Phase I/II Study of Metastatic Melanoma Patients Treated with Nivolumab Who Had Progressed after Ipilimumab

The checkpoint inhibitor nivolumab is active in patients with metastatic melanoma who have failed ipilimumab. In this phase I/II study, we assessed nivolumab's safety in 92 ipilimumab-refractory patients with unresectable stage III or IV melanoma, including those who experienced grade 3-4 drug-related toxicity to ipilimumab. We report long-term survival, response duration, and biomarkers in these patients after nivolumab treatment (3 mg/kg) every 2 weeks for 24 weeks, then every 12 weeks for up to 2 years, with or without a multipeptide vaccine. The response rate for ipilimumab-refractory patients was 30% (95% CI, 21%-41%). The median duration of response was 14.6 months, median progression-free survival was 5.3 months, and median overall survival was 20.6 months, when patients were followed up for a median of 16 months. One- and 2-year survival rates were 68.4% and 31.2%, respectively. Ipilimumab-naïve and ipilimumab-refractory patients showed no significant difference in survival. The 21 patients with prior grade 3-4 toxicity to ipilimumab that was managed with steroids tolerated nivolumab well, with 62% (95% CI, 38%-82%) having complete or partial responses or stabilized disease at 24 weeks. High numbers of myeloid-derived suppressor cells (MDSC) were associated with poor survival. Thus, survival and long-term safety were excellent in ipilimumab-refractory patients treated with nivolumab. Prior grade 3-4 immune-related adverse effects from ipilimumab were not indicative of nivolumab toxicities, and patients had a high overall rate of remission or stability at 24 weeks. Prospectively evaluating MDSC numbers before treatment could help assess the expected benefit of nivolumab.

Immunosuppression in murine renal cell carcinoma. I. Characterization of extent, severity and sources

Four cell-mediated immunological responses related to tumor elimination have been examined in mice injected with a transplantable renal cell carcinoma (Renca). Lymphokine-activated killer (LAK) cells generated in vitro from spleen cells of normal mice were capable of attacking Renca, EL-4, P815 and YAC-1 targets, but those from mice bearing Renca for 3 weeks could not. Natural killer activity, stimulated in vivo by administering poly(I) poly(C), was less than 50% of normal in Renca-bearing hosts. In addition, development of cytotoxic T lymphocytes to allogeneic targets was markedly inhibited in mice possessing the renal tumor. Finally, the delayed hypersensitivity response to a dermally applied hapten was approximately 70% less than normal in tumor-bearing mice, no matter whether the tumor existed subcutaneously or intrarenally. A kinetic study of the development of non-responsiveness using the LAK assay showed onset of poor response at 1 week, which became maximal within 3 weeks following receipt of tumor subcutaneously. The immunological depression was seen to be attributable in part to suppressor cells present among spleen cells but not bone marrow cells of tumor-bearing hosts. The suppressor cells prevented in vitro LAK generation by normal spleen cells and, when adoptively transferred to normal mice, they inhibited natural killer stimulation and delayed hypersensitivity generation. Another source of immunological down-regulation was provided by Renca cells themselves. Incorporation of Renca cells that had been X-irradiated with 30,000 rad into cultures of normal and Renca-derived splenic cells suppressed replication of both almost completely. Furthermore, the presence of X-irradiated Renca cells in cultures of normal spleen cells prevented development of LAK cells. Thus, the suppression seen in Renca-bearing mice derives from multiple sources and whether each is in any way related to the other has been discussed. Identification of the phenotypes of cells responsible for the lymphoid cell-mediated suppression and examination of its elimination are communicated in the companion paper.

Properties of mouse melanoma antigen and its secretion mechanism from the cell surface

We analyzed the biochemical properties and biological significance of the melanoma antigen secreted in the culture supernatants of B16 melanoma cells. The 80 kilodalton (kd) molecule bearing the epitopes of mouse melanoma antigen was found to associate noncovalently with an 18 kd moiety in the culture supernatants as well as on the cell surface. Tunicamycin treatment of B16 cells did not affect the expression of the 69 kd nonglycosylated form of the 80 kd molecule but did abolish the association between the two molecules on the cell surface. We could not detect this antigen as a soluble form when the N-linked glycosylation was inhibited. Therefore, the glycosylation of the 80 kd molecule is essential for the formation of the 80 kd/18 kd complex and also for the secretion. Moreover, the affinity-purified melanoma antigen from the supernatants could induce anti-melanoma suppressor cells which block the generation of cytotoxic T lymphocytes against melanoma cells. Thus, the 80 kd glycoprotein as a soluble melanoma antigen performed a pivotal function in the escape mechanisms of melanoma cells from the host immune surveillance system.

Suppressor cells in the effector phase of autologous cytotoxic reactions in cancer patients

Cytotoxicity was induced in lymphocytes (CL) from 10 out of 15 patients by autologous mixed lymphocyte tumor cell culture and further cultivation with recombinant interleukin-2. In cells from 3 of the 10 patients, cytotoxicity was suppressed by more than 50% when autologous peripheral blood mononuclear cells (PBMC) from the patients with large tumors were added to the autologous killing system. The cells responsible for suppressing the cytotoxicity in the effector phase were adherent or nonadherent to plastic depending on the patient examined. The T cell fraction from 1 patient significantly suppressed the cytotoxic activity, and this suppression was seen only in the autologous system. On the other hand, plastic adherent cells but not T cells from PBMC of 2 subjects suppressed the cytotoxic activity of CL. The reason why the main cell population suppressing the CL activity differed among the patients is unclear. However, the findings that the suppression was mostly abrogated following resection of the tumor mass suggested that suppressor cells, either of macrophage lineage or T cells, are induced in patients with a large tumor mass. This speculation is supported by the finding that the PBMC from a patient with tumor recurrence regained the suppressive activity.

Suppressor deletion therapy: selective elimination of T suppressor cells in vivo using a hematoporphyrin conjugated monoclonal antibody permits animals to reject syngeneic tumor cells

A MAb (B16G) which recognizes a constant epitope on TsC and their soluble factors in DBA/2 mice has been described previously. In this study, we show that when this MAb is covalently linked to the photoactivable molecule Hp, and injected i.v. into P815 tumor-bearing mice which were subsequently exposed to light, tumors undergo permanent regression in 10%-40% of these mice (depending on the individual experiment). All control animals died within an average of 22-24 days after tumor cell injection. It is suggested that tumor regression is attributable to immune mechanisms facilitated by the elimination of a population of TsC. When splenocytes of B16G-Hp-treated mice were assayed in vitro for the generation of CTL active against P815 tumor cells, it was found that 24 h after treatment, a significant increase in killer cell activity was noted but that this effect was gone by 48h. We also show that B16G-Hp conjugates are capable in vitro of specifically killing cells of a TsC hybridoma, A10 (which has been shown previously to secrete a T suppressor factor reactive with P815 cell surface antigens). This conjugate had no cytotoxic effect on P815 cells under conditions in which A10 cells were killed.

Characterisation of suppressor cells generated following cryosurgery of an HSV-2-induced fibrosarcoma

Cryosurgery of a primary HSV-2-induced hamster fibrosarcoma resulted in the generation of a population of suppressor cells. These cells were detectable in the spleen 1-10 days post-cryosurgery by their ability to suppress the proliferation of immunocompetent splenic T-lymphocytes following exposure to concanavalin A (Con A). The spleens of tumour-bearing (t.b.) animals which received cryosurgery 3 days previously displayed gross splenomegaly due to the generation of large numbers of highly proliferative erythroblasts. The erythroblast cells were unlikely to be the source of suppression since time course studies have demonstrated the presence of suppressor cells before and after their appearance in the spleen. The erythroblasts therefore probably reflected a response by the host to regenerate the erythrocytes lost during surgery and their presence was independent of the appearance of suppressor cells. Characterisation of the suppressor cell has revealed it to be non-adherent and esterase negative making it unlikely to be of macrophage (MO) lineage. This was confirmed by the ability of splenic MOs from day 3 t.b. cryosurgery-treated animals to completely restore Con A-dependent T-lymphocyte proliferation following MO depletion. As nylonwool column-eluted cells are able to suppress Con A-dependent T-lymphocyte proliferation, it seemed unlikely that B-lymphocytes play a role in cryosurgery-induced immunosuppression. These findings suggest that cryosurgery of a t.b. animal results in the generation of a population of T-lymphocytes capable of suppressing Con A-dependent T-lymphocyte proliferation, and infers that these cells contribute to the inferior prognosis following cryosurgery as compared to excision of a metastatic tumour.

Suppression of murine NK activity induced by Corynebacterium parvum: further characterization and abrogation of suppressor cells

Formalin-killed Corynebacterium parvum (CP), given at a dose of 0.4-0.7 mg/mouse IV or IP, induced suppressor cells for NK activity in B6C3F1 mice. The suppressor cells belong to at least two different populations, plastic adherent and nonadherent, and were not depleted by antibodies specific for Thy-1.2, Iak, or NK-1.2 surface markers. Administration of p-I:C, an interferon-inducer, to animals 18 h before the assay did not affect the suppressor activity. Hypotonic shock treatment of splenocytes abrogated the in vitro suppressive activity, and subsequent reconstitution of the shock-treated cells with RBC failed to restore the suppressive activity. SJL/J mice, which have low NK activity, exhibited suppressor activity comparable to B6C3F1 mice following CP treatment, whereas CP-treated BALB/c athymic and euthymic mice showed a lower ability to generate suppressors for NK as compared to B6C3F1 mice.

In vitro selective effect of melphalan on human T-cell populations

In vitro treatment with 2 micrograms/2 X 10(6) cells melphalan (L-PAM: L-phenylalanine mustard) significantly decreased the total number of T lymphocytes from peripheral blood (PBL) of healthy human donors and of the OKT4 population (precursor suppressor/helper/inducer) T cells as defined by monoclonal antibodies OKT3 and OKT4, respectively. No changes in the OKT+8 lymphocyte population (cytotoxic/mature suppressor cells) were observed following the same treatment. Preincubation of PBL with L-PAM at concentrations that do not affect the rate of DNA synthesis in PHA-stimulated lymphocytes inhibited the generation of T suppressor lymphocytes by ConA, as shown by their effect on PHA stimulation. Treatment of allogeneic PBL with L-PAM had no effect on mature suppressor T cells already induced by ConA, as shown by incubation of PBL with L-PAM after incubation with ConA.

Suppression of generation of concomitant antitumor immunity by passively transferred suppressor T cells from tumor-bearing donors

Infusion of normal recipient mice with suppressor T cells from donors bearing a progressive Meth A fibrosarcoma results in a diminished capacity of the recipients to generate concomitant and postexcision antitumor immunity. The passive transfer of suppressor cells which prevented the generation of immunity to the Meth A fibrosarcoma did not affect the capacity of the recipients to reject an allogeneic tumor. The data provides direct evidence in support of the hypothesis that suppressor T cells, generated at later stages of growth of Meth A fibrosarcoma, function to down-regulate an already acquired mechanism of concomitant immunity.

Phase-I study of intravenous modified lipid A

Endotoxin and the lipid-A portion of the molecule have a variety of biological effects, including the induction of necrosis and regression of malignancy. To date extensive clinical trials of endotoxin as a potential therapeutic agent have been shunned due to the toxicity of the material. Several lipid-A analogues have been described which have reduced toxicity and retain antitumor activity. We have investigated in a phase-I trial the clinical toxicity and immunological effects of monophosphoryl lipid A prepared from Salmonella typhimurium and Salmonella minnesota. Patients entered on the study received IV monophosphoryl lipid A twice weekly for a total of 4 weeks. At least three patients were entered sequentially at each of the dose levels of 10, 25, 50, 100, and 250 micrograms/m2 body surface area. One patient was treated at the dose level of 500 micrograms/m2. The major clinical toxicity was fever, chills, and rigor, which occurred in over 50% of the treatments at doses of 250 micrograms/m2. Two instances of bronchospasm occurred in one patient who received 250 micrograms/m2. One patient received 500 micrograms/m2 and became hypotensive. Sequential clinical data showed no evidence of renal or hepatic toxicity. A transient decrease in the WBC and platelets occurred during the first 24 h after therapy. Immune function testing measured T cells, monocyte cytostasis, monocyte suppressor cell activity, and NK activity. These data suggested a shift in monocyte populations with activated cells moving into the tissue. Direct objective antitumor activity or necrosis was not observed in this group of patients. We conclude that monophosphoryl lipid A can be given to patients in a dose of up to 100 micrograms/m2 with acceptable toxicity. Its clinical activity as a single agent in combination with other immunomodulators remains to be demonstrated.

Supernatant derived from a human hepatocellular carcinoma cell line (PLC/PRF/5) activates a population of T-suppressor cells

Harvest fluid derived from a primary hepatocellular carcinoma cell line (PLC/PRF/5) inhibited the incorporation of 3H-thymidine into PHA-activated human lymphocytes. A similar effect was observed when lymphocytes were pre-incubated with the tumour supernatant and washed prior to mitogen activation. Not only did the tumour supernatant inhibit 3H-thymidine incorporation by mitogen-activated lymphocytes, but it also inhibited production of the lymphokine leucocyte inhibitory factor (LIF). In experiments designed to establish whether a component of the tumour harvest fluid was activating a population of suppressor cells, normal mononuclear (MN) cells were treated with the PLC/PRF/5 or embryonic fibroblast supernatant for 48 h, after which they were washed and added to normal mitogen-activated lymphocyte cultures. Only cells pretreated with the PLC/PRF/5 supernatant suppressed mitogenesis. The cell responsible for the suppressor effect was a T cell, which after a further 24 h in culture liberated a suppressor factor responsible for inhibiting lymphocyte function. Although the nature of the factor/s in the PLC/PRF/5 supernatant responsible for activation of the T-suppressor cell population is unknown, it is suggested that this mechanism may be important in protecting the tumour from the immune response.

Cellular suppression of murine ADCC and NK activities induced by Corynebacterium parvum

Administration of a single dose of C. parvum (CP) induces depression of splenic NK activity in mice after a lag period of 3-5 days and this depression lasts about 2 weeks. The depressed levels of NK activity noted in this study depended on time of CP administration and were associated with the induction of suppressor cell activity. Neonatally thymectomized or sublethally irradiated mice had unimpaired ability to generate suppressor cells following CP treatment. Depletion of adherent/phagocytic cells by carbonyl iron plus magnetism, Sephadex G-10 filtration, or both neither enriched NK activity nor removed suppressor activity from the spleens of CP-treated mice. Antibody-dependent cellular cytotoxicity (ADCC) against lymphoma targets was also depressed in CP-treated mice, accompanied by a concomitant appearance of suppressor cells that interfere with ADCC at the effector level.

Serial immune testing in surgically resected lung cancer patients

Immunoregulation of phytohemagglutinin (PHA) responsiveness by glass-adherent cells and prostaglandin-synthesizing cells was serially monitored in the peripheral blood mononuclear cells (PBMC) of surgically resected stage I and stage II lung cancer patients entered on a trial of adjuvant immunotherapy. The relationship between immunoregulatory cell function, immunocompetence, and disease relapse was determined. Immunoregulatory activity was measured in PHA-stimulated cultures in the presence and absence of 2 micrograms/ml indomethacin and in the presence and absence of glass-adherent cells. In each instance of disease relapse seen, an increase in immunoregulatory cell function to a level significantly different from normal was observed 3 months prior to or coincident with clinical confirmation of disease recurrence. This was usually associated with a decline in PHA responsiveness. In the patients who did not relapse, the levels of PHA responsiveness and immunoregulatory function persisted within normal limits throughout the course of study. Percentages and numbers of leukocytes and leukocyte subsets and delayed cutaneous hypersensitivity were also monitored in this study, but could not be consistently correlated with early changes in clinical disease status. These data suggest that the development of indomethacin-sensitive and glass-adherent suppressor cells may precede and predict for tumor recurrence in surgically resected lung cancer patients.

Absence of virus-induced lymphocyte suppression and interferon production in multiple sclerosis

Lymphocytes from normal adult donors exposed in vitro to inactivated measles virus were found to exert significant suppression (33.9%) of the concanavalin A responses of cryopreserved, autochthonous responder cells. In marked contrast, lymphocytes from multiple sclerosis patients exhibited significantly reduced suppression (1.5%), and in 80% of cases failed to suppress at all. The degree of suppression increased slightly with age of the patient but did not vary with the clinical stage of disease. There was no apparent genetic restriction of suppressor activity. Although specificity of this phenomenon for measles virus has not been established, no differences in the responses of lymphocytes from normal or multiple sclerosis patient donors were found with subacute sclerosing panencephalitis, Sendai, canine distemper, mumps, or influenza viruses. Supernates of measles-treated lymphocytes from normal donors possessed both suppressive and antiviral activities. Both activities were resistant to pH 2 treatment and were neutralized by an anti-human leukocyte interferon antiserum, strongly suggesting that interferon (probably type I) was the mediator of suppression. Consistent with their inability to suppress concanavalin A responses, lymphocytes from multiple sclerosis patients failed to produce significant amounts of interferon in response to measles challenge in vitro. These results extend previous observations that multiple sclerosis patients are unable to respond appropriately to measles virus antigen in vitro.