Effect of Smoking on Human Natural Killer Cell Activity

BARBARA PHILLIPS, MD,’ M. ERNEST MARSHALL, MD,t STEPHEN BROWN, MS,* AND JOHN S. THOMPSON, MDS

Natural killer (NK) cells play a central role in immune surveillance against tumors and viral infections. NK activity is depressed in patients who have a wide range of carcinomas, including carcinomas of the lung. Peripheral blood NK activity was measured in 22 nonsmokers, 15 light/moderate smokers, 12 heavy smokers, and 19 patients witb carcinoma of the lung. Patients with carcinoma of the lung had marked depression in NK activity compared with nonsmokers. Light/moderate smokers had NK activity comparable to that of nonsmokers, whereas heavy smokers had marked depression in NK activity that was comparable to that of patients with carcinoma of the lung. These results suggest that smoking-induced alterations in NK activity may have a role in the pathogenesis of smoking-associated carcinoma of the lung.

Cancer 56:2789-2792,1985.

EAVY CIGARETTE SMOKING causes a number of al- H terations in both the humoral’.’ and cellula? components of the immune system. Such alterations in immunity are believed to be causally related to the in- creased incidence of infections and neoplasia found in heavy smokers.

Natural killer (NK) cells comprise a subset of lympho- cytes that play an important role in defense against certain infections, especially viral, and in immune surveillance against malignant tumors.’-’’ Peripheral blood NK activ- ity is depressed in patients with a wide range of malig- nancies, including carcinoma of the lung.’”’’ In human studies it is difficult to discern whether depressed NK ac- tivity preceded and contributed to the development of malignancy or whether the malignancy resulted in the depressed NK a~tivity.’~ Based on the observations that cigarette smoking is associated with an increased incidence of carcinomas of the lung and NK activity is depressed in patients with carcinomas of the lung, we prospectively examined the effects of smoking (in the absence of car- cinoma) on NK activity. In the current report comparisons in NK activity are made among healthy nonsmokers, smokers, and patients with carcinoma of the lung.

Methods

Study Population

Peripheral blood NK activity was measured in 68 in- dividuals (Table 1). Twenty-two healthy, nonsmoking

From the Divisions of *Pulmonary Medicine and tHematology/On- cology, $Department of Medicine, University of Kentucky Medical Centers and Veterans Administration Hospital, Lexington, Kentucky.

Address for reprints: Barbara A. Phillips, M.D., MN578, Pulmonary Division, University of Kentucky Medical Center, 800 Rose Street, Lex- ington, Kentucky, 402364084.

Accepted for publication March 15, 1985.

volunteers served as controls. The control subjects were nonpregnant and were taking no medications at the time of study. Although a number of patients in the study pop- ulations were taking medication(s) at the time of study, no patients were receiving steroids or antihistamines. Smokers were grouped according to the intensity of smoking. Light-to-moderate smokers were defined as those who had accumulated 49 or fewer pack-years ( 1 5 subjects), and heavy smokers were defined as who had 50 or more pack-years (12 subjects). Nineteen patients with carcinoma of the lung were studied (Table 2). None of these patients had been given chemotherapy or radiation therapy or had undergone surgical resection at the time of study.

NK Cell Assay

The NK cell activity was determined by a chromium 5 I (”Cr)-release assay as follows. Effector lymphocytes were isolated from peripheral blood by Hypaque (Win- throp Laboratories, Sterling Drug, New York, NY)-Ficoll (Sigma Chemical Co., St. Louis, MO) centrifugation. The human leukemic cell line, K562, was used as the target cells. The assays were performed in round-bottom well microtiter trays (Flow Labs, McLean, VA). Each well contained 5 X lo3 51Cr-labeled K562 target cells in 0.1 ml Roswell Park Memorial Institute (RPMI)-I640 medium that contained 10% fetal calf serum, 25 mM N-2-hydroxyethylpiperazine Nf-2-ethanesulfonic acid (HEPES), 2 mM glutamine, and penicillin-streptomycin (“medium”). Effector cells were suspended in medium at a concentration of 2 X lo6 cells/ml. Effector cells were added to individual wells in volumes of 0.1 ml, 0.05 ml, 0.025 ml, and 0.01 ml, which corresponded to effector: target cell (E:T) ratios of 40:l, 20:1, l O : l , and 4:1, re- spectively. The volume of each well was maintained at

2789

2790 CANCER December 15 1985 Vol. 56

TABLE I . Characteristics of Study Populations

Age (YE)

No. of mean Sex Group subjects (range) (M:F)

Nonsmokers 22 38.3 19:3

Light smokers* 15 42. I 11.4 (27-58)

(24-71)

(34-72)

(43-87)

Heavy smokers? 12 58.9 10:2

Lung carcinoma 19 64.6 I5:4

Smoking history (pack years) mean

(range)

0

26.7 (4-45) 79.0

(50-120) 7 I .o

(40- 120)

* 549 pack years. t 250 pack years.

0.2 ml with medium. Each E:T ratio was plated in tnp- licate, and the mean counts per minute (CPM) were used in calculations. Spontaneous s’Cr-release from target cells was measured in the absence of effector cells, and maxi- mum release was determined by treating the target cells with Triton (Sigma) X-100 detergent. The trays were in- cubated for 4 hours at 37°C in 5% COz. After they were incubated, the trays were centrifuged at 350 X g for 10 minutes. Supernatant 0. I-ml aliquots were removed from each well and counted in a gamma counter. The results were calculated according to the formula:

CPM (experimental release)

x 100 - CPM (spontaneous release)

TO lysis = CPM (maximum release)

- CPM (spontaneous release)

The results are given here as the percent lysis at E:T of 40: I .

Statistical Method

Comparisons in NK activity among the control and test groups were made using the unpaired student’s t test.

Results

As indicated in Table 1, there was some variability in age among the four study populations. Heavy smokers and patients with carcinoma of the lung were generally older than the nonsmokers and light/moderate smokers. This was not considered to be a significant variable, how- ever, since several author^'^*'^.*^ have noted no relation-

TABLE 2. Characteristics of Lung Cancer Patients

Histologic features No. Sex (M:F)

Small cell 8 5:3 Squamous 4 4:O Adenocarcinoma 4 3: 1 Large cell 2 2:o Anaplastic 1 I :O

70

- 0

Y

8

CONTROLS LIGHT SMOKERS

8

0

A HEAVY

SMOKERS CARCINOMA

OF LUNG

RG. 1. NK activity (96 lysis) in nonsmoking controls, light smokers, heavy smokers, and patients with carcinoma of the lung. Each point represents one subject. The long horizontal bar represents the mean, and the short horizontal bars represent f I standard deviation.

ship between NK activity and age, sex, or race. When age- related differences in NK activity were reported, however, NK activity was found to be higher in older individuak2’ Such age-related differences would only serve to enhance the differences between our control and study groups, since the older populations (heavy smokers and carcinoma patients) had low NK levels. Therefore, the only variables were considered to be smoking history and the presence or absence of carcinoma.

The smoking habits for the light smokers ranged from 4 to 45 pack-years (mean, 26.7), whereas the heavy smok- ers had a range of 50 to 120 pack-years (mean, 79.0). The I9 patients with carcinoma of the lung were grouped to- gether without regard to histologic type, smoking history, or stage of disease.

When considered as a single group, smokers had a lower NK activity (29.5 2 23.8%) compared with the non- smoking controls (43.4 f 17.3%) (P < 0.02) (Fig. 1). When grouped according to smoking intensity, however, light smokers had an NK activity that was not significantly different from that of the nonsmokers (41.3 f 22.9%). Heavy smokers had an NK activity of 14.7 t I5.7%, which was significantly lower than that of light smokers ( P < 0.002). The mean NK activity for patients with carci-

No. 12 NK ACTIVITY IN SMOKERS Phillips el al. 279 1

noma of the lung was 19.3 f 13.770, which was signifi- cantly different from that of nonsmokers (P < 0.001) and light smokers ( P < 0.00 1 3 , but not significantly different from that of heavy smokers.

Discussion

Previous studies have shown that heavy cigarette smoking is associated with a number of alterations in both humoral and cellular immunity. Heavy smoking results in depressed serum levels of IgG, IgM, and IgE compared with nonsmokers.'-2 This depression in immunoglobulin levels is reversible upon cessation of smoking. Heavy smokers tend to have higher peripheral blood leukocyte counts and, in some studies,3d they have been shown to have higher absolute lymphocyte counts. In addition to the higher lymphocyte counts, Miller and co-workers)-6 found heavy smokers to have a significantly lower per- centage of OKT4+ (helper/inducer) lymphocytes and a higher percentage of OKT8' (suppressor/cytotoxic) lym- phocytes, with a resultant lower T4:T8 ratio, compared with nonsmokers or light/moderate smokers. These al- terations also may be reversed by cessation of smoking.

Ferson and associates" demonstrated that normal smoking subjects and melanoma patients who smoked had lower levels of NK activity compared with non- smokers. No correlation was made with the intensity of smoking. They subsequently showed that this suppression of NK activity was reversible upon cessation of smoking23 Ryu ef measured peripheral blood NK activity in nine smokers with a history of less than 50 pack-years and found that such smokers had, as a group, an NK activity that was 42.7 f 5.6% of the NK activity of the nonsmoking controls (P < 0.05). In the current report we have shown that depression in NK activity is associated with heavy smoking (250 pack-years), whereas light/ moderate smokers (549 pack-years) have NK levels that are comparable to those of nonsmokers. We have shown, further, that heavy smoking results in low levels of NK activity, which are comparable to the low levels found in patients with carcinoma of the lung. The current study deals with NK activity; however it does not address the issue of whether the depressed activity seen in heavy smokers is the result of a quantitative or qualitative defect of NK cells. We are currently extending the current study to measure absolute numbers of NK cells in smokers and nonsmokers with the use of two monoclonal anti-NK cell antibodies, GO22 and HNK-1. It is likely that the low NK activity observed in lung cancer patients has a mul- tifactorial etiology that includes the effects of smoking, the presence of carcinoma and, probably, other factors, such as nutritional status.

Natural killer cells have been the object of intense in- vestigation in recent years because of the demonstrated

role of these cells in antitumor immunity. Historically, it was believed that T e l l s played a central role in immune surveillance against neoplasia, but evidence has accu- mulated that indicates that T e l l s alone are probably in- capable of fulfilling this Experimental and clinical data point to NK cells as more attractive effector cells in immune surveillance. Patients with the rare Chediak-Hi- gashi syndrome are characterized by a genetic abnormality in leukocyte granulation and a resultant increased sus- ceptibility to infections. Such patients have normal hu- moral immunity and delayed hypersensitivity. Patients who do not die of infection often suffer from lethal lym- phoproliferative malignancies. Roder and c01leagues~~ demonstrated that patients with Chediak-Higashi syn- drome have a profound deficiency in NK cell function, which may be responsible, in part, for the development of malignancy. Further clinical evidence is derived from patients with familial melanoma. In such patients mela- noma develops at an earlier age; they also have a higher frequency of multiple lesions and a higher incidence of second malignancies compared with nonfamilial mela- noma patients. Hersey et ~ 1 . ~ ~ demonstrated that patients with familial melanoma and their relatives have low NK activity, whereas B- and Tcell immunity is intact. Sullivan and associatesz7 found deficient NK activity in patients with the X-linked lymphoproliferative syndrome. Lym- phoproliferative malignancies develop in approximately 40% of such patients. In the mouse melanoma model, Warner and Dennert" showed that NK cell deficiency promoted both tumor development and tumor metas- tases.

We have shown that cigarette smoking is associated with depression in NK activity in a fashion that correlates with the intensity of smoking and that such depressions in NK activity are present in the absence of carcinoma of the lung, yet are comparable to the levels found in carcinoma patients. These data suggest that depressed NK activity precedes the development of carcinoma; however, it will be important to follow serial NK values in a cohort of heavy smokers to establish a stronger cause/effect link, i.e., to determine in how many of such individuals car- cinoma of the lung will ultimately develop. Although the ultimate pathogenesis of smoking-associated carcinomas of the lung will have a multifactorial etiology, the current results and the data of suggest that smoking- induced alterations in NK cell activity will be an important factor in the development of malignancy and subsequent metastasis.

REFERENCES I . Gerrard JW. KO CG, Myers A, Heiner DC, Mink J, Dosman JA.

Immunoglobulin levels in smokers and nonsmokers. Ann Allergy 1980;

2. Bahna SL, Heiner DC. Myhre B. IgE elevation and suppression by M261-262.

tobacco smoking. J Allergy Clin Imrnunol 1980; 65:23 1-232.

2192 CANCER December 15 1985 Vol. 56

3. Miller L. Hunvitz C, Gillick L et a/. Analysis of T-lymphocyte subsets in asymptomatic smokers. Am Rev Resp Dis 198 I ; 123: 148.

4. Miller LG, Ginns LC. Reversible defect of immunoregulating T- cells in cigarette smokers. Am Rev Resp Dis 1982; 125:60.

5. Miller LG. Murphy M, Goldstein G, Ginns LC. Reversible alter- ations in immunoregulatory T cells in smoking. Analysis by monoclonal antibodies and flow cytometry. Chest 1982; 82526-529.

6. Ginns LC, Miller LG, Gillick L et a/. T lymphocyte subsets in smoking and lung cancer. Analysis by monoclonal antibodies and flow cytometry. Am Rev Resp Dis 1982; 126:265-269.

7. Onari K. Takaishi M, lkuta T el a/. Immunological studies on cigarette smokers. Part II. Cell-mediated immunity in cigarette smokers and the influence of the water-soluble fraction (WSF) on the immunity of mice. Hiroshima J Med Sci 1980; 29:29-34.

8. Anderson JB, DeLucia AJ, Bryant LR Jr, Whitaker JH. Inhibition of lymphocyte response in Stumptail Macaques chronically exposed to cigarette smoke. Fed Proc 198 I ; 40:793.

9. Herberman RB. Possible role of natural killer cells in host mistance against tumors and other diseases. Clin Immunol Allergy 1983; 3:479- 493.

10. Warner JF, Dennert G. Effects of a cloned cell line with NK activity on bone marrow transplants, tumor development and metastasis in vivo. Nature 1982; 300:31-34.

I I . Timonen T. Characteristics of fresh and cultured natural killer cells. Clin Immunol Allergy 1983; 3:465-477.

12. Stein-Streilein J, Bennett M, Mann D, Kumar V. Natural killer cells in mouse lung: Surface phenotype, target preference, and response to local influenza virus infection. J Immunol 1983; 131:2699-2704.

13. Stewart LD, Ades EW. Prospective study of natural cytotoxicity in peripheral blood of patients with nonlymphoid solid malignancies. Clin Immunol Allergy 1984; 31:78-86.

14. Saijo N, Shimizu E, lrimajiri N et a/. Analysis of natural killer activity and antibodydependent cellular cytotoxicity in healthy volun- teers and in patients with primary lung cancer and metastatic pulmonary tumors. JCancer Res Clin Oncol 1982; 102:195-214.

15. Vanky F. Williams J, Kreicbergs A et a/. Correlation between lymphocyte-mediated anti-tumor reactivities and clinical course. I.

Evaluation of 46 wtients with sarcoma. Cancer Immunol Immunother 1983; 1611 1-16.

16. Vanky F, Peterffy A, Book K, Willems J, Klein E, Klein G. Cor- relation between lymphocyte-mediated antitumor reactivities and the clinical course. II. Evaluation of 69 patients with lung carcinoma. Cancer Immunol Immunorher 1983; 16: 17-22.

17. Balch CM, Tilden AB, Dougherty PA, Cloud GA, Abo T. De- pressed levels ofgranular lymphocytes with natural killer (NK) cell func- tion in 247 cancer patients. Ann Surg 1983; 198:192-199.

18. Morales A, Ottenhof PC. Clinical application of a whole blood assay for human natural killer (NK) cell activity. Cancer 1983; 52:667- 670.

19. Takasugi M, Ramseyer A, Takasugi J. Decline of natural non- selective cell-mediated cytotoxicity in patients with tumor progression. Cancer Res 1977; 37:413-418.

20. Sugiyama E, Ito M, Suzuki H, Yamashita N, Maruyama M, Yano S. Natural killer cell activity associated with aging and sex in humans. Jap JMed 1983; 22:195-199.

2 I . Onsrud M. Agedependent changes in some human lymphocyte subpopulations. Acta Pathol Microbiol Scand 198 I ; 8 9 3 - 6 2 .

22. Ferson M, Edwards A, Lind A, Milton GW, Hersey P. Low MtUd killer-cell activity and immunoglobulin levels associated with smoking in human subjects. Cancer 1979; 23:603-609.

23. Hersey P, Prendergast D, Edwards A. Effects ofcigarette smoking on the immune system. Follow-up studies in normal subjects after ces- sation of smoking. Med J Ausl 1983; 2:425-429.

24. Ryu JH, Rogol PR, Ginns LC. Reduced natural killer activity in smokers (Abstr) Am Rev Resp Dis 1984; 129:A7.

25. Roder JC, Haliotis T, Klein M ef a/. A new immunodeficiency disorder in humans involving NK cells. Nature 1980; 284553-555.

26. Hersey P, Edwards A, Honeyman M, McCarthy WH. Low natural- killercell activity in familial melanoma patients and their relatives. Br J Cancer 1979; 4 0 1 13-122.

27. Sullivan JL, Byron KS, Brewster FE, Purtilo DT. Deficient natural killer cell activity in X-linked lymphoproliferative syndrome. Science 1980; 210543-545.