Patient Outcome

Clinical features of our patients are summarized in Table 1. Of the patients found to have cancer, two had positive bronchial cytologies, three had positive trans- bronchial lung biopsies, six had surgical confirmation, and two had positive bone scans with diagnostic bone biopsies. Only two of our patients were nonsmokers, and one was found to have adenocarcinoma of the bronchoalveolar cell type. Surprisingly, the majority of our cancer patients were found to have an advanced stage at time of diagnosis: stage I (N = 2), stage III (N = 4), and stage IV (N = 2). The only cancer patient with a marked elevation in slgA had stage III disease.

Cancerous Compared with Contralateral Noncancerous Lung

We attempted to find a focal increase in BAL slgA by obtaining samples from each lung despite the fact that all patients had unilateral nodules. Results are shown in Figure 1. There was not a statistically significant difference in either absolute or corrected slgA concentration between cancerous lung and con­tralateral noncancerous BAL samples obtained from patients eventually found to have lung cancer. kamagra soft tablets

Table 1—Study Patient Characteristics and Diagnoses

Lung Cancer

Noncancer

Patients

Patients

Mean age, у

65.6

67.3

Mean nodule size, cm

3.1

2.3

Percentage of smokers

87.5

83.3

Diagnoses

Adenocarcinoma

Nodule disappeared

(N = 6)

(N = 2)

Squamous cell

Lung abscess (N = 1)

carcinoma

Coccidioidomycosis

(N = 2)

(N = 1)

Mycobacterium avium- intraceUulare

(N = 1) No diagnosis (N = 1)

The 95 percent confidence interval (CI) for the ratio of the corrected BAL (slgA) between the cancerous and noncancerous lung in a cancer patient (CI = 1.76) suggests that if cancer is associated with increases in slgA levels, those increases are unlikely to be substan­tial and one can rule out doubling of the level.

Cancer Patients Compared with Noncancer Patients

If slgA levels are to be useful in diagnosing malig­nant neoplasms, we would need to observe measurable increases in those patients whose nodules were even­tually found to be malignant as compared with those with nonmalignant lesions. Such was not the case, as illustrated in Figure 2. Although the corrected slgA appears to be higher in the cancer patients, the difference is largely due to one markedly elevated value and is not statistically significant. The BAL “corrected” slgA for our noncancer patient whose condition remained undiagnosed was 3.74 on the side of the nodule and 2.36 on the contralateral side. buy cialis super active

FIGURE 1. A, upper: Absolute secretory IgA (slgA) in micrograms per milliliter from cancerous bronchoalveolar lavage (BAL) (33.3 ±8.8 |xg/ml) and from noncancerous contralateral BAL (25.0 ±6.6 ц-g/ml). B, lower: Corrected slgA from cancerous BAL (6.9 ±3.3) and from contralateral noncancerous BAL (5.1 ±1.1). Values from the same patient are connected by a line. Numbers in parentheses represent mean ± SEM shown in figure by horizontal line and shaded bar. Neither difference was statistically significant (p>. 1 in both cases).

If one were to arbitrarily designate that a “normal” value for slgA should not exceed the mean -1-2 SDs (as derived from those with noncancerous lesions), estimates of sensitivity, specificity, and predictive values can be made. For our data, a noncancer patient s slgA/albumin ratio should be ^6.4, and a “positive” test would be a BAL slgA/albumin >6.4. Using this value, measurement of corrected slgA has a 25 percent sensitivity, 100 percent specificity, positive predictive value of 1, and negative predictive value of 0.5. The ratio of positive tests to total tests performed was only 14 percent.

FIGURE 2. A, upper: Absolute secretory IgA (slgA) in micrograms per milliliter from bronchoalveolar lavage (BAL) done ipsilateral to a roentgenographs lesion eventually found to be cancer (33.3 ±8.8 M-g/ml) or noncancer (29.7 ± 10.4 g/ml). B, lower: Corrected slgA from cancer patient (6.9 ±3.3) and noncancer patient (3.8 ±0.5). Neither difference was statistically significant (p>. 1 in both cases).