Cell-mediated immunity in breast cancer patients

Biomed & Pharmacother 0 Elsevier, Paris

1996;50:337-343

Cell-mediated A Nicolini’*, ‘Inscituces

P Ferrari’,

of Clinical

Medicine

immunity

in breast cancer patients

R Spinelli’,

A Carpi’, A Sagripanti’,

University of Pisa, Pisa; ‘Unit of Allergology University of Pisa, Pisa Italy

F Ambrogi2

and Clinical

Immunology.

Summary - In clinical practice, reports have been made on immunosuppression after surgical excision of primary tumor or at relapse. However, the relationship between undefined or overt metastases and the host immune system has not been sufficiently examined over a prolonged period. These aspects were investigated in 160 breast cancer patients followed up post-operatively with serial controls over a long period. One hundred and thirty-four cases (91 node negative, (N-), 43 node positive (N+)) were disease-free and 26 relapsed. In all patients, serum T cell populations, serum B lymphocytes and skin reaction of delayed hypersensitivity (SRDH) were serially determined for 39 f 12 months (m f SD). The reference values for these parameters were assessed as follows: T populations were evaluated in 24 healthy donors and SRDH in 95 healthy females. In non-relapsed patients, constant CD8’ T cell decrease and T4/‘I8 ratio increase were observed; the T4/T8 ratio was significantly higher (ranging from P <0.05-P< 0.001) than in the control group. The mean values of NK cells and B lymphocytes, the former parameter being highly significant (P < O.OOl), were higher than in controls. In the 26 metastatic patients, the T4/T8 ratio from 20 months before to 30 months after the first sign of relapse decreased from 3.2 to about 1 (r = -0.256, P < 0.05) and from 30 to 92 months after relapse progressively increased to 2. Similarly, in the former subinterval a progressive decrease in the number of positive antigens and score was found (from 2.4 to 0 and from 10 to zero respectively). A significant inverse correlation between these two parameters and observation time occurred (P < 0.01 and P < 0.001 respectively). From 30 to 86 months after relapse, a progressive increase in the number of positive antigens and scores up to 2 and 12 were observed. A significant direct correlation (P < 0.05) was noted. In conclusion, these data indicate significant changes in T populations during the disease-free interval in breast cancer patients. The decrease in circulating CD8+ T cells is compatible with the hypothesis of CD8+ T cell localization at the site of the micrometastases. The increase in circulating B lymphocytes and NK cells suggests activation of aspecific humoral immunity and NK function. In addition, they show that progressive deficiency in cell-mediated immunity appears many months before and that recovery continues for a long time after overt metastatic disease. breast

cancer

/ relapse

/ immunosuppression

INTRODUCTION

prolonged latency observed in many instances, the frequent observation of immunosuppression at relapse and the occurrence of secondary tumors are main clinical data supporting this hypothesis. On the other hand, many experimental data suggest that host immunity might play a dual role in its response to metastases. In some tumor systems, immune cells can inhibit tumor spread, while in others lymphocytes may actively increase metastases [6, 201. Recent clinical reviews suggest the dual role of CM1 even in cases of breast cancer. In fact, in a few subsets treated in a dose-dependent fashion with drugs capable of inducing immunosuppression, dose-dependent regression of metastatic breast cancer has been

The concept of immunological surveillance against neoplasia was initially proposed by Ehrlich [5] in 1909 and later developed by Burnet [4]. It states that the normal function of the immune system, or more precisely cell-mediated immunity (CMI) provided by thymus-derived (T) lymphocytes, is to recognize and destroy newly appearing tumor cells in situ. The hypothesis of immunological surveillance has been critically reviewed by several authors and the role of CM1 at the onset of malignancies remains controversial. There is some evidence about its importance in the spread of cancer. The

* Correspondence

and reprints:

A Nicolini,

Istituto

Clinica

Medica

2a, Spedali

Riuniti

S Chiara,

Via

Roma

67, 56126

Pisa,

338

A Nicolini et al

Table I. T populations, T4/T8 ratio, NK and B lymphocytes (means k SD) in the peripheral blood of non-relapsed breast cancer patients and controls. pTNM

n

N- (total)

91

CD4+ (crlls/mLJ

CD8+ (cellslmL)

771

338f

i256

NS N+ (total)

43

749k292

24

777f241

Table II. SRDH non-relapsed controls. n

pTNM

N- (total)

Positi1fe antigens (mean + SD)

90. 1.64 i 0.9 NS

N+ (total) 41 Controls

95

6.40

5.7

+

6.7

0.62

177 f 72

NS

NS

x30* 143 NS

1,252f400

415e139

1.364i

Score (mean 2 SD)

i 4.4 NS

11/90

k 2.6 NS

5141

f

3.4

patients

n

o/o 12

P <

0.01 12

NS 28195

29

observed. Moreover, in patients which had undergone adjuvant locoregional radiotherapy, an increase in overall survival was found [21, 221. There is now evidence that cytokines produced by cancer infiltrating mononuclear cells can either stimulate or down-regulate cancer growth [14, 151. However, as regards the relationship between the immune system of the host and tumor metastasization, a number of problems are still unsolved. In this study, the time course of some important parameters connected with the immune system and occurring in the post-operative follow-up of breast cancer patients has been analyzed. In particular, all breast cancer patients have been serially followed up after surgery by clinical, instrumental and humoral examinations [ 18, 191. More recently, lymphocyte populations have been assessedby flow cytometry at the same time

as routine

examinations

were

carried

out.

Also CM1 function was examined by skin reaction of delayed hypersensitivity (SRDH) using the IMC Multitest (MCrieux). The aim of this paper is to retrospectively evaluate T cell populations, serum B lymphocytes and SRDH in the post-operative follow-up of breast

cancer patients.

2OOk85

NS

Anergic

CD19+ (cellslmL)

1,242 i- 404

P < 0.05 386

161+_67

MATERIALS

breast cancer patients and

1.6 3~0.6 NS 1.80

141

P < 0.05

NS Controls

CD2 + (rells/mL)

T4ITR (cellsfmL)

+ 1.0 P < 0.001

2.53

2.33

f

1.3

P < 0.05 1.82

f

CD56+ CDR(cells/mL) 132fl6

P < 0.001 132298

P
0.48

41f

13

AND METHODS

Patients From June 1989 to June 1993, 160 breast cancer patients post-operatively subjected to serial controls for a long period also underwent immunological evaluation. Consistent with axillary lymphnode involvement and pTNM classification [l], breast cancer patients were divided into the following classifications: 95 N-, 51 N+ and 14 Ml which showed distant metastases at the beginning of the study. Twelve patients suffered a relapse during the study period. Eight of them were N+ and four were N-. In 24 (92%) of the 26 relapsed patients, distant metastases and in the remaining two locoregional repetitions occurred: the time lapse was 69 +45 months (mean&SD) since mastectomy and 26 + 12 months since the beginning of the study. No adjuvant systemic therapy was administered to the Ngroup except those with locally advanced cancer (pT3 pT4). The latter patients and N+ patients were submitted to hormone and/or chemotherapy according to the presence or not of positive receptors or menopausal condition. Throughout the immunological study. T populations, B lymphocytes and SRDH were serially determined in the 160 breast cancer patients. The same parameters were also evaluated in two control groups, the former including 24 healthy donors and the latter 95 healthy females without any known condition capable of interfering with CM1 (tables I, II). In particular, either primary or secondary immunodeficiencies (bacterial and viral infections, corticosteroids or immunosuppressive drugs, tumors, diabetes, renal and liver failure, sarcoidosis) were reasons for exclusion from controls. All subjects over 70 years old were excluded from the study and the mean age (mean + SD) in the 160 breast cancer patients, the 24 healthy donors and the 9.5 healthy females was 55 * 9, 40 f 12 and 48 + 10 years, respectively.

339

Cell-mediated immunity in breast cancer

Follow-up Control visits were carried out every 4 or 6 months according to whether disease-free patients showed a higher or lower risk of relapse. All breast cancer patients were followed-up by a tumor marker panel with carcinoembryonic (CEA), tissue polypeptide (TPA) and 115 D8/DF3 (Ca 15-3) antigens. Routine laboratory examinations, skeletal X-ray, chest X-ray, bone scan, liver echography as well as a precise case history and clinical examinations were also regularly carried out, consistent with a previously reported fixed schedule [18, 191. During post-operative follow-up in most patients, tumor marker increase indicated relapse; lead time to radiological confirmation was 5.3 * 7 months (mean f SD). The mean follow-up time was 90 + 47 months (mean + SD) since mastectomy and 39 * 12 months since the beginning of the immunological study.

Methods Peripheral heparinized blood samples were taken in fasting patients and immediately transferred to the laboratory for measurement. Phenotypic analysis of the main lymphocytic populations such as CD2+ (total T cells), CDs+ (T suppressor/cytotoxic cells), CD4+ (T helper/inducer cells), CD19+ (B lymphocytes), CD56+ CD8- (NK cells) was determined by flow cytometry (FACStar, Becton-Dickinson, Mountain View, CA using the following monoclonal antibodies combined with phycoerythrin (PE) or fluorescein isothiocyanate (FITC): anti-CD2 (FITC), anti-CD8 (PE), anti-CD4 (FITC), anti-CD19 (PE), anti-CD56 (PE). The parameters were expressed as cells/ml. In addition, CD4+/CD8+ T cell ratio (T4/T8 ratio) was evaluated. All monoclonal antibodies were provided by BectonDickinson. SRDH was evaluated by IMC Multitest (Merieux) and three parameters were taken into account: number of positive antigens; anergic patients; score. IMC Multitest was performed on the forearm. It included seven recall antigens (tetanus, diphtheria, Streptococcus, tuberculin, Proteus, tricopyton, Candida) and glycerin as control. The expected SRDH at the antigen site was read 48 hours after the test was carried out and result calculated as half the addition of the value of the maximum diameter of the palpable infiltration to the perpendicular one. Two millimeters was considered the lowest measurable response. Erythema without palpable infiltration was not considered, and in the case of skin reaction at the control site, the test was invalidated. The score consisted of the total obtained by adding the results from all recall antigens and

anergy was defined when the score of all determinations was zero or when in few of them it did not exceed 2 mm. In controls in order to improve reliability of SRDH results, all values were obtained as the mean of three readings made by professional personnel working independently of one another. In the study group, the same staff alternated in reading serial determinations.

Statistical analysis All determinations except for those affected by any known condition capable of interfering with CM1 other than breast cancer were taken into account. In the 134 (91 N- and 43 N+) non-relapsed breast cancer patients, lymphocytic populations and SRDH were computed as the mean value (mean f SD) of all measurements performed and in each patient the mean value of all parameters was considered. They were compared with those of controls by unpaired r-test and significance was at the P < 0.05 level. The percentages of anergic patients were compared with those of controls by the x2 test. In the 26 metastatic patients, the SRDH and lymphocytic populations were evaluated by considering the significance of the regression lines obtained from all determinations carried out from 20 months before to 86 and 92 months respectively after the first sign of relapse. These intervals were selected because during this time the variations in lymphocytic populations and SRDH were more relevant. The whole interval was divided into two subintervals and two regression lines were taken into account: from 20 months before to 30 months after the first sign of relapse; regarding SRDH and lymphocytic populations, from 30 to 86 and from 30 to 92 months after relapse respectively. A graphic report was made: an interpolation line of all determinations (mean * SD) was carried out, then regression lines were drawn (figs l-3).

RESULTS T populations, NK, B lymphocytes and SRDH in non relapsed N- and N+ groups As shown in table I, when all 91N- and 43 N+ patients were taken into account, no significant lymphocytic CD4+ and CD2+ T cell variations occurred. In the N- group, significant CD8+ T cell decrease (P < 0.05) and in the N+ group significant

CD19

cell

increase

(P < 0.05)

were

found. In both groups, significant T4/T8 (P < 0.001 in N- and P < 0.05 in N+) and NK increases (P c 0.001) were found. As shown in

340

A Nicolini

et al

table II, regarding SRDH in the N- and N+ groups a slight statistically non-significant decrease in positive antigens and a score higher in N+ than in the N- group occurred. In both, a strong reduction in anergic patients which was significant in N- (P < 0.01; x2 test) was found.

creased, with a positive correlation but lower significance than in the former subinterval (r =0.449; P < 0.05, and r=0.496; P-co.05 respectively) (fig 3).

T populations, NK, B lymphocytes and SRDH at relapse

T populations, NK, B lymphocytes and SRDH in non relapsed N- and N+ groups

From 20 months before to 30 months after relapse, a significant inverse correlation (ranging from P < 0.05-P < 0.01) between CD4+, CD2+ T cells, T4/‘T8 ratio and the interval time was found, while a slight non-significant decrease in CD8+ T cells occurred (figs 1, 2). From 30 to 92 months after relapse, a significant positive correlation (ranging from P < 0.0% P < 0.01) between CD4+, CD8+, CD2+ and time interval was observed, while no variation in the T4/T8 ratio occurred (figs 1, 2). NK cells showed a significant decrease (P < 0.05) in the former subinterval, while B lymphocytes increased (P < 0.05) in the latter. With regard to SRDH from 20 months before to 30 months after relapse, a highly significant inverse correlation both of positive antigens (r= -0.338; P < 0.01) and score (Y = -0.378; P < 0.001) was found. From 30 to 86 months after relapse, both parameters in-

In the N- and N-t groups a significant T4/T8 ratio increase, and although not significant, a decrease in CD4+ and CD2+ T cells were observed. Therefore a constant CD8+ T cell decrease which has already been reported in early stages of different cancers [7], although slight and significant only in the N- group, is likely to explain the T4/r8 ratio increase and the tendency toward CD2+ T cell reduction (table I). The localization of circulating T suppressor/cytotoxic lymphocytes at the site of micrometastases by antigenic determinants of cancer cells could be the reason for the constant T4/T8 ratio increase in these apparently disease-free patients. A slight increase in B lymphocytes was constant although it became significant (P < 0.05) only in the N+ group. Conversely, NK cell increase was highly significant (P < 0.001) in both N- and N+ groups (table I).

DISCUSSION

OJ

Fig 1. CD4+ 0: time

,

,

and CD8+ T cells from 20 months before to 92 months after the first at the first sign of relapse; -: subinterval regression line; I: means f SE.

,

,

AND CONCLUSION

,

,

sign of relapse

,

,

,

,

,

in the 26 relapsed

,

(

patients.

Cell-mediated

OJ

r

,

,

-20

.

0

,

.

20

,

.

,

4)

immunity

,

60

,

.

x0

(

in breast

OJ

,

IW

341

cancer

.

-31

,

*

0

month.t h-om relapse

Fig 2. CD2+ T cells and T4/T8 ratio from patients. 0: time at the first sign of relapse;

1

.

20

months

20 months before to 92 months after the first -: subinterval regression line; I: means f SE.

1

.

40

sign

I

.

ho

fnm

I Ml

.

, Icnl

rclap\e

of relapse

in the 26 relapsed

16

12

4

0 -20

0

20

40

60

80

100

-20

0

2U

40

60

60

100

months from re1up.w

Fig 3. Number patients.

of positive antigens and score from 20 months before to 86 months after the first sign of relapse 0: time at the first sign of relapse; -: subinterval regression line; I: means f SE.

Activation of aspecific humoral and CM1 by minimal cancer disease, with the latter more strongly involved than the former, can be hypothesized. These findings have been partially confirmed in the experimental model where

in the 26 relapsed

NK/LAK cytotoxic activity is well represented in peripheral blood likely to prevent metastases rather than destroy the primary tumor [7]. A significant NK increase was also found in metastasesfree patients with malignant melanoma [2]. With

342

A Nicolini

regard to SRDH in the N- and N+ group, a constant decrease, even if non-significant, in positive antigens and score was observed. This finding suggests a tendency of non-relapsed breast cancer patients towards a lower CM1 than in controls. As CD4+ T cells are also involved in this reaction [12, 231, it indicates a reduction in T helper/inducer function in these patients. Neverthless, the reduction in number of anergic cases in the Nand N+ breast cancer patients could mask a normal or even higher SRDH response than in controls.

T populations, at relapse

NK, B lymphocytes

and SRDH

In the subinterval from 20 months before to 30 months after the first sign of relapse, the progressive and significant CD4+ T cell decrease which was higher than that of CD8+ T cells was responsible for the significant and progressive CD2+ T cell and T4/T8 ratio decreases (figs 1, 2). These findings are consistent with those previously reported that show a progressive decrease in T lymphocytes, CD4+ and CD8+ subsets with the spread of cancer [2, IO]. Towards the end of this subinterval and soon after, the mean value of the T4/T8 ratio dropped from 3.1 to about 1 (fig 2). Interestingly, the T4/T8 ratio of around 1 is reported by the analysis of T lymphocytes infiltrating murine virus-induced and human solid tumors [3, 161. From 30 to 90 months after the first sign of relapse in spite of CD4+, CD8+ and CD2+ T cells which showed significant positive regression lines (ranging from P < 0.05-P < O.Ol), no correlation between T4/T8 ratio and interval time occurred (figs 1, 2). These findings could be explained by a progressive replacement of CD4+ and CD8+ T cells from the sites of production to peripheral blood after their migration to the regions metastatic. Nevertheless, they suggest that CD4+ migration from the peripheral blood to the site of relapse is higher than that of CD8+ T cells, and that the opposite occurs from the sites of production to peripheral blood. Unexpectedly, NK cell decrease and B lymphocyte increase were only slightly significant (P < 0.05) and restricted to the former and the latter subintervals respectively. At least as regards B lymphocytes, this seems to agree with previously reported findings. In fact, in breast cancer patients at relapse, only an occasionally short significant increase in

et al

serum immunoglobulins. particularly IgG, IgA and IgM, has been found [ 171. The inverse correlation of positive antigens and score in the former subinterval is more significant than their positive correlation in the latter (fig 3). This indicates that at relapse there is a strong decrease in SRDH followed by recovery, possibly mainly in patients with prolonged response to therapy. Defects in immunoregulation in cancer have been widely reported [ 111. In particular, post-operative immunosuppression is well known [8] and has often been observed with increase in tumor mass and metastases [9, 131. Yet findings in this study indicate that in breast cancer at relapse, variation in CM1 is highly significant and that progressive immunosuppression appears many months before and continues many months after the appearance of overt metastatic disease. In conclusion, data from this study show that during prolonged post-operative follow-up in non relapsed breast cancer patients, a slight aspecific activation of humoral and a much higher activation of T cytotoxic/suppressor and NK cells occurs, perhaps in response to antigenic determinants of micrometastases. A few months before the appearance of overt metastatic disease the progressive reduction, and in some instances the inversion of the T4/T8 ratio with a highly significant progressive decrease in SRDH, indicates an important alteration of T helper/inducer and T suppressor/cytotoxic cells with relevant cellmediated immunosuppression. However, the relationship between these variations with regard to latency and spread of malignant disease is unclear, and further investigations are necessary. In particular, it seems important to define whether they play a role in the progression of disease. In the first instance, inclusion of suitable immunomodulators could be used in addition to conventional treatment. REFERENCES I American Joint Committee on Cancer Staging and EndResults Reporting. In: Beahrs OH, Myers MA, eds. Manual for &aging of Cancer. Second Edition. Philadelphia: Lippincott, 1983;127-133 2 Bernengo MG, Doveil GC, Appino A, Maasobrio R. Thymus hormones in malignant melanoma: personal experiences with thymostimulin (TP-1) and thymopentin (TP-5). In: Nagel G, Schioppacassi G, Schuff-Werner P, eds. Thymus Hormones in Oncology. Rome: Sedac srl, 1988;133-48