either postoperative hemopericardium, tumoral and inflammatory pericarditis in our study.

Mediastinal lymphadenopathies were found in 20 out of 47 patients, 13 of them with tumoral pericardial lesions – Case 3. Regarding the association between lymphadenopathy and the high-density value of pericardial effusion (above 30 HU), we found a highly sensitive diagnostic value for tumoral pericardial involvement – 85.7%, and specificity of 61.5%. Using high DPE as single factor in a classification model, we achieved a 70.4% probability of correct classification. When the presence of lymphadenopathy was added along the DPE, the percentage of correct classified cases increased to 74.1%.

When adding tumoral mass as factor along the DPE, both sensitivity and specificity increased from 71.4% and 69.2% to 78.6% and, respectively, 100%. Moreover, the overall percentage of correct classified cases jumped from 70.4% to 88.9%, showing thus the importance of tumoral mass factor in the diagnosis of tumoral pericarditis versus inflammatory pericarditis.

In our patients, the combination between a large pericardial effusion (more than 2 cm) and a tumoral mass had a specificity of 100% and a sensitivity of 78.6% for the diagnosis of tumoral pericarditis versus inflammatory pericarditis (Case 3).

Discussion

Tumoral pericardial mass, nodular or irregular pericardial thickening, moderate-to-large pericardial effusions and hemopericardium represent the most common imaging findings in the case of a patient with neoplastic pericardial involvement^(4,8).

Regarding the density value of pericardial effusion, more specific, a high-density level correlated well with postoperative hemopericardium (Case 1), but when there was an association with lymphadenopathy or a pericardial tumoral mass (Case 3), we found a high sensitivity diagnostic value for tumoral pericardial involvement (85.7%) and a specificity of 61.5%.

In a study from 2010, Sun et al. found a good correlation between the association of pericardial thickening and mediastinal lymphadenopathy with malignant pericardial effusion⁽⁹⁾. In our study, pericardial thickening alone was not associated with tumoral pericardial involvement (Case 2), one reason being an increased number of patients (10/14) with large tumoral masses.

Although tumors represent the main cause of large pericardial effusions⁽¹⁰⁾, in our study there was no statistically significant correlation between the type of pericarditis and the size of the pericardial effusion, seven out of 14 patients with tumoral pericarditis (50%), respectively seven out of 17 patients with postoperative hemopericardium (41%) presenting large quantity of pericardial fluid (Cases 1 and 3).

In our study, we found no correlation between pulmonary lesions or pleural effusion and the type of pericardial involvement (Cases 2 and 3), even in cases with lung metastases. This is due to particularities of the cases: an increased number of patients with postoperative CT evaluations with lung consolidations and pleural effusion, respectively only one of 14 patients with tumoral pericardial involvement and lung metastases.

The limitation of this study is represented by the small number of patients included. The conclusions require a larger number of patients with different types of pericardial effusion.

Conclusions

The high density level of pericardial effusion and mediastinal lymphadenopathies or pericardial tumoral mass represent the best imaging biomarkers in our study, statistically significant for tumoral pericarditis. Although cytological analysis represents the gold standard method to exclude tumoral pericardial effusions, CT imaging biomarkers can be an effective tool in the etiological search, providing different elements that can be used to guide clinicians in the diagnostic work-up of this entity (the type and distribution of the pericardial effusions, the presence of a pericardial tumoral mass, mediastinal and lung lesions).  

Conflicts of interests: The authors declare no conflict of interests.