ASILI - UCLA Radiation Oncology

24 downloads 97498 Views 2MB Size Report
Radiological Sciences, UCLA School of Medicine, Los An- geles, California. ..... auto all0 allo allo allo allo allo allo allo allo allo allo allo auto allo auto auto auto auto allo auto ..... tumor response and eventual restoration of reac- tivity to recall ...
A PHASE I STUDY OF ACTIVE SPECIFIC INTRALYMPHATIC IMMUNOTHERAPY (ASILI) GUYJ. F. JUILLARD, MD, PAMELA J. J. BOYER,P H D ,

AND

CHARLES H. YAMASHIRO, MU

Twenty-one patients with advanced malignancies who had exhausted or refused conventional modalities of treatment were entered in a Phase I toxicology trial of active specific intralymphatic immunotherapy (ASILI). The patients were immunized with 1 X l o 7 to 1.2 X 10' viable autochthonous or allogeneic irradiated tumor cells intralymphatically each month and received no other antineoplastic treatment. T o date, 274 intralymphatic injections have been performed and except for one case of bacterial lymphangitis, no adverse side effects have been observed. ASILI did not significantly alter peripheral blood lymphocyte counts, absolute E-rosette forming cell levels, or EA-rosette forming cell levels. PHA reactivity of peripheral blood lymphocytes increased slightly in all but one patient tested. Seven out of nine patients who had not had delayed hypersensitivity to recall antigens developed positive reactions following ASILI. Sixteen out of twenty patients tested also developed reactivity to their immunizing cells after treatment. Objective regression (greater than 50% reduction of tumor mass) was observed in five out of nineteen evaluable patients. Six patients showed stabilization of tumor growth and eight patients continued to progress under treatment. Cancer 41:2215-2225, 1978.

A

CTIVE SPECIFIC IMMUNOTHERAPY FOR HUMAN

malignancies was first attempted by Von Leyden in 1902. '' Although subsequent efforts made during the following seventy-five years yielded only sporadically encouraging results, recent advances in understanding tumor-host immune relationships have led to renewed interest in the therapeutic possibilities of active specific immunization. The success or failure of any active immunization procedure is now known to be dependent on many interrelated factors such as the immunologic status of the recipient, the agent(s) employed, dose schedule, and the route of administration. In the case of specific tumor vaccines intended to induce or augment immune From the Division of Radiation Therapy, Department of Radiological Sciences, UCLA School of Medicine, Los Angeles, California. Supported by NCI grant CA 12800, UCLA Cancer Center Core Grant USPRS CA 16042-03, the Cassie Low Cancer Fund and various donors of the Division of Radiation Therapy, UCLA, and the Leo G. Rigler Center for the Department of Radiological Sciences, UCLA. Address for reprints: Guy J. F. Juillard, M.D., Division of Radiation Therapy, Department of Radiological Sciences, UCLA School of Medicine, Center for the Health Sciences; Los Angeles, CA 90024 The authors thank Dr. E. A . Langdon and Dr. J. L. Fahey for their support and cooperation, and J. Ambersley, B. Stanley, P. Painter, and the nursing staff of the UCLA Clinical Research Center for their excellent assistance. Accepted for publication October 3, 1977.

responses against putative weakly immunogenic tumor associated antigens, the localization of immunogenic materials at anatomic foci of potentially reactive lymphoid cells may be of particular importance. For this reason, the intralymphatic route was originally chosen for investigation of the induction of immune responses in an effort to concentrate weakly immunogenic materials in regional lymph nodes without disrupting the normal nodal architecture. The application of the intralymphatic route to active specific immunotherapy in canine malignancies showed significant clinical benefits with negligible toxicity. 7 ~ 8 , 9Concurrent animal investigations showed that the- intralymphatic route could be used safely and effectively to induce high levels of cellular immunity to membrane antigens.' A clinical Phase I toxicology trial of active specific immunotherapy (ASILI) was initiated in May 1976. Results of the first year of the study are reported in this communication.

MATERIALS A N D METHODS Patient Population A total of 21 patients were treated in this study. Criteria for patient entry to the protocol were: 1) failure of surgery, radiation therapy, and chemotherapy to control the progression of

0008-543X-78-0600-2215-0115

221 5

@ American Cancer Society

2216

CANCER June 1978

neoplastic disease, 2) a life expectancy of two to three months, 3) lack of any other concomitant antineoplastic treatment, and 4) informed consent of the patient. Exception to the criterion of having failed established modalities of treatment was made for two patients who had not received previous treatment; one patient because of her poor general condition, and the other patient because of her refusal of surgery and radiation therapy. Clinical Evaluation

Prior to initiation of treatment, each patient was clinically evaluated including physical examination, hematologic studies, electrolyte levels, SMA-12 and creatinine levels. Tumor status was evaluated by clinical and radiologic techniques. Immunoglobulin determination by immunoelectrophoresis, direct a n d indirect Coombs' tests, antinuclear antibody (ANA) by immunofluorescence and rheumatoid factor determination by latex fixation and sensitized sheep cell (SSC) assays were also performed to establish a baseline for the possible induction of autoimmune disease. After initiation of treatment, tests were repeated monthly when possible for two to twelve months. Skin Tests Patients were tested for delayed hypersensitivity to a battery of recall antigens prior to and serially after the initiation of immunotherapy. The following antigen preparations were used: Candida (Hollister-Stier), Mumps (Lily), Streptokinase-Streptodornase (Lederle), PPD5TU (Park-Davis), and Trichophyton (Hollister-Stier). Patients were sensitized to DNCB according to a modification of the methods of Bleumink et a1.4 Patients were skin tested with tumor cells by intradermal injection of 0.5 X 10' cells and 1 X 10' cells in 0.1 ml of saline. Skin tests were read as positive if erythema and induration were present at 24 hours with diameters of 5 to 8 mm for 0.5 X 10' cells and 7 to 15 mm for 1 X 10' cells. Preparation of tumor cell suspensions for skin tests is as described below for vaccine preparation. In Vitro Immunologic Evaluation

T and B cell enumeration and PHA stimulation of peripheral blood lymphocytes was performed prior to and serially after initiation of treatment according to the method of Elhilali et al.' Blood was generally drawn for zn vztro testing immediately prior to treatment which may have

Val. 41

prevented detection of transient changes in measured parameters following treatment. Normal ranges for these studies are based upon values for a control population of 203 healthy working individuals between the ages of 40 and 70. Since this population may not adequately control for the patient population under study, primary emphasis was given to longitudinal changes of these values. Normal ranges used are two standard deviations from the mean as follows: percent E rosette-forming cells: 64 f 5; absolute count; 1602 k 635 per mm3; percent EA rosetteforming cells: 18 f 6; absolute count: 455 f 205 per mm3. PHA: 1:lOO: 155,000 f 71,000; 1:2500: 15,266 f 8,604; No PHA: 1,500 k 600. Tumor Cell Vaccines Autochthonous and allogeneic cells were obtained from fresh sterile surgical specimens. Cell suspensions were frozen in a Cryo-med Programable freezer and stored in liquid nitrogen until used. Prior to use all preparations were tested for bacterial and fungal contaminants by conventional microbiological techniques. Donors of surgical specimens to be used for allogeneic vaccine preparation and donors of human AB serum were tested for the presence of Australia antigen3 and were excluded if they had had blood transfusions within three months or had a prior history of hepatitis. T h e dosage of irradiation which would prevent further cell proliferation was determined by W-thymidine incorporation studies in vitro. Immediately prior to use, the frozen cells were thawed rapidly, passed through sterile Nitex screens to remove cell clumps, washed, counted, and suspended to a final concentration of 1 X lo' viable cells per ml in sterile Hartmann's solution (Travenol) and irradiated to 10,000 rad in a Gammacell 220 over approximately two minutes. Immunotherapy Cannulation of the afferent lymphatic vessels was pqrformed under local anesthesia (1 ml of 2% xylocaine) according to the technique of Kinmonth (10). The incision was kept as small as possible (no larger than 1 cm) in anticipation of multiple treatments. The site of cannulation was on the dorsum of the feet and hands where the lymphatic vessels which drain the upper and lower extremities are readily accessible. Although the number of sites infused per treatment varied from one to four,

ASILI

No. 6

9

Juillard et al.

most treatments involved infusion of two to three sites (average: 2.2 sites per treatment). T h e concentration of the infused material was 1 X 10’ viable cells per ml for all treatments. The total volume infused varied throughout the study from 1 to 5 ml per site. An average of 2.5 ml per site was given. Treatment was given monthly on the basis of extrapolation from preclinical investigation (refs. 7, 8 and unpublished data). When disease could no longer be detected in responding patients, treatments were given at two month intervals. Seven patients received autochthonous tumor cells (Table 3) which had been processed and stored following prior surgeries. T h e other fourteen patients received histologically similar allogeneic tumor cells from the same anatomic location. RESULTS Toxicity of ASILI

The 21 patients treated in this study are summarized in Table 1. Assigned patient numbers are consistent throughout this report. At the beginning of the study, all patients were admitted to the hospital for monitoring after ASILI treatment. During the last six months of the study, however, patient demand and the previous lack of hazardous or difficult side effects led to the treatment of most participating patients on a n outpatient basis. The progression of the vaccine material along the lymphatic vessels of the limbs was felt in 174 infusions (66%). A sensation described as “tingling, ” “aching, ” or “pressure, ” usually occuring near the end of the infusion, was reported by the patients. All patients reported that the sensation stopped abruptly when the infusion was discontinued. Twenty of the 21 patients found the discomfort of the procedure to be minimal. One patient reported the infusion to be “painful,” but this pain disappeared at the end of the infusion (when pressure was discontinued). T h e average duration of each procedure was one hour for the infusion of two sites. None of the patients required additional analgesics for the procedure and patient discomfort never resulted in early termination of a procedure or withdrawal from the study. Temperature, blood pressure, CBC, electrolytes, SMA-12 and creatinine levels were monitored on all patients and did not show alterations subsequent to immunotherapy. O n e case of lymphangitis occured following a pa-

2217

tient’s fifth treatment, but subsided rapidly upon administration of antibiotics and did not recur after subsequent treatments. No other complications of treatment have been observed or reported in the course of 274 infusions (1 17 treatments) in this study. No late developing side effects of treatment have been observed in ten patients with 3 to 7 months follow-up and six patients with 8 to 13 months follow-up. In an effort to monitor the possible induction of autoimmune effects, direct and indirect Coombs’ tests, serum immunoglobulin determinations, latex fixation and sensitized sheep cell assays for rheumatoid factor and immunofluorescent tests for antinuclear antibody were performed on patients before initiation of treatment and serially during treatment and followup. These tests showed that prior to immunotherapy, one patient (3) had detectable antinuclear antibody and two patients (15 and 16) had positive indirect Coombs’ tests. Five other patients (2, 6, 7, 9 and 17 were initially negative for antinuclear antibody and became positive at varying intervals after initiation of immunotherapy. Patients 2 and 5 who were positive during treatment have reverted to negative after 4 and 8 treatments. Additionally, patients 7 and 9 also developed positive latex agglutination tests. All Coombs’ and SSC tests which were negative prior to treatment remained negative and no serum immunoglobulin abnormalities were detected. Only one of the five patients who developed positive tests had been immunized with autochthonous tumor cells. These findings suggest the need for close follow-up, but are not felt to be indicative of active autoimmune disease and no clinical manifestations of disease have been found. However, the possibility of immunotherapeutic autoimmunization cannot be excluded. Autopsies were performed on four patients who died during the course of the study. Causes of death in all cases appeared to be tumor related, e g . , intestinal fistula, bilateral renal fistula, and in two cases, cachexia. No pathologic changes could be found grossly or microscopically which appeared to be referable to immunotherapy. Immunologic Parameters

Delayed hypersensitivity to recall antigens: Fourteen patients were skin tested with Candida, mumps, PPD, SK-SD and trichophyton both prior to and after immunotherapy. As shown in Table 2, prior to immunotherapy, five patients

CANCER June 1978

2218

Vol. 41

TABLE 1. Summary of Patient Population Patient Sex

Age (yrs.)

1

F

56

2

F

52

3

F

41

NO.

61 58 50 7

F

36 60 69 64 49

12

M

67

13

F

73

14

M

64

15

F

52

16

F

56

17

F

33

18

M

62

19

F

67

20

M .

68

21

F

19

Diagnosis and measurable tumor mass Disseminated hypernephroma Retroperitoneal8 X 17 X 14 cm (3) Supraclavicular 5.5 X 3.5 X 3.5 cm ( 1) Recurrent serous cystic carcinoma ovary. Pelvic 10 X l 0 c m (1) Multiple peritoneal mets (5) Recurrent mucinous ovarian carcinoma R U Q 7 X 5 X 5 cm (1) Fixed pelvic mass (1 ) colostomy Recurrent epidermoid carcinoma cervix, vagina. 5 X 5 X 3 cm (1 ) Epidermoid carcinoma maxilla, antrum, orbit. 6 X 4 X 6 cm (2) Adenoid cystic carcinoma parotid, lung & bone mets. Multiple pulmonary mets (2) Serous cystic carcinoma ovary (poorly differentiated). Pelvic7X7XGcm(l) Serous cystic carcinoma ovary 18 X 15 X 15 cm (1, 2) Serous cystic carcinoma ovary 20 X 25 X 15 cm (1, 2) Recurrent epidermoid carcinoma tongue,4X5X3cm(l) Recurrent epidermoid carcinoma cervix. Fixed pelvic mass & colostomy, inguinal lymphadenopathy 3 X 2 X 2cm(l) Disseminated hypernephroma Scalp & skull 10 X 15 X 8 c m ( l ) Recurrent endometrial carcinoma. Fixed pelvic mass (1 ) colostomy, cystostomy Recurrent breast carcinoma. Skin nodules 1.5-1.8 cm ea (1) Serous cystic carcinoma ovary 30 X 25 X 20 cm (2, 2) Recurrent retroperitoneal leiomyosarcoma. 16 X 12 cm (1) Serous cystic carcinoma ovary 25X23X15cm(l,3,4) Disseminated Iymphocytic lymphoma. Multiple lymphadenopathies, 2 to 4 cm (1, 2) Leiomyosarcoma (jejunum & widespread metastases). Peritoneal nodules 2 cm (1, 5) Recurrent epidermoid carcinoma oftongue. 6 X 5 X 4 c m (1) Recurrent embryonal rhabdomyosarcoma. Pulmonary mets 3 X 3 X 3 (2)

Previous treatment

No. ASILI

No. Infusions LE

UE

3

3

3

Surgery, R.T., Chemotherapy 2 mo.*

12

9

17

Chemotherapy 3 mo.*

4

0

8

Sugery, R . T .

6

1

12

11

9

17

6

5

11

Chemotherapy 1 mo.*

in

11

16

Chemotherapy 3 mo.*

1

0

chemotherapy 1 mo.*

8

9

R.T.

8

6

Surgery, R.T. 1 mo.*

4

1

Surgery, R.T. 2 years*

6

5

6

Surgery, R.T. Chemo., Hormones 1 mo. *

2

2

4

Surgery, chemotherapy. 15 mo.* Chemotherapy 2 weeks*

3

2

8

8

1

1

8

14

Chemotherapy, R.T. 2 mo.*

5

7

Surgery, chemotherapy. 3 mo.*

4

3

in

R.T., chemotherapy 1 % mo.* R.T. surgery, chemotherapy 9 mo.*

3

4

4

4

4

4

None (due to poor general condition)

None

Surgery, chemotherapy, R.T.

Surgery, chemotherapy. 1 mo.* Chemotherapy 1 mo.*

.~

(1) = Clinical measurements; (2) = x-rays; (3) = Ultra-sound scans; (4) = radioactive liver scan; (5) = laparotomy. * Interval between last chemotherapy and/or radiation therapy and ASILI. Volumes were calculated using the formula for an ellipsoid (V = 4/3 ?r abc).

No. 6

Asm

2219

Juillard et al.

TABLE 2. Delayed Hypersensitivity Reactions of ASILI Patients ~-

Before ASILI Patient No.

Recall antigens*

DNCB

Vaccine!

1

2 3 4

5 6 7 8 9

10 11 12 13 14 15 16 17 18

-

-

not done

-

19 20

21

Vaccine Source auto all0 allo allo allo allo allo allo allo allo allo allo allo auto allo auto auto auto auto allo auto

After initiation of ASILI Recall antigens* DNCB Vaccine

+ + + + + + + not repeated + +-

+ + +

+ -

+

+ + + + + +-

+

+

+ ++ + +

-

-

not done

not done

+

-

c

-

+ +

not repeated

+ +

+

+-(3 months)

+ -

-4

+-

-

-

( 3 months)

* (+)

indicates that the patient had positive reactions to two or more antigens of the skin test battery described in Malerials and Methods. (-) indicates reactivity to less than two antigens. + Tested at time of first ASILI.

had positive delayed hypersensitivity to two or more recall antigens and nine failed to respond to two or more of the antigens used. Those patients who had positive reactions to recall antigens retained their reactivity throughout treatment and follow-up with the exception of patient 21, who progressed under treatment and lost reactivity to recall antigens at three months. Of the nine patients who failed to respond to recall antigens before ASILI, seven regained delayed hypersensitivity to two or more antigens one to three months after initiation of treatment. Similarly, all patients, except patient 21, who could be sensitized to DNCB prior to treatment, retained their reactivity. Of the ten patients who could not be sensitized to DNCB before treatment, four responded to challenges given after initiation of treatment. In this study, patients were not given repeat sensitizing doses of DNCB after starting immunotherapy which may in part account for the observed lack of responsiveness upon rechallenge. None of the 20 patients who were skin tested with their own tumor cells or allogeneic cells of the same histologic type before initiation of immunotherapy showed any signs of reaction to the cell vaccines. Of these 20 patients, 16 developed reactivity to their vaccine tumor cells after

immunotherapy. Average time of development of delayed hypersensitivity after the first ASILI treatment was 2 months or 2 treatments. Eleven of the patients who developed positive skin tests to tumor cells had received and been tested with allogeneic tumor cells, and the observed delayed hypersensitivity may have been due to sensitization to disparate histocompatibility antigens of the immunizing tumor cells. Several of these patients also reacted to other allogeneic tumor cells of the same histologic types; however, data on the specificity of these reactions are currently insufficient to draw any conclusions. Of greater interest is the observation that five of the patients who developed positive delayed hypersensitivity to vaccine cells had been treated and skin tested with autochthonous tumor cells. Controls for these skin tests included intradermal testing with serum-containing freezing medium and with the saline solution used to suspend the cells for injection which were negative. Several of these patients also developed reactivity to other tumor cells of similar histologic type, but the greatest diameters of erythema and induration were usually obtained with their own cells. These skin test observations suggest the development of antitumor delayed hypersensitivity responses following ASILI which is

2220

CANCER June 1978

further supported by the observation that all five of these patients showed objective tumor response and clinical improvement which will be discussed in the next section. All patients in this study who regained delayed hypersensitivity to recall antigens also acquired reactivity to their immunizing materials; however, not all of the patients who developed positive reactions to tumor cells were able to respond to recall antigens. Enumeration of patients’ E and EA rosetteforming lymphocytes was performed serially before and during treatment in an effort to evaluate the effect of ASILI on peripheral blood presumptive T and B levels. Mean values of absolute numbers of E-RFC per mm3 for the total patient population both before and after ASILI are significantly lower (p < 0.00s) than the arbitrary control population described in Materials and Methods; however, the percentages of E-RFC remain within normal limits. This depression of absolute E-RFC is related to the fact that the mean total lymphocyte count for this patient population during the course of the study was 924.6 f 479.4 lymphocytes/mm3 as compared to 2,497.0 f 840.1 lymphocytes/mm3 for the control population. The reduction of the total lymphocyte count of the patient population under study may reflect extended histories of prior antineoplastic treatment, or poor general physical condition as depression of total lymphocyte counts was also observed in two patient who had not had any prior treatment. Determination of absolute number of EARFC per mm3 of peripheral blood also revealed significantly reduced levels (p < 0.005) in the patient population before and during ASILI as compared to normal controls. These studies indicate that ASILI did not appreciably alter either the absolute or relative levels of peripheral blood B and T cells in this patient population. When E-RFC and EA-RFC values of the patients who responded to treatment and patients who did not respond were compared, neither group was found to change significantly with treatment. Responding patients differed from nonresponding patients only in absolute E-RFC levels before treatment, where it was observed that responders had higher (p < 0.025) absolute E-RFC levels than nonresponders. It should be noted, however, that even the responding patients with less lymphocyte depletion are more likely to respond to immunotherapy. The difference between responders and nonresponders following treatment was not significant. PHA reactivity of unfractionated patients’

Vol. 41

lymphocytes was performed before and after ASILI. Although responses are generally low, all patients except patient 6 showed an increase at 1/ 1000 during ASILI treatment. There were no significant differences between responding and non-responding patient populations.

Evaluation of Tumor Response T h e current status of the 21 advanced cancer patients involved in this study is shown in Table 1. Early deaths occurred in two patients two and three weeks following initiation of intralymphatic immunotherapy. The causes of death were found to be bilateral ureteral obstruction (patient 8) and pyelonephritis with multiple fistulas (patient 16). No pathologic changes referable to intralymphatic immunotherapy could be found at the time of death. The remaining 19 patients are evaluable for tumor response following ASILI. It should be emphasized once again that all patients had advanced disease and an average life expectancy of approximately two to three months a t the time of initiation of treatment. Eight of the patients experienced continued progression of disease with little or no objective reduction of tumor mass. Five of the eight patients (4, l l , 12, 13 and 15) died between two and five months after initiation of immunotherapy with a mean survival of 3.2 months. Patient 4 in this group, though not meeting the criterion of objective response, showed a marked slowing of disease progression and died of sepsis after pelvic exenteration which followed the immunotherapy. Two patients (20 and 21) are alive with progressive disease after four and three months of immunotherapy. Neither patient has developed delayed hypersensitivity to immunizing materials as shown in Table 2. One patient (7) had progressive disease with restored skin test reactivity to both recall antigens and her vaccine materials. After five months of immunotherapy, concomitant chemotherapy at low doses was initiated and tolerated. At nine months, the patient is alive with progressive disease. Stabilization of a tumor which was progressing prior to immunotherapy was defined as the failure to detect any objective change in tumor size for three months, or less than SO% objective response, associated with subjective improvement or status quo. Such stabilization of tumor growth was observed in patients 1, 3, 6, 9, 14, and 18 for periods of two, three and one-half, six, eight, seven and three months, respectively. These patients all acquired delayed hypersensitivity to vaccine materials and to recall an-

ASILI

No. 6

222 1

Juillard et al.

TABLE 3. Tumor Response to Active Specific Intralymphatic Immunotherapy (ASILI) Patient

Vaccine source

Tumor response

1

auto

2 3

allo allo

4

allo

5

allo

Partial response (50% decrease)

6

allo

7 8 9 10

allo allo allo allo

11 12 13 14

allo allo allo auto

15 16 17

allo auto auto

18 19 20 21

auto auto allo auto

Stabilization 6 m, then slow progression Slow progression Not evaluable Stabilization Partial response (75% decrease) Progression Progression Progression Stabilization (30% decrease) Progression Not evaluable Partial response (50% decrease) Stabilization 3 m complete response Progression Progression

Follow-up (months)

Stabilization (30% decrease, temporary) Complete response Stabilization (30% decrease) Slow progression

tigens following ASILI. Four of them are still alive (Table 3). Tumor response was defined as an objectively measurable decrease in the size of the tumor mass by at least 50%. Five patients (26%) demonstrated such a response as well as subjective signs of improvement such as weight gain, increased activity, relief of pain, etc. after three to five treatments of ASILI. Additionally, all responding patients developed delayed hypersensitivity to the immunizing materials. T o date, they have received ASILI for four to thirteen months with an average treatment time of 7.8 months. Patient 5, originally presenting with a large epidermoid carcinoma of the maxillary antrum invading the orbit with ptoptosis and diplopia, had refused surgery and radiation therapy. Figures 1 and 2 show computerized tomograms of patient 5 immediately prior to initiation of ASILI and document the response to treatment after 2 95 and 4 months. During the following three months, immunotherapy was discontinued due to a lack of suitable vaccine and the tumor began to regrow in the infrastructure of the maxillary antrum. At this time, the patient consented to radiation therapy

2 m, dead, hypercalemia, extensive disease 13 m, alive, NED 3 % m, dead, intestinal fistula 5 m, dead, complications of pelvic extenteration done because of slow progression. 9 m, alive. Radiation therapy added at 7 m to infrastructure of maxillary antrum, orbit was not irradiated. Apparently free of disease. 9 m, alive 9 m, alive, chemotherapy added after 5 m 2 w, dead, bilateral ureteral obstruction 8 m, alive, with disease 8 m, alive with disease

2.5 m, dead, cachexia 3 m, dead, cachexia

2.5 m, dead, cachexia 7 m, alive with disease 3.5 m, dead, intestinal obstruction 3 w, dead, pyelonephritis with multiple fistulae 6 m, alive with disease

5 m, alive, chemotherapy added after 3 m 4 m, alive, NED 4 m, alive with disease 3 m, alive with disease

of the rnaxillary antrum, sparing the orbit. Irradiation was performed with an excellent response and the orbit, which was not irradiated, has remained free of tumor at nine months follow-up. T h e patient currently has no evidence of d’isease. Patient 10 presented with recurrent epidermoid carcinoma of the tongue, adjacent to an area which had been irradiated one year previously with complete response. Supplemental interstitial radiation therapy was refused at this time. The patient returned for treatment five months later because of pain, loss of weight and recurrent episodes of bleeding. The tumor volume was found to have quadrupled and was no longer amenable to treatment by interstitial method. Chemotherapy was not recommended for this 75 lb patient with abnormal liver functions studies. Two months after initiation of ASILI, the patient began to show signs of subjective improvement, gained 8 lbs and had no further bleeding. A 75% decrease in tumor size was observed by the third month of immunotherapy and has been maintained a t 8 months. Two patients (2 and 19) are apparently free of disease at 13 and 4 months respectively. Patient

2222

CANCER June 1978

Vol. 41

FIG. 1. Epidermoid carcinoma of the maxillary antrum. Patient 5. Computerized axial tomographs on the left show a large tumor of the left maxillary antrum and proptosis before treatment. Pictures on the right show reduction of the tumor and almost complete reduction of the proptosis after 2 % months of ASILI.

2 who presented with advanced ovarian cystadenocarcinoma showed remarkable improvement and decreases in pelvic and abdominal masses during the first six months of immunotherapy. After nine months of immunotherapy, an unrelated cholecystectomy for lithiasis was performed. At the time of surgery, less than 1 g of residual tumor was found and removed for pathologic examination. The specimen was heavily infiltrated with lymphocytes and histiocytes and contained less than 5% viable tumor cells. Additional multiple biopsies revealed no evidence of tumor. The patient continues to be well with no evidence of disease at 13 months. Patient 19 who presented with palpable pelvic metastatic leiomyosarcoma confirmed by laparotomy and biopsy, also responded rapidly to treatment and at four months has no evidence of disease by any available criteria. Patient 17 was seen 18 months after a Stage IV serous cystadenocarcinoma of the ovary had

been diagnosed by exploratory laparotomy and biopsy. Chemotherapy was successful initially, but the tumor began to progress despite various chemotherapeutic combinations which were finally abandoned. Autochthonous tumor cells were obtained from ascitic fluid and the tumor mass was estimated as approximately 5 kg after paracentesis. O n e month after initiation of immunotherapy, the patient began to pass normal sized stools for the first time in one year. During the following five months, the tumor mass shrank to 50% of its original size. Ascities has continued to be present, but its cytology has changed markedly, showing predominantly small and medium size lymphocytes with a 70% decrease in tumor cell concentration. Direct cell mediated cytotoxicity of the patient’s peripheral blood lymphocytes toward autochthonous tumor cells was detectable. All patients who demonstrated a positive tumor response developed positive delayed hyper-

No. 6

ASILI

sensitivity to immunizing materials prior to indications of clinical response. It is of interest that four of the five patients who responded to immunotherapy had positive skin test reactions to recall antigens prior to treatment and one did not. All the responding patients were able to demonstrate positive skin test reactivity to recall antigens after treatment. In this study, patients received either allogeneic or autochthonous tumor cell for immunization. Two responses (33%) were observed in six patients who had received autochthonous tumor cells and three (23%) in thirteen patients who had received allogeneic cells. The two responding patients who had received autochthonous cells developed vigorous reactivity to the cells, but not to control, which strongly suggests an immune basis for the observed responses. Tumor burden which usually correlated with the general condition of the patient, was found to be indicative of the potential efficacy of ASILI. Three objective responses and one stabilization were observed in 6 patients who had less than 100 g of tumor at initiation of treatment. The other two patients who did not benefit from treatment interestingly failed to develop positive skin tests to the tumor vaccine. Only two objective responses occurred in 13 patients presenting with large amounts of tumor ranging from 500 g to 5000 g. However, one of these (patient 2) has no evidence of disease at 13 months. A striking correlation was noted between the response to immunotherapy and previous response to other treatments. Nineteen patients had previous chemotherapy and/or radiation therapy. N o response was observed in eleven patients who had not responded to previous treatments, although two stabilizations and two

FIG.2. Pictures of coronal computerized axial tomography on the left show the orbital involvement prior to treatment. Patient 5. Picture on the right shows no tumor in the orbit and return of the globe to a symmetrical position 4 months after ASILI as the sole treatment.

Juillard et al.

2223

slow progressions were observed. Four responses and three stabilizations were observed in the seven patients who had temporarily responded to previous treatment.

DISCUSSION The results of this study suggest that some of the observations made in preclinical investigations are applicable to humans. These include: 1) Intralymphatic injections of immunogenic materials are relatively safe and free from deleterious or painful side effects, as long as no viable microorganisms are injected. The only acute complication which was observed in 117 treatments in this study was of bacterial origin and although the level of contamination which had escaped routine screening was probably minimal, it was sufficient to cause bilateral lymphangitis. The hazards of bacterial lymphangitis are sufficient to exclude the use of the intralymphatic route for the intentional introduction of viable microorganisms for any purpose. 2) No acceleration or enhancement of tumor growth has been observed secondary to ASILI. The mean survival of the 7 patients whose neoplasms continued to progress under treatment was 2.6 months which is not inconsistent with the expected survival of such a patient population with very advanced and intractable disease and includes two early deaths resulting from unmanageable complications of their previous clinical status. 3) The viability of the immunizing tumor cell preparation was of major importance in predicting the efficacy of active immunotherapy. In previous animal ~tudies~*~+’ we had observed that cell preparations containing low percentages of viable tumor cells (e.g., 5-15% viable) failed to

2224

CANCER June 1978

stimulate a tumor reductive response in the immunized animal although the total number of viable cells was not reduced. As a consequence, in the human study, efforts were made to provide the highest possible viability in cell preparations for immunization. One exception was patient 21 who for technical reasons was immunized with F cell preparation which was approximately 5% viable. Although the patient received the same total number of viable cells as other patients, her failure to respond, despite positive skin tests and low tumor burden, may be analogous to the lack of response exhibited by experimental animals receiving low viability cell preparations. 4) Active specific intralymphatic immunotherapy can elicit the regression of large amounts of tumor. Although most responding patients in this study initially had relatively greater tumor burdens than had been studied in animal experiments, their response was predicted by previous correlations of response vs. tumor burden. Since most attempts at active specific immunotherapy have not been effective in the presence of a large tumor burden in humans, the difference in efficacy may be attributable to the unconventional route of immunization. Animal studies have shown t h a t immunization via the intralymphatic route preferentially stimulates high levels of cellular immunity in normal animals'as compared to other routes of immunization. Similar phenomena may be operative in the immunotherapeutic application of this route. The fate of the intralymphatically injected vaccine materials has not been investigated in humans; however, a variety of radioactively labelled cellular materials, e.g., autologous, allogeneic, xenogeneic, normal and malignant cells has been injected via different routes in animals and have shown consistently different distribution patterns dependent upon the route of injection. Unique to the intralymphatic route is the observation that when up to 10' viable cells are injected via the afferent lymphatics, the injected cells are almost quantitatively trapped by the first echelon of regional lymph nodes and remain in these nodes for several weeks Uuillard et al., manuscript in preparation). Subsequent serial histologic examination of these nodes shows a predominantly lymphoblastic reaction. The only exception to this pattern following intralymphatic immunization was observed when the lymph nodes were involved by malignant disease, in which case the injected material was found as high as the second and third echelons of lymph nodes. In addition to information which could be considered to be analogous to previous observa-

Vol. 41

tions in animal experiments, the present study has provided information which could not be extrapolated from preclinical investigations and which may have bearing on the immunologic assessment of advanced cancer patients. In this study, the presence of delayed hypersensitivity to recall antigens appears to be related to the amount of tumor burden which has been observed by other investigators and interpreted as a favorable prognostic sign. An unexpected observation in this study was the apparent restorative action of ASILI on recall antigen skin test reactivity in previously anergic patients, and the generation of positive reactions to both allogeneic and autocfithonous tumor cells. This observation suggests that both specific (anti-tumor) and nonspecific (recall antigen) immune stimulation or restoration may result from this form of immunotherapy. T h e frequent association of the development of delayed hypersensitivity to tumor cells with the reqppearance of reactivity to recall antigens suggests that immune mechanisms in both types of reactivity may be closely linked. It is most likely that in responding patients, restoration of reactivity to recall antigens is secondary to a n antitumor response and subsequent redrction of tumor burden. For example, patients 2 and 3 developed reactivity to vaccine materials prior to signs of tumor response and eventual restoration of reactivity to recall antigens. Patients 11 and 12 both failed to regain delayed hypersensitivity to recall antigens and continued to progress, although patient 12 did develop a positive reaction to the vaccine material in the absence of reactivity *tb recall antigens. Similarly, patient 21 who had recall antigen reactivity prior to treatment, never developed reactivity to vaccine materials and as her disease progressed, eventually lost reactivity to recall antigens with increased tumor burden. More frequent skin testing is needed to determine the sequence of restoration of reactivity. Data from this study on the prognostic utilization of skin test reactivity and restoration of recall antigen delayed hypersensitivity are statistically inconclusive. Of the seven patients who both regained delayed hypersensitivity to recall antigens and developed positive skin tests to tumor vaccines, one showed partial response, three stabilized and two progressed slowly despite very large tumor burdens. Investigation of a larger number of patients is obviously required to determine if any correlation exists. The results of E and EA rosette-forming cell enumeration in this study reflect primarily the low levels of circulating lymphocytes in the pa-

No. 6

ASILI

Juillard et al.

tient population. Although responding patients were found retrospectively to have slightly higher absolute levels of E rosette-forming cells (p < 0.025), no significant changes in any parameters measured could be associated with either treatment or response to treatment. In view of the observed tumor reductive responses, these findings can only suggest that such measurements may not be accurately reflective of changes in patient clinical status and that additional techniques should be used to more precisely define the immune status of advanced tumor patients. Similarly, although the PHA reactivity of peripheral blood lymphocytes tended to increase slightly with treatment, these data also reflect consistently depressed immune function in the patient population throughout treatment. These in vitro parameters appear to be a more limited measure of patient immune resources than in vivo parameters, e.g., skin tests, which may be more accurately representative of the qualitative aspects of patient immunocompetence. The results of the rather limited evaluation for autoimmunity in this patient population warrent further comment. Although some advanced cancer patients have detectable circulating autoantibody’.’’ seven of the patients evaluated in this study developed detectable antinuclear antibody after the initiation of immunotherapy. Of these seven patients, an objective reductive response was observed in three, stabilization in three and slow progression of tumor growth in one. While the development of autoantibody may be associated with the immunization procedure, it is also possible that immunization by nuclear debris occured as a result of immune

2225

destruction of tumor cells and the development of autoantibody may be indicative of an ongoing cytolytic immune response. This hypothesis might explain the obvservation that two of the patients who were positive for antinuclear antibody (patients 2 and 5) became negative when they no longer had an clinically evident disease. It was noted, however, that the sensitized sheep cell (SSC) test, the direct and indirect Coombs’ tests remained negative in all previously negative patients. Additionally, no clinical evidence of autoimmune disease was ever detectable in any of the six patients. Nevertheless, a possible correlation between these parameters and tumor reductive responses warrents further investigation. The possibility of a correlation between beneficial response to immunotherapy and previous responses to chemotherapy and/or radiation therapy also deserves further investigation. It is well known that patient response to chemotherapy or radiation therapy is unpredictable on a n individual basis and may be related to the patient’s ability to generate or sustain a concomitant antitumor immune response. If concurrent immune responses do play a role in chemotherapy or radiation induced remissions, then immune potentiation by active immunization may augment regressions and should be considered as a therapeutic adjunct. Although the observed responses to ASILI have been encouraging, we could not propose its use as a single modality of treatment in any patient population having established treatment alternatives, but rather as a low-morbidity form of immunotherapy to be used in conjunction with established modalities.

REFERENCES 1. Ablin, R. J., and Soanes, W. A , : Immunohistologic studies of carcinoma of the prostate. Ann. Clin. Res. 3:226230, 1971. 2. Baker, M . A,, and Taub, R. N.: Immunotherapy of human cancer. Prog. Allergy 17:227-229, 1973. 3. Baur, J. D., Ackermann, P. G., and Toro, G.: Clinical Laboratory Methods, St. Louis, C. V. Mosby Co., 1974; pp. 321-322. 4. Bleumink, E., Nater, J. P., Koops, S., and The, T. H.: A standard method for DNCB sensitization testing in patients with neoplasia. Cancer 33:911-921, 1974. 5. Boyer, P. J. J., Juillard, G. J. F., Yamashiro, C. H., and McCarthy, T. : Characterization of immune responses to tumor cells following intralymphatic immunization. Proc. Am. Assoc. Cancer Hes. 17:69, 1976 (Abstract). 6. Elhilali, M . M., Britton, S., Brossman, S., and Fahey, ,J. L. : Critical cvaluation of lymphocyte function in urological cancer patients. Cancer Res. 36:132-137, 1976. 7. Juillard, G. J. F., Boyer, P. J. J., and Snow, H. D.: Intralymphatic infusion of autochthonous tumor cells in ca-

nine lymphoma. Int. .7. Rad. Oncol. Bid. Phys. 1:497-503, 1976. 8. Juillard, G. J. F., Boyer, P. J. J., Yamashiro, C. H., Snow, H. D., Weisenburger, T., McCarthy, T., and Miller, R. J.: Regional intralymphatic infusion of irradiated tumor cells with evidence of systemic effects. Cancer 39:126-130, 1977. 9. Juillard, G. J. F., and Boyer, P. J . J.: Intralymphatic immunization: present status. Eur. J . Cancer 13:439-441, 1977. 10. Kinmonth, ,J. B., Taylor, G. W., and Harper, R. H . : Lymphansiography technique for its clinical use in lower limbs. Br. Med. 3. 1:940-942, 1955. 11. Von Leyden, E., and Blumenthal, F. : Vorlgufige mittheilungen Sber einige Ergebnisse der Krebsforshung auf der 1. Medizinschen Klinik. Dtsch. Med. Wschr. 28:637-638, 1902. 12. Whitehouse, J. M . A,, and Holborow, E. J . : Smooth muscle antibody in malignant disease. Br. Med. 3. 4:511513, 1971.