Oncologic implications of laparoscopic and open surgery

Surg Endosc (2002) 16: 441±445 DOI: 10.1007/s00464-001-8112-z Ó Springer-Verlag New York Inc. 2001 Oncologic implications of laparoscopic and open surgery C. A. Jacobi ,1 H. J. Bonjer,2 M. I. Puttick,3 R. O'Sullivan,4 S. W. Lee,5 P. Schwalbach,6 H. Tomita,7 Z. G. Kim,8 P. Hewett,9 P. Wittich,2 J. W. Fleshman,10 P. Paraskeva,3 T. Geûman,8 S. J. Neuhaus,11 P. Wildbrett,1 M. A. Reymond,12 C. Gutt,8 R. I. Whelan5 1 Surgical Department, University of Berlin ChariteÂ, Schumannstar. 20 / 21, D-10098 Berlin, Germany Department of Surgery, University Hospital Rotterdam, The Netherlands 3 Minimal Access Surgical Unit, Imperial College School of Medicine, St. Mary's, London, United Kingdom 4 Department of Surgery, Cork University Hospital, Cork, and Beaumont Hospital, Dublin, Ireland 5 Surgical Department, Columbia Presbyterian Medical Center, New York, NY, USA 6 Surgical Department, University of Heidelberg, Heidelberg, Germany 7 Department of Colorectal Surgery, Cleveland Clinic, Cleveland, OH, USA 8 Department of Surgery, Goethe University of Frankfurt, Frankfurt, Germany 9 Department of Surgery, Queen Elizabeth Hospital, Woodville, Australia 10 Department of Surgery, Section of Colon and Rectal Surgery, Washington University, School of Medicine, St. Louis, MO, USA 11 University Department of Surgery, Royal Adelaide Hospital, Adelaide, Australia 12 Digestive Surgery, University Hospital of Geneva, Geneva, Switzerland 2 Received: 18 January 2001/Accepted: 24 January 2001/Online publication: 10 December 2001 Abstract Although instrumental manipulation and mechanical tumor cell spillage seem to play the major role in portsite metastases from laparoscopic cancer surgery, minimally invasive procedures are used more and more in the resection of malignancies. However, port-site metastases also have been reported after resection of colon cancer in International Union Against Cancer (UICC) stage I [2, 14]. Therefore, changes in the peritoneal environment during laparoscopy also might in¯uence intra- and extraperitoneal tumor growth during laparoscopy and pneumoperitoneum. Di€erent results of experimental studies presented at the Third International Conference for Laparoscopic Surgery are analyzed and discussed. Key words: Laparoscopic surgery Ð Tumor growth Ð Pathophysiology Ð Therapy Pathophysiologic changes in laparoscopic tumor models Carbon dioxide (CO2), shown to stimulate tumor cell growth in di€erent animal models [1, 9, 12], seems to be partly responsible for metastases mainly in the trocar Presented at the Third International Laparoscopic Physiology Conference, New York, NY, 1999 Correspondence to: C. A. Jacobi sites. Schwalbach et al. [27] demonstrated that CO2 stimulates the growth of Morris hepatoma 3924 A cells in vitro. Furthermore, helium and xenon caused signi®cant suppression of tumor cell growth in these experiments. Nevertheless, in this model, all cells were incubated only with the di€erent gases in one experiment without repetition in each group. Therefore, statistical analyses cannot be performed in this experiment. Thus, results may be accidental ®ndings rather than a re¯ection of real di€erences between the gas groups. Jacobi et al. [11] investigated the in¯uence of di€erent gases (CO2 helium, and xenon) on tumor growth in a colon cancer rat model. They found signi®cantly increased intraperitoneal tumor growth in the CO2 group, as compared with helium group, whereas xenon did not di€er with the carbon dioxide group. Thus xenon does not seem to be an alternative gas for laparoscopic cancer surgery. In these experiments, a tumor cell suspension model was used. Although this model can simulate intraoperative tumor spillage during laparoscopy, the number of free tumor cells after instrumental manipulation in patients still is unknown. The problems with tumor cell suspension models have been evaluated and discussed by di€erent authors. Wittich et al. [29] could demonstrate that the time between the ®rst and last samples and the resuspension of the cell probes signi®cantly in¯uenced the viability and number of injected tumor cells in the so-called cell seeding models. Thus, intraperitoneal tumor take and development of port-site metastases also might be dif- 442 ferent between groups, independently of the gas used during laparoscopy. Furthermore, the number of injected cells certainly does in¯uence the tumor weight and development of port-site metastases, as shown by Fleshman [3]. Signi®cantly higher tumor development at port sites and intraperitoneal tumor weights was found after laparoscopy with CO2, as compared with the condition in the control group, when 2 ´ 106 colon cancer cells were injected into the peritoneal cavity of hamsters. However, using decreasing amounts of tumor cells in a second study, the di€erences between the groups were found to be without statistical relevance. Thus, di€erent numbers of injected tumor cells in cell seeding models might be one reason for controversial results in the literature. Di€erent tumor cells lines and the speci®c tumor cell biology certainly are other factors that also in¯uence the results in experimental models. Besides the in¯uence of di€erent gases on tumor cell growth, changes in humidity and temperature also might cause di€erences in tumor growth between laparoscopic and open procedures. The peritoneal surface drying out during laparotomy can cause cell damage and necrosis, which might rather explain increased surgical trauma than the larger incisions in the abdominal wall, as compared with those in laparoscopic surgery. Puttick et al [22] could demonstrate using a rat model that a laparotomy in a warm humid environment caused less postoperative intraperitoneal tumor growth than a laparotomy in room air. Furthermore, the expression of tumor necrosis factor-alpha by peritoneal macrophages was better preserved in the humidi®ed closed environment, whereas there were no di€erences in postoperative adhesion formations between the groups [22]. It seems that the provision of a warmed, humidi®ed enclosed space during laparoscopy may account for some bene®ts of laparoscopic surgery reported in the literature. Lee et al. [16] further demonstrated that laparotomy is associated with an increase in serum levels of a heparin-binding growth factor consistent with platelet-derived growth factor (PDGF). Incubation of C26 colon cancer cells with the serum of mice that had undergone surgery further led to a signi®cant increase in tumor cell growth, as compared with that of control serum. This increase could be neutralized with anti-PDGF antibody. Unfortunately, a comparison with a laparoscopic procedure was not performed in this experiment. Although CO2 pneumoperitoneum has been demonstrated to increase tumor growth in di€erent animal models, the exact pathomechanisms are not known at this writing. One possible mechanism might be the changes in tumor cell adherence and invasion during pneumoperitoneum. Puttick et al. [23, 24]. therefore investigated the cell invasion after incubation with air, helium, and CO2 as well as the expression of matrixmetalloproteinase genes in di€erent cancer cell lines. The tumor cell invasiveness of CC531s rat colon carcinoma through arti®cial basement membrane (Matrigel) was increased after helium and CO2 incubation, as compared with air. Nevertheless, a signi®cant di€erence was found only between CO2 and air (p < 0.05) [23]. The expression of matrix-metalloproteinase genes was further investigated in four di€erent cell lines (CC531s rat colonic cancer, SW480 human colon cancer, T3M4 pancreatic cancer, PSN-I pancreatic cancer). Although the authors presented early results of an ongoing study, it seemed that CO2 modulates the expression of MMP-2 and MMP-9, two of the most important matrix metalloproteinases in tumor invasion [24]. Instrumental manipulation and mechanical tumor cell spillage have been demonstrated to play the major role in postoperative tumor metastases in the peritoneal cavity or the port sites [17]. Nevertheless, intraperitoneal tumor cell spillage is connected not only to laparoscopic procedures, but also to conventional open surgery. Hewett et al. [6] investigated the movement of malignant cells in the peritoneal cavity as well as intraluminal tumor cell movements in the colon during laparoscopic and open colectomy in a pig model. Cells in a disseminated tumor model dispersed throughout the abdominal cavity, the intruments, the trocars, and the gloves in both laparoscopic and open surgery. There was no difference between the two groups. Intraluminal cells did not contaminate the operative ®eld in either open or laparoscopic resection [6]. Thus, no di€erence was found in any tumor models between laparoscopic and open procedures. Besides peritoneal tumor cell spillage, mechanical manipulation has further been discussed as a cause of increased hematogenous spread of malignant cells during laparoscopic procedures. Therefore, the hematogenous spread of tumor cells was investigated additionally in a second study by the aforementioned group [7]. Pigs underwent either laparoscopic or open colectomy, and blood samples were collected before, during, and after the resection. A disseminated intraperitoneal tumor model as well as intraluminal and intramural tumor injections were used in this study. False-positive results of hemotogenous tumor cells were found in three pigs, one before the introduction of tumor cells. There was no further proof of hematogenous spread of tumor cells in either open or laparoscopic procedures. Mechanical manipulation therefore did not play a major role in hematogenous tumor cell spread in this model. The in¯uence of pneumoperitoneum and laparotomy on the perioperative development of liver metastases and growth of existing hepatic micrometastases was investigated by Tomita et al. [28] in a rat model. The rats underwent infusion of 2 ´ 106 colon cancer cells in the cecal vein during laparoscopy with CO2 or air as well as laparotomy. The number and weight of hepatic tumors then were investigated 5 weeks after the operation. In a second experiment, cancer cells were injected into the cecal wall, and rats underwent CO2 pneumoperitoneum, laparotomy, or no further intervention in the control group 4 weeks after tumor cell injection. In this experiment, rats were explored 2 weeks after surgical intervention. No di€erence in either tumor incidence or weight and number of tumors was found in any of the groups. The authors concluded that pneumoperitoneum does not enhance liver tumor growth and metastases during cancer surgery. Nevertheless, no di€erences were found between the animals that underwent surgery and the control group. Thus, the number of injected tumor 443 cells may have been too high to detect di€erences between the surgical procedures in the animal model used. Further investigations with a decreased number of tumor cells are needed to prove whether the model used is appropriate for comparing liver metastases and growth between open and laparoscopic surgery. The di€erent in¯uences of surgical procedures on hepatic tumor growth of colonic cancer cells were investigated in two experimental models by Gutt [4, 13]. In the ®rst experiment, rats underwent tumor cell inoculation in the portal vein and laparoscopy with either helium or CO2 at 7 mmHg [4]. No signi®cant di€erence in total and hepatic tumor growth was found between the two groups 4 weeks after intervention. The authors concluded that elevated intraperitoneal pressure seems to have a greater in¯uence on in vivo tumor growth than the insu‚ation gas itself. Nevertheless, a control group and di€erent intraperitoneal pressures were not further investigated, so this theory remains unproved. In a second experiment, rats were injected with laparoscopic intrasplenic tumor cells to induce hepatic carcinomatosis [13]. After 7 days, the animals underwent either laparoscopy with CO2 laparotomy or gasless laparoscopy. Hepatic tumor growth was measured 4 weeks after the surgical interventions. Whereas laparotomy and laparoscopy with CO2 showed no signi®cant di€erence in tumor growth, gasless laparoscopy was associated with a signi®cant decrease in tumor growth, as compared with the procedures used with the two other groups. Therefore, it seems that gasless laparoscopy might be superior to either laparotomy and CO2 pneumoperitoneum. Nevertheless, the rat model is hardly comparable with the clinical situation. The abdominal wall of the rat can be lifted without any tissue trauma or local ischemia. In human patients, lifting systems still are causing local pressure and tissue trauma at the abdominal wall, and technical problems are known to make laparoscopic resections dicult. Thus, technical improvements in lifting systems for gasless surgery must be realized before this technique can become an alternative laparoscopic procedure in cancer patients. Prevention of metastases in laparoscopic surgery Besides the pathomechanisms of intraoperative tumor cell attachment and growth, little is known about possible therapeutic interventions to prevent tumor metastases in laparoscopic surgery. It has been demonstrated that tumor cell attachment is suppressed by binding domains of the extracellular matrix after intraperitoneal instillation of heparin in a murine model [5]. Neuhaus et al. [18] con®rmed these results using a tumor cell suspension model in the rat. These authors further demonstrated that intraperitoneally applying 2 ml of additional blood from a syngeneic donor rat caused a signi®cant increase in tumor growth in both the heparin and control groups. Again, a signi®cant decrease in tumor growth was found after heparin instillation in the rats undergoing intraperitoneal blood infusion, as compared with the control animals. Hyaluronate, the natural ligand of the broadly distributed adhesion molecule CD-44, also has been investigated as a prevention of postoperative tumor cell invasion by Paraskeva et al. [21]. Viscous hyaluronate solution was found to decrease cancer cell invasion through reconstituted basement membrane (Matrigel) and systemic ®lters, with signi®cant di€erence in the control group in vitro. Although intraperitoneal administration of viscous hyaluronate solution may reduce tumor cell invasion, further experiments must be performed to con®rm these ®ndings in vivo. Local application of cytotoxic agents has been discussed as preventing tumor metastases by killing spilled tumor cells after laparoscopic interventions. Intraperitoneal instillation of povidone-iodine reportedly causes a signi®cant decrease in port-site metastases after laparoscopy with CO2 in di€erent tumor cell suspension models [8, 19]. Furthermore, Lee et al [15] con®rmed the cytotoxic e€ect of povidone-iodine in a solid tumor model of the spleen. Splenic tumors were established via a subcapsular splenic injection of 105 C-26 colon adenocarcinoma cells in female Balb/C mice. Then 7 days later, the animals with isolated splenic tumors underwent intraperitoneal crushing of the tumor via laparoscopy and extracorporal splenectomy with saline irrigation, povidone-iodine irrigation, or no irrigation in the control group. Whereas povidone-iodine signi®cantly reduced the rate of port-site metastases, saline irrigation had no bene®cial e€ect. Recent experimental studies have shown a signi®cant decrease in tumor growth after intraperitoneal instillation of taurolidine (Taurolin; Hoechst, Germany), a derivative of the amino acid taurine, in combination with heparin, using CO2 for the establishment of pneumoperitoneum [8]. It is thought that inhibition of tumor growth after intraperitoneal application of taurolidine might be caused by inhibited IL-1b production of intraperitoneal macrophages. O'Sullivan et al. [20] additionally demonstrated signi®cant attenuation of postoperative increase in proangiogenic factors vascular endothelial growth factor [VEGF], vascular cell adhesion molecule-1 [VCAM-1] by taurolidine in patients [20]. Furthermore, it seems that taurolidine acts directly on the tumor cells and inhibits tumor cell growth itself. The combination of taurolidine and heparin produced signi®cant synergistic e€ects on suppression of tumor growth in vivo without any side e€ects [8]. Nevertheless, all antiadherent or cytotoxic agents were used in animals undergoing CO2 insu‚ation, although this gas might stimulate tumor growth itself. The combination of taurolidine, heparin, and povidone-iodine with di€erent insu‚ation gases such as helium or xenon were evaluated by Jacobi et al. [11] in rats. They found a signi®cant decrease in tumor growth after intraperitoneal instillation of either taurolidine or taurolidine±heparin, as compared with control animals in all gas groups. Whereas povidone-iodine caused signi®cant lower tumor growth in the CO2 group, the combination of helium and xenon with povidone-iodine caused no reduction in tumor growth, as compared with the control groups. Further investigations demonstrated that systemic application of these agents did not cause sig- 444 ni®cant inhibition of intraperitoneal tumor growth [10]. Combined intravenous and intraperitoneal application of these therapeutic agents did not lead to signi®cantly lower tumor weights, as compared with single intraperitoneal application. A new method of intraperitoneally applying therapeutic substances was introduced by Reymond et al. [25]. A micropump connected to a trocar allows microdroplets of di€erent therapeutic agents to be vaporized during laparoscopy. Early results in pigs showed that bowel resection is feasible without smog formation. Additionally, general distribution of the substances to all exposed intraperitoneal surfaces was demonstrated. Further clinical investigations are needed to prove the feasibility of this new device during laparoscopic procedures in patients. Besides additive intraperitoneal instillation, local treatment of port sites with tumoricidal agents has been reported to reduce metastatic tumor growth in laparoscopic experimental models [3]. Local application of 1% silver sulfadazine or 10% povidone-iodine at the trocar sites signi®cantly reduced metastatic tumor growth in a hamster model. Nevertheless, tumor incidence was still higher (75% after silver sulfadazine and 78% after povidone-iodine treatment) than in the control group (93%). The combination of di€erent protective local measures (trocar ®xation, prevention of gas leakages, rinsing of instruments with povidone-iodine, minilaparotomy protection, rinsing of wounds with povidone-iodine) was investigated in a tumor xenograft model in pigs [26]. After injection of 107 human HeLa cells, the pigs underwent laparoscopic sigmoid resection either with protective measures or without additional procedures. Whereas port-site metastases developed in 63.8% of the animals in the control group, only 13.8% of all the animals with protective measures had tumor growth at the trocars. Conclusion Laparoscopic procedures in cancer patients and the development of port-site metastases have raised substantial questions of general importance for oncologic surgery. Although the problem of port-site metastases is related mainly to the surgeon, the technique and manipulation of the tumor-bearing organ as well as some other factors related to laparoscopy itself have been demonstrated to in¯uence tumor growth. The possible stimulation of tumor cell growth and the suppression of local immune defense by CO2, as shown in many experimental studies, can be avoided by the alternative use of helium also in clinical trials. New therapeutic strategies, including instillation of cytotoxic and immune modulating agents in combination with laparoscopy, were reported to inhibit tumor growth strongly in experimental investigations. Nevertheless, perioperative pathophysiologic and immunologic changes caused by either open or laparoscopic procedures must be further evaluated to gain a better understanding of how surgical approach (i.e., open vs laparoscopic) e€ects postoperative tumor growth. References 1. Bouvy ND, Giu€rida MC, Tseng LN, Steyerberg EW, Marquet RL, Jeekel H, Bonjer HJ (1998) (E€ects of carbon dioxide pneumoperitoneum, air pneumoperitoneum, and gasless laparoscopy on body weight and tumor growth. Arch Surg 133: 652±656 2. Fingerhut A (1996) Laparoscopic-assisted colonic resection: the French experience. In: Jaeger R, Wexner SD (eds) Laparoscopic colorectal surgery. Churchill Livingstone New York, pp 253±257 3. Fleshman JW (1999) Tumor implantations in trocar sites in hamsters using a human colon cancer model (GW-39). Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA 4. Geûman T, Kim ZG, Gutt CN (1999) Tumor growth after CO2 versus helium pneumoperitoneum in a laparoscopic liver metastases model. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA 5. Goldstein DS, Lu ML, Hattori T (1993) (Inhibition of peritoneal tumor cell implantation: model for laparoscopic cancer surgery. J Endourol 7: 237±241 6. Hewett P, Texler M, Anderson D, Chatterton B (1999) The movement of malignant cells within the peritoneal cavity during open and laparoscopic colectomy: an animal model. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA 7. Hewett P, Texler M, Hardingham (1999) Haematogenous spread of malignant cells during laparoscopic colectomy: are animal models e€ective? Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA 8. Jacobi CA, Ordemann J, Bohm B, Zieren HU, Sabat R, Muller JM (1997) (Inhibition of peritoneal tumor cell growth and implantation in laparoscopic surgery in a rat model. Am J Surg 174: 359±363 9. Jacobi CA, Sabat R, BoÈhm B, Zieren HU, Volk HD, MuÈller JM (1997) (Pneumoperitoneum with carbon dioxide stimulates growth of malignant colonic cells. Surgery 121: 72±78 10. Jacobi CA, Wildbredt P, Braumann C, MuÈller JM (1999) In¯uence of intraperitoneal or intravenous application of taurolidine/ heparin in laparoscopic surgery on peritoneal tumor growth. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA 11. Jacobi CA, Wildbredt P, MuÈller JM (1999) In¯uence of gases and intraperitoneal instillation of di€erent agents in the prevention of metastases during laparoscopy. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA 12. Jones DB, Guo LW, Reinhard MK, Soper NJ, Philpott GW, Connet J, Fleshman JW (1995) (Impact of pneumoperitoneum on trocar site implantation of colon cancer in hamster model. Dis Colon Rectum 38: 1182±1188 13. Kim ZG, Gutt CN (1999) The impact of open versus laparoscopic small bowel resection in the presence of hepatic micrometastases: a pilot study. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA 14. Lauroy J, Chaumpault G, Risk N, Boutelier P (1994) (Metastatic recurrence at cannula site: should digestive cancers still be managed by laparoscopy? Br J Surg 81(Suppl): 31 15. Lee SW, Gleason C, Zhai C, Bessler M, Whelan RL (1999) The e€ects of wound painting and peritoneal irrigation with povidoneiodine solution on port wound recurrence following laparoscopic splenectomy in a murine model. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA 16. Lee SW, Gleason NR, Zhai C, Allendorf JDF, Huang EH, Bessler M, Whelan RL (1999) Increased platelet-derived growth factor (PDGF) release after laparotomy stimulates systemic tumor growth in mice. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA 17. Lee SW, Southall JC, Allendorf JD, Bessler M, Whelan RL (1998) Traumatic handling of the tumor independent of pneumoperitoneum increases port-site implantation rate of colon cancer in a murine model. Surg Endosc 12: 828±834 18. Neuhaus SJ, Ellis T, Jamieson GG, Watson DI (1999) An experimental study of intraperitoneal heparin on tumour implantation following laparoscopy. Data presented of the Third 445 19. 20. 21. 22. 23. 24. International Laparoscopic Physiology Conference, March 1999, New York, USA Neuhaus SJ, Watson DI, Ellis T, Dodd T, Rofe AM, Jamieson GG (1998)( Ecacy of cytotoxic agents for the prevention of laparoscopic port-site metastases. Arch Surg 133: 762±766 O'Sullivan R, Bouchier-Hayes D, Redmond HP (1999) Antiendotoxin therapy attenuates angiogenesis in the perioperative period. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA Paraskeva PA, Puttick MI, Rigg A, Lemoine N, Darzi A (1999) Viscous hyaluronate solution: a potential agent for the prevention of postoperative tumour cell invasion following laparoscopic surgery for malignant disease. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA Puttick MI, Jacklin R, Nudka CC, Vaughan NV, Gul Y, Yong LY, Darzi A (1999) Full laparotomy in a warm humid environment causes less postoperative tumour growth than full laparotomy in room air. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA Puttick MI, Rigg AS, Lemoine N, Nudka CC, Darzi A (1999) Carbon dioxide pneumoperitoneum increases tumour cell invasiveness. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA Puttick MI, Rigg AS, Paraskeva PA, Lemoine N, Nudka CC, Darzi A (1999) Expression of matrix-metalloproteinase genes is 25. 26. 27. 28. 29. modulated by the CO2 pneumoperitoneum: e€ect on tumour invasion and preliminary results. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA Reymond MA, Bo H, Hess J, KoÈckerling F, Morel PH (1999) A micropump allowing locoregional therapy during cancer laparoscopy. Data presented of the Third International Laparoscopic Physiology Conference (March 1999), New York, USA Schneider C, Reymond MA, Tannapfel A, Franklin ME, Morel PH, KoÈckerling F (1999) Surgical technique is a major factor in the incidence of port-site recurrences. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA Schwalbach P, Reinshagen S, Kuntz C, Glaser F, Herfarth C (1999) Potential in¯uence of laparoscopic gases (xenon, helium, air, CO2, N2O) on tumor cell growth. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA Tomita H, Marcello PW, Milsom JW (1999) Pneumoperitoneum does not enhance tumor growth and metastasis in the liver study in rat models: preliminary results. Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA Wittich P, Hazebroek EJ, Marquet RL, Bonjer HJ (1999) Problems arising with the cell seeding model? Data presented of the Third International Laparoscopic Physiology Conference, March 1999, New York, USA