Ann Thorac Surg 1995;59:568-572
© 1995 The Society of Thoracic Surgeons
Expression of Blood Group Antigen A by Stage I Non–Small Cell Lung Carcinomas
Thomas W. Rice, MD,
Raymond R. Tubbs, MD,
Gerald A. Hoeltge, MD,
Thomas J. Kirby, MD,
David P. Meeker, MD,
Sharon V. Medendorp, MPH,
Ronald M. Bukowski, MD
Departments of Thoracic and Cardiovascular Surgery, Clinical Pathology, Hematology and Medical Oncology, Pulmonary and Critical Care Medicine, and Biostatistics and Epidemiology, The Cleveland Clinic Foundation, Cleveland, Ohio
 |
Abstract
|
|---|
To clarify the significance of blood group antigen A (BAA) expression by neoplastic cells, we studied patients who had curative resections of stage I non–small cell lung carcinomas. Immunohistochemical staining using monoclonal antibodies was used to detect BAA expression by paraffin-embedded carcinoma cells. One hundred three patients were studied; mean age was 62.6 years, and 70 (68%) were male. Histologic types were as follows: adenocarcinoma, 52 (50.5%); squamous cell, 25 (24.3%); large cell, 24 (23.3%); and adenosquamous, 2 (1.9%). Histologic grades were as follows: I, 13 (12.6%); II, 26 (25.3%); and III, 64 (62.1%). All patients had American Joint Committee on Cancer stage I tumors: 65 patients (63.1%) had T1 tumors, and 38 (36.9%) had T2 tumors. Recurrences developed in 25 (24.3%) and metachronous malignancies in 4 (3.9%). Survival was 75% ± 4.8% at 3 years and 66.6% ± 7.5% at 5 years. Eighty-nine patients (86.4%) were blood group A and 14 (13.6%) were AB. Ninety-five (92.2%) were secretors of BAA and 8 (7.8%) were not. The expression of BAA by neoplastic cells was not detectable in 34 (33%), trace (1% to 5% of neoplastic cells) in 10 (9.7%), 1+ (6% to 25%) in 8 (7.8%), 2+ (26% to 50%) in 12 (11.7%), 3+ (51% to 75%) in 12 (11.7%), and 4+ (76% to 100%) in 27 (26.2%). The pattern of neoplastic cell staining was homogeneous in 14 patients (20.3%) and heterogenous in 55 (79.7%). Carcinoma recurrence, overall survival, and event-free survival were not related to secretor status, BAA expression, or pattern of staining. Based on recurrence and survival analyses we conclude that secretor status, expression of BAA, and pattern of staining were not of prognostic significance.
|
Introduction
|
|---|
The most important predictor of survival for patients with non–small cell lung carcinoma is the stage of the disease at diagnosis and treatment. However, within a staging group there is no means of identifying those patients who are at risk for development of local recurrences, distant metastases, or metachronous malignancies. Additional prognostic factors must be found so that within a staging group, patients who have a poor prognosis may be identified and have their therapy modified and their outcome improved.
The expression of blood group antigens in normal epithelium and loss or reduction of this expression in neoplastic cells has been documented in many tissues [1–6]. This phenomenon also has been found in lung carcinomas [7, 8]. It has been reported that the loss of tumor expression of blood group antigens in patients with lung carcinomas may be associated with a poor prognosis [9, 10]. In a recent report, loss of expression of blood group antigen A (BAA) in patients with stage I and stage II non–small cell lung carcinomas was found to be an independent predictor of reduced survival, and loss of expression of BAA was associated with a more rapid disease progression (recurrence, secondary primary tumors, or death due to tumor or unknown causes) [11]. Because patients with stage I non–small cell lung carcinomas have the best prognosis of all stages, a survival difference should be most apparent in this group of patients. To clarify the significance of BAA expression by neoplastic cells, we studied blood group A and AB patients who underwent curative resections of stage I non–small cell bronchogenic carcinomas.
|
Patients and Methods
|
|---|
Patient Selection and Analysis
Patients seen at the Cleveland Clinic Foundation between January 1, 1986, and July 1, 1991, who underwent a curative resection of a pathologic stage I primary non–small cell lung carcinoma and who were blood group A or AB were included in the study. Patients who had received induction treatment before resection of the primary carcinoma or who had a previous malignancy were excluded from the study.
The patient's sex, age at operation, date and type of operative procedure, date and location of recurrence, and date of and disease status at last follow-up or death were recorded. The date and location of metachronous malignancy also were recorded. All patients were followed up prospectively at the Cleveland Clinic Foundation or by the referring physician. A standard follow-up visit consisted of a history, a physical examination, and a chest roentgenogram, conducted 6, 12, 18, 24, 36, 48, and 60 months postoperatively. Typically a hemogram and SMA-16 were drawn annually.
Histologic cell type of the carcinomas was classified according to World Health Organization criteria [12]. Tumors were divided into one of three histologic grades: I (well differentiated), II (moderately differentiated), and III (poorly differentiated). Undifferentiated large cell carcinomas all were classified as grade III. All tumors were American Joint Committee on Cancer pathologic stage I tumors [13]. The classification of T1 and T2 tumors was based on the size of the tumor, endobronchial position of the tumor, and visceral pleural invasion.
Pathologic Analysis
Immunohistologic analysis was performed using the labeled streptavidin biotin system. The procedure was automated using an automated immunostainer (Ventana 320; Ventana Medical Systems, Tucson, AZ). Four-micrometer sections were prepared and affixed to electrostatically charged slides. After deparaffinization and rehydration in xylene and graded alcohols, endogenous peroxidase was blocked with methanolic hydrogen peroxide, and nonspecific immunoglobulin binding sites were saturated with normal rabbit serum. Mouse monoclonal anti-blood group A (clone 81 FR2.2), an immunoglobulin M class murine monoclonal antibody, was incubated with the section at a concentration of 0.02 mg/mL. Negative control consisted of equivalently diluted nonimmune mouse immunoglobulin M. Internal positive controls were used in this study, consisting of complete labeling of capillary and other vascular endothelium, erythrocytes, and normal pneumocyte, bronchial epithelium, and bronchiolar epithelium. Secretory status was defined as the presence or absence of blood group A expressed in bronchial or bronchiolar epithelium. Sections were incubated sequentially with biotinylated anti-mouse immunoglobulin G1 streptavidin-horseradish peroxidase, the chromogenic substrate, 3, 3-diaminobenzadine/H2O2, and hematoxylin counterstaining, with intervening buffer washes. All steps in the procedure were automated using the immunostainer (Ventana 320). The percentage of neoplastic cells expressing blood group antigen A was determined by light microscopy by one pathologist (R.R.T.). Interpretative criteria were as follows: 0, no tumor cells positive; trace, 1% to 5% of tumor cells positive; 1+, 6% to 25% of tumor cells positive; 2+, 26% to 50% of tumor cells positive; 3+, 51% to 75% of tumor cells positive; and 4+, 76% to 100% of tumor cells positive.
Statistical Analysis
The 
2, exact 2, and Kruskal-Wallis tests were used to examine associations among factors [14]. Survival time, time to recurrence, and event-free survival (disease progression was defined as death, recurrence, or metachronous primary malignancy) [11] were calculated from the date of the surgical procedure to the event of interest or last follow-up. Survival, recurrence, and event-free distributions were estimated using the Kaplan-Meier method for categorical factors [15] and compared using the logrank test where appropriate [16]. Results with p values less than 0.05 were considered statistically significant.
|
Results
|
|---|
The study population consisted of 103 patients. Patient and carcinoma characteristics are listed in Table 1
. Ninety-five patients (92.2%) were secretors of BAA, and 8 patients (7.8%) were not (Figs 1, and 2). Thirty-four patients (33.0%) did not have expression of BAA by neoplastic cells, 10 patients (9.7%) had trace expression, and 59 patients (57.3%) had 1+ to 4+ expression of BAA by neoplastic cells (Table 2). The pattern of neoplastic cell staining in the 69 patients with BAA expression was homogeneous in 14 patients (20.3%) and heterogenous in 55 (79.7%) (Figs 3, 4). Patients with no or trace expression of BAA had fewer grade I tumors (2.3% versus 20.3%; p = 0.008) and more large cell undifferentiated tumors (36.4% versus 13.6%; p = 0.004) (Table 3). Sex, blood group, and T status did not differ significantly by BAA expression.
View larger version (2K):
[in this window]
[in a new window]
|
Fig 1. . Photomicrograph of adenosquamous carcinoma. Tissue immunostained for blood group antigen A demonstrates absence of staining by the neoplastic cells. Absence of blood group antigen A expression by nonneoplastic bronchiolar epithelium (small arrows) indicates that this patient is a nonsecretor of blood group antigen A. Staining of endothelium and red blood cells (large arrows) serves as an internal control. (Avidin biotinylated peroxidase technique using diaminobenzidine/H2O2 as chromogenic substrate and hematoxylin counterstaining; x400 before 50% reduction.)
|
|
View larger version (2K):
[in this window]
[in a new window]
|
Fig 2. . Photomicrograph of moderately to poorly differentiated adenocarcinoma. Tissue immunostained for blood group antigen A demonstrates absence of staining by the neoplastic cells. Blood group antigen A expression by nonneoplastic bronchiolar epithelium (arrows) indicates this patient is a secretor of blood group antigen A. (Avidin biotinylated peroxidase technique using diaminobenzidine/H2O2 as chromogenic substrate and hematoxylin counterstaining; x400 before 50% reduction.)
|
|
View larger version (2K):
[in this window]
[in a new window]
|
Fig 3. . Photomicrograph of poorly differentiated adenocarcinoma. Tissue immunostained for blood group antigen A demonstrates 4+ heterogeneous staining by the neoplastic cells. (Avidin biotinylated peroxidase technique using diaminobenzidine/H2O2 as chromogenic substrate and hematoxylin counterstaining; x400 before 50% reduction.)
|
|
View larger version (2K):
[in this window]
[in a new window]
|
Fig 4. . Photomicrograph of large cell undifferentiated carcinoma. Tissue immunostained for blood group antigen A demonstrates 4+ homogeneous staining by the neoplastic cells. (Avidin biotinylated peroxidase technique using diaminobenzidine/H2O2 as chromogenic substrate and hematoxylin counterstaining; x400 before 50% reduction.)
|
|
Follow-up was complete as of August 1, 1993. Sixty-nine patients (67.0%) were alive at follow-up. Median follow-up of survivors was 47.6 months (range, 26.0 to 88.2 months). Of 103 patients, 25 (24.3%) had a documented recurrence of their carcinoma. The recurrence rates at 1, 3, and 5 years were 7.3% ± 2.6%, 19.3% ± 4.1%, and 28.9% ± 5.5%, respectively. Recurrence of the carcinoma was not related to BAA expression (p = 0.84), blood group (A or AB) (p = 0.70), T status (p = 0.31), or histologic cell type (p = 0.22). Four patients (3.9%) had metachronous malignancies (1 leukemia, 1 myeloma, 1 ovarian, 1 bronchogenic).
Overall survival for all patients was 75.0% ± 4.8% at 3 years and 66.6% ± 7.5% at 5 years. Event-free survival was 71.6% ± 4.48% at 3 years and 58.6% ± 5.64% at 5 years. Expression of BAA by carcinoma cells was not a significant predictor of overall survival (p = 0.57) or event-free survival (p = 0.54) (Fig 5). This was true for each level of expression and for the following groupings: no expression versus any expression, trace and 1% to 5% versus 6% to 100%, and the combinations less than 5%, 6% to 25%, 26% to 50%, and greater than 51%. Overall and event-free survival were not influenced by expression of BAA when nonsecretors were included or excluded. Secretor status was not of prognostic significance (p = 0.75 overall, p = 0.92 event-free) and was not related to T (p = 1.0). The pattern of staining of carcinoma cells was not a prognosticator (p = 0.14 overall, p = 0.18 event-free).
Outcome was not related to any patient or carcinoma characteristics except T status. The overall 5-year survival for T1 N0 and T2 N0 patients was 72.3% ± 9.0% and 57.2% ± 13.2%, respectively (p = 0.06). The event-free survival for T1 N0 and T2 N0 patients was 70.1% ± 6.15 and 41.0% ± 9.62% (p = 0.031) (Fig 6). There was a trend for a worse overall survival for large cell carcinomas than for squamous cell carcinomas or adenocarcinomas (p = 0.06), but this trend was not seen in event-free survival (p = 0.17). Patients of AB blood group had a worse overall survival than patients of blood group A (p = 0.04); however, this trend was present but not statistically significant in event-free survival (p = 0.11).
|
Comment
|
|---|
The loss of blood group antigen expression by lung carcinomas is a well-known phenomenon [7, 8]. The loss of expression and its influence upon survival vary widely in different studies. The reported significance of loss of expression in non–small cell lung carcinoma is inconsistent and in our experience is of no predictive value.
Ogawa and colleagues [9] in a study of 83 patients with resected adenocarcinomas and squamous cell carcinomas of the lung divided the study patients into two groups, those with a good prognosis (survival of more than 4 years) and those with a poor prognosis (survival of less than 2 years). The loss of ABH antigen was seen in 12% (4/33) of good prognosis patients and 38% (19/50) of poor prognosis patients (p < 0.01). However, in the good prognosis group 18% of patients and in the poor prognosis group 14% of patients had tumor antigen expression incompatible with the patients' blood group. All stages were included (25 stage I, 8 stage II, 42 stage III, and 8 stage IV); however, loss of ABH antigen was predictive of reduced survival only in stage I patients. This group consisted of only 25 patients of all blood types: 7 without ABH expression, 3 with incompatible blood group expression, and 15 with ABH expression. Thirteen percent (2/16) of the good prognosis group lost ABH antigen in tumors and 56% (5/9) of the poor prognosis group lost ABH antigen (p < 0.05).
Matsumoto and colleagues [10], in a study of 89 patients with all stages of lung cancers, stated that loss of ABH blood group antigens in lung carcinomas correlated with metastatic potential and poor survival. This reduced survival was seen in the group as a whole (p < 0.05) and for blood group B patients (p < 0.05). Loss of antigen expression was not associated with reduced survival in patients with A or H antigen. The carcinomas included small cell carcinoma, adenoid cystic carcinoma, mucoepidermoid carcinoma, and carcinosarcoma as well as non–small cell lung carcinomas. Pathologic lymph node status (N) was determined in only 64 patients (72%), and therefore staging was incomplete. It is difficult to interpret these findings in this heterogenous group of incompletely staged patients.
Lee and colleagues [11], in a study of 253 patients with stage I, II, and III non–small cell lung carcinoma, found that loss of expression of blood group antigen B and H did not correlate with survival. In a subgroup of patients, those with stage I and II non–small cell lung carcinomas, loss of BAA was predictive of a decreased survival. This advantage was not found in stage III patients. In this study the individual groups were small and the significant conclusions of BAA expression in stage I and II disease were based on only 43 patients.
In Lee and colleagues' study progressive disease was seen in 39.5% (17/43) of patients with BAA-positive tumors and 50% (14/28) of patients with BAA-negative tumors. The median survival was 6 months versus 18 months, respectively (p = 0.001). Recurrence, second primary malignancies, and deaths due to carcinoma or unknown causes were classified as disease progression. However, the groups were not identical. T2 tumors, which have a poorer prognosis than T1 tumors, were more frequent in patients with BAA-negative tumors (p = 0.04), and there were more large cell tumors in patients with BAA-negative tumors (p = 0.05). The survival advantage attributed to loss of expression of BAA may be due to the predominance of T2 tumors in this group of patients.
Presently, there are no means of identifying those patients within a staging group who will have relapse or development of distant disease. Based on survival analyses we conclude that expression of BAA by neoplastic cells (including and excluding nonsecretors), the pattern of staining of neoplastic cells, and secretor status were not of prognostic significance in stage I non–small cell lung carcinomas. Only primary tumor status (T) was associated with overall and event-free survival; patients with T2 tumors had a worse overall (p = 0.06) and event-free survival (p = 0.03) than those with T1 tumors.
|
Footnotes
|
|---|
Presented at the Poster Session of the Thirtieth Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 31–Feb 2, 1994.
Address reprint requests to Dr Rice, The Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195.
|
References
|
|---|
- Lloyd KO. Blood group antigens as markers for normal differentiation and malignant change in human tissues. Am J Clin Pathol 1987;87:129–39.[Medline]
- Juhl BR, Hartzen SH, Hainau B. A, B, H antigen expression in transitional cell carcinomas of the urinary bladder. Cancer 1986;57:1768–75.[Medline]
- Sakamoto Y, Kamada M, Irahara M, et al. Difference in clinical significance of blood group antigens A, B, and H in carcinoma tissue in uterine cervix. Gynecol Oncol 1986;25:195–203.[Medline]
- Itzkowitz SH, Yaun M, Ferrell LD, et al. Cancer associated alterations of blood group antigen expression in the human pancreas. J Natl Cancer Inst 1987;79:425–34.
- Hirohashi S, Shimosato Y, Ino Y, et al. Distribution of blood group antigens and CA19-9 in gastric cancers and non-neoplastic gastric mucosa. Gann 1984;75:540–7.[Medline]
- Schoentag R, Williams V, Kuhn W. The distribution of blood group substance H and CEA in colorectal carcinoma. Cancer 1984;53:503–9.[Medline]
- Davidsohn I, Ni LY. Loss of isoantigens A, B, H in carcinoma of the lung. Am J Pathol 1969;57:307–34.[Medline]
- Alvarez-Fernandez E, Carretero-Albinana L. Expression of blood group antigens by normal bronchopulmonary tissues and common forms of pulmonary carcinomas. Arch Pathol Lab Med 1991;115:42–9.[Medline]
- Ogawa J, Iwazaki M, Inoue H, Koide S, Kawada S, Shohtsu A. Relation of ABH blood group antigen expression to prognosis in lung carcinoma. Tokai J Exp Clin Med 1987;12:49–54.[Medline]
- Matsumoto H, Muramatsu H, Shimotakahara T, et al. Correlation of expression of ABH blood group carbohydrate antigens with metastatic potential in human lung carcinomas. Cancer 1993;72:75–81.[Medline]
- Lee JS, Ro JY, Sahin AA, et al. Expression of blood-group antigen A-a favorable prognostic factor in non–small-cell lung cancer. N Engl J Med 1991;324:1084–90.[Abstract]
- The World Health Organization. Histological typing of lung tumors. Am J Clin Pathol 1982;77:123–36.[Medline]
- Beahrs OH, Henson DE, Hutter RVP, Kennedy BJ. Manual for staging cancer. 4th ed. Philadelphia: Lippincott, 1992: 115–21.
- Kruskal WH, Wallis WA. Use of ranks in one-criterion variance analysis. J Am Stat Assoc 1952;47:583–621.
- Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–81.
- Peto R, Peto J. Asymptotically efficient rank invariant test procedures. J Royal Stat Soc, Series A 1972;135:185–98.
This article has been cited by other articles:
|
|
|
|
 
J. C. Nesbitt, J. B. Putnam Jr, G. L. Walsh, J. A. Roth, and C. F. Mountain
Survival in Early-Stage Non--Small Cell Lung Cancer
Ann. Thorac. Surg.,
August 1, 1995;
60(2):
466 - 472.
[Abstract]
[Full Text]
|
|
|
Top
Abstract
Introduction
