Surgery is the cornerstone of EOC, and it is considered curative in most of the cases of women with early-stage tumors. Early-stage ovarian cancer includes FIGO stages Ia, Ib, and Ic [64]. An adequate cytoreductive surgery should be offered both to remove the disease and to provide accurate staging, which is a key factor also for the decision to undergo an adjuvant therapy. The prognosis of early-stage ovarian cancer is good, with a 5-year survival rate of 70–90 % [1]. Although surgery is curative in many cases, about 30 % of the patients recur and require further therapy.

The rationale of adjuvant chemotherapy after complete removal of the disease and adequate surgical staging is to eradicate any residual microscopic deposits of cancer cells responsible of potential recurrence of disease. In order to make an appropriate treatment choice, it is, therefore, important to understand what factors in early stages of ovarian cancer influence the prognosis and the risk of recurrence and, therefore, which patients would benefit most from additional treatment. Uncontrolled retrospective studies identified prognostic factors in this setting. A multivariate analysis of 1545 patients with stage I EOC confirmed tumor grade to be the single most important determinant of survival [78]. Other important independent prognostic factors are age, FIGO stage, substage (capsular involvement or cyst rupture, FIGO stage Ic), histological subtype (e.g., worse prognosis in the undifferentiated tumors), and the presence of ascites. According to literature data, early stages of epithelial ovarian cancer are divided into three different risk categories. Low-risk patients have tumors at FIGO stages Ia and Ib with well-differentiated disease and with no clear-cell histology; in such conditions, surgery is resolutive in 95 % of cases in the absence of evidences that demonstrate a benefit of a subsequent adjuvant chemotherapy [39]. The patients with FIGO stage Ia–Ib moderately differentiated belong to intermediate risk. Poorly differentiated tumors or stages Ic–II are considered high risk as associated with a recurrence rate of 25–40 % and, therefore, are candidates for adjuvant chemotherapy [75]. The patients with intraoperative rupture of the tumor within the abdomen fall into this group. A number of evidences support the use of adjuvant chemotherapy in early-stage ovarian cancer,however the optimal regimen, and duration is still debated [16, 29, 70]. A Cochrane meta-analysis assessed the survival advantage of adjuvant chemotherapy in early-stage ovarian cancer [79]. Five randomized clinical trials were included in the analysis. While the earliest three trials lacked enough events to demonstrate a possible impact of adjuvant treatment, on the contrary, the ICON1 and the ACTION trial [75] included a larger number of patients and had a sufficient power to demonstrate a treatment effect. These two trials randomized patients with serous, mucinous, endometrioid, clear-cell, and undifferentiated ovarian carcinomas to receive platinum-based chemotherapy or no chemotherapy. The 5-year and the updated 10-year overall survival rate was significantly better for women receiving adjuvant chemotherapy compared with the control group. Similar findings were reported for progression-free survival [79]. Additionally, these two trials address the question of which patients with early-stage ovarian cancer benefit more from adjuvant chemotherapy. The ICON1 trial stratified patients in low/intermediate risk (FIGO stage Ia, G1–G2, or FIGO stages Ib–Ic, G1) and high risk (FIGO stage Ia, G3, and Ib/Ic, G2–G3, any clear cell). Both overall and progression-free survival were better in high-risk patients, but no difference was observed among treated and non-treated low- to intermediate-risk patients at 5- and 10-year follow-up [72]. Conversely, the ACTION trial strongly recommended a complete surgical staging and planned a subgroup analysis on suboptimally and optimally staged patients. Patients with FIGO stage Ia/Ib, G2–G3, and FIGO stage Ic/IIa were included. Among the suboptimally staged women, adjuvant chemotherapy increased the overall (OS) and disease-free survival (DFS), whereas in the optimally staged patients, no difference was observed between the treated group and the control group. At a median follow-up of 10 years, earlier data were confirmed [79]. These results may suggest that there is a subgroup of good-prognosis patients who apparently do not benefit from adjuvant chemotherapy. In particular, it is hypothesized that chemotherapy impacts only on occult disease in suboptimally staged patients. Nonetheless, a benefit in optimally staged tumor cannot be excluded. In summary, adjuvant chemotherapy may be avoided only for low-risk, optimally staged, stage I patients (FIGO stage Ia/Ib, G1–G2); chemotherapy is indicated after surgery for patients with high-risk stage I disease (FIGO stage Ic, G3). In case of suboptimal surgical staging of low-risk stage I patients, benefits and effects of adjuvant chemotherapy should be discussed with each individual patient [1, 53, 54]. Regarding the chemotherapy regimen in this setting, studies comparing carboplatin plus paclitaxel versus carboplatin alone are not available. Indirect suggestions of greater efficacy of the carboplatin–taxane combination are gathered by results in advanced-stage disease. In the absence of clear recommendations, single-agent carboplatin can be considered a reasonable alternative to the doublet in intermediate- and high-risk early-stage ovarian cancer patients [1, 39, 53, 54]. The optimal duration of adjuvant chemotherapy remains a matter of investigation. A randomized trial which compared three versus six cycles of platinum plus paclitaxel for early-stage ovarian cancer revealed a nonstatistically significant 24 % reduction in recurrence rate in patients who underwent six courses of chemotherapy [6]. A subgroup analysis stratifying patients on the basis of clinical and pathological features showed a statistically significant benefit of six versus three cycles of chemotherapy only in serous tumors, while outcome for non-serous was not influenced by the duration of chemotherapy [14]. Again, there may be a subgroup of patients who do not benefit from intensive adjuvant chemotherapy and future research is needed to confirm these hypotheses. Recently the Cochrane review has been updated including mature data (10-year follow-up) of the same trials [36]. The conclusion is that adjuvant platinum-based chemotherapy is effective in prolonging survival in women with early-stage (FIGO stage I/IIa) epithelial ovarian cancer. It remains uncertain whether women with low- and intermediate-risk early-stage disease will benefit as much from adjuvant chemotherapy as women with high-risk disease. Decisions to use adjuvant chemotherapy in these women should be individualized to take into account individual factors.

Also, the choice to prescribe adjuvant chemotherapy in the early stages of the type I tumors including the less common clear-cell, mucinous, and low-grade serous carcinoma [34] is object of debate. Comparing the outcome for different histologic subtypes in large randomized trials evaluating paclitaxel/carboplatin in advanced ovarian cancer, it seems that clear-cell and mucinous carcinomas are more chemoresistant than serous carcinomas [25, 57]. Adjuvant chemotherapy is usually not given for stage I mucinous tumors, while this topic is still debated for clear-cell cancers. On the contrary, stage Ic tumors, that have poorer prognosis, are usually treated as for the other hystotypes.

Both the American and European guidelines recommend surgery as the initial approach to ovarian malignancies [1, 39, 53, 54]. The actual treatment of ovarian cancer at an advanced stage (FIGO stages III–IV) is based on the proper integration of surgery and medical approach. The goal of primary surgery, defined as optimal cytoreduction, is the absence of residual cancer. For more than 20 years, tumor debulking surgery followed by platinum/taxane combination chemotherapy has remained the standard first-line treatment. Currently, the first-line treatment of ovarian cancer consists of the combination of carboplatin AUC 5.0–7.5 and paclitaxel 175 mg/m² with three-weekly schedule. This choice comes from the results of several trials that demonstrated the superiority of paclitaxel-based chemotherapy regimens and the equal effectiveness of carboplatin compared to those containing cisplatin (GOG111, GOG 114, GOG 158, AGO-OV.2, AGO-OV.104) [47, 52, 55, 60]. The shift from cisplatin to carboplatin was based on safety considerations and easier handling use. As expected, paclitaxel/carboplatin was associated with significantly different toxicity profiles consisting in less nausea, vomiting, and neuropathy, but greater myelosuppression that was overall manageable. Furthermore, the analysis of quality of life, assessed by the EORTC QLQ-C30 questionnaire, demonstrated significantly better scores for the paclitaxel/carboplatin arm when compared with the cisplatin-based treatment. Given the evidence of a more favorable toxicity profile and ease of delivery, the carboplatin/paclitaxel combination has been considered for years the standard of care in epithelial ovarian cancer treatment. However, although the initial response rates are high, most patients (75–80 %) with advanced-stage ovarian cancer relapse within 18 months and eventually die from the disease within a median of 32–57 months.

New strategies have been studied to improve the outcome of this first-line therapy. Moreover, a safety profile of carboplatin and paclitaxel combination, mainly alopecia and neurotoxicity, has prompted researchers to look for new doublets potentially better tolerated and equally or more effective. Several efforts have focused on further intensifying the chemotherapy regimens with the addition of a third or fourth cytotoxic agent; however, attempts in this regard have consistently been disappointing, producing suboptimal responses with no survival advantage and higher toxicities [9, 28]. The SCOTROC1 trial [77] was a Scottish randomized phase III trial, which included more than 1000 women with stage Ic–IV EOC; it compared the carboplatin/docetaxel doublet to the combination of carboplatin and paclitaxel. No efficacy differences were founded between both arms at final analysis. The biggest difference that emerged was related to the toxicity profile of the two treatment regimens. A higher rate of myelosuppression (including complicated grade 3–4 neutropenia) was found with carboplatin/docetaxel, whereas a higher rate of neurotoxicity has been described in patients treated with carboplatin/paclitaxel. According to these data, docetaxel and carboplatin could be considered as a valid alternative to carboplatin/paclitaxel in some selected cases. In the MITO-2 trial [61], carboplatin AUC 5 in combination with pegylated liposomal doxorubicin (PLD) at a dose of 30 mg/m² every 3 weeks for six cycles was compared to the standard carboplatin/paclitaxel combination in 820 patients with FIGO stage Ic–IV EOC. Once again the alternative doublet (carboplatin/PLD) failed to show a greater efficacy compared to standard-of-care chemotherapy in terms of PFS, OS, and response rate (PFS, 19 vs. 16.8 months; p = 0.58; median OS, 61.6 vs. 53.2 months; p = 0.32; response rate, 57 vs. 59 %; p = 0.76). Regarding toxicity profile, thrombocytopenia, anemia, stomatitis, and skin toxicity were significantly more frequent in the PLD arm, whereas neuropathy, diarrhea, and hair loss were most frequent in those treated with paclitaxel. These results let us to consider carboplatin/PLD as a reasonable alternative to carboplatin/paclitaxel, particularly in patients who experienced hypersensitivity to paclitaxel, asked to avoid alopecia, or are at risk of peripheral neuropathy.

Dose-dense therapy has been investigated as another alternative to improve first-line efficacy treatment. Literature data show that weekly administration of antineoplastic agents may have potentially enhanced antitumor activity and decreased toxicity due to extended exposure but fairly low concentration of the drugs [43].

Studies in breast cancer demonstrated a greater activity of paclitaxel administered at low doses every week compared to standard three-weekly schedule. This difference was supposed to be also related to an antiangiogenic effect associated to dose-dense administration. These results prompted to the JGOG 3016 trial [33]. This multicentric Japanese phase III study randomized more than 600 women with FIGO stage II–IV EOC to receive weekly paclitaxel (80 mg/m², days 1, 8, and 15) in combination with three-weekly carboplatin (AUC6, day 1) or a standard three-weekly regimen into first-line ovarian cancer treatment. The results showed an impressive superiority of the dose-dense paclitaxel and carboplatin regimen in terms of PFS (28.2 vs. 17.5 months; HR, 0.76; 95 % CI, 0.62–0.91; p = 0.0037) and OS (100.5 vs. 62.2 months; HR, 0.79; 95 % CI, 0.63–0.99; p = 0.039), but with increased toxicity leading to more frequent early discontinuation of treatment in the dose-dense arm (53 vs. 37 %). The role of weekly chemotherapy was further evaluated in a European study, the Multicenter Italian Trials in Ovarian Cancer (MITO-7) randomized trial [62]. In this study, over 800 women with stage Ic to IV EOC were treated with a total of six cycles using carboplatin and paclitaxel on either a standard (every 3 weeks) or on a weekly schedule, with both agents administered on days 1, 8, and 15 every 21 days. At a median follow-up of 22 months, similar PFS (18 versus 17 months; HR, 0.96; 95 % CI, 0.80–1.16) was found in the experimental arm compared with standard treatment, with no statistically overall survival differences between two arms (77 versus 79 % of probability of survival at 24 months, respectively, of HR 1.20 and 95 % CI 0.90–1.61), but a better tolerability profile and quality of life founded in the weekly arm. The MITO-7 trial differs from JGOG 3016 in several aspects. In contrast to the Japanese one, the European trial the total dose of the drugs is similar between the arms, but both carboplatin and paclitaxel are administrated weekly. The most interesting advice that at least emerged from these trials concerns the race of the enrolled population (Asian versus Caucasian), which may explain the discrepancy of results probably related to distinct profiles of response and tolerability, often imputed to genetic polymorphisms involved in drug metabolism and chemosensitivity.

Another smaller European trial did not find a benefit from a regimen including induction therapy with three cycles of dose-dense weekly paclitaxel and weekly platinum over a standard three-weekly regimen [76]. The JGOG 3016 dose-dense regimen was also evaluated in the USA through the GOG 262 phase III trial, where a similar schedule was analyzed but with the optional addition of bevacizumab (at physician choice) [15]. The use of bevacizumab in most patients enrolled (85 %) and in both arms prevents a strict comparison between GOG 262 and JGOG 3016, except in the small subset of patients (15 %) where bevacizumab was not added to chemotherapy. No significant difference in efficacy was found between both arms, with a difference in favor of the dose-dense weekly regimen observed only in patients who did not receive bevacizumab. The limited series of patients in this trial precludes any definitive conclusion. There is currently no data suggesting that the weekly dose-dense regimen is beneficial to non-Japanese women. Thus, the three-weekly carboplatin/paclitaxel regimen remains the standard in Caucasian populations, although we await further data; the ICON8 (https://​clinicaltrials.​gov/​ct2/​show/​NCT01654146), which is ongoing, is a phase III trial, comparing three possible schedules of paclitaxel/carboplatin combination (three-weekly carboplatin and paclitaxel, three-weekly carboplatin plus weekly paclitaxel, weekly carboplatin and paclitaxel), and might help to definitively clarify the role of dose-dense chemotherapy.

As the majority of clinical recurrences are generally confined to the peritoneal cavity, there is a strong rationale for administering cytotoxic drugs directly into the abdomen (intraperitoneal chemotherapy), thus increasing the dose intensity delivered to any residual tumor while avoiding additional systemic toxicity. Intraperitoneal chemotherapy (IP) is unable to penetrate deeply into tissues so it is only likely to be suitable for patients who have undergone optimal cytoreductive surgery with minimal residual disease.

There have been three large phase III trials comparing intraperitoneal (IP) chemotherapy with IV chemotherapy. In the first study, 654 patients were randomly assigned to receive six cycles of IV cyclophosphamide in combination with either IV or IP cisplatin. The median OS was significantly longer in the IP arm (49 versus 41 months; p = 0.02; HR, 0.76) [2]. This questionable trial was carried out before the introduction of taxanes; however, it formed the basis for further trials of IP chemotherapy. The second study, GOG 114, incorporated IV paclitaxel in each arm. Patients were randomly assigned to receive either IV paclitaxel 135 mg/m² over 24 h followed by IV cisplatin 75 mg/m² every 3 weeks for six courses or IV carboplatin (AUC 9) every 28 days for two courses, then IV paclitaxel 135 mg/m² over 24 h followed by intraperitoneal cisplatin 100 mg/m² every 3 weeks for six courses. In the 462 assessable patients, a substantial improvement in PFS was observed but the difference in OS was not significant (p = 0.05). Again toxic effects were higher in the experimental arm and 18 % of patients received <2 courses of IP treatment [48]. The use of IP chemotherapy as a viable alternative to intravenous (IV) treatment is supported by European and American guidelines on the basis of a large phase III randomized trial (the GOG172) that randomized 415 among the 429 women with stage III EOC treated with optimal debulking surgery (residual disease ≤1 cm) [3] to receive IV paclitaxel and cisplatin or IV paclitaxel followed by IP cisplatin (d 1) and paclitaxel (d 8). In the experimental arm, the patients received IP cisplatin (100 mg/m²) on day 1 and IP paclitaxel (60 mg/m²) on day 8; in the control arm, both cisplatin (75 mg/m²) and paclitaxel (135 mg/m²) were administered intravenously on days 1 and 2, respectively, every 3 weeks. The experimental arm showed a significant benefit compared with the control arm in PFS (23.8 versus 18.3 months, p = 0.05). A significant improvement in OS was also demonstrated, 65.6 vs. 49.7 months, respectively, in the IP arm compared to IV arm (p = 0.03). It is worth to be mentioned that only 42 % of patients was able to complete six cycles of IP chemotherapy. Grade 3–4 toxicity was greater and quality-of-life scores were significantly worse in the IP arm. The three trials described above all recorded favorable PFS or OS results for the IP arms. Data from the GOG 114 and 172 trials were recently retrospectively analyzed by Tewari and colleagues [74] to determine 10-year long-term survival and prognostic factors linked to IP therapy. After a median follow-up of 10.7 years, the survival benefit for IP chemotherapy was confirmed, with a median survival with IP therapy of 61.8 months (95 % CI, 55.5–69.5), compared with 51.4 months (95 % CI, 46.0–58.2) for intravenous therapy. Factors associated with poorer survival included: clear, mucinous versus serous histology (AHR, 2.79; 95 % CI, 1.83–4.24; p < .001), gross residual versus no visible disease (AHR, 1.89; 95 % CI, 1.48–2.43; p < .001), and fewer versus more cycles of IP chemotherapy (AHR, 0.88; 95 % CI, 0.83–0.94; p < .001). In this setting, also, an interesting field of study is related to tumor genomic alterations. Recently, an analysis of somatic loss of BRCA1 in patients participating in GOG-172 demonstrated a profound effect on overall survival [40]. In this analysis of 393 patients (94 % of GOG-172 participants), somatic loss of BRCA1 was observed in 48 %, and the median overall survival for IP vs. IV therapy was 84 vs. 47 months (p = .0002), amounting to a 33 % reduction in the hazard for death.

Despite such encouraging and undoubted findings on the impact of overall survival, IP chemotherapy still raises many concerns among oncologists and has been adopted only in a few countries [10]. The additional toxicity (grade 3–4 pain, fatigue, and hematologic toxicities were significantly more common in the IP cohort), lack of familiarity with placement and use of peritoneal catheters, the worse quality-of-life scores observed in the experimental arms, and concern over the design of trials have hindered its widespread use. In the last months conflicting results from two important phase II [59] and phase III trials (PERTROC-OV21 and the GOG 252 trial) [26] on adjuvant IP treatment in OC patients have been presented at the 2016 Annual Meeting of the Society of Gynecologic Oncology and at the 2016 ASCO Annual Meeting. To further elucidate the role of IP in the treatment of ovarian cancer and better define which patients will truly benefit from this approach, other studies [31] are ongoing, the researchers are also conducting correlative studies on collected tissue samples to determine whether certain biologic characteristics may be associated with improved outcomes using IP versus IV chemotherapy.