Correspondence *Abramson Family Cancer Research Institute, 551 BRB II/III, 421 Curie Boulevard, Philadelphia, PA 19104, USA

Email rhv@mail.med.upenn.edu

The case

A 47-year-old woman with a 6-year disease-free remission from stage III ovarian cancer presented to her gynecological oncologist with dyspnea and increased abdominal girth in April 2002. Following her initial diagnosis, she had undergone debulking surgery with total abdominal hysterectomy and bilateral salpingo-oophorectomy. Pathological investigation had revealed a serous tumor of low malignant potential, with no evidence of stromal invasion; adjuvant chemotherapy was, therefore, not administered. Disease recurrence manifested as ascites and a left pleural effusion, with malignant cytology observed in fluid samples from both locations. The patient was treated with weekly carboplatin and docetaxel from April 2002 to April 2003 and then switched to tamoxifen 20 mg daily from April 2003 to September 2003, but dyspnea returned and she was found to have an increased left pleural effusion with mediastinal shift requiring a large-volume thoracentesis for symptom control.

Tamoxifen was discontinued, and the patient was briefly treated with oral etoposide 50 mg daily in December 2003, but her disease progressed and a second large-volume thoracentesis was required for a recurrent, symptomatic left pleural effusion. Postprocedure imaging demonstrated entrapment of her left lung by visceral pleural tumor with an ex vacuo pneumothorax (Figure 1A), precluding her from a simple pleurodesis as treatment for recurrent malignant effusion.

Figure 1 Evaluation of tumor response to therapy.

CT scans of the chest (A) before and (B) 60 days after interferon therapy show improvement of the left pleural effusion. 2-[¹⁸F]fluoro-2-deoxyglucose PET scans (C) before and (D) 60 days after therapy show complete resolution of abnormal tracer activity in the abdomen (arrows). Normal tracer activity in the kidney is circled.

Full figure and legend (24K)Figures & Tables indexDownload Power Point slide (228K)

The patient was then enrolled in a phase I clinical trial of a novel adenoviral vector encoding interferon (IFN-) (BG00001; Biogen Idec, Cambridge, MA), and in February 2004 she received a single infusion of 9 10¹¹ viral particles delivered via a tunneled pleural catheter in the left chest. Toxicities included grade 1 chills, fatigue, hypotension and grade 2 body aches. Swabs of the left chest wall and samples of pleural fluid and plasma were obtained before treatment and at regular intervals for 8 weeks afterwards to assess gene transfer. Pleural fluid culture for replication-incompetent adenovirus and adenovirus polymerase chain reaction (PCR) were negative at baseline, but each test became positive within 3 days of treatment (Table 1). Virus was detectable for up to 1 week after treatment by culture and for up to 6 weeks by PCR. Chest wall and plasma samples were always negative by culture and PCR. Replication-competent adenovirus was not detected in any sample before or after treatment. No IFN- protein was found before treatment using an enzyme-linked immunosorbent assay, but within 24 h of vector instillation there was a rapid and marked elevation of intrapleural IFN- (to 11.5 ng/ml), which declined to undetectable levels (<0.5 ng/ml) over 1 week (Table 1). Intrapleural IFN-, IFN-, and interleukin (IL-) 10 protein were also monitored, but none reached a level, either before or after treatment, sufficient for detection by the assays used (Table 1).

Table 1 Evaluation of adenoviral shedding and interferon gene transfer.

Full tableFigures & Tables indexDownload Power Point slide (246K)

Restaging studies 2 months after treatment showed re-expansion of the left lung (Figure 1B) and regression of tumor lesions in the abdomen. Although tumor regression did not meet the Response Evaluation Criteria In Solid Tumors (RECIST) definition of a partial response, a 2-[¹⁸F]fluoro-2-deoxyglucose (FDG)-PET scan performed 60 days after vector infusion revealed a complete resolution of the intense abnormal tracer uptake in the abdomen that had been noted on the baseline FDG-PET scan (Figure 1C, D). Tracer uptake in the left chest wall was minimal. Serum cancer antigen 125 dropped from 140 U/ml at baseline to 107 U/ml 2 months after treatment.

To investigate the immunological impact of IFN- gene therapy in this patient, alterations in lymphocyte populations in pleural fluid were analyzed before and after treatment. Cell count and differential analysis of pleural fluid cells demonstrated an increase in pleural fluid leukocytes beginning 2 days after vector instillation (Table 2). At 1 week, the concentration of total pleural leukocytes had increased 25-fold over baseline, including a 13-fold increase in pleural lymphocytes (baseline <70 cells/l). At each time point after treatment, nearly all pleural fluid lymphocytes were CD45⁺ CD3⁺ T cells, for which the CD4:CD8 ratio remained roughly 7:1, as determined by flow cytometry (Figure 2A). For both CD4⁺ and CD8⁺ T cells the dominant phenotype was CD45RA^- CD45RO⁺, and a notable fraction of these T cells expressed CD69, a marker of activation (Figure 2B).

Figure 2 Cellular composition of pleural fluid lymphocytes before and after treatment.

(A) The absolute number of cells per l for each of six cell populations is shown as a function of time after treatment: total lymphocytes, total CD3⁺ T cells, CD3⁺ CD4⁺ T cells, CD3⁺ CD8⁺ T cells, CD19⁺ B cells, and CD56⁺ natural killer (NK) cells. (B) Pleural effusion T cells obtained 1 week after treatment were analyzed with multiparameter flow cytometry and found to be largely CD45RA^- CD45RO⁺, with a notable fraction of cells expressing the activation marker CD69.

Full figure and legend (40K)Figures & Tables indexDownload Power Point slide (245K)

Table 2 Cellular composition of the patient's pleural fluid.

Full tableFigures & Tables indexDownload Power Point slide (227K)

In vitro analysis of pleural effusion T cells obtained 1 week after treatment showed that more than 50% of CD4⁺ T cells and more than 70% of CD8⁺ T cells produced the effector cytokines IFN- and tumor necrosis factor following stimulation with phorbol 12-myristate 13-acetate and ionomycin, as determined by intracellular cytokine flow cytometry (data not shown). Fewer than 0.5% of pleural effusion CD4⁺ or CD8⁺ T cells produced the suppressive cytokine IL-10. Together, these findings suggested an infiltration and accumulation of activated T cells into the pleural space following treatment.

Although the patient received the adenoviral vector in the thorax only, her clinical response in both the thorax and abdomen was consistent with a systemic cellular immune response to therapy. When stimulated in vitro with autologous tumor RNA, post-treatment CD8⁺ T cells (isolated 60 days after therapy) lysed SKOV3 ovarian carcinoma cells that matched the patient's MHC class I allele of human leukocyte antigen (HLA)-A3 (Figure 3A). This ability was not noted for cells isolated before treatment, and neither pretreatment nor post-treatment CD8⁺ T cells stimulated with green fluorescent protein RNA (negative control) lysed SKOV3 ovarian tumor cells (Figure 3A). These results suggested that IFN- gene therapy induced functional antitumor T cells in this patient that contributed to clinical regression.

Figure 3 Evaluation of tumor-specific cellular immune response to therapy.

(A) Peripheral blood T cells obtained before (open circles) or after (closed circles) treatment were stimulated with RNA-loaded autologous dendritic cells, following methods previously described.¹⁷ Cells were stimulated either with total tumor RNA prepared from the autologous tumor cell line or with green fluorescent protein RNA, as described.¹⁷ After 2 stimulations over 14 days, T cells were evaluated for cytotoxicity against SKOV3 ovarian tumor cells or an autologous tumor cell line that was generated from ascites obtained after disease progression. Standard chromium release assays were performed as described.¹⁷ Lysis was observed only against SKOV3 using post-treatment T cells stimulated with total autologous tumor RNA. Standard deviation at each time point was <5%. (B) Flow cytometry was used to measure total MHC class I expression (HLA-ABC) or HLA-A3 expression on an autologous tumor cell line, shown in comparison to the appropriate negative control. The autologous tumor cell line generated after treatment was MHC class I-positive but HLA-A3 negative. SKOV3 cells were positive for HLA-ABC and HLA-A3 (not shown). Neither autologous tumor nor SKOV3 cells expressed MHC class II molecules (not shown). (C) Flow cytometry was used to analyze MHC class I expression on pleural fluid cells at baseline versus ascites cells at disease progression. Dead cells that were labeled with 7-amino-actinomycin were excluded, and patient leukocytes and tumor cells were identified by expression of CD45 or Ber-EP4, respectively. Leukocytes labeled with HLA-A3 (and HLA-ABC) mAb in all cases, as expected. Tumor cells labeled with HLA-ABC mAb at each time point, but showed greatly reduced levels of HLA-A3 expression after treatment when compared with baseline. Abbreviations: HLA, human leukocyte antigen; mAb, monoclonal antibody.

Full figure and legend (65K)Figures & Tables indexDownload Power Point slide (270K)

Four months after vector instillation, the patient developed worsening ascites, and restaging studies showed progressive disease, with new lesions and an increase in the size of existing lesions. FDG-PET showed multiple new foci of increased tracer uptake within the abdomen consistent with peritoneal carcinomatosis, and serum cancer antigen 125 increased to 296 U/ml. There was no change in metabolic activity within the left hemithorax, nor any reaccumulation of the malignant pleural effusion, as determined by CT scan.

To investigate whether the patient's disease progression was related to an immune escape mechanism, an autologous tumor cell line was generated from ascites obtained at the time of disease progression and evaluated as a target in cytotoxicity assays. In contrast to the patient's lymphocytes, the autologous tumor cell line expressed only low levels of HLA-A3 (Figure 3B). The expression of other HLA molecules was intact. Moreover, although post-treatment CD8⁺ T cells stimulated with autologous tumor RNA efficiently lysed HLA-A3-matched allogeneic ovarian carcinoma cells, these T cells failed to lyse autologous tumor cells (Figure 3A). Flow cytometric analysis of unmanipulated samples of malignant fluid obtained before treatment and at the time of disease progression further demonstrated allelic loss of HLA-A3 on the patient's tumor cells (Figure 3C). Overall, these findings suggested a mechanism of tumor immune escape that could have contributed to disease progression. Rescuing T-cell cytotoxicity in vitro by reintroducing HLA-A3 into the autologous tumor would be an interesting area for future study.

The patient was treated with bevacizumab 15 mg/kg every 3 weeks from August 2004 to April 2005 and then bevacizumab 15 mg/kg every 3 weeks with oral cyclophosphamide 50 mg daily from December 2005 to April 2006. The best clinical response during this time was stable disease by RECIST. In May 2006, she commenced liposomal doxorubicin 40 mg/m² every 4 weeks because of disease progression.

Discussion of diagnosis

Ovarian cancer is the fifth most common malignancy in women, with the majority of patients presenting with advanced-stage disease.¹ Surgery and chemotherapy readily produce remission, but relapse is common, typically in the peritoneal cavity. Risk of relapse is highest in patients with advanced-stage, high-grade, and suboptimally debulked tumors.¹ The 5-year survival rate for all stages is less than 40%.¹ The diagnostic evaluation of a woman presenting with simultaneous pleural effusion and ascites is markedly influenced by a prior history of an ovarian tumor, which, even in the setting of a 6-year remission and an initial pathology of low malignant potential, could in some patients progress to advanced, invasive carcinoma. Radiographic evidence of peritoneal and pleural disease in the absence of parenchymal disease in the liver, lung, and other organs is a classic presentation of invasive stage IV ovarian cancer. Nevertheless, pathological diagnosis is mandatory before treatment, obtained in this case by cytology. In light of ongoing studies regarding tumors with low malignant potential as potential precursor lesions to invasive carcinoma, molecular genetic analysis of this patient's initial and subsequent tumor specimens might have helped to distinguish whether the patient had one neoplastic process or two.²

Differential diagnosis

The differential diagnosis of simultaneous development of pleural effusion and ascites includes inflammatory pleuroperitonitis from drug reaction, pancreatitis, or intra-abdominal abscess; cirrhotic liver disease with portal hypertension and hepatic hydrothorax; and congestive heart failure with passive hepatic congestion. Laboratory and cytologic evaluation of pleural or ascitic fluid is the first step in narrowing the differential diagnosis.

Treatment and management

Increasing evidence that ovarian cancer can be recognized and attacked by the immune system has fueled efforts to develop novel immunotherapies for this disease.³ The presence of CD3⁺ intraepithelial T cells strongly correlates with increased progression-free survival and overall survival in patients with ovarian cancer,⁴ and tumor-specific T cells can be identified in the peripheral blood of patients with ovarian cancer.⁵ T-cell immunosurveillance in ovarian cancer has been most dramatically illustrated by the devastating syndrome of paraneoplastic cerebellar degeneration, which is mediated by CD8⁺ T cells specific for the cdr2 antigen expressed by both tumor cells and cerebellar Purkinje neurons.⁶ Negative immune regulatory cells that typically accompany robust cellular immune responses are also prominent in ovarian cancer.⁷ Regulatory T cells function to downregulate antitumor immune responses, and the presence of these cells in ovarian tumors is associated with reduced survival.⁷ Novel strategies that block negative immune mechanisms, including anti-CTLA4 monoclonal antibodies, have shown great promise in early clinical trials.⁸ Strategies for vaccination and adoptive T-cell therapy are also under active clinical investigation.³

Cytokines including IFN- IFN-, IFN-, and IL-12 have been previously explored as immunostimulants for the treatment of ovarian cancer.³ IFN- mediates complex antitumor properties, including direct induction of cancer cell apoptosis and potent enhancement of antitumor immune responses through the stimulation of macrophages, dendritic cells, natural killer cells, and T cells.⁹ Phase I trials of recombinant IFN- have, however, been limited by toxicity, poor efficacy, and limited drug half-life.^{10, 11} In preclinical models, cytokine-encoding viral vectors achieve higher and more sustained cytokine levels at the tumor site than those resulting from systemic or regional administration of recombinant cytokine proteins.¹² Clinically, intracavitary catheters facilitate delivery of these vectors to the tissues and, although our initial clinical experience has been with the use of intrapleural catheters for patients with malignant pleural effusions, intraperitoneal catheters have already proven to be an important drug delivery device in ovarian cancer.¹³ The prospect of adenoviral gene therapy in ovarian cancer using intraperitoneal catheters represents an exciting future approach, especially given early promising results with intraperitoneal administration of recombinant IFN- in this disease.¹⁴

In preclinical studies, intracavitary delivery of a replication-deficient adenoviral vector encoding IFN- in tumor-bearing mice elicits immune activation and remarkable tumor regression.¹² Delivery of a control adenoviral vector without the IFN- insert has minimal antitumor effect in this model—the same effect as a saline injection.

Approval was obtained from the Institutional Review Board, the FDA, and the Recombinant DNA Advisory Committee, for a phase I trial of a novel recombinant serotype 5 adenoviral vector encoding IFN- for patients with malignant pleural effusions. The patient in this case had a promising initial response to this therapy, but after 4 months her disease progressed, and investigations suggested that a mechanism of tumor immune escape might be involved. Downregulation of MHC class I expression on tumors is a well-characterized mechanism of tumor immune escape and remains a notable obstacle for immunotherapies that induce a robust T-cell immune response.¹⁵

Although preclinical studies in mice argue against adenovirus itself as a major contributor to the induced immune response,¹² this issue would be ultimately best settled in humans by a randomized study comparing the use of an adenoviral vector encoding IFN- and an adenoviral vector alone.

Given the inflammatory and immunoreactive response to treatment experienced by the patient, it was somewhat surprising that there was no detectable increase in host IFN- in the pleural fluid. Although the host response to physiologic viral infection is often a positive feedback for type I IFN secretion,¹⁶ the lack of detectable IFN- in the patients in the phase I trial could represent a response to pharmacologically driven, high-level secretion of IFN-.

Nine other patients with malignant pleural effusions (seven patients with mesothelioma and two with lung carcinoma) were treated in the phase I study. Overall, stable disease by RECIST criteria was observed in four patients following treatment. Tumor-specific humoral immune responses were observed in eight patients. Only the patient examined in this Case Study had a documented CD8⁺ T-cell response following treatment. The authors plan to submit the full trial data from the phase I study for publication.

Conclusion

Immunotherapy represents a promising approach for ovarian cancer, but numerous challenges remain. In a phase I clinical trial of adenoviral-mediated IFN- gene therapy given as a single injection via a pleural catheter, clinical benefit was observed in an ovarian cancer patient with refractory malignant pleural effusion and ascites. Laboratory studies revealed the induction of antitumor T cells following treatment, a mechanism of action that had been predicted from preclinical mouse models. Clinical benefit was transient, however, and tumor progression occurred in the setting of loss of tumor MHC expression, a likely means of immune escape. A trial testing repeated doses of IFN- gene therapy is underway.

Acknowledgments

This work was supported by National Institutes of Health grant P01 CA 66726.

References

1. Cannistra SA (2004) Cancer of the ovary. N Engl J Med 351: 2519–2529 | Article | PubMed | ISI | ChemPort |
2. Shih LM and Kurman RJ (2004) Ovarian tumorigenesis: a proposed model based on morphological and molecular genetic analysis. Am J Path 164: 1511–1518 | PubMed |
3. Coukos G et al. (2005) Immunotherapy for gynaecological malignancies. Expert Opin Biol Ther 5: 1193–1210 | Article | PubMed | ChemPort |
4. Zhang L et al. (2003) Intratumoral T cells, recurrence, and survival in epithelial ovarian cancer. N Engl J Med 348: 203–211 | Article | PubMed | ISI | ChemPort |
5. Schlienger K et al. (2003) TRANCE- and CD40 ligand-matured dendritic cells reveal MHC class I-restricted T cells specific for autologous tumor in late-stage ovarian cancer patients. Clin Cancer Res 9: 1517–1527 | PubMed | ISI | ChemPort |
6. Albert ML et al. (1998) Tumor-specific killer cells in paraneoplastic cerebellar degeneration. Nat Med 4: 1321–1324 | Article | PubMed | ISI | ChemPort |
7. Curiel TJ et al. (2004) Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival. Nat Med 10: 942–949 | Article | PubMed | ISI | ChemPort |
8. Hodi FS et al. (2003) Biologic activity of cytotoxic T lymphocyte-associated antigen 4 antibody blockade in previously vaccinated metastatic melanoma and ovarian carcinoma patients. Proc Natl Acad Sci USA 100: 4712–4717 | Article | PubMed | ChemPort |
9. Le Bon A and Tough DF (2002) Links between innate and adaptive immunity via type I interferon. Curr Opin Immunol 14: 432–436 | Article | PubMed | ISI | ChemPort |
10. Rambaldi A et al. (1985) Intraperitoneal administration of interferon- in ovarian cancer patients. Cancer 56: 294–301 | Article | PubMed | ChemPort |
11. Markman M et al. (2004) Phase 2 trial of interferon-beta as second-line treatment of ovarian cancer, fallopian tube cancer, or primary carcinoma of the peritoneum. Oncology 66: 343–346 | Article | PubMed | ChemPort |
12. Odaka M et al. (2001) Eradication of intraperitoneal and distant tumor by adenovirus-mediated interferon- gene therapy is attributable to induction of systemic immunity. Cancer Res 61: 6201–6212 | PubMed | ISI | ChemPort |
13. Armstrong DK et al. (2006) Intraperitoneal cisplatin and paclitaxel in ovarian cancer. N Engl J Med 354: 34–43 | Article | PubMed | ISI | ChemPort |
14. Berek JS et al. (1999) Intraperitoneal interferon- in residual ovarian carcinoma: a phase II gynecologic oncology group study. Gynecol Oncol 75: 10–14 | Article | PubMed | ISI | ChemPort |
15. Seliger B et al. (2002) HLA class I antigen abnormalities and immune escape by malignant cells. Semin Cancer Biol 12: 3–13 | Article | PubMed | ChemPort |
16. Biron CA (2001) Interferons and as immune regulators: a new look. Immunity 14: 661–664 | Article | PubMed | ISI | ChemPort |
17. Coughlin CM et al. (2004) RNA-transfected CD40-activated B cells induce functional T-cell responses against viral and tumor antigen targets: implications for pediatric immunotherapy. Blood 103: 2046–2054 | Article | PubMed | ISI | ChemPort |

Contact the journal about this article

Subject areas under which this article appears: Immunotherapy