Table 1 . Distribution of clinical characteristics of patients across the study groups
Clinical Attributes
Group 1 Lung Cancer
Group 2 Malignancy other than Lung Cancer
mentioned patient characteristics and CT features were compared between group 1 vs groups 2 and 3 combined . Chi-square test was used to determine statistical difference . A p value of less than 0.05 was considered statistically significant .
RESULTS
A total of 280 charts were reviewed excluding cases with incomplete information available . 49 of these patients had a lung mass ( 18 %). The median age of patients with a lung mass was 58 years ( IQR 52-67 ); about 35 % were females ; and 7 ( 14 %) were never-smokers . The median size of the mass was 5 cm .
Table 1 provides the distribution of clinical characteristics across each group . 31 of 49 ( 63 %) lung masses were primary lung cancer . Nine ( 18 %) cases represented malignancies which were not lung cancer . Five of these 9 masses were lymphoma . Benign tumors and infectious processes accounted for the 9 remaining cases .
Table 2 provides the comparison of the presence of specific CT features among the groups . These variables were compared between lung cancer and all other pathologies ( group 2 and 3 combined ). A mass that crossed anatomical boundaries on CT was 5.5 times more likely to represent lung cancer than any other etiology ( p = 0.01 ; 95 % CI , [ 1.47-20.86 ]). The relation between the presence of satellite lesions and the likelihood of the mass to be lung carcinoma did not achieve a statistical significance ( p = 0.08 ; OR 2.77 , 95 % CI [ 0.82-9.3 ]). Among the clinical attributes compared , age and
Group 3 Infectious / Benign
P-Value * Odds Ratio *
|
Median Age
58 ( IQR 52-67 )
|
58 ( 52-65 ) |
53 ( 32-68 ) |
64 ( 50-67 ) |
0.4829 |
1.52 ( 0.47-4.88 ) |
Smoking Status Never smoker |
0 |
4 |
3 |
|
|
Former smoker |
11 |
1 |
5 |
|
|
Current smoker |
13 |
1 |
0 |
|
|
Unknown |
7 |
3 |
1 |
|
|
Gender Female |
10 |
4 |
3 |
0.6383 |
1.33 ( 0.4-4.49 ) |
Male |
21 |
5 |
6 |
|
|
* p-values and Odds ratio were obtained by comparing lung cancer to all other pathologies combined ( group 1 |
vs group 2 and 3 ). Smoking status could not be compared as there were 0 ( zero ) never smokers in lung cancer |
group . IQR denotes Interquartile range . |
sex did not have any correlation with the pathology of the mass .
DISCUSSION
How often is a lung mass a lung cancer : the overall outlook
Size is one of the most important predictors of malignancy in a lung nodule . 5 , 7-8 The presence of a lung mass , which , by definition is more than 3 cm in size , is very concerning for malignancy . However , much of our knowledge about prevalence of lung cancer among lung masses stems from older studies . 3-4 Greater than 90 % of lung masses were lung cancer in both of these studies . The large reduction in proportion of lung masses representing lung cancer in our study compared to these studies from the 1980s could be explained by the significant decrease in smoking rates . 9 Another significant change in epidemiology of lung mass pathologies has been a higher rate of infectious processes . Human immunodeficiency virus ( HIV ) infection and immunosuppression are significant risk factors for many infectious processes that could present as nodules or masses . 10 Some infectious processes can lead to development of large lung masses even in an immunocompetent host . 11-12 About every one in five lung masses were benign or infectious in our study . The histologic appearances of some of these masses from our study are illustrated in figures 1a-c . Most benign and infectious masses had good outcomes with resection and appropriate antimicrobial therapy respectively . Also , most of the cases of non-lung cancer malignancies in our study were lymphomas , which generally have a better prognosis compared to primary lung carcinoma . These findings somewhat reduce the grim outlook usually associated with large lung nodules and masses . The concept of “ false despair ,” which has been described mostly after diagnosis of a terminal illness , applies equally well to the phase of diagnostic workup . As opposed to the generalization “ a lung mass is almost always malignant ,” knowledge of this data will aid pulmonary physicians and internists to appropriately counsel their patients until the pathological diagnosis is established and hopefully reduce patient anxiety .
Table 2 . Comparison of the presence of specific computed tomography ( CT ) features across the groups
CT Findings
Group 1 Lung Cancer
Group 2 Malignancy other than Lung Cancer
Group 3 Infectious / Benign
P-Value * Odds Ratio *
|
Median Size in cm
5 ( IQR 4-6 )
|
6 ( 4-7 ) |
5 ( 4-6 ) |
4 ( 4-5 ) |
|
|
Air Bronchogram |
10 |
3 |
3 |
0.66 |
0.95 ( 0.28-3.28 ) |
Central Cavitation |
8 |
2 |
3 |
0.69 |
0.9 ( 0.25-3.35 ) |
Central Necrosis |
15 |
4 |
3 |
0.37 |
1.47 ( 0.45-4.78 ) |
Crosses Anatomical Boundaries
19 2 2 0.01
5.54 ( 1.47-20.86 )
Lymphadenopathy |
24 |
5 |
8 |
0.47 |
1.32 ( 0.35-4.99 ) |
Pleural Effusion |
9 |
4 |
4 |
0.92 |
0.5 ( 0.15-1.72 ) |
Satellite Lesions |
18 |
2 |
4 |
0.08 |
2.77 ( 0.82-9.3 ) |
Well-Defined Borders |
13 |
6 |
5 |
0.94 |
0.46 ( 0.14-1.5 ) |
* p-values and Odds ratio were obtained by comparing lung cancer to all other pathologies combined ( group |
1 vs group 2 and 3 ). IQR denotes Interquartile range . |
NUMBER 9 MARCH 2018 • 211