Four hundred fifty-four cases were identified from the Marshfield Clinic Tumor Registry. These cases had been diagnosed between October 1990 and August 1992. Of these cases, the initial chest radiograph was available in 345 (76%) instances, and these formed the basis of our analysis.
The population included 116 (34%) female patients. The age range was 37 to 88 years with a median of 68 years. Frequency of each cell type is shown in Table 2. In the Marshfield Clinic series, adenocarcinoma was the most frequent type, followed by squamous cell, small cell, and large cell carcinoma. These frequencies are in marked contrast to the Mayo Clinic series. After grouping the infrequent bronchoalveolar and ade-nosquamous cases into an “other’ categoiy, the differences in relative frequencies of the cell types for the two series were highly significant (p<0.001).
Table 3 is a statistical comparison between Mayo Clinic and Marshfield Clinic for relative frequency by cell type and gender. In both men and women, there is a significant change in distribution of cell types. Adenocarcinoma has an increased frequency in men, whereas there is a relative increase in both squamous cell and small cell carcinoma in women.
The radiographic patterns for the four principal cell types from the Marshfield Clinic series are summarized in Table 4. Percents do not total 100% due to a number of cases showing more than one radiographic pattern. The definitions of the “central mass only” and “central mass plus” classifications are given in the “Materials and Methods” section above.
Table 5 compares the principal radiographic findings by cell type between Marshfield Clinic and Mayo Clinic. Adenocarcinoma showed an increased frequency of central mass compared with the Mayo Clinic series. This was true for the group with central mass as the only abnormality (14% Marshfield vs 5% Mayo; p=0.008) as well as all central masses combined (49% Marshfield vs 17% Mayo; p<0.001). Squamous cell was the only other cell type to show an increased frequency of central masses, but only for the “central mass plus” group (61% Marshfield vs 40% Mayo; p<0.001). This group contains both primary central tumors and peripheral primaries with lymphadenopathy. No significant differences were noted for small cell or large cell for frequency of presentation as central mass compared with the Mayo Clinic data.
Adenocarcinoma had a striking decrease in frequency of presentation as a parenchymal mass (45% Marshfield vs 71% Mayo; p<0.001). When parenchymal and apical masses were combined, adenocarcinoma showed a decreased frequency of presentation as a peripheral primary tumor (49% Marshfield vs 72% Mayo; p<0.001). Squamous cell showed significant change in the opposite direction for both parenchymal masses (39% Marshfield vs 28% Mayo; p=0.046) and all peripheral primary tumors (43% Marshfield vs 31% Mayo; p=0.033). There was no important change for small cell and large cell with regard to central and peripheral masses in comparing the Marshfield Clinic with the Mayo Clinic series.
Finally, Table 6 is a summary of the Marshfield Clinic data comparing radiographic site of origin for all four cell types. Central tumors included the following: mediastinal mass or adenopathy, hilar mass or adenopathy, perihilar mass, and obstructive findings. All of these cases had no associated parenchymal or apical mass. Peripheral origin included the cases with parenchymal or apical masses. Fourteen cases could not be classified as to site of origin. We were not able to abstract information from the published Mayo Clinic study to form an older comparison group for the central primary tumors. However, from classifiable Marshfield Clinic patients, we find no statistically significant difference between adenocarcinoma and squamous cell with regard to peripheral vs central sites of origin (p=0.35).
Table 1—Principal Radiographic Findings Recorded on Data Forms
|2.||Location: hilar, perihilar (within 4 cm of hilum), parenchymal, apical|
|3.||Hilar adenopathy or mass|
|4.||Mediastinal adenopathy or mass|
|6.||Obstructive findings: atelectasis, pneumonia, mucoid impaction|
Table 2—Comparison of Cell Types at Marshfield and Mayo Clinics
|Cell Type||rMarshfield (n=345)||IMayo (n=600)|
|Large cell||22 (6)||97|
Table 3—Comparison of Cell Types at Mayo and Marshfield Clinics by Gender
|1Mayo, No. (%)||iMarshfield,
|1Mayo, No. (%)||lMarshfield, No. (%)|
|Squamous||251 (48.5)||75 (34.4)||12 (14.6)||23 (20.4)|
|Adenocarcinoma||74 (14.3)||77 (35.3)||52 (63.4)||48 (42.5)|
|Small cell||105 (20.3)||50 (22.9)||9(11.0)||36 (31.9)|
|Large cell||88 (17.0)||16 (7.3)||9(11.0)||6 (5.3)|
|Total||518 (100.0)||218(100.0)||82 (100.0)||113 (100.0)|
Table 4—Radiographic Patterns for the Four Principal Cell Types in Marshfield
|Adenocarcinoma(n=125)||Squamous Cell (n=98)||Small Cell (n=86)||Large Cell (n=22)|
|Parenchymal mass||56 *||38||29||15|
|Central mass only||18||17||11||3|
|Central mass plus||61||60||66||9|
|or adenopathy Normal||3||4||0||0|
Table 5—Comparison of Principal Radiographic Findings by Cell Type
|RadiographicPattern||Adenocarcinoma||Squamous||Small Cell||Large Cell|
|Central mass only|
|Central mass plus|
|Peripheral primary tumor|
Table 6—Site of Origin and Cell Type at Marshfield Clinic
|Adenocarcinoma (n=125)||57 *||61||7|
|Squamous cell (n=98)||51||42||5|
|Small cell (n=86)||52||32||2(2)|
|Large cell (n=22)||7||15||0(0)|