CO2 pneumoperitoneum increases systemic but not local tumor spread after intraperitoneal murine neuroblastoma spillage in mice
- First Online:
- Cite this article as:
- Metzelder, M., Kuebler, J., Shimotakahara, A. et al. Surg Endosc (2008) 22: 2648. doi:10.1007/s00464-008-9778-2
- 80 Views
Minimally invasive techniques are increasingly used for biopsy and resection of neuroblastoma, but the impact on the behavior of spilled tumor cells is unknown. We aimed to investigate whether CO2 pneumoperitoneum can affect local or systemic tumor manifestation after spillage of neuroblastoma cells into the peritoneal cavity.
Murine neuroblastoma cells (Neuro2a, 1x106) were inoculated into the peritoneal cavity of 25 male A/J mice, which subsequently underwent CO2 pneumoperitoneum (n = 12) or laparotomy (n = 13) for 1 h. At the 28th postoperative day, local (peritoneal and surface of the gut) and systemic (liver, lung, spine) tumor spread was graded in a blinded manner (1–4 point scale) and specimens were histologically examined for tumor manifestation (hematoxylin and eosin stain) and tumor cell proliferation rate (Ki-67-stain). In the case of no visible lesion, five random sections were histologically examined. Peritoneal carcinosis was graded macroscopically.
Tumor manifestations were detected in 10 out of 12 (83%) animals after CO2 pneumoperitoneum, and in 9 out of 13 (69%) after laparotomy (n.s.). Incidence of liver metastasis was higher after CO2 pneumoperitoneum versus laparotomy (83% versus 31%; p < 0.05). Incidence and grading of peritoneal carcinosis was not significantly different between the groups (n.s.). Intrapulmonary metastasis was found in one mouse of each group, but no metastasis of the spine. However, the grading of liver metastasis was higher after CO2 pneumoperitoneum compared to laparotomy (p < 0.05). Tumor cell proliferation (Ki-67 stain) in the liver did not differ between both groups. Moreover, proliferation always exceeded 50% of tumor cells, irrespective local or systemic tumor manifestation.
CO2 pneumoperitoneum increased intrahepatic metastasis, but not local peritoneal carcinosis in a murine neuroblastoma model. This suggests that laparoscopy could promote systemic dissemination of intraperitoneally spilled tumor cells when no chemotherapy is applied. It remains to be determined whether this is due to local immune suppression or direct modulation of tumor cell behavior.
KeywordsEndoscopyCO2 pneumoperitoneumTumor spillageLocal metastasisSystemic metastasisNeuroblastoma
Laparoscopic surgery is increasingly accepted in pediatric malignancies [12, 38]. Today, almost all solid pediatric tumors have been reported to undergo minimally invasive biopsy and, to a lesser extent, minimally invasive resection [12, 16, 31, 38]. However, the impact of CO2 used for pneumoperitoneum on the behavior and potential progression of malignant tumors is controversial . It has been postulated for adult tumors that the immune system plays a key role in this respect, but the results of various experimental studies were not conclusive [1, 2, 4, 6, 7, 13, 18, 22, 27, 40]. The wide range of pediatric malignancies has not been investigated in this respect.
Neuroblastoma is the most common solid tumor in childhood, characterized by a diversity of clinical behavior ranging from spontaneous remission to rapid tumor progression and death [29, 39]. It now represents the most frequent indication for minimally invasive tumor biopsy and resection compared to other pediatric tumors [8, 12, 31, 33, 38, 42].
The aim of our study was to investigate whether application of a CO2 pneumoperitoneum could affect local or systemic tumor spread of isolated intraperitoneally inoculated neuroblastoma cells. To simulate a clinical setting, we used a mouse model with reproducible tumor spillage mimicking conditions as during minimally invasive bioptive and resective surgery for neuroblastoma.
Materials and methods
Cell culture and vitality
The Neuro-2a murine neuroblastoma cell line (DSMZ, Braunschweig, Germany) is characterized by local and progressive tumor growth, low metastatic potency until the death of the animal , and is known for its similarity to human neuroblastoma in biological, physiological, and immunological characteristics .
Neuro-2a cells were cultured in Dulbecco’s modified Eagle’s medium + 1000 mg/l glucose + L-glutamine + pyruvat (D-MEM, Gibco Invitrogen GmbH, Karlsruhe, Germany) supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS; PAA Laboratories, Pasching, Austria), 1x non-essential amino acids (Gibco), 100 U/ml penicillin, and 100 μg/ml streptomycin (PAA Laboratories, Pasching, Austria). Cells were cultured in a humidified incubator with 5% CO2 at 37°C. The cultures were split 1:3 once a week using trypsin/EDTA (PAA Laboratories, Pasching, Austria). Confluent grown cell cultures were used for the application. Vitality of Neuro-2a cells was confirmed to be above 90% throughout the experiments.
Animals and study design
Male A/J mice were purchased from the Jackson Laboratory (Bar Harbor, ME, USA) and were housed in a specific pathogen-free barrier facility with a 12 h dark–light cycle at the central animal laboratory of the university. Before the experiments mice were allowed to acclimatize for at least 1 week. The animals had access to standard laboratory feed and were provided with tap water ad libitum. All research and animal care procedures were approved by the local government committee.
In preliminary experiments, we determined the minimal number of tumor cells needed to induce intraperitoneal tumor manifestation in approximately 50% of the animals. Moreover, we observed the progression of the tumor disease and determined day 28 as the optimal time point to allow assessment of tumor spread without impairment or suffering of the mice used in this experiment.
All animals were fasted 1 h prior to the main experiments. Following a short induction with isoflurane, anesthesia (85 mg/kg ketamine, 15 mg/kg xylazine) was administered in the right thigh muscle. The abdomen was shaved and disinfected. Mice were divided into two groups: CO2 pneumoperitoneum (n = 12) or laparotomy (n = 13). Operation time was approximately 80 min and was equal for both groups. Mice in the CO2 group underwent a small umbilical incision followed by a purse-string suture to fixate an 18 gauge cannula in the abdominal cavity. Mice in the laparotomy group underwent a midline abdominal incision extending from the xiphoid to the suprapubic region. All mice were subsequently inoculated with 1 × 106 Neuro-2a cells into the peritoneal cavity, either through the umbilical 18 gauge cannula or via open abdominal application. Mice in the CO2 group were insufflated with a CO2 pneumoperitoneum at 3 mmHg pressure using a CO2 insufflator (Olympus Winter & Ibe, Hamburg, Germany) via inserted cannula for 1 h, while mice in the laparotomy group were left with unclosed abdomen for 1 h. Prior to the end of anesthesia, the laparotomy wound, respectively the incision for cannula were closed. After the procedure, all mice were housed in a specific pathogen-free barrier facility for 28 days and were observed twice a day.
At day 28 after tumor cell inoculation, all mice were sacrificed and local (peritoneal and surface of the gut) and systemic (liver, lung, spine) tumor spread was assessed and graded in a blinded manner by a histopathologist (1–4 point scale: 1 = no tumor lesion; 2 = single tumor lesion; 3 = multiple tumor lesions; 4 = generalized tumor).
To evaluate microscopic dissemination of Neuro-2a tumor cells, harvested organs (skinned peritoneum, gut, liver, lung and spine) were fixed in neutral buffered formaldehyde, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E stain). Care was taken to differentiate between metastatic liver disease and peritoneal tumor cells attached to the liver. Skinned peritoneum, the harvested gut, liver, lung, and spine specimens were processed in 1-mm layers, and lesions were histologically examined for tumor manifestation (H&E stain) and tumor cell proliferation rate (Ki-67-stain, MIB 1, Dako, Hamburg, Germany). Proliferation was assessed by counting the number of Ki-67 positively stained nuclei from total number of cancer cells in high-power field microscopy (HPF) in a representative region at 400 × magnification. In the case of no macroscopically visible lesion, five random sections were embedded and histologically examined. Proliferation of more than 50% of cells was regarded as high tumor cell proliferation.
The incidence of local versus systemic tumor spread was analyzed by the chi-squared test. Incidence and degree of tumor spread were compared by the Mann–Whitney U test. p < 0.05 was considered to be significant.
There was no morbidity during the observation period in any animal of either group. Tumor manifestations were detected after 28 days in 10 out of 12 (83%) animals after CO2 pneumoperitoneum, and in 9 out of 13 (69%) after laparotomy (n.s.).
Tumor cell proliferation rates were investigated by Ki-67-staining (Fig. 4) and were similar in the compared groups, i.e., revealed a high tumor cell proliferation rate of >50% of tumor cells in all tumor lesions, irrespective of local or systemic tumor manifestation in CO2 and laparotomy animals.The results of the grading of peritoneal carcinosis was not significantly different [CO2: no tumor lesion (n = 2); single tumor lesion (n = 6); multiple tumor lesions (n = 1); generalized tumor (n = 3); versus laparotomy: no tumor lesion (n = 4); single tumor lesion (n = 3); multiple tumor lesions (n = 2); generalized tumor (n = 4) between the groups (p = 0.42)]. In contrast, the grading of liver metastasis was significantly higher [CO2: no tumor lesion (n = 2); single tumor lesion (n = 2); multiple tumor lesions (n = 5); generalized tumor (n = 2); versus laparotomy: no tumor lesion (n = 9); single tumor lesion (n = 1); multiple tumor lesions (n = 3); generalized tumor (n = 0)] after CO2 pneumoperitoneum compared to laparotomy (p < 0.05) (Fig. 3).
Numerous authors reported an acceptable feasibility of minimally invasive techniques for biopsy and resection of various malignant tumors in children [12, 31, 38]. However, there is an ongoing debate on potential deleterious effects of CO2 on tumor cell behavior and on specific cells dealing with tumor spillage.
Experimental studies confirmed that CO2 pneumoperitoneum alters the behavior of tumor cells derived from colon carcinoma, adenocarcinoma and breast cancer [21, 27, 30]. However, the results of numerous in vitro and animal studies were not conclusive [6, 9, 10, 11, 22]. Some authors reported an increase in cell proliferation and tumor growth [13, 18, 20, 30], whereas others found beneficial effects of CO2 exposition [1, 27, 37]. We showed previously  that exposure to CO2 decreased the mitochondrial activity of several neuroblastoma cell lines, and of lymphoma and hepatocellular carcinoma cells in vitro. This phenomenon persisted over 4 days.
In addition, it has been shown that CO2 alters the functions of macrophages and polymorphonuclear cells in the peritoneal and thoracic cavity [19, 32, 35, 36, 41]. We have confirmed a downregulation of the release of various cytokines, free oxygen radicals, nitric oxide, and of the mitochondrial activity of macrophages and polymorphonuclear cells during and after CO2 exposition via alteration of the extracellular pH [23, 25]. This might interfere with the clearance of tumor cells, which are spread during laparoscopic or thoracoscopic operations.
In the present study, CO2 did not alter survival over 28 days in a mouse model with peritoneal spillage of Neuro-2a neuroblastoma cells during pneumoperitoneum. However, we observed a significantly increased spread of neuroblastoma cells in the liver after CO2 pneumoperitoneum compared to laparotomy. The peritoneal tumor manifestation and rate of metastasis in other organs was not different. This differs from the findings of Iwanaka et al. [14, 15], who investigated neuroblastoma C1300 and TBJ-NB cells in mice. In an initial study , the authors determined the impact of CO2 pneumoperitoneum versus laparotomy 4 and 11 days after retroperitoneal tumor inoculation. There was no difference in survival, tumor growth, and the incidence of metastasis. However, in this model, the tumor remained untouched without spillage of tumor cells. In a consecutive study , the authors performed laparoscopic tumor biopsy using CO2 versus gasless pneumoperitoneum. Again, the use of CO2 did not affect port-site recurrence, but laparotomy with air exposition mimicking clinical conditions was not performed in this study and a follow-up period of a maximum of 11 days postoperatively might be too short to detect differences.
Although murine Neuro-2a neuroblastoma cell line is known for its similarity to human neuroblastoma in biological, physiological, and immunological characteristics [43, 44], the behavior of neuroblastoma cells derived from cell lines may differ from that of neuroblastomas in children. Thus, our results cannot easily be transferred to the clinical setting. Furthermore, current tumor protocols for neuroblastoma in children include postoperative chemotherapy. This had not been performed in our experiments.
Ki-67 immunohistochemistry staining is an accepted instrument to determinate tumor cell proliferation of neuroblastoma , although there is no consensus about which marker is most appropriate . There was a homogeneous rate of proliferation in peritoneal and systemic tumor maniferations, and in CO2 and laparotomy animals. These results suggest that the increased systemic tumor spread was not due to an increase in tumor proliferation by CO2.
An increased rate of port-site metastasis of osteogenic sarcoma has been observed after minimally invasive resection . Besides these reports, there is no further evidence that the application of CO2 has an impact on survival or tumor behavior in patients. Several controlled clinical studies could not confirm an increase in tumor spread in adult patients [5, 26]. Survival rates and tumor or port-site recurrences were not affected by the use of CO2 in patients with colonic cancer. In children, Iwanaka et al.  found no port-site recurrence after endosurgery for malignancies in 129 children. We reported on similar results in a series of 90 children undergoing minimally invasive biopsy or resection of various types of tumors including neuroblastoma . However, a systematic follow-up remains mandatory after biopsy or resection of neuroblastoma and other tumors in children. The rates of survival and recurrences after minimally invasive tumor surgery should be investigated systematically in larger series of children.
In conclusion, we observed an increased hepatic spread of neuroblastoma cells after CO2 pneumoperitoneum without increase of peritoneal carcinosis. This suggests that laparoscopy could promote systemic dissemination of intraperitoneally spilled tumor cells when no chemotherapy is applied. It remains to be determined whether this is due to local immune suppression or to modulation of tumor cell behavior.