Introduction

Epithelial ovarian cancer (EOC), of which serous is the predominant histology [1], is primarily a disease of postmenopausal women and is the most lethal of all gynecologic malignancies. Annually, EOC is diagnosed in greater than 22,000 women in the USA with more than 14,000 deaths being attributable to this disease [2, 3]. Based on International Federation of Gynecologists and Obstetricians (FIGO) criteria, EOC is surgically staged (Fig. 1).

Fig. 1
figure 1

Simplified staging for ovarian cancers

Full size image

Following surgical cytoreduction or staging, the treatment of choice for some stage I and all stage II–IV cancers is carboplatin–paclitaxel combination chemotherapy [4]. This can be administered as adjuvant and/or neoadjuvant chemotherapy depending on the patients’ ability to undergo successful cytoreductive surgery [5].

The origins of ovarian cancer are unclear. Continuous or “incessant” ovulation, uninterrupted by pregnancy or oral contraceptive use, was once thought to be the main etiologic factor [6]. However, more recent studies suggest a fallopian tube origin of serous ovarian cancers [6, 7]. Genetic factors such as family history, mismatch repair gene abnormality (Lynch syndrome) and BRCA status also contribute to an individual woman’s risk.

The chief reason for the high mortality rate from OC is late presentation with 75% of cases presenting with advanced stage disease [8]. OC has insidious onset as well as non-specific signs and symptoms. These include pelvic or abdominal pain, early satiety, urinary frequency, constipation and abdominal distension [9]. Delayed detection is also a function of a lack of an accepted general screening test. For these reasons, OC is referred to as “the whispering disease” and the “silent killer” [10, 11]. The main cause of death is metastatic growth that impinges on visceral organ function. After what may have been initially exquisite sensitivity to standard platinum chemotherapies, EOC recurs in the majority of patients and progressive resistance to treatment develops leading to exhaustion of available treatment options [12]. Therefore, the emphasis of treatment in the recurrent setting is palliative, with focus on symptom control, prolongation of survival and quality of life improvement.

Since 1990, the speed of tumor recurrence during a chemotherapy-free interval (CFI) after carboplatin–paclitaxel, empirically divided into 6-month blocks of time, forms the basis for classification of the disease, prediction of clinical course and subsequent management. Per Gynecologic Oncology Group (GOG), classification is as follows (Fig. 2):

Fig. 2
figure 2

Platinum resistance definition by Gynecologic Oncology Group (GOG). Platinum sensitivity is classified as resistant, partially sensitive or sensitive, according to the time elapsed since finishing first-line treatment. Probability of re-treatment response is shown

Full size image

  • Refractory Relapse during or within 4 weeks following platinum-based chemotherapy

  • Resistant Relapse under 6 months from last platinum therapy as platinum resistant

  • Partially platinum sensitive Relapse between 6 and 12 months

  • Platinum sensitive Relapse after more than 12 months

Platinum-resistant and platinum-refractory disease, which is uniformly fatal [13], carries the worst prognosis compared with platinum-sensitive disease. Median OS decreases over subsequent lines of therapy [14] so that by the fifth relapse, according to Hanker et al. [15] the OS is 5 months (although platinum sensitivity is not specified).

According to NCI Surveillance, Epidemiology and End Results (SEER) statistics, the lifetime risk of developing ovarian cancer is 1.3% [16], which meets requirements for orphan drug designation. Since one quarter (25%) of these patients are platinum resistant [17] (although all patients with recurrent disease eventually develop platinum resistance), the prevalence for platinum-resistant/platinum-refractory ovarian cancer (PRR-ROC) is approximately 50,000 patients, which fits the definition of a rare disease since fewer than 200,000 patients are affected. Current management and treatment options for platinum-resistant and platinum-refractory recurrent ovarian cancer are limited and discussed herein. Accordingly, this review is limited by a paucity of randomized studies in platinum-resistant and platinum-refractory recurrent ovarian cancer.

Treatment options for recurrent resistant or refractory disease

Chemotherapy

While no standard second-line treatment stands out as superior with regard to safety or efficacy, chemotherapy options are non-platinum monotherapy, including paclitaxel, docetaxel, pegylated liposomal doxorubicin (PLD), topotecan and gemcitabine [18]. Overall response rates range from 10 to 35% in phase II studies with relatively short-lived responses of less than 8 months [19]. Combination chemotherapy has been shown to increase toxicity without a clear survival benefit over sequential monotherapy.

Taxanes

In phase II and III trials, single-agent paclitaxel has shown activity as a salvage therapy in recurrent refractory/resistant disease, resulting in objective responses of 22–30% [20]. Low-dose weekly administration of paclitaxel results in equal (PFS), but reduced overall toxicity compared to every-3-week dose administration [21]. Single-agent docetaxel is more toxic than paclitaxel, as grade 4 neutropenia is reported in 75% of docetaxel-treated patients [22].

Pegylated liposomal doxorubicin (PLD)

Polyethylene glycol (PEG)–liposomal doxorubicin is a formulation of the anthracycline doxorubicin encapsulated in PEG-coated liposomes [23]. Because PLD is infused once every 28 days, hematologic and gastrointestinal (GI) toxicity is decreased; cardiomyopathy is also decreased up to a cumulative dose of 440–840 mgs/m2 [24]. In the trial supporting its approval in ovarian cancer, in which PLD was compared with topotecan in platinum-sensitive and platinum-resistant recurrent disease, PLD was associated with an 18% decreased risk of death and lower rates of neutropenia, anemia and thrombocytopenia [25]. However, higher rates of hand–foot syndrome and stomatitis were observed.

Topotecan

The activity of the topoisomerase I inhibitor, topotecan, in platinum-resistant recurrent ovarian cancer is similar to paclitaxel and PLD; however, myelosuppression, particularly neutropenia, limits its use [26]. In a large, multicenter, phase II study, topotecan demonstrated response rates of 5.9% in patients with recurrent resistant disease and 17.8% in patients with recurrent sensitive disease [27]. A phase III randomized trial of topotecan versus paclitaxel in recurrent disease showed similar response rates (20.5% for topotecan and 13.2% for paclitaxel). Median survival time was 61 weeks for topotecan and 43 weeks for paclitaxel (the difference was non-significant; P = 0.515) [28]. No cross-resistance was observed after crossover for progressive disease with response rates of 13.1% for topotecan and 10.2% for paclitaxel [29].

Gemcitabine

In twelve monotherapy studies with 411 mostly platinum-resistant patients, the overall response rates to the nucleoside antimetabolite, gemcitabine ranged from 14 to 22% with a median duration of response from 4 to 10.6 months. Grade 3 or 4 neutropenia was the most common dose-limiting adverse event, which underscores the relative contraindication to treatment with multiagent chemotherapy in the setting of platinum-resistant recurrence, given the increased rates of toxicity [30].

Others

Oral etoposide at a dose of 50 mgs/m2 has the advantage of improved convenience over I.V. administration but is disadvantaged by predominant myelosuppressive effects and a modest response rate of 25% [31]. Single-agent ifosfamide and cyclophosphamide have limited neoplastic efficacy (response rate of 10–15%) with significant renal, CNS or bladder toxicities [32]. The response rate of vinorelbine at a dose of 25–30 mgs/m2 on a 3 weekly schedule is 15–20% with the caveat that it has not been associated with improved quality of life or performance status [33]. A summary of salvage chemotherapy options is shown in Table 1.

Table 1 Salvage chemotherapy options in recurrent platinum-resistant/platinum-refractory platinum patients (adapted from Lorusso et al. [30].)
Full size table

Hormonal therapy

Antiestrogens, aromatase inhibitors and gonadotropin releasing hormone analogs have been evaluated in small randomized trials and while the correlation between hormone receptor positivity and response is inconsistent (even though the presence of estrogen and progesterone receptors would be expected to correlate with response to hormonal therapy) and benefits over chemotherapy unclear, these agents have the advantage of low toxicity and high patient acceptance. In the 1991 GOG study with 105 recurrent ovarian cancer patients including those with platinum-sensitive and platinum-resistant cancer, the complete response rate to second-line oral tamoxifen 20 mg/d was 17% and the disease control rate (DCR) including stable disease (SD) was 55% [34]. Overall response rates are not improved significantly by the addition of tamoxifen to cytotoxic agents [35].

Clinical studies with letrozole have provided conflicting results: one study showed an objective response of 15%; the other showed no objective responses [36].

Because hormonal therapy may have some activity in a subset of endocrine-sensitive patients, and is generally a far less toxic treatment than chemotherapy, it continues to be administered in the resistant/refractory setting, despite inconsistent and unclear efficacy data.

Targeted therapies

Bevacizumab

Bevacizumab, the humanized monoclonal antibody against the vascular endothelial growth factor (VEGF), is the most extensively studied targeted agent in EOC since angiogenesis is an important hallmark of the disease. The ICON-7 and GOG-218 phase III trials established bevacizumab as a first-line treatment in platinum-sensitive ovarian cancer [37, 38]. Bevacizumab-related toxicities included hypertension >grade 2 and bowel perforation. Several phase II trials, which have evaluated bevacizumab specifically in platinum-resistant or refractory patients, demonstrated high rates of disease stabilization, an important endpoint in this population, where prolonged stable disease promotes improved quality of life and management as a chronic illness [39].

The FDA-approved bevacizumab at a dose of 10 mg/kg every 2 weeks or 15 mgs/kg every 3 weeks in combination with topotecan (days 1, 8, 15 every 4 weeks), pegylated liposomal doxorubicin (every 4 weeks) or paclitaxel (days 1, 8, 15, 22 every 4 weeks) on the basis of results from the phase III AURELIA trial, which randomized 361 platinum-resistant and platinum-refractory recurrent patients to bevacizumab combination therapy or these three single chemotherapy agents (i.e., topotecan, PLD and paclitaxel) alone. The combination with bevacizumab resulted in an ORR (RECIST) of 27.3% compared to 11.8% for chemotherapy alone (P = 0.001), and a median PFS of 6.7 months versus 3.4 months (HR 0.48, 95% CI 0.38–0.60, P < 0.001) without a significant OS improvement (combination with bevacizumab median OS of 16.6 months versus 13.3 months for chemotherapy alone, HR 0.85, 95% CI 0.66–1.08) [40]. Based on an exploratory analysis, maximum benefit was seen with paclitaxel.

PARP inhibitors

Through the concept of tumor-specific synthetic lethality, whereby the potency of PARP inhibition increases in the presence of a preexisting BRCA1/2 mutation, PARP inhibitors have demonstrated single-agent activity in BRCA-mutated ovarian cancer; in a multicenter phase II trial, which enrolled 193 patients with platinum-resistant ovarian cancer, the response rate was 31.1% (95% CI, 24.6–38.1%) and the rate of stable disease at 8 weeks or greater was 40% (95% CI, 33.4–47.7%). Median PFS and overall survival (OS) were 7.0 months and 16.6 months, respectively. Partially on this basis, the FDA-approved olaparib (400 mg twice daily) as monotherapy for patients with germline BRCA-mutated advanced ovarian cancer treated with 3 or more prior lines of chemotherapy [41]. In addition, rucaparib (Rubraca) was approved in 2016 for germline or somatic BRAC mutations treated with two or more chemotherapies. Accordingly, the race is on with several other PARP inhibitors, including niraparib, talazoparib and veliparib all currently in late phase clinical trial development [42].

Epigenetic inhibitors

The best characterized epigenetic therapies, which target proteins or pathways that tend to silence gene expression, are the DNMT inhibitors such as the FDA-approved azacitidine and decitabine and the HDAC inhibitors such as the FDA-approved vorinostat [also known as suberoylanilide hydroxamic acid (SAHA)], romidepsin, panobinostat and belinostat. Epigenetic therapies offer intriguing potential in resistant/refractory ovarian cancer through the amelioration of epigenetic repression of the genes that are responsible for chemoresistance, resulting in reversal of the resistant phenotype and sensitization/resensitization to first-line platinum doublets. A phase Ib–IIa study of decitabine and carboplatin in 29 platinum-refractory or platinum-resistant ovarian cancer patients demonstrated a disease control rate of 34% with one complete response, three partial responses, and ten stable diseases. However, these epigenetic therapies may maximally resensitize to platinum doublets when they are administered prior to rather than concomitantly with chemotherapy on the premise that it takes time (weeks or even months) to activate silenced genes and reverse chemoresistance [43]. However, this application will require more favorable safety profiles for both HDACs and DNMT inhibitors, as they are associated with toxicities, likely owing to their relative non-specificity [44].

Others

Several targeted therapies including pertuzumab, cediranib, gefitinib and sunitinib are under clinical investigation; however, the “genomic chaos” (i.e., the large number of chromosomal alterations), which characterizes ovarian tumors [45] and which worsens over time due to therapy selective pressures, raises doubts that targeted agents directed at only a few mutations will significantly alter outcomes, especially in resistant or refractory disease. To date, none of these agents have unequivocally demonstrated efficacy.

Immunotherapies

The rationale for the use of immunotherapy in ovarian cancer is that the presence of T cells in the tumor microenvironment correlates with improved progression-free and overall survival, while the presence of regulatory T cells and T cell inhibitory molecules is associated with poorer outcomes [46]. Vaccines, oncolytic viruses, immune checkpoint blockade and adoptive T cell therapy have been associated with clinical activity in small numbers of patients. While these strategies provide a platform for future progress, only FDA-approved checkpoint inhibitors are briefly covered below since preliminary analysis suggests that responses from them may be higher in patients with fewer prior lines of chemotherapy and platinum-sensitive disease [47].

Despite the success of “antibody-based” therapy in ovarian cancer with bevacizumab, the clinical activity reported with checkpoint inhibitors has been modest at best. In a phase II study of the anti-PD-1 antibody, nivolumab, the overall response rate was 15% (3 of 20 patients) with a median PFS of 3.5 months [48]. In another trial with nivolumab 2/8 patients (25%) with resistant ovarian cancer experienced a complete response [49]. Similar response rates (3 of 20 patients; 15%) were reported in a phase IB study of patients with PD-L1-positive ovarian cancer treated with the anti-PD-1 inhibitor, pembrolizumab [50]. In a phase 1b study of 75 patients with platinum-resistant or chemotherapy-refractory ovarian cancer regardless of PD-L1 expression treated with the PD-L1 antagonist, avelumab, the overall response rate was 10.7%, and the DCR was 54.7%, with no complete responses, 8 partial responses and 33 with stable disease [51].

Conclusion

As an invariably fatal disease, with complex mechanisms of resistance that are incompletely understood [52], the current primary intent of treatment for platinum-resistant/platinum-refractory ovarian cancer is to preserve quality of life and palliate symptoms with sequential single-agent chemotherapy. In contrast to the substantial number of innovative agents approved to treat a variety of solid tumors and hematologic malignancies over the past decade, advances in ovarian cancer, particularly platinum-resistant/platinum-refractory ovarian cancer, have been few and far between with a relative paucity of effective new drugs or strategies in the current pipeline.

Molecularly targeted therapies are an attractive alternative or adjunct to traditional chemotherapy. Bevacizumab and olaparib are currently FDA-approved and rucaparib was granted accelerated approval for advanced BRCA-mutated ovarian cancer treated with ≥2 chemotherapies [53]. In addition, multiple promising trials are underway (or will soon be underway), for other PARP inhibitors alone or in combination, which have the potential to improve treatment outcomes. Other potentially promising strategies include: the use of epigenetic therapies to restore/reverse platinum sensitivity and therefore restart end-stage patients on first-line treatment and immunotherapies to incite antitumor T-cell responses.

However, optimism remains tempered by the formidable therapeutic challenge of this recalcitrant disease subtype with poorly understood pharmacological mechanisms of resistance and a track record of failure with previous clinical candidates. Ultimately, vanquishing recurrent EOC will require a concerted effort to tackle platinum resistance with clinical trials featuring innovative strategies and therapies.