Ann Surg Oncol DOI 10.1245/s10434-015-4644-7

ORIGINAL ARTICLE – GASTROINTESTINAL ONCOLOGY

Intraperitoneal Vascular Endothelial Growth Factor: A Prognostic Factor and the Potential for Intraperitoneal Bevacizumab Use in Peritoneal Surface Malignancies Claramae Shulyn Chia, MBBS, MMed, FRCS1, Olivier Glehen, MD, PhD2,4, Naoual Bakrin, MD3,4, Evelyne Decullier, PhD5,6, Benoit You, MD, PhD4,7, Franc¸ois Noe¨l Gilly, MD, PhD2,4, and Guillaume Passot, MD2,4 Department of Surgical Oncology, National Cancer Centre Singapore, Singapore, Singapore; 2Department of Oncologic and General Surgery, Hospices Civils de Lyon, Centre Hospitalier Lyon Sud Pierre Be´nite, Lyon, France; 3Department of Gynaecology, Hospices Civils de Lyon, Centre Hospitalier Lyon Sud Pierre Be´nite, Lyon, France; 4UMR 37-38, Universite´ Lyon 1, Lyon, France; 5Hospices Civils de Lyon, Pole IMER, Lyon, France; 6Health Information and Clinical Research, Lyon, France; 7Department of Oncology, Hospices Civils de Lyon, Centre Hospitalier Lyon Sud Pierre Be´nite, Lyon, France 1

ABSTRACT Introduction. Intraperitoneal (IP) vascular endothelial growth factor (VEGF) levels have been shown to vary in the peritoneal cavity of patients with peritoneal surface malignancies. Our purpose was to correlate levels of IP VEGF with overall and disease-free survival to identify whether IP VEGF can be used to prognosticate patients and the possible role of IP bevacizumab. Methods. From February to October 2012, 97 consecutive patients with peritoneal carcinomatosis were treated with cytoreductive surgery and hyperthermic intraperitoneal chemotherapy. Intravenous (IV) VEGF levels were taken before surgery, whereas IP VEGF levels were taken at various time points during and after surgery. Results. Median follow-up was 19.48 months. On univariate analysis, a lower IP VEGF taken just after incision (T1) was associated with improved overall (P = 0.0004) and disease-free survival (P = 0.0006) at 2 years. A lower T1/IV VEGF ratio also was associated with improved overall (P = 0.004) and disease-free survival (P = 0.0051). On multivariate analysis, a lower T1 was associated with improved overall survival, whereas a lower T1/IV VEGF was associated with improved disease-free

Ó Society of Surgical Oncology 2015 First Received: 23 November 2014 C. S. Chia, MBBS, MMed, FRCS e-mail: [email protected]

survival. On subset analysis, these two variables were associated with improved survival in colorectal cancers. Conclusions. A lower IP VEGF level prior to surgery is associated with improved survival. The use of preoperative intraperitoneal bevacizumab for patients with a heavy disease load should be considered, especially in colorectal cancers.

Patients with peritoneal surface malignancies (PSM) were once considered terminal and treated palliatively. With the advent of cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC), these patients now have a means of cure. Multiple studies have shown that the morbidity and mortality of performing CRS and HIPEC are comparable to other major abdominal surgeries, as the improvement in surgical technique and anaesthetic ability continues.1 Survival rates also have been shown to be better than with intravenous (IV) chemotherapy.2 However, there is no consensus as to the best way to treat these patients in terms of neoadjuvant and adjuvant therapy, HIPEC agent and duration, as well as the use of preoperative, early postoperative and postoperative intraperitoneal (IP) chemotherapy. Vascular endothelial growth factor (VEGF) is a signal protein produced by cells that stimulates vasculogenesis and angiogenesis. Its role in the development of solid tumors has been documented, including breast, colorectal, pancreatic, esophageal, and ovarian.3 Bevacizumab is a recombinant humanized monoclonal antibody that blocks the activity of VEGF. It is used intravenously with multiple

C. S. Chia et al.

solid organ tumors to improve overall survival and has been used intraperitoneally as a palliative treatment for malignant ascites.4,5 However, few studies have investigated the use of bevacizumab intraperitoneally as a curative agent. In a previous study, we investigated the evolution of IP VEGF during CRS and HIPEC.6 We showed that IP VEGF increased after cytoreductive surgery but subsequently decreased after HIPEC and up to 24 h after surgery. Serum levels of VEGF have been shown to correlate with extent of disease in ovarian cancer, and the same may be true for intraperitoneal levels, although this has not been shown in studies.7 If it holds true, the increase after cytoreductive surgery is likely due to manipulation of the tumor, while the decrease after HIPEC is due to eradication of the microscopic disease by the HIPEC. Neoadjuvant IV bevacizumab was shown to significantly reduce the IP VEGF burden in our previous study. To further investigate the potential of intraperitoneal bevacizumab, we evaluated the value of IP VEGF at different time points with relation to overall and disease-free survival. METHODS Between February and October 2012, all patients who were planned for curative CRS and HIPEC for PSM in a single institution (Centre Hospitalier Lyon Sud) were included. Patients had complete preoperative workups, were medically suitable, and were discussed at a multidisciplinary tumor board. Primary tumors from various origins were enrolled. None of the patients received preoperative or postoperative intraperitoneal treatment. Patients underwent a laparotomy, and the extent of disease was assessed according to the Peritoneal Cancer Index (PCI). The goal was complete resection of disease; this was scored using the completeness of cytoreduction (CC) score. HIPEC was only performed if cytoreduction was complete (no visible disease (CC0) or residual nodules \2.5 mm (CC1)). HIPEC was performed with the closed abdomen technique. The cytotoxic agent varied according to the primary tumor and the preoperative decisions in the multidisciplinary meeting. The agents used were oxaliplatin, doxorubicin, mitomycin, and cisplatin, alone or in combination. Blood Draws and VEGF Analysis IV samples were drawn just before incision (T0). IP draws were performed at five time points: T1 was during abdominal cavity exploration, T2 after completion of CRS, T3 after completion of HIPEC, T4 was taken 1 h after abdominal closure, and T5 was taken 24 h after abdominal closure.

Because no prior studies have looked at intraperitoneal VEGF in association with CRS and HIPEC, we used a standardized method that we were familiar with that our unit had previously used to extract fluid when studying the levels of intraperitoneal chemotherapy concentration in relation to intravenous levels.8,9 If no peritoneal fluid was present at exploration to obtain T1, we performed peritoneal washings with 50 ml of saline into the pouch of Douglas. The pouch of Douglas was chosen as we know the spread of peritoneal disease works by gravity and is most likely to be found in the dependent areas. T2 and T3 samples were obtained with peritoneal washings in the pouch of Douglas, in a standardized manner with T1. Postoperative peritoneal fluid was drawn from the remaining drains. For patients with unresectable disease, only T1 and T2 were drawn. Plasma and peritoneal VEGF-A concentrations were determined using the human VEGF Quantikine ELISA kit (R&D Systems). Follow-up Patients were followed up at the outpatient clinics every 3 months for the first 2 years and 6 monthly thereafter with imaging and tumor markers. Events of mortality and recurrent disease were recorded. Statistical Analysis VEGF variables that were used include T0, T1, T2, T3, T4, T5, and the ratios of T1/T0 and T2/T1. Clinicopathological variables that were analyzed include gender, age, tumor markers (CEA, CA 125, CA 19-9), use of chemotherapy (neoadjuvant and adjuvant), number of chemotherapy cycles, use of radiotherapy, use of bevacizumab and number of cycles, performance status, PCI score, CC score, presence of ascites, duration of surgery, extent of surgery, use of HIPEC, intraperitoneal chemotherapy agent, development of complications, grade of complication, primary tumor origin, involvement of lymph nodes, differentiation of the tumor, presence of vascular emboli, and perineural invasion. The extent of surgery was considered as major if it lasted longer than 7 h, or if one colonic anastomosis was performed, or if more than 4 peritonectomies were done.10 The Clavien–Dindo classification was used to grade complications. Grades 1 and 2 were considered minor complications, whereas grades 3 and 4 were major complications. Overall survival curve and disease-free survival curve at 2 years were plotted and estimators calculated using the Kaplan–Meier method. Influence of baseline risk factors was assessed using univariate and multivariate Cox proportional hazards model. Factors with a significance degree B0.10 and with \20 % of missing data were introduced in multivariate

VEGF: Prognostic Factor in Peritoneal Malignancies

Ninety-seven consecutive patients with PSM were included. The median age was 59 (range 34–70) years. There were 66 women (68 %) and 31 men (32 %). Forty patients had colorectal tumor primaries (41 %), 23 had ovarian primaries (24 %), 12 had gastric primaries (12 %), 12 had pseudomxyoma peritonei or disseminated peritoneal adenomucinosis (DPAM) (12 %), and 2 had appendicael carcinomas. There were four peritoneal mesotheliomas, two primary peritoneal carcinomas, one urachal carcinoma, and one breast cancer. Seventy-nine patients received neoadjuvant systemic chemotherapy, 25 of whom had intravenous bevacizumab as part of their regimen. Of these 25, 24 had colorectal primaries and 1 had an appendiceal primary. The median PCI was 8 (range 0–39), and the median CRS time was 270 (range 135–600) min, excluding the HIPEC time. The CRS was complete (CC 0 or 1) for 80 patients, whereas 17 patients were considered unresectable. HIPEC was given to 64 patients: 24 patients received oxaliplatin, 23 had cisplatin, 7 had cisplatin and mitomycin, 6 had cisplatin and doxorubicin, and 4 had mitomycin. Median follow-up was 19.48 (range 0–37.8) months. At the time of last follow-up, 30 patients (31 %) were deceased and 67 were still alive. Median time to death for the 30 deceased patients was 12.5 (range 0–20.8) months. Of 97 patients, 63 patients had recurrences (65 %). Median time to recurrence was 8.8 (range 0–23.8) months. Figure 1 shows the Kaplan–Meier curves for overall survival and disease-free survival over 2 years. IP VEGF was drawn in patients at T1 [n = 84 (87 %), median 96.5 pg/ml (range 7–7400)], T2 [n = 73 (75 %), median 556 pg/ml (range 23–6420)], T3 [n = 61 (63 %), median 206 pg/ml (range 16–932)], T4 [n = 39 (40 %), median 250 pg/ml (range 27–1704)], and T5 [n = 31 (32 %), median 164 pg/ml (range 20–1579)]. Patients were divided into two groups based on the median T1. Figure 2a shows the Kaplan–Meier curve for overall survival according to this median. Although the curves separate, this difference was not statistically significant (p = 0.3308). We also analyzed the ratio of postCRS IP VEGF to preoperative IP VEGF (T2/T1). The median T2/T1 was 2.53, and the patients were divided into two groups based on this median. Figure 2b shows the

1.00

Overall survival probability

RESULTS

(a)

0.75

0.50

0.25

0.00

(b)

Disease free survival probability

models. P values \0.10 were retained in the final model using a stepwise procedure. After variables selection, results were adjusted with PCI (\15 or C15) and neoadjuvant chemotherapy (yes or no). All statistical analysis was performed with SASÒ 9.2 (SAS Institute Inc., Cary, NC).

0

3

6

0

3

6

9

12

15

18

21

24

1.00

0.75

0.50

0.25

0.00 9

12

15

18

21

24

FIG. 1 Kaplan–Meier curves for overall survival (a) and disease-free survival (b)

Kaplan–Meier curve for overall survival according to the two groups. Again, this difference was not statistically significant (p = 0.3743). Using the median of 55 for IV VEGF to divide the patients into two groups, there was no significant difference in overall survival at 2 years between the two groups. On univariate analysis of overall survival at 2 years, only the following clinical variables were significant at 10 %: lower age at diagnosis, CA 19-9 level C18, PCI score C15, CC score of 2, duration of surgery C270 min, use of chemotherapy, absence of HIPEC, and the presence of vascular emboli (Table 1). As for the VEGF variables, lower T1 and a lower ratio of T1/T0 were associated with improved

C. S. Chia et al.

(b) 1.00

1.00

0.75

0.75

Survival probability

Survival probability

(a)

0.50

0.25

0.50

0.25

logrank test : p = 0.3308

logrank test : p = 0.3743 96.5

18

2.53

18

24

Months after intervention

Months after intervention Alive at 1 year (%)

Alive at 2 years (%)

≤ 96.5

92.74

74.55

> 96.5

85.42

60.96

Alive at 1 year (%)

Alive at 2 years (%)

≤ 2.53

94.12

68.66

> 2.53

93.94

78.86

FIG. 2 Kaplan–Meier curve for overall survival according to median preoperative IP VEGF (a) and ratio of preoperative to postoperative IP VEGF (b)

TABLE 1 Analysis of factors predictive for overall survival at 2 years Variable

Multivariate analysisa

Univariate analysis P value (Cox model)

Hazard ratio (95 % CI)

P value

Hazard ratio (95 % CI)

Age at diagnosis

0.033

0.591 (0.364–0.958)

CA 19-9 C 18 (vs. \18)

0.044

2.221 (1.024–4.819)

Use of neoadjuvant chemotherapy yes (vs. no)

0.055

7.060 (0.962–51.836)

0.198

4.219 (0.471–37.819)

PCI C 15 (vs. \15)

0.036

2.316 (1.056–5.080)

0.508

1.603 (0.397–6.476)

\0.0001

6.948 (3.353–14.398)

0.044

4.868 (1.046–22.652)

Presence of ascites

0.596

0.823 (0.401–1.690)

Duration of surgery C270 mn (vs. \270 mn) Extent of surgery: minor (vs. major)

0.066 0.107

4.122 (0.913–18.610) 1.811 (0.879–3.731)

Use of HIPEC

0.001

0.300 (0.145–0.618)

Presence of vascular emboli

0.003

4.742 (1.694–13.270)

Preoperative IV VEGF(T0)

0.846

1.001(0.994–1.008)

Preoperative IP VEGF level (T1)

0.0004

1.004 (1.002–1.006)

0.025

1.003 (1.000–1.005)

CC [ 2 mm (vs. B2 mm)

Ratio of preoperative IP VEGF to IV VEGF (T1/IV)

0.004

1.013 (1.004–1.022)

1 h postoperative VEGF level (T4)

0.075

1.013 (0.999–1.027)

a

Results were adjusted with PCI and prior chemotherapy

VEGF: Prognostic Factor in Peritoneal Malignancies

overall survival at 2 years. In a multivariate analysis adjusted for the use of neoadjuvant chemotherapy and PCI, only the CC score and the T1 level remained significant. For disease-free survival at 2 years, the following variables were significant: lower age at diagnosis, CEA level C17.5, the presence of chemotherapy use, PCI score C15, CC score of 2, no use of HIPEC, minor extent of surgery, absence of complications, involvement of lymph nodes, and the presence of vascular emboli (Table 2). As for the VEGF variables, lower T0, lower T1, and a lower ratio of T1/T0 were associated with improved disease-free survival 2 years. In a multivariate analysis, PCI score, CC score, presence of chemotherapy, extent of surgery, involvement of lymph nodes, and the T1/T0 ratio remained significant. Because our group of patients is composed of patients with tumours of varying etiologies, we also did the analysis for each type of primary tumour. For patients with colorectal cancer, univariate analysis showed that lower T1 and lower ratio T1/T0 were associated with improved overall survival at 2 years (Table 3). PCI score [15 and CC score of 2, absence of cytoreductive surgery, duration of surgery C270 min, involvement of lymph nodes, presence of vascular emboli, CEA C 17.5, and CA125 C 20 were significantly associated with poorer survival. On multivariate analysis, only PCI [ 15, absence of chemotherapy, and a higher T1/T0 ratio remained significant. None of the VEGF values were significant for survival at 2 years for patients with ovarian cancer, gastric cancer, and Pseudomyxoma peritonei.

DISCUSSION Bevacizumab is used in conjunction with standard regimes in metastatic colorectal cancer.11,12 In ovarian cancer, bevacizumab is used in combination in the upfront setting as well as with recurrences.13,14 However, IV bevacizumab is associated with an increase in complications and mortality, including hemorrhage (23.5 %), neutropenia (12.2 %), and gastrointestinal perforation (7.1 %).15 It also interferes with wound healing.16 Due to the high risk of complications of IV bevacizumab, it should be considered for use intraperitoneally. Intraperitoneal chemotherapy is designed to deliver higher doses of chemotherapy into the peritoneal cavity with limited deep tissue penetration. In a review of the use of intraperitoneal bevacizumab, Kobold et al. summarized the preclinical evidence in the literature for the use of targeted therapy in malignant ascites.17 In vitro experiments have shown that an anti-VEGF antibody can neutralize VEGF in its role to induce vascular hyperpermeability.18,19 Various animal studies have shown that the intraperitoneal application of anti-VEGF antibodies is safe and can lead to complete removal of fluid accumulations in mice inoculated with various carcinoma cell lines, decreased vascular permeability of microvessels, inhibition of tumor growth to a lesser extent, and prolonged survival.18,20,21 Various case reports and small case series have shown success with intraperitoneal bevacizumab for malignant

TABLE 2 Analysis of factors predictive of disease free survival 2 years Variable

Multivariate analysisa

Univariate analysis P value (Cox model)

Hazard ratio (95 % CI)

P value

Hazard ratio (95 % CI)

Age at diagnosis

0.048

CEA C 17.5 (vs. \17.5)

0.044

2.272 (1.022–5.052)

Use of neoadjuvant chemotherapy

0.0134

2.901 (1.246–6.754)

0.0415

4.203 (1.057–16.718)

0.079

1.643 (0.944–2.860)

0.0381

2.711 (1.056–6.957)

79.988 (10.609-603.067)

0.0474

21.151 (1.036–431.768)

PCI C 15 (vs. \15) CC [ 2 mm (vs. B 2 mm) Presence of ascites Use of HIPEC Extent of surgery: minor (vs. major)

\0.0001 0.072

0.796 (0.635–0.998)

0.634 (0.386–1.041)

\0.0001

0.293 (0.174–0.490)

0.0833

0.512 (0.240–1.092)

0.016

1.850 (1.123–3.047)

0.016

2.644 (1.203–5.808)

0.0116

2.760 (1.254–6.072)

0.0398

1.014 (1.001–1.027)

Development of major complication

0.0230

0.571 (0.344–0.947)

Involvement of lymph nodes

0.059

1.708 (0.979–2.979)

Presence of vascular emboli

0.0256

2.667 (1.128–6.306)

Preoperative IV VEGF level (T0)

0.037

1.041 (1.002–1.082)

Preoperative IP VEGF level (T1)

0.0006

1.004 (1.002-1.006)

Ratio of preoperative IP VEGF to IV VEGF (T1/T0)

0.005

1.013 (1.004–1.022)

Immediate postoperative IV VEGF (T2)

0.064

1.062 (0.997–1.131)

a

Results were adjusted with PCI and prior chemotherapy

C. S. Chia et al. TABLE 3 Analysis of factors predictive of overall survival 2 years for colorectal cancer only Variable

Multivariate analysisa

Univariate analysis P value (cox model)

Hazard Ratio (95 % CI)

CEA C 17.5 (vs. \17.5)

0.039

5.342 (1.085–26.303)

CA 125 C 20 (vs. \20)

0.035

3.654 (1.093–12.217)

PCI C 15 (vs. \15)

0.004

4.998 (1.667–14.987)

CC [ 2 mm (vs. B2 mm)

0.015

5.045 (1.371–18.559)

Cytoreductive surgery

0.081

0.154 (0.019–1.259)

Use of neoadjuvant chemotherapy

0.767

0.731 (0.093–5.783)

Duration of surgery C270 min (vs. \ 270) Involvement of lymph nodes

0.020 0.091

6.569 (1.343–32.140) 2.936 (0.842–10.241)

Presence of vascular emboli

0.011

6.251 (1.536–25.438)

Preoperative IV VEGF level (T0)

0.673

1.002 (0.993–1.011)

Preoperative IP VEGF level (T1)

0.002

1.011 (1.004–1.019)

Ratio of preoperative IP VEGF to IV VEGF (T1/IV)

0.010

1.054 (1.013–1.097)

a

P value

Hazard ratio (95 % CI)

0.009

6.009 (1.557–23.186)

0.013

0.039 (0.003–0.502)

0.038

1.009 (1.000–1.017)

Results were adjusted with PCI and prior chemotherapy

ascites from various cancers, including colorectal, breast, uterine, and ovarian.4,5,22,23 We now understand that bevacizumab can be used intraperitoneally, but the question remains as to whether it should be used. To answer this question, we evaluated whether a change in VEGF levels correlates to improved survival. Our results show that a lower T1 was significantly associated with overall survival at 2 years. IV VEGF levels did not correlate with survival; this is not surprising, because the mode of dissemination of peritoneal disease is local and not vascular. Our results show that a lower T1 and T1/T0 ratio can contribute to better overall and disease-free survival on univariate analysis. A lower T1 remained significant for improved overall survival, whereas a lower T1/T0 ratio remained significant for disease-free survival on multivariate analysis. Lowering the preoperative IP VEGF can be done by considering preoperative IP bevacizumab. At the same time, IV bevacizumab may not be as useful preoperatively in view of IV VEGF levels not significantly affecting survival. This also addresses the problem that the use of IV bevacizumab is associated with a high rate of complications.24 Intraperitoneal bevacizumab, on the other hand, theoretically should have a safer profile. It has been shown in animal studies that it does not affect wound or anastomotic healing, decreases the adhesion rate, and does not increase the morbidity and mortality with surgery.25–28 On analysis of the clinicopathological variables, various variables, such as a lower PCI score, higher age at diagnosis, shorter duration of surgery, and use of HIPEC amongst others, were found to be associated with improved overall and disease-free survival. Older age seemed to be

associated with improved survival. We postulate that this could be to younger patients having more aggressive disease. The same argument could be made for those who did not receive chemotherapy having improved disease-free survival. On multivariate analysis, only CC score was remained significant for both overall and disease-free survivals. This once again highlights the importance of complete cytoreduction to achieve a CC score of 0 or 1 that has been shown in numerous studies. The use of bevacizumab was not significantly associated with survival, but this could be due to the small sample size. IP VEGF levels may be released due to the surgical trauma and hence may rise after CRS. To attempt to control for this factor, we included the extent of surgery in our analysis. While it was a significant factor for disease-free survival, the same was not seen in the analysis for overall survival. It is not possible to derive the exact impact the CRS itself has on the levels of IP VEGF. Larger studies may control for this factor by grouping the patients according to extent of surgery. However, the variables that did turn out to be significant in our study (e.g., T1 and T1/ T0) are both levels of IP VEGF that are taken prior to surgery and hence the extent of surgery should not affect these variables. When we did a subgroup analysis by tumour type, only colorectal cancers showed variables with significant results. This is not surprising as bevacizumab is used most frequently in colorectal cancers. One of the drawbacks of our study is that we have a study population that is made up of various tumour types and hence the levels of VEGF may vary greatly. To properly evaluate the impact of IP VEGF and the possibility of using IP bevacizumab, further studies

VEGF: Prognostic Factor in Peritoneal Malignancies

with a larger study population should be performed with a focus on colorectal cancers. Previous studies have looked at VEGF expression in a single tumour type. Jimenez et al. studied VEGFR-2 expression in 59 patients with peritoneal carcinomatosis from appendiceal carcinoma.29 Low levels of VEGFR-2 expression were associated with improved overall survival. Harlozinska et al. studied the levels of VEGF in ascitic fluid in 86 ovarian carcinoma patients.30 A lower ascitic fluid VEGF level was associated with longer disease-free survival and a lower serum and ascitic fluid VEGF levels were associated with longer overall survival. Our finding correlates with what they have shown and strengthens the need to investigate the use of intraperitoneal bevacizumab to target a lower VEGF level. In our current study, preoperative T1 was taken at laparotomy just prior to CRS. If the use of IP VEGF levels becomes clinically relevant, the results should be known before CRS. Because the use of IP bevacizumab would likely be targeted at patients who have a heavy disease load on preoperative imaging, we suggest taking the IP VEGF levels during a diagnostic laparoscopy or mini-laparotomy, which these patients may have prior to CRS to assess resectability. In these instances, IP bevacizumab potentially can be added to the neoadjuvant regime. Another possibility of the clinical use of IP VEGF would be during follow-up. As shown in our results, those with a higher IP VEGF level preoperatively had a poorer overall and disease-free survival. By knowing these levels preoperatively, we can identify these patients whose prognosis may be worse. They may be considered for more aggressive adjuvant therapy as well as closer monitoring and follow-up to pick up recurrences. Limitations of our study are due to the initial difficulties in collecting the intraperitoneal fluids and hence we only have a collection of rate of 75 %. One other issue was the collection of the fluid at T1. In patients who did not have ascites, we used peritoneal washings that may not be comparable to ascites. To attempt to control for this variable, the presence of ascites was added to the analysis and was not found to be a significant factor in overall or disease-free survival. However, we suggest that subsequent studies use a more standardized approach to the collection of IP VEGF and perhaps use only peritoneal washings rather than the ascites. Having a mixed cohort of patients of varying primary tumours also may have confounded the survival outcomes as different primary tumours may have different levels of IV and IP VEGF as well as inherent varying survival rates to begin with. We acknowledge that a small sample may have resulted in small hazard ratios, and the effect may appear to be clinically insignificant. However, because none of our patients were lost to followup, this study serves as an encouraging initial study into the

possibility of using intraperitoneal bevacizumab. Further clinical studies should be performed. CONCLUSIONS We have shown that a lower preoperative IP VEGF correlates with improved overall and disease-free survival. Further studies should be aimed at reducing the preoperative IP VEGF level. As such, the possibility of using preoperative IP bevacizumab should be studied. Future efforts to study IP VEGF and the use of IP bevacizumab should perhaps focus on colorectal cancer, because it appears to be the tumor type in which IP VEGF levels correlate most closely with survival. As only case reports and small case series have reported the use of IP bevacizumab, further clinical studies also need to evaluate the safety and efficacy of using IP bevacizumab. ACKNOWLEDGMENT The authors thank Adeline Roux for her contributions to the statistical analysis. DISCLOSURES The authors have no financial disclosures or conflict of interests.

REFERENCES 1. Elias D, Goe´re´ D, Dumont F, et al. Role of hyperthermic intraoperative peritoneal chemotherapy in the management of peritoneal metastases. Eur J Cancer. 2014;50(2):332–40. 2. Passot G, You B, Boschetti G, et al. Pathological response to neoadjuvant chemotherapy: a new prognosis tool for the curative management of peritoneal colorectal carcinomatosis. Ann Surg Oncol. 2014;21(8): 2608–14. 3. Harmey J. VEGF and cancer. Georgetown: Landes Bioscience/ Eurekah.com, Kluwer Academic/Plenum Publishers; 2004; ISBN0-306-47988-5. 4. Kesterson JP, Mhawech-Fauceglia P, Lele S. The use of bevacizumab in refractory ovarian granulosa-cell carcinoma with symptomatic relief of ascites: a case report. Gynecol Oncol. 2008;111:527–9. 5. Hamilton CA, Maxwell GL, Chernofsky MR, et al. Intraperitoneal bevacizumab for the palliation of malignant ascites in refractory ovarian cancer. Gynecol Oncol. 2008;111:530–2. 6. Passot G, Bakrin N, Garnier L, et al. Intraperitoneal vascular endothelial growth factor burden in peritoneal surface malignancies treated with curative intent: the first step before intraperitoneal anti-vascular endothelial growth factor treatment? Eur J Cancer. 2014;50(4):722–30. 7. Li L, Wang L, Zhang W, et al. Correlation of serum VEGF levels with clinical stage, therapy efficacy, tumor metastasis and patient survival in ovarian cancer. Anticancer Res. 2004;24:1973–9. 8. Panteix G, Beaujard A, Garbit F, et al. Population pharmacokinetics of cisplatin in patients with advanced ovarian cancer during intraperitoneal hyperthermia chemotherapy. Anticancer Res. 2002;22(2B):1329–36. 9. Cotte E, Colomban O, Guitton J, et al. Population pharmacokinetics and pharmacodynamics of cisplatinum during hyperthermic intraperitoneal chemotherapy using a closed abdominal procedure. J Clin Pharmacol. 2011;51(1):9–18.

C. S. Chia et al. 10. Passot G, Bakrin N, Roux AS, et al. Quality of life after cytoreductive surgery plus hyperthermic intraperitoneal chemotherapy: a prospective study of 216 patients. Eur J Surg Oncol. 2014;40(5):529–35. 11. Kabbinavar FF, Hambleton J, Mass RD, Hurwitz HI, Bergsland E, Sarkar S. Combined analysis of efficacy: the addition of bevacizumab to fluorouracil/leucovorin improves survival for patients with metastatic colorectal cancer. J Clin Oncol. 2005;23(16):3706–12. 12. Macedo LT, da Costa Lima AB, Sasse AD. Addition of bevacizumab to first-line chemotherapy in advanced colorectal cancer: a systematic review and meta-analysis, with emphasis on chemotherapy subgroups. BMC Cancer. 2012;12:89. 13. Burger RA, Brady MF, Bookman MA, et al. Incorporation of bevacizumab in the primary treatment of ovarian cancer. N Engl J Med. 2011;365(26):2473–83. 14. Pujade-Lauraine E, Hilpert F, Weber B, et al. Bevacizumab combined with chemotherapy for platinum-resistant recurrent ovarian cancer: the AURELIA open-label randomized phase III trial. J Clin Oncol. 2014;32(13):1302–8. 15. Ranpura V, Hapani S, Wu S. Treatment-related mortality with bevacizumab in cancer patients: a meta-analysis. JAMA. 2011;305(5):487–94. 16. Scappaticci FA, Fehrenbacher L, Cartwright T, et al. Surgical wound healing complications in metastatic colorectal cancer patients treated with bevacizumab. J Surg Oncol. 2005;91(3):173–80. 17. Kobold S, Hegewisch-Becker S, Oechsle K, Jordan K, Bokemeyer C, Atanackovic D. Intraperitoneal VEGF inhibition using bevacizumab: a potential approach for the symptomatic treatment of malignant ascites? Oncologist. 2009;14(12):1242–51. 18. Luo JC, Toyoda M, Shibuya M. Differential inhibition of fluid accumulation and tumor growth in two mouse ascites tumors by an antivascular endothelial growth factor/permeability factor neutralizing antibody. Cancer Res. 1998;58:2594–600. 19. Ke L, Qu H, Nagy JA, et al. Vascular targeting of solid and ascites tumours with antibodies to vascular endothelial growth factor. Eur J Cancer 1996;32A:2467–73. 20. Mesiano S, Ferrara N, Jaffe RB. Role of vascular endothelial growth factor in ovarian cancer: Inhibition of ascites formation by immunoneutralization. Am J Pathol. 1998;153:1249–56.

21. Shaheen RM, Ahmad SA, Liu W, et al. Inhibited growth of colon cancer carcinomatosis by antibodies to vascular endothelial and epidermal growth factor receptors. Br J Cancer. 2001;85:584–9. 22. Numnum TM, Rocconi RP, Whitworth J, et al. The use of bevacizumab to palliate symptomatic ascites in patients with refractory ovarian carcinoma. Gynecol Oncol. 2006;102:425–8. 23. El-Shami K, Elsaid A, El-Kerm Y. Open-label safety and efficacy pilot trial of intraperitoneal bevacizumab as palliative treatment in refractory malignant ascites. J Clin Oncol. 2007;25(18 suppl):9043. 24. Eveno C, Passot G, Goe´re´ D, et al. Bevacizumab doubles the early postoperative complication rate after cytoreductive surgery with hyperthermic intraperitoneal chemotherapy (HIPEC) for peritoneal carcinomatosis of colorectal origin. Ann Surg Oncol. 2014;21(6):1792–800. 25. Karanlik H, Kurt A, Kunduz E, et al. Effects of intraperitoneal bevacizumab administration on colonic anastomosis and early postoperative adhesion formation. Surg Innov. 2013;20(6):559–65. 26. Basbug M, Bulbuller N, Camci C, et al. The effect of antivascular endothelial growth factor on the development of adhesion formation in laparotomized rats: experimental study. Gastroenterol Res Pract. 2011;2011:578691. 27. Ignjatovic D, Aasland K, Pettersen M, et al. Intra-abdominal administration of bevacizumab diminishes intra-peritoneal adhesions. Am J Surg. 2010;200(2):270–5. 28. Passot G, Dupre´ A, Rivoire M, Mohamed F, Bakrin N, Glehen O. Intraperitoneal bevacizumab combined with cytoreductive surgery: a pre-clinical study of tolerance and pharmacokinetics in an animal model. Clin Transl Oncol. 2012;14(12):931–6. 29. Jimenez WA, Sardi A, Ledakis P, Nieroda C, Sittig M, Gushchin V. The role of VEGFR-2 expression in outcomes and survival of patients with peritoneal carcinomatosis from appendiceal cancer. Eur J Surg Oncol. 2013;39(12):1415–22. 30. Harlozinska A, Sedlaczek P, Kulpa J, et al. Vascular endothelial growth factor (VEGF) concentration in sera and tumor effusions from patients with ovarian carcinoma. Anticancer Res. 2004;24(2C):1149–57.

Intraperitoneal Vascular Endothelial Growth Factor: A Prognostic Factor and the Potential for Intraperitoneal Bevacizumab Use in Peritoneal Surface Malignancies.

Intraperitoneal (IP) vascular endothelial growth factor (VEGF) levels have been shown to vary in the peritoneal cavity of patients with peritoneal sur...
489KB Sizes 0 Downloads 6 Views