Psychological and immunological reactions of family members to patients undergoing bone marrow transplantation

The authors' goal was to evaluate the impacts of patients' bone marrow transplant (BMT) on their spouse/partner's (subjects) psychological and immunological status at four key points in the course of their transplant. Subjects' (N = 24) psychological and immunological status was prospectively evaluated at four key points in the patient's BMT which included: at patients' admission to hospital and 0-, 20-, and 34-day intervals after BMT infusion. Psychological variables examined included: a) general psychological distress and negative affect; b) tendency to respond in a socially desirable manner; c) state negative affect; and d) coping style, specifically if escape-avoidance coping was used. Immune variables examined included: percentages of total T cells and of CD4+, CD8+ cells, B cells, and natural killer (NK) cells, and NK cytotoxicity. Greatest abnormality in immune variables was detected before the initiation of BMT (i.e., between admission and day 0) with normalization between days 21 and 34 thereafter. During the waiting period before BMT, the subjects had the highest scores on negative affects, escape-avoidance coping, and psychological symptoms. These progressively declined after the BMT procedure. Significant correlations were found among trait anxiety, escape-avoidance coping, and total percentage of T cells and of CD4+ cells. Escape-avoidance coping was reliably correlated with percentage of B cells. The greatest psychological and immunological impacts on spouse/partners of BMT patients were found in the period directly after hospital admission and before BMT infusion. Alterations in immune values occurred in anticipation of BMT in the spouse/partners. Psychological symptoms followed this same pattern, being most elevated before BMT and decreasing in the successive evaluations post-BMT for the spouse/partners. The most significant and consistent psychological variable in predicting immune changes was escape-avoidance coping, with less escape-avoidance coping predicting better immune functioning.

Methods for studying intestinal sensitivity and compliance: in vitro studies of balloons and a barostat

The aim of this study was to compare in vitro various methods for recording intestinal sensitivity and compliance. Relationships between volume and pressure were determined in segments of penrose tubing and pig gut ("artificial intestine') using pressure increments of 2 mmHg (0-24 mmHg). We tested two direct methods of distension of the entire segments (by syringe inflation and the Mayo barostat); we also used three different balloon devices for indirect distension (a 10 cm polyethylene barostat bag, a 10 cm latex condom balloon and a 6 cm latex condom balloon). Maximal distending diameters of the recording systems were measured by injecting from 0 to 160 mL of air. The elastic properties of the balloons were also tested by distensions in air and in rigid tubes. All recording systems accurately detected a lesser compliance of the penrose drain as compared to pig gut. In absolute terms, only the compliance measured with a polyethylene barostat bag distended with a syringe was not different from the compliance of the segment as measured directly. The bellows of our barostat and the latex balloons had significant intrinsic compliances which interfered with the recorded pressure-volume curves. On the other hand, highly compliant plastic bags recorded most faithfully the compliance of artificial gut and that of non-compliant rigid tubes. For comparable volumes of distension, external diameters were larger with the 6 cm latex balloon than with the 10 cm latex balloon or the 10 cm polyethylene barostat balloon. A polyethylene bag distended with a non-compliant air injector (syringe) reflected most accurately the pressure-volume relationships of tubular structures. The different maximal diameters assumed by the three distending devices may explain, in part, why lower volumes of distension are required to elicit symptoms with smaller distending balloons in vivo.

Lupus anticoagulant masquerading as an acute abdomen with multiorgan involvement

A 45-yr-old male patient developed acute abdominal pain, ileus, and microscopic hematuria with biochemical evidence of pancreatitis and a marked increase in liver alkaline phosphatase; CT demonstrated swelling of the pancreas, bilateral adrenal hemorrhage, and a suggestion of renal hemorrhage. ERCP was negative and renal arterial and venous blood flow normal. A coagulation profile demonstrated the presence of lupus anticoagulant, but tests for anticardiolipin antibodies and collagen vascular diseases were negative. Treatment with corticosteroids and anticoagulation resulted in improvement in clinical and all biochemical indices. Thus, lupus anticoagulant syndrome may masquerade as an acute abdominal illness with multiorgan involvement.

Visceral perception in irritable bowel syndrome. Rectal and gastric responses to distension and serotonin type 3 antagonism

We wished to determine if visceral perception in the rectum and stomach is altered in patients with irritable bowel syndrome and to evaluate the effects on visceral sensation of 5-HT3 receptor blockade. Twelve community patients with diarrhea-predominant irritable bowel syndrome and 10 healthy controls were studied in a double-blind, randomized, placebo-controlled study. Using two barostats, the stomach and rectum were distended, with pressure increments of 4 mm Hg, from 10 to 26 mm Hg; visceral perception was measured on an ordinal scale of 0-10. Personality traits were measured using standard psychological methods, and somatic pain was evaluated by immersion of the nondominant hand in cold water. The effect of 5-HT3 antagonism was tested with a single intravenous dose of ondansetron at 0.15 mg/kg. Gastric perception was higher in irritable bowel syndrome, but rectal distension was perceived similarly in irritable bowel syndrome and controls. Pain tolerance to cold water was also similar in irritable bowel syndrome and controls. Ondansetron induced rectal relaxation and increased rectal compliance but did not significantly alter gastric compliance or visceral perception. Psychological test scores were similar in patients and controls. We conclude that in this group of psychologically normal patients with irritable bowel syndrome, who were not chronic health-care seekers, visceral perception was normal. Ondansetron did not alter gut perception in health or in irritable bowel syndrome.

Abnormalities in the shape of the duodenal loop on X-ray. Associated with idiopathic antral hypomotility?

We aimed to determine if abnormalities in the shape of the duodenal loop would be useful in identifying patients with gastroduodenal dysmotility. Retrospectively, 126 consecutive patients with suspected functional abdominal symptoms who underwent upper gastrointestinal barium x-ray studies and gastrointestinal manometry were independently evaluated. Twenty-seven patients (21%) had an abnormally shaped duodenal loop (two proximal and 25 distal) by x-ray. An abnormal duodenal loop was associated with female gender but the presenting symptoms were similar in patients with normal and abnormal loops. Antral hypomotility was significantly more common in patients with distal duodenal malrotations compared to those with a normal x-ray (56% vs 27%, P < 0.01); intestinal dysmotility was not associated with the shape of the duodenal loop. The presence of an abnormally shaped duodenal loop in patients presenting with functional gastrointestinal symptoms may be a useful marker for idiopathic antral hypomotility.

Response of gastric fundus to rectal distension in healthy persons

We aimed to record fundic motor activity in man using the barostat to ascertain if fundic motility is affected by rectal distension. The distal ends of two barostat tubes were placed in the gastric fundus and rectum in 10 healthy volunteers. The gastric bag was first inflated to a constant pressure level that recorded phasic motor activity as changes in volume of the air-filled bag. Baseline motor activity was recorded before, during, and after a 15-min period of constant rectal distension that was clearly perceived by all subjects but was not painful. In all subjects, continuous phasic volume changes, reflecting fundic motor activity, were recorded at a rate of 1-3/min. During rectal distension, a consistent change in mean contractile force of these phasic volume events was not detected; a decrease of more than 30% occurred in only three subjects. We conclude that fundic phasic volume changes are recordable by the barostat, but these are not substantially inhibited by rectal distension.

Primary colonic lymphoma. Clinical presentation, histopathologic features, and outcome with combination chemotherapy

Primary colonic lymphomas are rare, but we identified 15 cases at our institution between 1973 and 1992. They comprised 5.8% of all cases of gastrointestinal lymphoma (15 of 259) and 0.16% of all cases of colon cancer (15 of 9,193) during the last 20 years. The most common presenting symptoms were abdominal pain and weight loss (40% each). In seven patients (47%), a palpable abdominal mass was noted on the initial physical examination. The most frequent site of involvement was the cecum (73%). Histologically, six (40%) were classified as high-grade and nine (60%) as intermediate-grade non-Hodgkin's lymphoma. The tumors usually presented at an advanced stage: in 13 of 15 patients (87%), the lymphoma had spread to the adjacent mesentery, the regional lymph nodes, or both when first diagnosed. The 5-year survival rate was 27% for all patients and 33% (4 of 12) for patients treated with combination chemotherapy. Two patients relapsed after 8 years of complete remission. Primary colonic lymphomas have an aggressive behavior and only a marginal response to surgery and combination chemotherapy.

Endoscopic laser coagulation of radiation-induced mucosal vascular lesions in the upper gastrointestinal tract and proximal colon

Endoscopic laser coagulation effectively controls bleeding from radiation-induced rectal vascular lesions.

OBJECTIVE

To assess the outcome of endoscopic treatment of radiation-induced bleeding due to vascular lesions located proximal to the sigmoid colon.

METHODS

We identified 15 consecutive patients with such proximal radiation enteropathy treated at our Institution with Nd:YAG laser between 1984 and 1991. Ten patients (66%) had gastric and/or small bowel involvement, and five (33%) had colonic involvement with or without more proximal lesions. Bleeding first appeared at a mean of 21.2 +/- 12.5 months after completion of radiotherapy. Mean duration of gastrointestinal bleeding before laser treatment was 7.6 +/- 4.6 months.

RESULTS

After completion of laser therapy, bleeding ceased in nine (60%) patients, decreased in three (20%), and persisted in three (20%). The mean hemoglobin level increased from 8.4 +/- 0.5 g/dl to 10.4 +/- 0.6 g/dl after completion of laser treatments (p < 0.02). The mean number of transfusions per patient per year decreased from 10.5 +/- 2.8 to 0.9 +/- 0.7 (p < 0.01). No treatment-related complications or deaths occurred.

CONCLUSIONS

Endoscopic laser coagulation of radiation-induced mucosal vascular lesions in the upper gastrointestinal tract and proximal colon appears to be safe and, in most cases, effective.

Endoscopic Nd:YAG laser coagulation of bleeding from radiation proctopathy

Hematochezia from mucosal vascular lesions usually confined to the rectum represents an uncommon but problematic late complication of pelvic radiotherapy. We studied 47 patients with medically refractory hematochezia resulting from radiation-induced rectosigmoid mucosal vascular lesions. All lesions were endoscopically coagulated with Nd:YAG laser. Median duration of hematochezia before laser therapy was 11 months, despite previous medical treatment (98%) or bypass colostomy (6%). Within 3 to 6 months after laser treatment, the number of patients with daily hematochezia fell from 40 (85%) to 5 (11%; p < 0.001), and the median hemoglobin level increased from 9.7 gm/dl to 11.7 gm/dl (p < 0.001). Complications occurred in three patients (6%); no deaths occurred. The condition in six patients (12.8%) was not improved by laser treatment. Two patients (4%) ultimately required surgical treatment for bleeding control. On the basis of symptomatic, hematologic, and endoscopic responses, Nd:YAG laser photocoagulation controlled bleeding from radiation proctopathy in most patients with an acceptably low morbidity. Patients with sigmoid colon involvement responded less favorably. Endoscopic laser photocoagulation should be considered before surgical intervention for treatment of hematochezia from radiation proctopathy.

Successful medical management of a patient with multiple hepatic abscesses due to Edwardsiella tarda

Isolation of Edwardsiella tarda in humans has been associated with an asymptomatic carrier state as well as mild, self-limited diarrheal illness. Extraintestinal manifestations have included soft-tissue infections, meningitis, osteomyelitis, cholangitis, and sepsis. Only three cases of patients who had documented hepatic abscess due to E. tarda have been reported in the English-language literature; two patients died, and the third required a laparotomy and drainage. We report what is, to our knowledge, the first autochthonous case of hepatic abscess due to E. tarda in the United States and the first case that was successfully managed with antibiotic therapy alone.

Prostatic abscess due to Histoplasma capsulatum in a patient with the acquired immunodeficiency syndrome

Disseminated histoplasmosis is a systemic fungal infection that may occur in previously healthy or immunocompromised patients. The condition is being recognized with increasing frequency in persons infected with the human immunodeficiency virus. The most common organs involved include the lung, bone marrow, lymph nodes, liver, adrenals and central nervous system, with genitourinary involvement being exceedingly unusual. We describe a Histoplasma capsulatum prostatic abscess occurring after therapy for pulmonary histoplasmosis in a patient with the acquired immunodeficiency syndrome. The prostate may be a difficult focus from which to eradicate disseminated fungal infection in immunocompromised patients.

Synergy is documented in vitro with low-dose recombinant tumor necrosis factor, cisplatin, and doxorubicin in ovarian cancer cells

Ovarian carcinomas have been shown to be sensitive or insensitive to the in vitro exposure of several cytotoxic drugs and cytokines. Because of the potential for cytokines to enhance the efficacy of chemotherapeutic agents and to improve their therapeutic index, the optimal dose and schedule of the combination of these agents have been studied. We examined the cytotoxic effect of a combined modality using a variety of concentrations of recombinant tumor necrosis factor (rTNF) (a cytotoxic cytokine) with Adriamycin (ADR) and cisplatin (CDDP) on human ovarian carcinoma cell lines. Cytotoxicity was determined in a 24-hr 51Cr-release assay and confirmed in a 5-day viability culture assay. Five cell lines were used: PA-1, 222, OVCAR-3, SKOV-3, and OVCAR-8. Doses of rTNF that were minimally cytotoxic resulted in significant cytotoxicity and synergy when used with optimal or suboptimal concentrations of ADR or CDDP. This synergy was observed in four cell lines. Interestingly, the rTNF- and drug resistant SKOV-3 cell line was sensitive to the synergistic effect of Adriamycin and rTNF. The synergistic effect that was obtained was specific to rTNF, while the combined use of ADR and CDDP or recombinant interleukin-2 and cytotoxic drugs had no synergistic effect on tumor cell lysis. Further, the addition of anti-TNF antibody abrogated the synergistic effect seen with rTNF and the cytotoxic drugs. These studies demonstrate clearly that significant synergistic antitumor cytotoxic activity against human ovarian carcinoma cell lines can be achieved with combinations of low doses of rTNF and ADR or CDDP, suggesting their possible adaptation in vivo for cancer therapy. Further, the studies suggest that rTNF and the cytotoxic drugs tested may share a common lytic pathway. Since rTNF used alone has been relatively inactive in clinical trials, its potential activity may be apparent only when combined with conventional cytotoxic chemotherapeutic agents and when administered in relatively low concentration.

Cycloheximide-induced modulation of TNF-mediated cytotoxicity in sensitive and resistant ovarian tumor cells

The mechanism of sensitivity and resistance of various ovarian carcinoma lines to recombinant tumor necrosis factor (rTNF)-mediated cytotoxicity has been investigated using a 24-h 51Cr-release assay. The cell line PA-1 is sensitive to TNF in a dose-dependent manner, whereas the cell line SKOV-3 is resistant to TNF even at high concentrations. The simultaneous addition of TNF and cycloheximide (CHX) in the assay converted the resistant SKOV-3 line into a sensitive line, but no detectable change was observed with PA-1. rTNF inhibited DNA, RNA, and protein synthesis of the sensitive PA-1 line, whereas it had no effect on SKOV-3. This finding was not due to differences in the expression of TNF receptors, as both cell lines expressed equivalent numbers of receptors. The addition of CHX to TNF resulted in suppression of DNA, RNA, and protein synthesis in both the sensitive and the resistant cell lines. Pretreatment of the cell line with TNF for 3 h and subsequent washing resulted in significant cytotoxicity of the sensitive PA-1 line and some cytotoxicity against SKOV-3. However, if the cells were pretreated with CHX for 3 h followed by rTNF for 24 h, a significant decrease in cytotoxicity was observed in both cell lines. Under these conditions, there was no significant inhibition of DNA, RNA, or protein synthesis. Pretreatment of cells for 24 h with TNF and 24 h with CHX resulted in augmentation of the cytotoxicity of PA-1 and SKOV-3, whereas pretreatment for 24 h with CHX followed by 24 h with TNF resulted in no cytotoxicity. Cells pretreated with CHX for 24 h showed poor binding of [125]I-TNF and poor internalization, whereas cells pretreated for 24 h with TNF showed marked enhancement of internalization. The sensitivity of freshly derived ovarian carcinoma lines to TNF and CHX demonstrated that TNF-resistant cells became more sensitive if treated with CHX. These results demonstrate the potential use of metabolic inhibitors in increasing the sensitivity of fresh ovarian tumor cells to TNF.

Augmentation of cytotoxicity of lymphokine activated killer cells on ovarian tumor cells by various biological response modifiers

The effect of combining lymphokine activated killer (LAK) cells with either recombinant interleukin-2 (rlL-2), recombinant interferon-alpha (rIFN-alpha), recombinant tumor necrosis factor (rTNF) or streptococcal preparation OK-432 were assessed, using Raji, 2 kinds of cultured ovarian lines (PA-1 and SKOV-3), and 7 fresh ovarian tumor lines. LAK cells were generated by culturing peripheral blood lymphocytes (PBL) with rIL-2 for 3-5 days. The simultaneous combination of LAK cells with rIFN alpha or OK-432 augmented the cytotoxicity of LAK cells. The susceptibility of tumor cells to LAK cells also increased after the pretreatment of tumor cells with OK-432. These results suggest that the simultaneous injection of LAK cells with rIFN-alpha or OK-432 and the intralesional injection of OK-432 prior to the adoptive transfer of LAK cells may be a beneficial combination treatment for LAK treatment.

Human neutrophil-mediated lysis of ovarian cancer cells

Because of recent questions concerning the sensitivity of human tumor cells to neutrophil-induced oxidative injury, we studied six freshly obtained human ovarian cancer (OC) specimens. Stimulation of neutrophils (PMNs) by phorbol myristate acetate (PMA) did not result in OC cytolysis during the first nine hours of incubation. However, three of six specimens were significantly lysed by stimulated PMNs when assay length was increased to 18 hours. Cytotoxicity was mediated by PMN production of reactive oxidative intermediates (ROIs). Presentation of ROIs to OC targets as preformed or enzymatically generated molecules in cell-free systems duplicated the enhanced lysis at 18 hours (as compared with six hours). Since addition of catalase at three or six hours did not inhibit enhanced lysis at 18 hours (achieved by PMNs or in cell-free systems), it appears that an initial ROI-mediated lethal event occurs early, but longer incubations are required for the event to become manifested as cell death. These data suggest that shorter assays may underestimate the potential of PMNs as effector cells against human tumor cells.

Cytotoxic and cytostatic effects of the streptococcal preparation OK-432 and its subcellular fractions on human ovarian tumor cells

Previous studies have indicated that OK-432 is a potent biologic response modifier (BRM) and that it augments immune responses to tumor cells. We studied the direct effect of OK-432 on tumor cells. Established and freshly derived human ovarian carcinoma lines were examined for their susceptibility to OK-432 or its subcellular fractions in direct cytotoxicity, cytostatic activity, and inhibition of metabolic activity. OK-432 was cytotoxic to 13 of 15 freshly derived ovarian carcinoma lines in a 24-hour chromium-51 (51Cr) release assay. The optimal effect was noticed at OK-432 concentrations between 0.1 and 1.0 Klinishe Einhert (KE) per milliliter. The cytostatic effect on two established lines and one fresh line correlated with the cytotoxic activity. In all three lines, however, the metabolic activities (DNA, RNA, and protein synthesis) were inhibited by OK-432, suggesting that cell lysis by OK-432 may not be directly correlated with the inhibition of metabolic activities. Several subcellular fractions were derived from OK-432 and only the cytoplasmic and protoplast membrane fractions showed cytotoxic activity against the OK-432-sensitive tumor cell lines, although the cytotoxicity obtained was greatly less than the whole microorganism OK-432. The direct binding of 14C-OK-432 to tumor cells was examined. Binding took place rapidly after 1 hour of incubation and reached a maximum activity at 37 degrees C. Binding in all three lines ranged between 1.7 and 2.7 pg/cell. These results demonstrate the direct cytotoxic effect of OK-432 and some subcellular fractions on human ovarian carcinoma lines. These results also show that the BRM OK-432 may exert its effect by both potentiating the antitumor response and directly inhibiting tumor cell growth.

Lymphocyte activation by recombinant interleukin-2 in ovarian cancer patients

Peripheral blood lymphocytes from 42 patients and peritoneal cavity lymphocytes from eight patients with advanced ovarian carcinoma were tested for lymphokine-activated killer lymphocyte cytotoxicity against several ovarian carcinoma lines before and after exposure to recombinant interleukin-2 in vitro for 3-5 days. Only four of 42 (9.5%) of peripheral blood lymphocytes and zero of eight (0%) of peritoneal cavity lymphocytes had spontaneous cytotoxicity (greater than 20%) against the ovarian carcinoma lines. After in vitro exposure to recombinant interleukin-2, 41 of 42 (98%) of patients' blood lymphocytes showed a two- to fivefold increase in cytotoxicity against K562 (a lymphoblastoid human target), and 40 of 42 (95%) demonstrated lymphokine-activated killer cytotoxicity (greater than 20%) to the ovarian carcinoma lines. Lymphokine-activated killer activity against fresh allogeneic cell lines was variable, although most patients' peripheral blood lymphocytes (70%) had significant cytotoxicity. By contrast, incubation of patients' peritoneal cavity lymphocytes with recombinant interleukin-2 in vitro did not result in the generation of lymphokine-activated killer cell activity against K562 or ovarian cell lines. Peritoneal lymphocytes did produce lymphokine-activated killer cells in the presence of OK432 in half of the patients tested. The presence of autologous serum during recombinant interleukin-2 activation with blood lymphocytes had an augmentative effect on the resulting lymphokine-activated killer cytotoxicity in two of 20 patients, a suppressive effect in four of 20, and no effect in the other 14 of 20 patients tested.(ABSTRACT TRUNCATED AT 250 WORDS)

Intraperitoneal administration of human recombinant interferon-alpha in patients with ovarian cancer: effects on lymphocyte phenotype and cytotoxicity

Eleven patients with persistent Stage III ovarian cancer, documented at second look laparotomy, received i.p. human recombinant interferon-alpha (5-50 x 10(6) units/week). Prior to immunotherapy, patients' peritoneal cell lymphocytes (PCLs) contained decreased proportions of Leu-7+ and Fc-receptor+ cells and almost nondetectable natural killer (NK) and antibody-dependent cell cytotoxic (ADCC) activity. In contrast, patients' peripheral blood lymphocytes (PBLs) contained normal proportions of lymphocyte subsets and cytotoxic activity compared to control donor PBLs. During therapy, there was a concurrent increase in PCL Leu-7+ cells and NK lysis. Both peaked predictably at 24 h after each treatment, regardless of the dose injected, and usually returned to baseline by Day 7 of each weekly cycle. PCL NK enhancement was striking, usually increasing from 2-6% (effector:target ratio, 25:1) to over 30% lysis. Enhancement of PCL ADCC was less impressive. PCLs of several patients developed lytic activity towards NK-resistant Raji targets. During therapy, patients' PBLs demonstrated: (a) modestly enhanced NK lysis at Day 4 of each cycle, and; (b) no development of Raji lysis. These data clearly demonstrate the efficacy of i.p. interferon in activation of peritoneal NK activity. However, increased NK lysis did not correlate with individual tumor responses in this cohort of patients.

Sensitivity of fresh and cultured ovarian tumor cells to tumor necrosis factor, interferon-alpha 2, and OK-432

The in vitro sensitivity to rTNF, rIFN-alpha 2, and OK-432 of 11 freshly derived human ovarian tumors and 2 established tumor cell lines was examined in a cytotoxic assay using the 51 Cr release test. Nine fresh lines were sensitive to rTNF, 8 to OK-432, and only 2 were sensitive to rIFN-alpha 2. Cytotoxicity by rIFN-alpha 2 was of lesser magnitude than the cytotoxicity mediated by rTNF or OK-432. The time of exposure and the concentration of BRM required for maximal cytotoxicity varied from line to line. Two fresh tumor cell lines and 1 established cell line (PA-1) were sensitive to all 3 BRMs, while 2 other FOCs and 1 cell line (SKOV-3) were resistant to all BRMs. The remaining FOC showed an intermediate degree of sensitivity. These results demonstrate the existence of heterogeneity of ovarian carcinoma tumor cell lines to lysis by BRMs. Among the FOCs, the 2 endometrioid carcinomas tested were highly sensitive to rTNF, whereas the serous carcinomas were more sensitive to OK-432. Low grade tumors were more sensitive to BRM than high grade tumors, and tumor extension did not correlate with sensitivity to the BRM. When tumor targets were exposed to more than 1 BRM added either simultaneously or sequentially, the net cytotoxic effect achieved was usually inferior to the sum cytotoxicity obtained by each BRM alone. Furthermore, rTNF and OK-432 were cytostatic to most ovarian tumor cell lines examined. The results of this study demonstrate that certain BRMs exert a direct effect on fresh ovarian tumor cells independently of host factors. These findings suggest that in vitro screening of a patient's tumor cells for sensitivity to a particular BRM prior to therapy could be beneficial for the proper identification of patients most likely to benefit from the treatment.

Antigen-specific suppression of the in vitro cytotoxic response by a soluble factor produced by Corynebacterium parvum-induced allospecific suppressor cells

The adjuvant Corynebacterium parvum, when administered intravenously during an ongoing alloimmunization, induces alloantigen-specific splenic suppressor cells which inhibit primary and secondary in vitro sensitizations. We have previously shown that these cells produce a soluble suppressor factor in culture. We now further characterize this factor and its mechanism of action. Release of this suppressive factor is dependent upon specific restimulation of the splenic suppressor cell with the sensitizing alloantigen for 24-48 hr in culture. The suppressor factor inhibits primary, but not secondary, in vitro sensitizations in an antigen-specific, genetically unrestricted manner. The suppressive activity is not absorbed by passage through immunoadsorbent columns containing anti-mouse immunoglobulin. The factor does not lyse tumor cells bearing the sensitizing alloantigen. Delay in addition to primary cultures of as little as 4 hr after culture initiation leads to loss of suppressive activity, suggesting that this antigen-specific allosuppressor factor inhibits an early step in the sensitization of precursor cytotoxic T lymphocytes.

Monoclonal antibodies against squamous cell carcinoma

We report on a monoclonal antibody reactive with squamous carcinoma cell lines and with frozen sections from squamous cell carcinomas of the oral cavity but not with frozen sections of normal squamous epithelium of the oral cavity, esophagus, and skin. In addition, we identified other monoclonal antibodies with restricted specificity as reflected by their binding to a panel of established tumor cell lines. Subsequent testing of these monoclonal antibodies against a panel of normal and tumor tissue sections revealed two types of reactivity patterns: binding to normal tissue and binding to normal and tumor tissue.

Synergistic effects of combination sequential immunotherapies in a murine ovarian cancer model

The antitumor effects of Corynebacterium parvum in a murine ovarian teratocarcinoma model depend upon a sequential activation of neutrophils and macrophages within the peritoneal cavity. We studied the sequential administration of biological response modifiers that independently activate each phase of the response. Tumor-challenged mice treated by i.p. injection of a pyridine-extracted fraction of cell-free Propionibacterium acnes (PA-PE, 1400 micrograms) demonstrated prolonged survival in less than 20% of the cases. An i.p. injection of a detoxified Salmonella endotoxin (DSE) preparation (150 micrograms) had no effect on tumor outgrowth. However, i.p. treatment with PA-PE (1400 micrograms), followed by 150 micrograms of DSE 1 day later, resulted in long-term survival (greater than 100 days) in 40 to 60% of mice. This antitumor effect was only evident when PA-PE was administered first (before DSE) and optimal when DSE was administered 24 h after PA-PE. The synergistic antitumor effect could be duplicated when tumor-challenged mice were first treated i.p. with peritoneal polymorphonuclear leukocytes, elicited by injection of PA-PE, and then treated with DSE 18 h later. These data indicate that appropriately timed injection of biological response modifiers with complementary effects can result in a synergistic prevention of tumor growth.

Treatment of acute myelogenous leukemia. A prospective controlled trial of bone marrow transplantation versus consolidation chemotherapy

In a prospective controlled trial, the relative effectiveness of allogeneic bone marrow transplantation and postremission chemotherapy was assessed for adult patients with acute myelogenous leukemia in first complete remission. Twenty-three patients, 15 to 45 years of age, who had an HLA-identical sibling donor were designated to receive bone marrow transplantation. Forty-four patients who either lacked an HLA-identical sibling or were over 45 years of age were designated to receive intensive consolidation chemotherapy. The actuarial rate of leukemia relapse was significantly lower in the transplantation group than in the chemotherapy group (40 +/- 25% [95% confidence interval] compared with 71 +/- 14%, p = 0.01). Actuarial survival at greater than 4 years was not significantly different (40 +/- 21% compared with 27 +/- 14%, p greater than 0.4). These data show that bone marrow transplantation is more effective than consolidation chemotherapy in preventing leukemia relapse, but overall survival was not improved in this study.

Natural killer inhibitory substance produced by the peritoneal cells of patients with ovarian cancer

Peritoneal cells obtained from 8 patients with minimal residual ovarian cancer produced a substance during in vitro culture that markedly inhibited the expression of natural killer (NK) cell-mediated lysis. Its molecular weight was less than 2,000, the same size as the NK-inhibiting substance (NK-IS), a similar NK-suppressive molecule produced by the peritoneal cells of rats. Human NK-IS suppressed the expression of antibody-dependent cell cytotoxicity as well as NK lysis, but it had no effect on erythrocyte-rosette formation and was not cytotoxic to peripheral blood lymphocytes or cell fractions enriched for large granular lymphocytes. NK-IS inhibited lysis mediated by interferon-activated lymphocytes and completely prevented NK activation when used in a preincubation. During intraperitoneal immunotherapy with Corynebacterium parvum, an agent that can activate peritoneal cytotoxic effectors, the production of NK-IS by peritoneal cells decreased considerably. Human peritoneal cells produce an NK-IS similar to the peritoneal cells of rats, and this material may create an environment within the peritoneal cavity that is permissive to the growth of NK-sensitive tumor cells.

Lymphocyte cytotoxicity in the peritoneal cavity and blood of patients with ovarian cancer

After an intensive course of combination chemotherapy, 16 patients with minimal residual ovarian cancer that was documented at second-look laparotomy, had an indwelling Tenckhoff catheter placed and underwent multiple peritoneal saline lavages. Lymphocyte-enriched populations from the peritoneal cavity and peripheral blood were obtained by density gradient centrifugation and examined for cell-surface phenotype and a variety of immune functions, including natural killer cytotoxicity and antibody-dependent cell-mediated cytotoxicity. Phenotypic characterization revealed that peritoneal lymphocytes consisted primarily of T cells and cells bearing receptors for the crystallizable fragment of immunoglobulin G (IgG) (crystallizable fragment-receptor), and contained a very low number of B cells. Peritoneal natural killer lymphocyte cytotoxicity and antibody-dependent cell-mediated cytotoxicity were very low in all but two patients. Incubation of peritoneal lymphocytes with Corynebacterium parvum and interferon in vitro did not result in augmented cytotoxicity against susceptible targets. Supernatants from cultured peritoneal cells of all patients markedly inhibited natural cytotoxic activity of normal donor blood lymphocytes. These results suggest that lymphocytes collected from the peritoneal cavity of patients with minimal residual ovarian cancer are deficient in natural and antibody-dependent cytotoxic effector function. This deficiency may influence the host's ability to control the spread and proliferation of tumor cells in the peritoneal cavity.

Role of inflammatory neutrophils in antitumor effects induced by intraperitoneal administration of Corynebacterium parvum in mice

We studied the role of inflammatory neutrophils in the antitumor effects that follow i.p. injection of Corynebacterium parvum (1400 micrograms) into C3HeB/FeJ mice challenged with the murine ovarian teratocarcinoma. Peritoneal neutrophils, obtained from mice 6 hr after injection of C. parvum, exerted significant antitumor effects when injected admixed with murine ovarian terato-carcinoma cells into the peritoneal cavities of normal mice. Treatment of recipient mice with whole-body irradiation or repeated injections of silica prevented the antitumor effects, indicating that neutrophils were activating a second effector mechanism in recipient mice. Peritoneal cells obtained at 24 or 72 hr or at 7 or 11 days following C. parvum injection were considerably less effective in activation of this effector mechanism. Heat-killed C. parvum (6 hr)-induced neutrophils activated antitumor responses, but thioglycolate-induced cells were without effect. Antitumor responses in mice receiving peritoneal neutrophils were not due to simple transfer of C. parvum organisms in the inocula. These results indicate that inflammatory neutrophils, elicited into the peritoneal cavity by injection of C. parvum, play an important role in the activation of subsequent antitumor effects.

Activation of peritoneal lymphocyte cytotoxicity in patients with ovarian cancer by intraperitoneal treatment with Corynebacterium parvum

Following an intensive course of combination chemotherapy, patients with minimal residual ovarian cancer were treated with increasing concentrations of intraperitoneal Corynebacterium parvum to assess whether or not natural killer (NK) cells could be activated. Immunotherapy was administered every 2 weeks, initially with a dose of 250 micrograms/m2, which was progressively escalated as tolerated. Each treatment induced a peritoneal cellular exudate which consisted primarily of neutrophils 48h after injection and of lymphocytes and macrophages at 7 and 14 days after injection. Peritoneal NK cytotoxicity increased during treatment in six of nine patients tested. NK precursor cells susceptible to in vitro activation with either C. parvum or interferon became detectable during treatment in all five patients tested. In four of these five, precursors were detected prior to the development of enhanced spontaneous NK activity. In four patients serially studied, peripheral blood NK activity increased during therapy. These results indicate that regional intraperitoneal therapy with C. parvum can enhance nonspecific antitumor cytotoxic mechanisms within the peritoneal cavity.

Successful immunotherapy with intraperitoneal Corynebacterium parvum in a murine ovarian cancer model is associated with the recruitment of tumor-lytic neutrophils into the peritoneal cavity

The rejection of a murine ovarian teratocarcinoma (MOT) after i.p. injection of Corynebacterium parvum was investigated. Treatment with C. parvum (1400 micrograms) 24 hr after i.p. inoculation of a lethal number of tumor cells (10(5] induced an antitumor effect that cured 75 to 95% of the mice. Morphologic analysis and an in vivo cytotoxicity assay that measured the rate of disappearance of radioactivity from the peritoneal cavity after injection of 125IUdR-labeled tumor cells indicated that the antitumor effect was initiated during the first 24 hr after C. parvum injection. During this period of time, host effector cells retrieved from the peritoneal cavity prevented tumor growth in a Winn assay and lysed radiolabeled MOT targets in a 4-hr Cr-release assay. After separation of peritoneal inflammatory cells on a Percoll gradient, neutrophil-enriched fractions demonstrated significant in vitro tumor lysis, but neutrophil-depleted populations were ineffective. Microscopic analysis of lysis at the single cell level confirmed that neutrophils were binding to and lysing MOT targets. Further characterization of these tumor cytolytic neutrophils revealed that they are nylon wool-adherent, not generated in indomethacin-pretreated mice (but effectively generated in whole body-irradiated mice), and achieve lysis within 30 min after binding to MOT targets. These results indicate that neutrophils must be considered potential antitumor effectors that can be recruited by treatment with biologic response modifiers.

Successful immunotherapy with intraperitoneal Corynebacterium parvum in a murine ovarian cancer model is associated with the recruitment of tumor-lytic neutrophils into the peritoneal cavity

The rejection of a murine ovarian teratocarcinoma (MOT) after i.p. injection of Corynebacterium parvum was investigated. Treatment with C. parvum (1400 micrograms) 24 hr after i.p. inoculation of a lethal number of tumor cells (10(5] induced an antitumor effect that cured 75 to 95% of the mice. Morphologic analysis and an in vivo cytotoxicity assay that measured the rate of disappearance of radioactivity from the peritoneal cavity after injection of 125IUdR-labeled tumor cells indicated that the antitumor effect was initiated during the first 24 hr after C. parvum injection. During this period of time, host effector cells retrieved from the peritoneal cavity prevented tumor growth in a Winn assay and lysed radiolabeled MOT targets in a 4-hr Cr-release assay. After separation of peritoneal inflammatory cells on a Percoll gradient, neutrophil-enriched fractions demonstrated significant in vitro tumor lysis, but neutrophil-depleted populations were ineffective. Microscopic analysis of lysis at the single cell level confirmed that neutrophils were binding to and lysing MOT targets. Further characterization of these tumor cytolytic neutrophils revealed that they are nylon wool-adherent, not generated in indomethacin-pretreated mice (but effectively generated in whole body-irradiated mice), and achieve lysis within 30 min after binding to MOT targets. These results indicate that neutrophils must be considered potential antitumor effectors that can be recruited by treatment with biologic response modifiers.

Peripheral blood lymphocytes with receptors for a determinant common to B-lymphoblastoid cell lines and acute myelogenous leukemia blasts

We investigated whether the rosetting of B-lymphoblastoid cell lines (B-LCL) by peripheral blood lymphocytes (PBLs) reflected possible interactions between lymphoid cells and immature cells of the hematopoietic system. Rosette formation could be blocked by the addition of soluble antigen extracted from B-LCL or blasts obtained from patients with acute myelogenous leukemia (AML). This inhibition was specific for AML blasts (similarly extracted material from melanoma lines had no inhibitory effect) and for the B-LCL receptor (leukemic extracts had no effect on surface receptors for sheep red blood cells (E) or antibody-sensitized red blood cells (EA)). The B-LCL receptor is present on leukemic Sezary T-cells as well as normal T-cells and its sensitivity to various enzymatic treatments is markedly different from that of E and EA receptors. In addition, B-LCLs derived from in vitro EB-viral infection of a normal donor's B lymphocytes were significantly rosetted by that donor's autologous PBLs. These data suggests the B-LCL receptor, present on mature T-cells, can recognize self determinants on myeloblasts and B-LCL. Further investigation will determine whether this interaction can affect the function of rosetted target cells.

Immunotherapy with biochemically dissociated fractions of Propionibacterium acnes in a murine ovarian cancer model

The antitumor effect of two strains of Propionibacterium acnes (PAI and PAII) and chemically derived fractions from the whole bacterial cell were studied using a murine ovarian teratocarcinoma (MOT) model. When injected i.p. in high doses (700 to 1400 micrograms/mouse), both strains produce survival of a significant proportion of tumor-bearing mice (30 to 90%). On a weight to weight basis, however, PAI was significantly more effective than PAII. PAI and PAII were extracted using pyridine, which yielded four fractions, i.e., pyridine-extracted strains PAI and PAII (PA-PEI and PA-PEII, respectively) which are composed of the cell wall material extracted by pyridine, and the residues of PA-PEI and PA-PEII (PA-RI and PA-RII, respectively) which are composed of the residue material following the chemical extraction. The chemical composition of PA-PEI was different from that of PA-PEII (the latter had proportionately three times as many carbohydrates and one-third of the protein content of the former) and so were their antitumor properties in the MOT model. PA-PEI had markedly reduced antitumor effect when compared to the untreated cell on a per weight basis. Furthermore, curability was only seen when using a high dose (1400 micrograms/mouse). By contrast, the cell wall components extracted by pyridine from PAII (PA-PEII) had powerful antitumor effects, i.e., greater than 50% of mice given 1400-micrograms injections survived. The material contained in PA-PEII was further fractionated on the basis of its organic solubility in chloroform:methanol solvent. The water-soluble and solvent-insoluble fractions retained most of the antitumor effects of PA-PEII, while the water-insoluble and solvent-soluble fractions were only moderately effective, suggesting that the active moiety(ies) was associated with the nonlipid components of this fraction. Both residue fractions (PA-RI and PA-RII) were as effective on a per weight basis in controlling the growth of 10(5) tumor inoculum as were whole untreated cells. However, periodate oxidation of PA-RI resulted in complete loss of its antitumor effects. When surviving mice that had no evidence of tumor persistence following a tumor challenge (10(5) MOT cells) and i.p. treatment with PA were subsequently rechallenged with 10(4) tumor cells, survival was significantly prolonged, as compared to tumor-challenged (10(4) MOT) naive mice. In addition, 10 to 20% of these rechallenged mice had complete eradication of the tumor inoculum (no evidence of disease for greater than 120 days).(ABSTRACT TRUNCATED AT 400 WORDS)

Prolonged survival in acute myelogenous leukaemia without maintenance chemotherapy

46 adults with previously untreated acute myelogenous leukaemia who achieved complete remission with 6-thioguanine, cytarabine, and daunorubicin (TAD) received two courses of intensive consolidation chemotherapy. The first cycle consisted of 5-azacytidine and doxorubicin followed by a second consolidation cycle with TAD. Maintenance chemotherapy was not administered. Median remission duration was 14 months and 26% (95% confidence interval, 11%-41%) remained in continuous remission at 5 years. Actuarial 5 year survival was 31% (+/- 15%). Results were most favourable in patients who achieved complete remission within 60 days of chemotherapy being initiated. These data indicate that prolonged disease-free survival can be achieved in patients treated with intensive induction and consolidation treatment alone without maintenance chemotherapy.

AMSA: a phase II trial in resistant and recurrent acute myelogenous leukemia

27 patients with resistant acute myelogenous leukemia (AML) were treated with AMSA. Three achieved a complete remission, and two a partial remission. Median survival for all patients was 12 weeks and was 16 weeks for responders.

Inhibition of natural killer cell lysis by natural killer-inhibitory substance: mechanism of suppression

The mechanism of suppression of NK-mediated lysis by a soluble product of peritoneal cells (NK-IS, natural killer-inhibitory substance) was investigated. Pretreatment of effector cells resulted in depressed NK lysis while pretreatment of targets had no effect, indicating suppression is due to alterations in effector cell function rather than changes in target cells. NK-IS had no effect on the formation of conjugates between effectors and NK-susceptible targets. When NK-IS was added to effector-target cell mixtures after the binding step had been successfully completed, ensuing lysis was significantly depressed, confirming that NK-IS inhibited a postbinding lytic event. The degree of suppression caused by NK-IS was directly related to the duration of exposure to the inhibitory molecule. In addition, a preliminary temperature-dependent step of binding to and/or intracellular entry of NK-IS into effectors is required before suppression can occur. NK-IS prevents the activation of NK cell lysis by interferon and Corynebacterium parvum and effectively inhibits lysis mediated by already activated effectors. The potent suppression of NK lysis and prevention of interferon and C. parvum-mediated activation of NK lysis by a soluble product of peritoneal cells may explain the extremely low level of NK effector cell function within the peritoneal cavity.

Regulation of cytotoxic reactivity to minor histocompatibility antigens by administration of Corynebacterium parvum

B10.BR(H-2k) mice were primed with H-2-identical allogeneic CBA/J(H-2k) spleen cells and restimulated in vitro 14 days later to generate specific secondary cytotoxic lymphocytes. A single intravenous injection of primed mice with 700 micrograms of Corynebacterium parvum 7 days after alloimmunization markedly inhibited the subsequent secondary in vitro generation of cytotoxic cells. In addition, regulatory spleen cells were detected in alloimmunized C-parvum-injected mice that prevented the restimulation of primed control spleen cells. Suppressive activity could not be abrogated by treating regulatory cells with anti-theta antibody and complement or by removing phagocytic cells, but it was overcome by treatment with mitomycin C. Unfractionated regulatory cells suppressed responses in an antigen nonspecific fashion. However, cells remaining after carbonyl and iron treatment (nonphagocytic) could no longer suppress responses to third party alloantigens while maintaining significant suppression of anti-CBA responses. These data suggest the generation of two distinct suppressor cell types that can control the cytotoxic response to minor histocompatibility antigens: an antigen-nonspecific phagocytic cell and an antigen specific nonphagocytic, non-theta-bearing cell.

Inhibition of natural killer cell activity by a soluble substance released by rat peritoneal cells

We describe here a soluble substance released by nonadherent cells from the peritoneal cavity of W/Fu rats that markedly inhibits the activity of mouse, rat, and human natural killer (NK) cells. The NK-inhibiting substance (NK-IS) has low molecular weight (less than 1000), is heat resistant (100 degrees for 15 min), and is insensitive to nonspecific proteases. NK-IS is produced in the presence of indomethacin (1 to 10 micrograms/ml), suggesting it is not prostaglandin. The inhibitory effect was seen on unstimulated as well as on cells activated in vivo or in vitro by Corynebacterium parvum. The activity of cells mediated antibody-dependent cell cytotoxicity (K-cells) was also inhibited by NK-IS although to a lesser degree. In sharp contrast, the substance had little effect on lysis mediated by murine or human cytotoxic T-lymphocytes. Production of NK-IS from rat peritoneal cells was significantly greater than by spleen cells. since the peritoneal cavity is relatively deficient in base-line NK activity compared to spleen, these data suggest that NK-IS may play an in vivo role in the expression of NK cytotoxicity.

Augmentation of NK activity by Corynebacterium parvum fractions in vivo and in vitro

Biochemically modified whole cell preparations and derived fractions of Corynebacterium parvum (C. parvum) were evaluated for the ability to augment natural killer cell cytoxicity in vivo and in vitro in rats. Unfractionated C. parvum enhanced peritoneal cell (Pc) NK activity in a dose dependent fashion. This activity appeared to be enriched in insoluble light residue material obtained from hot phenol water extraction. Enhancement of Pc cytotoxicity was significantly greater at all time points tested in rats injected with light residue when compared to rats injected with comparable amounts (by dry weight) of unfractionated organisms. In addition, pyridine extractable material and HCl modified preparations were capable of boosting Pc NK activity following intraperitoneal (I.P.) injection. Periodate treatment abrogated C. parvum's ability to boost Pc cytotoxicity and insoluble residue material obtained from pyridine extraction was likewise devoid of NK-enhancing properties. Culture of rat spleen cells overnight with unfractionated C. parvum, light residue and pyridine residue materials enhanced NK cytotoxicity whereas HCl and periodate modified whole cell preparations as well as phenol and pyridine extractable material were incapable of boosting cytotoxicity in vitro. In vitro augmentation by culturing with light residue was dependent on the presence of adherent cells in rat spleen cell populations. Pyridine extracts boost cytotoxicity in vivo and have no effect in vitro while the opposite is true of pyridine residue material suggesting different mechanisms of NK augmentation by C. parvum between in vitro and in vivo systems.

Effects of different fractions of Corynebacterium parvum on the cytotoxic T-cell response to alloantigens in mice

The effects of several biochemically derived fractions of Corynebacterium parvum and chemically treated intact organisms on the generation of cytotoxic T-lymphocytes (CTL) were assessed in a tumor allograft model with the use of C57BL/6 and DBA/2 mice. The acid-modified, active and inactive fragmented preparations and pyridine extract and residue were all capable of inhibiting primary spleen cell allocytotoxicity. Only the active fragmented, pyridine residue and unfractionated preparations caused splenomegaly and prevented the secondary in vitro generation of CTL. Periodate treatment, acid modification, and pyridine extraction abrogated the ability of C. parvum to activate suppressor macrophages. Although incapable of activating suppressor macrophages, the phenol and pyridine extracts significantly enhanced suppressor T-cell activity. The effect of whole C. parvum organisms on T-cell-mediated cytotoxicity is thus extremely complex and due to the direct activation or inhibition of various cell types that interact with each other.

Intensive chemotherapy is the treatment of choice for elderly patients with acute myelogenous leukemia

One-hundred and seven patients with acute myelogenous leukemia (AML) ranging in age from 15 to 82 yr who were previously untreated, received a 70 day high-dose remission induction regimen consisting of daunorubicin, cytarabine, and thioguanine (TAD). Identical complete remission rates of 65% were observed for 33 patients 60 yr of age and older and for 74 patients age 15-59 yr. Median remission duration and survival were 14 mo and 22 mo for patients 60 yr and older, and 16 mo and 22 mo for patients 15-59 yr. These differences are not significant. These data indicate that older patients respond to intensive chemotherapy in a similar manner to younger patients with this disease.

Intensive chemotherapy for acute myelogenous leukemia

A complete remission rate of 82% was obtained in a group of 68 patients with acute myelogenous leukemia treated with a high-dose induction chemotherapy (TAD) consisting of 7-day courses of 6-thioguanine, cytarabine, and daunorubicin. The patients who achieved remission received intensive consolidation chemotherapy and were randomized to receive maintenance chemotherapy with or without immunotherapy. Median remission duration was 13 months and median survival, 21 months. Neither central nervous system prophylaxis nor the addition of immunotherapy to the maintenance regimen prolonged remissions or improved survival. Age, sex, and subclassification of acute myelogenous leukemia had no effect on the remission rate or survival. These data indicate that a large proportion of patients with acute myelogenous leukemia can achieve remission with intensive induction chemotherapy. Attempts to prolong remission have been less successful.

Peripheral blood lymphocyte receptors for B-lymphoblastoid cell lines (B-LCL)

Because interactions between B cells and T lymphoyctes are of fundamental importance in the generation of the immune response to most antigens, we attempted to identify the cells capable of binding B-lymphoblastoid cell lines (B-LCL), their tissue distribution, and their presence in other species. Cells bearing a surface receptor for B-LCL were found in human peripheral blood, tonsil, and bone marrow, as well as mouse and rat spleen. Binding cells were phenotypically heterogeneous. The majority are T cells as defined by their ability to bind sheep red blood cells (E-rosettes). However, a subpopulation of non-T-lymphocytes were capable of binding B-LCL. This was demonstrated by depleting T cells with an E-rosette centrifugation technique and then performing a double rosette assay. The wide distribution of T lymphocytes with receptors for B-lymphoblastoid cells within peripheral lymphoid organs and their presence in several species suggest that these surface molecules may represent one of the means by which T cells and B cells interact in the induction of the immune response to T-dependent antigens.

cis-Diamminedichloroplatinum(II) alone and in combination in the treatment of testicular and other malignancies

A total of 124 patients with malignancy were treated with cis-diamminedichloroplatinum(II) (CDDP) alone (58 patients, 121 courses) or in combination with other drugs (66 patients, 176 courses). Toxic effects of single-agent therapy were primarily nausea and vomiting; mild renal hematopoietic, and auditory toxic effects occasionally occurred. Positive responses were seen in ovarian and testicular tumors. Rare responses were also seen in melanoma and bladder carcinoma. CDDP combined with vinblastine and bleomycin resulted in a 39% complete response rate in testicular germ cell tumors. All complete responders and ten patients treated in an adjuvant fashion remain alive and disease-free. Failures to achieve a complete response were associated with bulky extrapulmonary disease, pulmonary disease greater than 2.5 cm in diameter, and past failure to respond to vinblastine and bleomycin. CDDP combination regimens for acute leukemia and for melanoma were of little value. Combination chemotherapy for head and neck cancer showed activity, including a complete response in one of two patients with nasopharyngeal lymphoepithelioma.