Immunoprofile studies for patients with bronchogenic carcinoma

Immunoprofile studies for patients with bronchogenic carcinoma I. Correlation of pretherapy studies with survival The general immune competence of 146 patients with bronchogenic carcinoma was measured, prior to irradiation therapy, by determining dinitrochlorobenzene (DNCB) reactivity, delayed cutaneous hypersensitivity (DCH) response to microbial antigens, peripheral lymphocyte counts, peripheral T and B lymphocyte counts, and the response of patient's lymphocytes to stimulation by phytohemagglutinin (PHA), concanavallin A (Con A) and pokeweed mitogen (PWM). Analyses were performed by the life-table method to determine the correlation of the immune status of these patients with survival rates. Statistically significant differences in survival were noted between the groups of patients with normal values when compared with the patients with abnormal values for the majority ofthe tests of general immunity. A stage of disease correlation with survival rate was noted for all groups of patients with abnormal immune measurements, but it was absent for many of the immune parameters when patients with normal values were compared. The effects of histology, age, and sex did not appear to influence the survival data as significantly as did the immune status of the patient. These data indicate that measurements of general immune competence may be of significant prognostic value for the management of patients with bronchogenic carcinoma. The measurement of DNCB reactivity shows the strongest correlation with survival rate.

George A. Liebler, M.D.* (by invitation), Joseph P. Concannon, M.D.** (by invitation), George J. Magovern, M.D.,* Milton H. Dalbow, M.S.** (by invitation), Pittsburgh, Pa., and Sheila E. Hodgson, M.D.*** (by invitation), Columbus, Ohio

T

he past decade has witnessed an ever-increasing interest in the relationship of the immune system to malignant disease. This interest stems from the widely accepted principle that the majority of well-studied From the Department of Surgery and the Division of Radiation Oncology of the Clinical Radiation Therapy Research Center, Allegheny General Hospital, Pittsburgh, Pa., and the Radiation Therapy Department, Ohio State University, Columbus, Ohio. Supported by Public Health Service Research Grant No.5 SOIRR05709 and by Grant No. CAI0438-09 from the National Cancer Institute, National Institutes of Health, Bethesda, Md. Read at the Fifty-seventh Annual Meeting of the American Association for Thoracic Surgery, Toronto, Ontario, Canada, April 18, 19, and 20,1977. Address for reprints: J.P. Concannon, M.D., Division of Radiation Oncology, Clinical Radiation Therapy Research Center, Allegheny General Hospital, 320 E. North Ave., Pittsburgh, Pa. 15212. *Department of Surgery, Allegheny General Hospital. **Division of Radiation Oncology, Clinical Radiation Therapy Research Center, Allegheny General Hospital. ***Radiation Therapy Department, Ohio State University.

506

animal tumors have been shown to be antigenic, and frequently immunogenic as well, in the syngeneic host. There are substantial data to support the concept that malignant neoplasms in man have unique antigens and, therefore, might be expected to be immunogenic. Numerous investigators have identified tumor antigens to be associated with malignant tumors of the lung 1-22 There is little doubt that the ideal approach to the evaluation of the host-tumor relationship is with studies that measure tumor-specific immunity. The controversy over the speciflcity-"?" and interpretation of such tests, as well as the problems associated with this direct approach, however, have led many investigators to evaluate general immune competence in patients with lung cancer rather than tumor-specific immunity. The tests of general immune competence used most frequently in the study of patients with lung cancer have been mitogen stimulation of patient's lymphocytes,26-34 skin testing with microbial antigens.v': 35-39 skin testing for delayed cutaneous hypersensitivity

Volume 74 Number 4 October. 1977

Bronchogenic carcinoma

5 07

(DCH) to a primary haptenic sensitizing agent such as 2-4 dinitrochlorobenzene (DNCB), distribution of peripheral T and/or B lymphocytes.w :" and multiples of these tests." 13, 45-51 The interpretation and usefulness of these tests in the management of patients with cancer also remains controversial, but many investigators have shown a correlation between immunologic deficiency, as measured by one or more of these tests, and some prognostic criteria, such as disease staging.": 31, 35. 41, 46. 47, 49 or response to treatment.9, 42. 44 Others have reported a correlation between the immune status of these patients and survival rate.P" 36, 39. 43-45, 48

vided into two age groups (40 to 59 and 60 to 80 years) for evaluation of the effect of age on general immune competence as a correlate with rate of survival. Immunologic studies. The immune parameters measured in this study were de novo DCH response to dinitrochlorobenzene (DNCB); in vivo skin testing for DCH to five microbial antigens; peripheral blood lymphocytes per cubic millimeter; percentage of T cells and T cells per cubic millimeter in peripheral blood; percentage of B cells; response of lymphocytes to the mitogens phytohemagglutinin (PHA), concanavallin A (Con A) and pokeweed mitogen (PWM). The methods used for these studies have been described previ-

The present study was performed to determine the correlation between various tests of general immune competence as measured prior to irradiation therapy and the survival rates of patients with bronchogenic carcinoma. The relationship of other factors-such as stage of the disease, histologic type, age, and sexwith patient survival rates and immune competence also have been submitted to analysis.

ously."

Patients and methods Patient selection. One hundred forty-six patients with a histopathologically confirmed diagnosis of bronchogenic carcinoma referred to the Department of Radiation Oncology by several thoracic surgeons were entered into this study, regardless of age, stage of disease, or histologic type. Forty-five patients had thoracotomies (24 resections) performed before the immunoprofile studies. The period intervening between the operation and the immunologic studies was more than 2 weeks for all patients. Staging of the disease was based on clinical and surgical findings. Most patients had mediastinoscopy as well as bone, brain, and liver scans performed prior to the immunoprofile studies. The clinical and surgical staging for each patient was based on the Radiation Therapy Oncology Group's modification of the American Joint Committee's Clinical Staging System for Carcinoma of the Bronchus. Following staging, the patients were assigned to one of three groups: localized disease-Stages I and II, regional disease-Stages III and IV (Stage IV without metastases), and remote disease-Stage IV (patients with metastases). There were 71 patients with squamous cell carcinoma, 17 with adenocarcinoma, 33 with undifferentiated small cell carcinoma, and 25 with undifferentiated large cell carcinoma. Twenty-one patients had localized disease, 68 had regional disease, and 57 had remote disease. There were 32 female patients and 114 male patients entered into the study. Patients were di-

Briefly, peripheral blood lymphocytes collected aseptically in acid citrate dextrose (ACD) solution or heparin were separated by the Ficoll/Hypaque density gradient method of Boyum. 52 The lymphocyte mitogen stimulation studies were prepared in Microtest II plates (Falcon) with 2 x 105 mononuclear cells per well. The cells were suspended in 200 JLl of Roswell Park Memorial Institute-1640 (RPMI-I640) media with Hepes (N'2-hydroxyethyl piperazine-N' -ethane sulfonic acid) buffer. Antibiotics (streptomycin, 100 JLg per milliliter; penicillin, 100 U per milliliter; and gentamicin, 50 JLg per milliliter), 10 percent human AB serum and glutamine (2 mM) were added to the RPMI-1640 Hepes-buffered media. The tissue culture reagents were obtained from Grand Island Biological Co., Grand Island, N. Y. (GIBCO). Ten microliters each of mitogen-PHA (Difco, 100 JLg per milliliter), PWM (GIBCO, 5 mg. per milliliter, and Con A (Sigma, 500 JLg per milliliter)-was added to quadruplicate wells, and four wells were used as unstimulated controls. The cultures were incubated for 20 hours at 37° C.-5 percent carbon dioxide and were pulsed for 4 hours with 2 JLCi (in 10 JLI) of 3H-uridine (SA 5 Ci per millimole, Amersham/Searle) at 37° C. in a 5 percent carbon dioxide atmosphere. The cultured lymphocytes were harvested onto glass fiber filters with a MASH-II aparatus (Microbiological Associates). The disclike impressions containing the cells were removed from the filter strip and placed in mini scintillation vials (Fisher Scientific Company, Pittsburgh, Pa.). Two milliliters of a simple primary scintillation cocktail (PPO, 5.28 Gm. per liter in toluene) were added, and the radioactive uptake by the lymphocytes was counted. A stimulation ratio was calculated by dividing the mean disintegrations per minute measured for the four stimulated cultures of each mitogen by the mean disintegrations per minute obtained for the four unstimulated cultures.

The Journal of

5 a8

Liebler et al.

Thoracic and Cardiovascular Surgery

Prerodiolion SIooe of DiseaseEffecl All Polients

Slag!

# PIs.

Localized 0 21 Reoionoi 0 68 Remote b. 57

..J

50~

Survival 9.2 6.7 3.7

Lac VB Reg Lac vs Rem

p a c

Reo VI Rem

c

~ sa: 60 ::3 en ~

40

0

20 0

Fig. 1. Survival curves by the life-table method for each of the disease staging groups. The percentage of peripheral thymic dependent lymphocytes (percentage of T cells) was determined by a modified spontaneous E-rosette technique similar to that described by Wybran, Carr, and Fudenberg. 53 Approximately 106 Ficoll/Hypaque separated mononuclear cells were suspended in 100 /Ll of Hanks balanced salt solution (GIBCO), and approximately 1.2 x 107 washed three times sheep erythrocytes (Sacks Farms, Pittsburgh, Pa.) in 200 /Ll of Hanks balanced salt solution were added. This preparation was spun at 100 x G for 5 minutes and allowed to stand at room temperature for 2 hours. The supernatant was decanted and the cells were resuspended in 500 /LI of phosphate-buffered saline (GIBCO). Trypan blue (100 /L1-4 percent solution, GIBCO) was added and the percentage (percentage of T cells) of rosette-forming cells was determined by averaging the number of rosette-forming cells observed per 100 viable mononuclear cells in three counts for duplicate reaction tubes. (Only mononuclear cells with three or more adherent erythrocytes were counted as rosette-forming cells.) The technique used in this study measures late (total) T cells under suboptimal conditions and not active (early) rosettes. The peripheral T cell levels (T cells per cubic millimeter of peripheral blood) were determined by multiplying the percentage of T cells by the number of lymphocytes per cubic millimeter of peripheral blood obtained from white and differential blood cell counts. The percentage of B cells (bone marrow dependent or bursaequivalent lymphocytes) was determined by the fluorescent antibody technique. Fifty microliters of polyvalent antihuman immunoglobulin (lg) serum conjugated with fluoresceinisothiocyanate, diluted I: 3,

was added to 12 by 75 mm. tubes containing 1061ymphocytes suspended in 100 /LI of Hanks balanced salt solution. The tubes were placed in an ice bath for 60 minutes and held overnight at 5° C. The tubes were spun at 100 G for 10 minutes and the cell pellet was washed two times, followed by resuspension in 500 /LI of phosphate-buffered saline. The total number of mononuclear cells was counted in the various grids of a phase-contrast counting chamber under phase contrast and the fluorescent B cells then were counted in the same grids under incident light fluorescence microscopy. The percentage of B cells was calculated by dividing the number of fluorescent cells by the total number of mononuclear cells counted. Cells with an obvious configuration of monocytes or macrophages were not counted in either the T or B cell assays, but Latex ingestion was not used to identify the monocytic cells positively. The reactivity of healthy volunteers to DNCB was not determined in this study. A modification of the method for DNCB sensitization described by Catalona and co-workers'< was used to determine the reactivity of the cancer patients. Two DNCB sensitizing doses (2,000 and 100 /Lg) were used in this study. A grading scale between I and 5 was established for the evaluation of DCH reactivity to de novo sensitization with DNCB. A DNCB response was scored as a Grade 3 if the patient had an erythema with moderate induration (5 to 10 mm.) at day 14 to the 2,000 /Lg sensitizing dose. Erythema with marked induration (2=20 mm.) at the 2,000 /Lgsensitizing site or erythema with moderate induration at both the 2,000 /Lg and the lao /Lg sensitizing sites were scored as Grade 4 reactions. Vesiculation of the sensitizing sites at day 10 and/or 14 was scored as a Grade 5 response. Patients who failed to respond to the sensitizing doses of DNCB were Challenged at day 14 with 200, 100,50, and 25 /Lg doses of DNCB. Patients with a Grade 3 response also were challenged at day 14 with 50 and 25 /Lg of DNCB. An induration response (> 10 mm.) in 48.hours to the two highest challenge doses (200 and 100 /Lg) were scored as Grade 3 reactions. Any degree of induration at the 50 /Lg site was scored as a Grade 4 reaction and any induration at the 25 /Lg site as a Grade 5 reaction. Patients with a minimal induration reaction to the 200 /Lg site and patients who did not respond to the sensitizing or challenge doses were scored as Grade 2 or I, respectively, and were considered to be abnormal for this study. Patients with responses of Grade 3, 4, or 5 to either the sensitizing or challenge doses were considered to be normal. The DCH response to five microbial antigens was tested in 22 healthy volunteers with a median age 57

Volume 74 Number 4 October, 1977

Bronchogenic carcinoma

100

\

Localized Disease

\

80

\

'* Pts. -Time to 50% Survival \ \

'~

60

\ \

RegIonalDisease

\

'* Pis. - Time to 50% Survival \ \ \

\

\

\

\

\

~""

\

~ 20

.41 .27

\

\

....J

,~

19.0

.za

1:111.8 P<005

\

40

s

.18 .3

\

\

\ ~

.,

"

"W-

10.2 4.6 1:156 P
- ...

01--+--+-+--+-+--+---1--+--+-++-+--+--If-+--+-+--+--11--+-......-f

UlIOO \

'if!:

5 09

'* Pts.- Time

\ 80 ..,

.81 .65

\

60

40

\

..

All Stages to 50% Survival

11.2

~

1:177 P
,

\

\

~

\

20

\

\

\

,

\

3

6

9

12 15 18 21 24 27 30

3

6

SURVIVAL (MONTHS)

9 12

15 18 21 24 27 30

Fig. 2. Survival curves by the life-table method, comparing the survival of patientswith normal versus abnormal response to dinitrochlorobenzene for each disease stage category. years and a range of 48 to 67 years. All the healthy volunteers responded to two or more of these antigens. The induration was measured in two diameters. The minimum induration, in the largest diameter, observed in any of the healthy volunteers reacting to purified protein derivative, mumps, Candida, or dermatophytin was 10 mm. The minimum induration to Varidase was 20 mm. in the largest diameter. A grading scale (zero4) for DCH to microbial antigens was established. A minimal induration of greater than 5 mm. but less than 10 mm. to multiple skin test antigens was a Grade 3. Grade 4 was an induration for any of the other skin test antigens. Grades 3 and 4 were considered to be normal DCH responses for the purposes of this study. Statistical methods. Immune profile data for healthy volunteers were used to establish upper and lower limits of normal values for the various parameters of the immune system studied. The immunoprofile data for cancer patients are not expressed quantitatively in this paper, but as normal or abnormal dependent upon whether the measurements performed are within (normal) or outside (abnormal) the values established for the healthy volunteers. The data obtained from the healthy volunteers (median age 51 years and range of 41 to 68 years) for circulating lymphocytes, percentage of T cells, T cells per cubic millimeter, and percentage of B cells closely

approximated normal distributions and the limits of normal were defined as the mean value ± 2 standard deviations (lymphocytes: 60 studies, mean 1,683 ± 674; percentage of T cells: 57 studies, mean 65.2 ± 12.4; T cells per cubic millimeter: 57 studies, mean 1,208 ± 482; and percentage of B cells: 57 studies, mean 18.8 ± 13.8). There is a 95 percent confidence level that these values were representative of the upper and lower limits of normal. The data for the responses of the lymphocytes from healthy volunteers to the plant lectins (63 studies for 46 individuals, mean age 51 years and range 41 to 68 years) were nonparametric, with a pronounced skew observed to the upper region of the distributions for all three mitogens. The limits of normal, therefore, were determined by percentile ranks to exclude the lower 5 percent of the calculated distribution. Normallymphocyte stimulation ratios, based on these data, were determined to be ::::4.0 for PHA, ::::3.0 for Can A, and ::::2.3 for PWM. The time to median survival (time to 50 percent survival) was obtained by the life-table technique" for groups of patients with normal values and for groups of patients with abnormal values for each parameter studied. The statistical significance between the survival data for groups of patients with normal values versus the patients with abnormal values for each parameter of

5 10

The Journal of Thoracic and Cardiovascular Surgery

Liebler et al.

Table I. Time to median survival in months with difference (.l) between median survival of patients with normal versus abnormal values by stage before radiation therapy Stage Immune test

Immune status

Localized

DNCB

Normal Abnormal

19.0 7.2 a 11.8"'

DCH

Normal Abnormal

10.6 7.3 ~

Lymphocytes/c.mm.

Normal Abnormal

a

9.8 6.0 3.8"

% T cells

Normal Abnormal

10.6 7.9 a 2.7

T cells/c.mm.

Normal

11.3 6.6 a 4.7"

Abnormal % B cells

Normal Abnormal

PHA

Normal Abnormal

Con A

PWM

Normal Abnormal

a

9.7 8.3

LfT4

a

9.3 8.7 0.6

a

9.8 4.4 5.4

Normal Abnormal

10.4 7.4 3.0

I

Regional

I

Remote

I

All stages

10.2 4.6 a 5.6e

11.2 2.2 a 9.0e

11.2 3.5 a 7.7e

a

7.4 2.9 4.5 b

a

4.8 2.3 2.5 b

a

7.0 2.8 4.2 e

a

7.5 4.4 3.1"

a

4.7 2.6 2.lb

a

7.3 3.5 3.8 e

a

9.9 2.8 7.le

10.4 4.1 a 6.3"

a

9.0 3.8 5.2"

a

7.8 3.2 4.6 e

a

8.8 3.9 4.9"

a

8.4 3.6 4.8 e

a

7.7 2.9 4.8"

10.9 4.9 a 6.0<

a

8.7 4.5 4.2"

a

6.7 2.9 3.8 b

a

8.1 4.3 3.8 b

a

5.1 2.3 2.8 b

a

7.9 5.1 2.8"

10.0 2.6 a 7.4"

a

8.0 4.1 3.9 b

a

9.0 2.8 6.2"

a

8.1 3.3 4.8 e

a

4.9 2.6 2.3 b

Legend: DNCB. Dinitrochlorobenzene. DCH, Delayed cutaneous hypersensitivity. PHA, Phytohemagglutinin. Con A, Concavavalin A. PWM, Pokeweed mitogen. 'Superscript letters indicate probability (p) that the two groups compared (normal versus abnormal) are of different populations based on survival (a0.05 > P > 0.01, b-O.OI > P > 0.001, c-p < 0.001).

the immune system studied in this investigation was determined by the generalized Wilcoxon test. 56 This statistical method (generalized Wilcoxon test) tests whether any two groups that are compared come from the same population and the statistically significant differences between any two groups reported in this study relate to the probabilities that these two groups are from different populations. The time to median survival (50 percent survival rate) also is used for convenience in reporting and in discussing the comparisons between various groups.

Results Survival curves generated by the life-table method for 146 patients with bronchogenic carcinoma in the various stages of disease are shown in Fig. 1. A statistically significant difference in survival rate was observed between each of the staging groups, indicating

that there were indeed three separate populations based on survival (stage of disease effect). The time to median survival was greater for the patients with localized disease than for the patients with regional disease and for the patients with remote disease. Increased median survival time also was noted when the patients with regional disease were compared to the patients with remote disease. Similar survival curves were generated by the lifetable method to determine the difference in time to median survival between groups of patients with normal and with abnormal values for each of the immune parameters measured in this study. The numbers of patients, the time to median survival for the groups of patients with either a normal or abnormal response to DNCB, and the months of difference in median survival between the two groups are shown for each stage of the disease and all stages combined in Fig. 2. The

Volume 74 Number 4

Bronchogenic carcinoma

October, 1977

statistically significant differences between the survival data observed for patients with a normal versus an abnormal response to DNCB in each of the disease staging categories also are shown in Fig. 2. Similar data obtained from survival curves generated for each of the parameters of immune competence studied are shown in Table I. A statistically significant difference in survival rate and a greater time to median survival were observed for the patients with normal values in the regional, remote, and all stages combined categories as compared to patients with abnormal values for each of the measures of immune competence. Statistically significant differences in survival also were noted between the patients with normal values versus patients with abnormal values in the localized disease group for the immune parameters DNCB, T cells per cubic millimeter, and lymphocytes per cubic millimeter. The effect of stage of the disease in relationship to the immune status and survival was examined for each parameter of the immune system studied. A stage effect was observed for each of the immune parameters when the survival of patients with abnormal values within each staging group were compared, i.e., the time to median survival of the abnormal localized group was greater than either the abnormal regional or abnormal remote groups. The time to median survival also was greater generally for the patients with abnormal values in the regional group, as compared to the abnormal remote group. A stage effect was not observed, however, for many of the immune parameters when the time to median survival of patients with normal values within the various staging groups were compared, i.e., the time to median survival for patients with normal measurements to percentage of T cells, T cells per cubic millimeters, PHA, and Con A were similar for each stage group. The patients with a normal response to DNCB in the localized group had a longer median survival time than did the normal DNCB responders in the regional and remote groups, but this stage effect in DNCB response was not observed between the normal regional and normal remote groups. The effect of the immune status in relationship to the stage of the disease was examined further by comparing median survival times of the patients with normal values in the regional and remote disease groups with patients with abnormal values in the localized disease group for each of the parameters studied (Table II). Patients with regional disease with normal values had a survival time similar to that of patients with localized disease with abnormal values for all parameters studied. The median survival times of the patients with

5II

Fig. 3. Survival curves by the life-table method for each of the major histologic categories.

Table II. Comparison of time to median survival in localized-abnormal patients versus regional-normal and remote-normal patients Stage-immune status Test of immune competence

Localizedabnormal

DNCB DCH Lymphocytes/c. mm. % T cells T cells/c.mm. % B cells PHA Con A PWM

Remotenormal

7.2 7.3 6.0 7.9 6.6 7.4 8.3 8.7 4.4

10.2 7.4 7.5 10.9 8.7 8.1 7.9 8.0 8.1

11.2 4.8 4.7 9.9 6.7 5.1 10.0 9.0 4.9

For legend see Table I.

Table III. Comparison of time to median survival for regional-abnormal patients with remote normal patients Stage-immune status Test of immune competence DNCB DCH Lymphocytes/c.mm. % T cells T cells/c.mm. % B cells PHA Con A PWM

Regionalabnormal

Remotenormal

4.6 2.9 4.4 4.9 4.5 4.3 5.1 4.1 3.3

11.2 b* 4.8 4.7 9.9 b 6.7 2.3 10.0 9.0 4.9

For legend see Table I. *Superscript letters indicate probability (p) that the two groups compared are of different populations (b-O.OI > P > 0.001).

The Journal of Thoracic and Cardiovascular Surgery

5 I 2 Liebler et al.

Table IV. Time to median survival in months with differences (a) between median survival of patients with normal versus abnormal values by histology before radiation therapy Histologic category Immune test

ONCB

Immune status

Adenocarcinoma

Normal Abnormal

8.6 2.5 ~6.lb*

OCH

Normal Abnormal

6.7 3.0 ~3.7

Lymphocytes/ c. mm.

Normal Abnormal

% T cells

Normal Abnormal

T cells/c.mm.

Normal Abnormal

6.8 4.4 ~2.2

8.2 4.8 ~3.4

% B cells

Normal Abnormal

8.4 3.7 M.7" 8.6 3.5 ~5.lb

PHA

Normal Abnormal

Con A

Normal Abnormal

PWM

Normal Abnormal

8.0 3.6 M.4" 9.4 3.7 ~5.7b

8.6 2.5 M.lb

I

Large cell

I

Small cell

9.6 3.7 ~5.9"

5.7 4.5 ~1.2

~

6.7 1.8 ~ 4.9 b

5.8 5.0 ~0.8

~

7.4 3.4 4.0"

~

10.2 4.2 6.0 b

~

8.1 4.4 3.7"

~

7.6 3.5 4.1"

~

8.6 3.5 5.3 b

~

8.8 3.3 5.5 b

~

6.7 2.6 4.1"

4.8 4.7 ~6.1"

9.0 4.5 ~4.5"

6.0 4.9 ~2.1

9.9 4.7 ~5.2"

8.3 4.4 ~3.9

5.7 4.5 ~1.2

9.4 2.7 6.7<

I

Epidermoid

12.8 2.8 ~1O.0"

~

8.5 2.9 5.6"

~

8.1 2.9 5.2'"

~

11.3 4.2 7.1"

~

10.1 3.8 6.3"

~

8.5 2.7 5.8<

~

9.7 3.8 5.9<

~

8.8 3.7 5.1"

~

9.2 2.9 6.1<

For legend see Table I. ·Superscript letters indicate probability (p) that the two groups compared (normal versus abnormal) are of different populations based on immune status and survival. (a-O.05 > p > 0.01; b-O.OI > P > 0.001; c-p < 0.001.)

normal values in the remote disease group also had a similar median survival when compared with the localized-abnormal group for six of the nine parameters studied, (DNCB, percentage of T cells, T cells per cubic millimeter, PHA, Con A, and PWM). Similar comparisons were made between the patients with remote disease who had normal values and patients with regional disease who had abnormal values (Table III). Patients with remote disease with normal values had survival rates similar to patients in the regional disease group with abnormal values for all parameters studied. In fact the survival data for patients with remote disease with normal values for DNCB and percentage of T cells were significantly different from the survival rates of patients with regional disease and abnormal values for these parameters (0.01 > P > 0.001). The effect of histology and the immune status on

median survival time was examined. There were too few patients in the various histology categories to perform this analysis by each stage of the disease. Chi square analysis indicated that the proportion of patients in the various stages of disease were similar for each of the histologic types. Consequently, this analysis on the effects of histology was performed on the data for all stages combined for each histologic category. The median survival time of patients was approximately equal for each of the four histologic groups (Fig. 3). The median survival times for the various cell type categories for those patients with normal and abnormal values for each parameter studied and the statistically significant differences observed in the survival times are shown in Table IV. Statistically significant longer survival was observed for the groups of patients with squamous cell carcinoma and with undifferentiated

Volume 74

Bronchogenic carcinoma

Number 4 October, 1977

5I3

Table V. Time to median survival in months with differences (Ll) between median survival of patients with

normal versus abnormal values by age groups and sex before radiation therapy Age group (yr.) Immune test DNCB

DCH

Sex group

Immune status

40 to 59

Normal Abnormal

10.9 4.4

10.3 3.1 a 7.2<

10.2 4.9 a 5.3 6

10.8 3.5 a 7.3 6

a

7.6 4.6 3.0"

a

7.0 2.5 4.5<

a

7.9 2.7 5.2"

a

7.2 2.8 4.4<

a

7.4 4.9 2.5

a

7.1 3.1 4.0<

a

7.2 3.0 4.2

a

7.3 3.6 3.7<

Normal Abnormal

.I63C*

I

60 to 80

Lymphocytes/c.mm.

Normal Abnormal

% T cells

Normal Abnormal

10.0 4.5 a 5.5"

10.6 3.7 a 6.9<

Normal Abnormal

10.0 4.6 a 5.4"

Normal Abnormal

a

8.4 5.2 3.2"

a

9.4 5.0 4.4 b

a

9.7 5.4 4.3 b

T cells/c.mm.

% B cells

PHA

Normal Abnormal

Con A

Normal Abnormal

PWM

Normal Abnormal

10.2 3.8 a 6.4 e

I

Female

I

Male

a

4.4 6.8 b

10.2 4 a 5.9<

a

8.7 3.7 4.0<

a

9.2 3.0 6.2 b

a

9.2 4.2 5.0<

a

7.9 2.5 5.4 e

10.1 3.3 a 6.8 e

a

7.6 3.4 4.2 e

a

8.7 3.2 5.5 e

10.2 3.7 a 6.5 b

a

8.5 4.0 4.5<

a

8.0 3.0 5.0<

10.6 3.7 a 6.g e

a

8.4 3.8 4.6<

a

7.8 2.6 5.2 e

7.1 3.8 3.3

a

8.1 2.8 5.3<

11.2

a

For legend see Table I. *Superscriptletters indicate probability (p) that the two groups compared (normal versus abnormal) are of different populations based on immune status and survival. (3--0.05> P > 0.01, b-O.OI > P > 0.001, c-p < 0.001.)

small cell carcinoma who had normal values as compared to those in the same histologic categories with abnormal values for each of the parameters studied. Similar findings were observed for the group of patients with adenocarcinoma who had normal values for six of the nine parameters studied, Patients with undifferentiated large cell carcinoma with normal values for DNCB, percentage of T cells, T cells per cubic millimeter, and lymphocyte response to PHA likewise had significantly greater survivals than did those with abnormal values. A chi square analysis showed that the distribution of patients in the various stages of disease was similar for the two age groups. The data regarding the effects of age, immune competence, survival, and statistical significance for the 146 patients (all stages of disease combined) by age group are shown in Table V. A

statistically significant increase in survival time was noted in both age groups for those patients with normal values, as compared to those with abnormal values for nearly all parameters studied. The data regarding the effect of sex, immune competence, survival, and statistical significance for all stages combined are shown in Table V. Statistically significant prolongation of survival time was shown in both sexes when the patients with normal values were compared to those with abnormal values for nearly all parameters studied. Discussion

Many investigators have shown a significant increase in the proportion of patients with cancer who fail to be sensitized to DNCB when compared with controls. It has been shown that a greater proportion of patients

5 14

Liebler et al.

with bronchogenic carcinoma with resectable disease respond to DNCB than patients with unresectable disease. 13. 45. 51 These investigators also noted a correlation of poor DNCB reactivity and shorter time from diagnosis to death. Although our data regarding the proportion of patients with DNCB reactivity in the resectable group are similar to those of others, 13. 51 the majority of patients with unresectable disease in our study also responded to sensitization with DNCB. Therefore, we were unable to confirm the observations of Wells and co-workers'" Krant and colleagues.v and Holmes and Golub'" that a poor response or anergy to DNCB sensitization may be a reliable criteria for unresectabilty. The median survival data reported in this study for DNCB reactivity in the various stages of disease, however, do indicate an increase in survival time of several months for those patients with a good response to DNCB. Similar results, in longer survival times, were observed when the data for DNCB reactivity were analyzed for the various histologic type, age, and sex categories. There are conflicting reports in the literature regarding the value of DCH to microbial antigens in relation to stage of disease and survival. Several investigators have shown that many patients with bronchogenic carcinoma were unresponsive to microbial antigens. 9. 13-35. 39. 45-51 Some of these investigators also showed increased survival rates in those patients who had good DCH to the microbial antigens. Other investigators, however, have noted that nearly all patients with bronchogenic carcinoma responded to the microbial antigens, nullifying the prognostic significance of these tests;"- 51 The median survival data in this investigation in relation to DCH to microbial antigens and to stage of disease, histologic type, age, and sex categories indicated a statistically significant increased survival for the groups with a good DCH response, with the exception of the groups with localized disease, adenocarcinoma, and undifferentiated large cell carcinoma. However, the increase in months to median survival time between patients with normal and abnormal DCH responses was less in most groups, as compared to similar DNCB groups. This would indicate that DNCB reactivity has greater prognostic significance than the DCH response. Riesco.s? and Bill and Morgan'" have demonstrated a correlation between cancer curability and the total number of peripheral blood lymphocytes. Although the increased time in survival observed in this investigation between patients with normal versus abnormallymphocyte levels was statistically significant for all of the

The Journal of Thoracic and Cardiovascular Surgery

various stage, histologic type, age, and sex categories, the difference in time to median survival was minimal. There is controversy regarding the value of studies of lymphocytes exposed to plant lectins. Several investigators observed that a significant number of patients with bronchogenic carcinoma had a deficient lymphocyte response to PHA. 26 • 29. 33. 34. 50 Although Braeman and Deelly'" have shown that patients with bronchial carcinoma may have a deficient lymphocyte response to PHA, they were unable to demonstrate a correlation between survival rate and lymphocyte transformation studies performed before treatment. On the other hand, Dalbow and associates'" have shown that there was a correlation between lymphocyte stimulation by PHA and survival time of patients with lung cancer. The correlation observed was due largely to those patients with a poor response who died shortly after testing and those patients with a good response who lived more than 12 months. However, many patients whose lymphocytes responded to PHA also died shortly after testing. Others have stated that lymphocyte mitogen stimulation by PHA in patients with lung cancer are of little value, since the response of the lymphocytes from patients with cancer are similar to those of aged-matched control subjects.s? 32. 51 Holmes and Golub" were unable to demonstrate a deficiency in the response of lymphocytes to PHA in patients with cancer, as cornpared to age-matched control subjects; however, they did show that lymphocyte stimulation by PWM and Con A was significantly lower in patients with lung cancer when compared to age-matched normal control subjects. A considerable number of patients in this study had abnormal mitogen stimulation studies to each of the plant mitogens. Statistically significant increased survival rates were observed in all staging groups with a normal response to PHA, except for those patients with localized disease. Similarly, an increased survival rate was noted for those with normal values to PHA in histologic type, age, and sex categories. The additional lymphocyte stimulation studies with Con A and PWM added no greater information of prognostic value than that obtained from the PHA studies alone. A stage-related reduction in T cells in groups of patients with bronchogenic carcinoma has been reported.": 43. 50 Anthony and co-workers" stated that five patients with reduced E rosettes (percentage of T cells) had unresectable disease at thoractomy whereas only one of four patients with normal or elevated percentage of T cells had unresectable disease. These authors reported that length of survival was shorter in patients with epidermoid carcinoma who had a reduced

Volume 74 Number 4

Bronchogenic carcinoma

5 I5

October, 1977

percentage of T cells than in those patients with a normal percentage of T cells. However, they were unable to demonstrate a correlation with low percentage of T cells and survival in patients with oat cell carcinoma. Although Anthony and associates" have observed a correlation with time of survival and percentage of T cells, they were unable to show a relationship between T cells per cubic millimeter of peripheral blood and time of survival. Dellon, Potvin, and Chretien," in reporting on studies for a group of patients with squamous cell carcinoma of the lung, observed a stage of disease correlation with T cell levels (T cells per cubic millimeter) with progressively diminishing numbers of T cells in the more advanced stages. They also reported a similar reduction of T cells in patients with metastatic undifferentiated large and small cell carcinoma, as compared to T cells in patients with localized disease of the same histologic type. Since disease staging at time of diagnosis shows a direct correlation with survival rate, it would be expected that a stage-related decrease in T cell levels, as described by Dellon, Potvin and Chretien" also would be reflected by a shorter length of survival for these patients. A correlation of 86 percent between the patient's values for percentage of T cells and T cells per cubic millimeter (normal-normal, abnormal-abnormal) was observed in the data used in this investigation. 59 Therefore, it might be anticipated that, if a correlation was observed between time of survival and T cells per cubic millimeter, a similar correlation would be observed for percentage of T cells. A similar increase in time to median survival was observed in this study for those groups of patients with normal values in each of the stage of disease, histologic type, age, and sex categories for both percentage of T cells and T cells per cubic millimeter. Ivins and associates'" have described an increase in B cells in patients with progressive osteogenic carcinoma. Elevations in percentage of B cells also have been described in patients with bronchogenic carcinoma.s": 61 A statistically significant increase in time to median survival was observed in this study for groups of patients with regional and remote disease and all stages combined who had a normal percentage of B cells, as compared to those with elevated values, but the improvement in survival rate was minimal. Disease staging in cancer is important in making decisions as to the appropriate form of treatment and is also of value in determining prognosis. The results of this investigation indicate that knowledge of the immune status of the patient with lung cancer may be as

important in determinimg prognosis as knowledge of the stage of the disease. The time to median survival for patients with advanced bronchogenic carcinoma with normal general immune competence as measured by most of the tests used in this investigation was as long as or longer than that in the groups of patients in earlier disease categories with evidence of impaired general immunity. Effects of histologic type, age, and sex on median survival time in the various stage-of-disease categories cannot be excluded totally. It has been suggested that sex-matched and age-matched control subjects are necessary because of observed immune deficiency in the aged. 62 - 64 However, separate analyses of median survival data by histologic type, age, and sex indicate that the groups of patients with normal values for nearly each parameter of immunity studied had longer median survival rates than did those groups with abnormal values. Therefore, it would appear that the effects of histologic type, age, and sex on the results described in this study are not as critically related to survival rate as is the immune status of the patient. The increased survival rate noted in various disease stages for many of the parameters of general immunity studied is equivalent to what medical oncologists might consider to be a worthwhile prolongation of life in patients with advanced cancer. It is reasonable to consider that manipulation of the immune system might convert some patients with bronchogenic carcinoma from subnormal to normal immune competence (as measured by the more significant immune tests) and that conceivably this might be reflected by improvement in the survival of these patients. An investigation has been initiated under the aegis of the Radiation Therapy Oncology Group to study the therapeutic efficacy of intravenous Corynebacterium parvum in the treatment of patients with localized and regional bronchogenic carcinoma. The results of such a study should demonstrate whether immunologic competence can be maintained or restored and whether correlations can be established between maintenance and/or restoration of general immune competence and survival. We wish to gratefully acknowledge the technical and clerical contributions of Elaine Callery, Dru Ann Heath, Lucinda Kline, Ann Lanious, Electra Markopoulas, and Roy Summers to this study. REFERENCES Nordquist, R.E.: Specific Antigens in Human Alveolar Cell Carcinoma, Cancer Res. 33: 1790, 1973. 2 Tillock, T. W., Rosa, 1., and Vervynck, D. J.: Irn-

The Journal of

5 16

Liebler et al.

munologic Glycoproteins Isolated from Cell Membranes of Human Lung Carcinomas, 1. Natl. Cancer Inst. 52: 1059, 1974. 3 Frost, M. 1., Rodgers, G. T., and Bagshawe, K. D.: Extraction and Preliminary Characterization of a Human Bronchogenic Carcinoma Antigen, Br. J. Cancer 31: 379, 1975. 4 Bell, C. E., Jr.: A Normal Adult and Fetal Lung Antigen Present at Different Quantitative Levels in Different Histologic Types of Human Lung Cancer, Cancer 37: 706, 1976. 5 Veltri, R. W., Mengoli, H. F., Maxim, P. E., and Burrell, R. G.: Search for New Tumor Associated Antigens in Carcinoma of the Lung, in Nieburgs, H. E., editor: Proceedings, 3rd International Symposium on Detection and Prevention of Cancer, New York, 1976, Marcel Dekker, Inc. In press. 6 Veltri, R. W.: Sprinkle, P. M., and Burrell, R. G.: Isolation and Identification of Human Lung Tumor Associated Antigens (LTAA), Proc. Am. Assoc. Cancer Res. 17 131, 1976. 7 Mengoli, H. F., Veltri, R. W., Huang, C. W., Maxim, P. E., and Sprinkle, P. W.: Cationic Proteins of Normal Lung Tisuse and Lung Carcinoma, in Neiburgs, H. E., editor: Proceedings 3rd International Symposium on Detection and Prevention of Cancer, New York, 1976, Marcel Dekker, Inc. In press. 8 Hellstrom, 1., Hellstrom, K. E., Sjogren, H. 0., et al.: Demonstration of Cell Mediated Immunity to human Neoplasms of Various Histological Types, Int. 1. Cancer 7: I, 1976. 9 Alth, G., Deneck, H., Fischer, M., etal.: Imminological Status Before and During Immunotherapy (lmmunopprofile), Cancer Chemother. Rep. 4: 275, 1973. 10 Mavligit, G. M., Gutterman, J. U., McBride, C. M., and Hersh, E. M.: Cell-Mediated Immunity to Human Solid Tumors: In Vitro Detection by Lymphocyte Blastogenic Responses to Cell-associated and Solubilized Tumor Antigens, Natl. Cancer Inst. Mono. 37: 167, 1973. II Takita, H., and Brugarolas, A.: Adjuvant Immunotherapy for Bronchogenic Carcinoma: Preliminary ResuIts, Cancer Chemother. Rep. 4: 293, 1973. 12 Wells, S. A., Jr., Burdick, J. F., Christiansen, C., et al.: Demonstration of Tumor Associated Delayed Cutaneous Hypersensitivity Reactions in Patients With Lung Cancer and in Patients With Carcinoma of the Cervix, Natl. Cancer Inst. Mono. 37: 197, 1973. 13 Wells, S. A., Jr., Burdick, 1. F., Joseph, W. L., et al.: Delayed Cutaneous Hypersensitivity Reactions to Tumor Cell Antigens and to Nonspecific Antigens, 1. THORAc. CARDIOVASC. SURG. 66: 557, 1973. 14 Hollinshead, A. C.; Stewart, T. H. M., and Herberman, R. B.: Delayed Hypersensitivity Reactions to Soluble Membrane Antigens of Human Malignant Lung Cells, J. Natl. Cancer Inst. 52: 327, 1974. 15 Boddie, A. W., Jr., Holmes, E. c.. Roth, J. A., and Morton, D. L.: Inhibition of Human Leukocyte Migration

Thoracic and Cardiovascular Surgery

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30 31

in agarose by KCI Extracts of Carcinoma of the Lung, Int. J. Cancer 15: 823, 1975. Roth, J. A., Holmes, E. C., Boddie, A. W., Jr., and Morton, D. L.: Lymphocyte Responses of Lung Cancer Patients to Tumor-Associated Antigen Measured by Leucine Incorporation, J. THORAc. CARDIOVASC. SURG. 70: 613, 1975. Stewart, T. H. M., Hollinshead, A. C., Harris, J. E., et aJ.: Immunochemotherapy of Lung Cancer, N.Y. Acad. Sci. 277: 436, 1976. Takita, H., Minowada, J., Han, T., et al.: Adjuvant Immunotherapy in Bronchogenic Carcinoma, N. Y. Acad. Sci. 227: 345, 1976. Urist, M. M., Boddie, A. W., Jr., Holmes, E. c., and Morton D. L.: Capillary Tube Leukocyte Adherence Inhibition: An Assay for Cell Mediated Immunity in Cancer Patients, Int. 1. Cancer 17: 338, 1976. Urist, M. M., Boddie, A. W., Jr., Townsend, C. M., and Holmes, E. C.: In vitro Evidence for Increased Cellular Immunity to Lung Cancer Antigen During Levamisole Immunotherapy, J. THORAC. CARDIOVASC. SURG. 73: 189, 1977. Oldham, R. K., Siwarski, D., McCoy, 1. L., et al.: Evaluation of a Cell-mediated Cytotoxicity Assay Utilizing 125 Iododeuxyuridine-Iabeled Tissue-culture Target Cells, Natl. Cancer Inst. Mono. 37: 49, 1973. Vose, B. M., Moore, M., and Jack, G. D.: Cell-mediated Cytotoxicity to Human Pulmonary Neoplasms, Int. J. Cancer 15: 308, 1975. McKhann, C. F., Cleveland, P. H., and Burk, M. W.: Some Problems Involving In Vitro Cellular Cytotoxicity Assays, Natl. Cancer Inst. Mono. 37: 37, 1973. Takasugi, M., Mickey, M. R., and Terasaki, P. I.: Qunatitation of the Microassay for Cell-mediated Immunity Through Electronic Image Analysis, Natl. Cancer Inst. Mono. 37: 77, 1973. Takasugi, M., Mickey, M. R., and Terasaki, P. 1.: Studies on Specificity of Cell-mediated Immunity to Human Tumors, 1. Natl. Cancer Inst. 53: 1527, 1974. Ducos, J., Migueres, 1., Colombies, P., et al.: Lymphocyte Response to PHA in Patients With Lung Cancer, Lancet 1: I I I I, 1970. Thomas, J. W., Coy, P., Lewis, H. S., and Yuen, A., Effect of Therapeutic Irradiation on Lymphocyte Transformation in Lung Cancer, Cancer 27: 1046, 1971. AI-Sarraf, M., Sardesai, S., and Viitkevicius, V. K.: Clinical Immunological Responsiveness in Malignant Disease, Oncology 26: 357, 1972. Han, T., and Takita, H.: Immunological Impairment in Bronchogenic Carcinoma: A Study of Lymphocyte Response to Phytohemagglutinin, Cancer 30: 616, 1972. Paty, D. W., and Bone, G.: Response to PHA in Cancer Patients, Lancet 1: 668, 1973. Mekori, T., Shulamith, S., and Robinson, E.: Suppression of the Mitogenic Response to Phytohemagglutinin in MaJignant Neoplasia: Correlation With Clinical Stage and Therapy, J. Natl. Cancer Inst. 52: 9, 1974.

Volume 74 Number 4 October, 1977

32 Barnes, E. W., Farmer, A., and Penhale, W. J., et al.: Phytohemagglutinin-induced Lymphocyte Transformation in Newly Presenting Patients With Primary Carcinoma of the Lung, Cancer 36: 187, 1975. 33 Braeman, J., and Deeley, T. J.: The Lymphocyte Response and Prognosis in Cancer of the Lung, Br. J. Radiol. 48: 668, 1975. 34 Dalbow, M. H., Concannon, J. P., Eng, C. P., et al.: Lymphocyte Mitogen Stimulation (LMS) Studies for Patients With Lung Cancer: Evaluation of Prognostic Significance of Pre-irradiation Therapy Studies, J. Lab. Clin. Med. 90: 295, 1977. 35 Hughes, L. E., and Mackay, W. D.: Suppression of the Tuberculin Response in Malignant Disease, Br. Med. 1. 2: 1346, 1965. 36 Israel, L., Mugica, 1., and Chahinian, P.: Prognosis of Early Bronchogenic Carcinoma: Survival Surves of 451 Patients After Resection of Lung Cancer in Relation to the Results of Pre-operative Tuberculin Skin Test, Biomedicine 19: 68, 1973. 37 Takita, H., and Brugarolas, A.: Skin Test in Bronchogenic Carcinoma. I. Correlation of the Immunological Status and the Extent of the Disease, J. Surg. Oncol. 5: 315, 1973. 38 Brugarolas, A., Takita, H., and Vincent, R. G.: Skin Test in Bronchogenic Carcinoma. II. Its value in the Differential Diagnosis, 1. Surg. Oncol. 5: 319, 1973. 39 Anthony, H. M., Templeman, G. H., Madsen, K. E., and Mason, M. K.: The Prognostic Significance of DHS Skin Tests in Patients With Carcinoma of Bronchus, Cancer 34: 1901, 1974. 40 Wybran, J., and Fudenberg, H. H.: Thymus-derived Rosette-forming Cells in Various Human Disease States: Cancer, Lymphoma, Bacterial, and Viral Infections, and Other Disease, 1. Clin. Invest. 52: 1026, 1973. 41 Dellon, A. L., Potvin, C.; and Chretien, P. B.: Thymusdependent Lymphocyte Levels in Bronchogenic Carcinoma: Correlations With Histology, Clinical Stage, and Clinical Course After Surgical Treatment, Cancer 35: 687, 1975. 42 Gross, R. L., Latty, A., Williams, E. A., and Newberne, P. M.: Abnormal Spontaneous Rosette Formation and Rosette Inhibition in Lung Carcinoma, N. Eng!. 1. Med. 292: 439, 1975. 43 Anthony, H. M., Kirk, T. A., Madsen, K. E., et a!.: E and EAC Rosetting Lymphocytes in Patients With Carcinoma of the Bronchus. I. Some Parameters of the Test and of Its Prognostic Significance, Clin. Exp. Immunol. 20: 29, 1975. 44 Anthony, H. M., Kirk, 1. A., Madsen, K. E., et al.: E and EAC Rosetting Lymphocytes in Patients With Carcinoma of Bronchus. II. A Sequential Study of Thirty Patients: Effect of BCG, Clin. Exp. Immunol. 20: 41, 1975. 45 Krant, M. J., Manskopf, G., Brandrup, C. S., and Madoff, M. A.: Immunological Alterations in Bronchogenic Cancer, Cancer 21: 623, 1968.

Bronchogenic carcinoma

5I7

46 Brugarolas, A., Han, T., Takita, H., and Minownada, 1.: Immunologic Assays in Lung Cancer, N. Y. State J. Med. 73: 747, 1973. 47 Rees,1. c., Rossio, J. L., Wilson, H. E., et al.: Cellular Immunity in Neoplasia Antigen and Mitogen Responses in Patient With Bronchogenic Carcinoma, Cancer 36: 2010, 1975. 48 Stefani, S., Kerman, R., and Abbate, J.: Immune Evaluation of Lung Cancer Patients Undergoing Radiation Therapy, Cancer 37: 2792, 1976. 49 Kopersztych, S., Rezkallah, M. T., Miki, S. S., et al.: Cell-mediated Immunity in Patients With Carcinoma Correlation Between Clinical Stage and Immunocompetence, Cancer 38: 1149, 1976. 50 Concannon, J. P., Dalbow, M. H., Eng, C. P., and Conway, 1.: Immunoprofile Studies for Patients With Bronchogenic Carcinoma. I. Correlation of Pre-therapy Studies With Stage of Disease, Int. J. Radiat. Bio!. 2: 447, 1977. 51 Holmes, E. C., and Golub, S. H.: Immunologic Defects in Lung Cancer Patients, J. THORAC. CARDIOVASC. SURG. 71: 161, 1976. 52 Boyum, A.: Isolation of Mononuclear Cells and Granulocytes From Human Blood, Scand. 1. Clin. Lab. Invest. 97: 77, 1968. 53 Wybran, J., Carr, M. C., and Fudenberg, H. H.: The Human Rosette Forming Cells as a Marker of a Population of Thymus-derived Cells, 1. Clin. Invest. 51: 2537, 1972. 54 Catalona, W. J., Taylor, P. T., Rabson, A. S., and Chretien, P. B.: A Method for Dinitrochlorobenzene Contact Sensitization, N. Eng!. 1. Med. 286: 399, 1972. 55 Armitage, P.: Survivorship Tables, in Statistical Methods in Medical Research, New York, 1974, John Wiley & Sons, Inc., chapter 4. 56 Gehan, E. A.: A Generalized Wilcoxon Test for Comparing Arbitrarily Singly-censored Samples, Biometrika 52: 203, 1965. 57 Riesco, A.: Five Year Cure: Relation to Total Amount of Peripheral Lymphocytes and Neutraphils, Cancer 25: 135, 1970. 58 Bill, A. H., and Morgan, A.: Evidence forimmune Reaction in Neuroblastoma and Future Possibilities for Investigation, J. Pediatr. Surg. 5: III, 1970. 59 Concannon, J. P.: Unpublished data. 60 Ivins, J. C., Ritts, R. E., Jr., Pritchard, D. J., et a!': Transfer Factor Versus Combination Chemotherapy: A Preliminary Report of a Randomized Post-Surgical Adjuvant Treatment Study in Osteogeneic Sarcoma, N. Y. Acad. Sci. 277: 558, 1976. 61 Ritts, R. E., Jr.: Personal communication. 62 Gross, L.: Immunological Defect in Aged Population and its Relationship to Cancer, Cancer 18: 201, 1965. 63 Roberts-Thomson, I. C., Whittingham, S., Youngchaiyud, U., and Mackay, I. R.: Aging, Immune Response, and Mortality, Lancet 2: 368, 1974. 64 Girard, J. P., Paychere, M., Cuevas, M., and Fernandes,

5 I8

The Journal of Thoracic and Cardiovascular

Liebler et al.

Surgery

B.: Cell-mediated Immunity in an Aging Population, Clin. Exp. Immunol. 27: 85, 1977.

Discussion DR. E. CARMACK HOLMES Los Angeles. Calif.

I would like to compliment Dr. Liebler on his presentation and on his work. Certainly this is an extraordinarily complicated area. At UCLA we have been interested in the immunobiology of lung cancer for several years, and our studies with a few exceptions are similar to the ones that Dr. Liebler has presented. For instance, in a study of 93 patients with lung cancer, we also have found a profound suppression in the reactivity to DNCB. However, interestingly enough, the reactivity to the recall antigens is not suppressed especially in patients with early disease. I would like to ask Dr. Liebler if he has seen this difference in reactivity to DNCB and the recall antigens in patients with relatively early disease. As this disease progresses, all patients lose reactivity to the recall antigens. We have interpreted the loss of DNCB reactivity with the maintenance of recall antigen reactivity to be a defect in the antigen-processing limb of the immune response. I would be very interested to know if Dr. Liebler's group has made the same observation. It has been proposed that sequential monitoring of the immune response might be useful in determining the therapy that we should give our patients with lung cancer. Can we use sequential DNCB skin testing to reflect the effect of immunotherapy? Our studies indicate that we cannot. These studies indicate that sequential DNCB skin testing is not an adequate way to measure immune competence, because by repeated antigenic stimulation alone one can stimulate the response to DNCB in both the intact immune system and in patients who have had immunosuppressive therapy. Finally, we also have noticed a suppression of in vitro lymphocyte function in patients with lung cancer, and we have noticed that in some instances the immunorestorative agent Levamisole is capable of reversing this immunosuppression in the in vitro lymphocyte studies. In addition, an important observation is that the sera from patients with lung cancer are capable of suppressing normal lymphocyte function, that is, suppressing the function of lymphocytes from normal people. These serologic factors, present in the sera of patients with lung cancer, that suppress the immune response are closely related to the stage of the disease, to resectability, and to survival. I would like to ask Dr. Leibler if they have any data examining the serologic immunosuppressive factors in patients with lung cancer, because I think that a sequential assay of these factors may be more appropriate in monitoring immunologic competence in response to immunotherapy in patients with lung cancer.

DR. LUCIUS D. HILL III Seattle, Wash.

We congratulate Dr. Liebler's group in their search for two very important needs in our management of lung cancer: tests to determine (I) the extent of the disease and (2) the patient's response to therapy as well as the response to tumor. Over the past 16 months we have been working toward the same goals. We have been conducting a three-arm doubleblind randomized study with intrapleural bacille Calmette Guerin (BCG) by the McKneally technique in the first arm, BCG plus Levamisole in the second, and a control arm. We have been conducting the same studies that Dr. Liebler and Dr. Holmes have referred to, both in vivo and in vitro. However, we have been unable to correlate the DNCB test with either the patient's response to tumor or to therapy. Interestingly enough, out of this entire battery, which includes all of the tests that Dr. Liebler included in his study, the purified protein derivative (PPD) test has been the most reliable test for us. In the in vitro test, the total lymphocyte count and the T lymphocyte count have been the most helpful. At the start of this study, 26 patients out of 56 who were available for evaluation had PPD negative reactions. Of these, 16 have shown conversion to PPD positive reactions. This has been quite specific for the BCG-Levamisole group, only one patient in the control group having shown conversion. Thus there does appear to be some correlation with the patient's response to therapy. Whether this will help us in regard to the patient's response to the tumor remains to be seen. Since the study has been in effect for only 16 months, it is too early to determine this. We can say at this point that of the 5 patients with recurrent tumor, only one is in the PPD conversion group. The in vitro study has been most helpful. The T lymphocyte count, with the usual rosette formation with T lymphocytes and the total lymphocyte count, has correlated very well with PPD conversion as well as with the patient's clinical course. If these figures continue to hold up, we may have a helpful set of tests. We congratulate Dr. Liebler and we too will continue our search for reliable monitors in our management of lung cancer. DR. LI E B L E R (Closing) In reply to Dr. Holmes, our best test was tlie DNCB test, although we did not find a strict correlation between the DNCB test and the delayed hypersensitivity cutaneous test. In reply to Dr. Hill, the DNCB test, the T lymphocytes, percentage of T lymphocytes that were measured were our best significant correlation and correlated in every stage of disease with survival and with immune competence as we followed the 146 patients throughout their course.