Mechanism of lcoal disease failure Cancer Res-2005-Höckel-2997-3002

The Hydra Phenomenon of Cancer: Why Tumors Recur Locallyafter Microscopically Complete Resection

Michael Hockel and Nadja Dornhofer

Department of Obstetrics and Gynecology, University of Leipzig, Leipzig, Germany

Abstract

After surgical resection with microscopically clear margins,solid malignant tumors recur locally in up to 50%. Although theeffect of a local tumor recurrence on the overall survivalmay below in common cancers such as carcinoma of the breast orprostate, the affected patients suffer from exacerbated fear andthe burden of the secondary treatment. With some tumorentities such as carcinoma of the uterine cervix or carcinoma ofthe head and neck, a local recurrence indicates incurability inthe majority of cases. The pathomechanisms of local tumorspread and relapse formation are still unclear and compara-tively little research has been devoted to their elucidation.Through the analysis of clinical andmolecular data, we proposethe concept of two pathogenetically and prognosticallydifferent local relapse types (i) in situ recurrences that arisein the residual organ//organ system not involved in the surgeryfor the primary tumor and (ii) scar recurrences that develop atthe site of previous tumor resection. Whereas field cancer-ization, the monoclonal or multiclonal displacement of normalepithelium by a genetically altered but microscopically undis-tinguishable homologue, may explain the origin of in siturecurrences, most scar recurrences are regarded as the result ofthe interaction of minimal residual microscopically occultcancer with the surgical wound environment inside a develop-mentally defined tissue or organ compartment. The therapeuticimplications derived from these concepts and areas of futureresearch aimed to reduce local relapses are discussed in thisperspective. (Cancer Res 2005; 65(8): 2997-3002)

Clinical Problem of Local Recurrences after SurgicalResection

According to the current WHO world cancer report, cancercontinues to be the second most important cause of death indeveloped countries killing 23% of their population (1). With manycommon cancers such as carcinoma of the breast and prostate,mortality from malignant disease is mainly due to the formation ofmetastases, and much effort is currently spent on investigatingdistant cancer dissemination. However, with some tumor entitiessuch as cancer of the uterine cervix or carcinoma of the head andneck a considerable number of patients die from uncontrolledprimary and recurrent local disease. Even if the effect on overallsurvival may be low, the occurrence of local tumor relapse increasesanxiety and the burden of treatment in the affected patient. Yet,research addressing the pathomechanisms of local tumor spreadand local relapse formation has been comparably sparse.

Most solid malignant neoplasms are treated with surgicalresection for local control and microscopically negative surgicalmargins after tumor removal (R0 resection) are crucial for theprevention of local recurrence. Nevertheless, local recurrence ratesafter wide surgical resection with microscopically clear marginsrange between 5% and 50% without adjuvant treatment (Table 1).(Neo)adjuvant radiation or chemoradiation may decrease theprobability of local recurrences after surgical resection and istherefore part of a patients individual treatment plan if the tumorexhibits specific histologic high risk factors. However, additionalmorbidity may result from multimodal local treatment.

Classification of Post-Surgical Local Relapses

Local recurrences develop at the site of the primary surgery/surgical scar (scar recurrences) or at some distance from thislocation in the residual organ that remained in situ after resection(in situ recurrences ; Fig. 1). Discrimination between these two typesof local recurrences is important as their different prognosis pointsto differences in their pathogenesis. A much inferior outcome forpatients with scar recurrences comparedwith in situ recurrences hasbeen shown with breast and vulvar cancer (24).The pathogenesis of in situ recurrences can be explained by the

phenomenon of field cancerization . To understand the pathogenesisof scar recurrences after microscopically complete tumor resection(R0), we need to look at the mechanisms of local tumor spread andthe interaction of minimal residual cancer with the surgical wound.

Field Cancerization

First proposed for oral cancer by Slaughter et al. (5), fieldcancerization describes clinically occult multifocal preneoplasticlesions of the epithelium within an anatomic region exposed to thesame carcinogen(s) [e.g., cigarette smoking, human papillomavirus(HPV) infection]. These lesions may not be apparent at histopath-ologic investigation but can be detected with molecular analyses forphenotypic or genetic alterations associated with carcinogenesissuch as p53 gene mutations, integrated viral DNA, loss ofheterozygosity, and microsatellite instability. Field cancerizationhas been described for lung, esophagus, vulva, cervix, anus, colon,breast, bladder, and skin in addition to the oral cavity, pharynx, andlarynx (6, 7).Both, monoclonal and polyclonal lesions have been shown in

those fields with genetically altered cells by X chromosomeinactivation analysis and comparison of distinct gene mutations.Polyclonality is explained by multiple genetic lesions producedindependently from each other by the same carcinogenic localenvironment. Monoclonal fields result from the lateral expansion ofa patch formed by a genetically altered stem cell exhibiting asignificant growth advantage over the neighboring stem cells.Cohesive lateral migration of these cells gradually displaces thenormal epithelium.

Requests for reprints: Michael Hockel, MD PhD, Professor and Chairman,Department of Obstetrics and Gynecology, University of Leipzig, Philipp-Rosenthal-Str. 55, 04103F, Leipzig, Germany. Phone: 49-341-9723400; E-mail:[email protected].

www.aacrjournals.org 2997 Cancer Res 2005; 65: (8). April 15, 2005

Perspectives in Cancer Research

Research. on January 9, 2015. 2005 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/M. Abraham Kuriakose

M. Abraham Kuriakose












After the complete resection of a carcinoma preserving part of theorgan in which it developed, microscopically normal but geneticallyaltered epithelium may remain in situ and acquire additionalmutations or epigenetic alterations that can initiate the develop-ment of a second tumor of the same or a different histologic type,representing an in situ recurrence. If the resection margin of thetumor operation is locatedwithin this genetically altered epithelium,a scar recurrence could also be a consequence of field cancerization(Fig. 1). However, the pathomechanism involving minimal residualcancer that will be described below seems to be more frequent (8).

Mechanisms of Local Tumor Spread

Local spread of a solid malignant tumor can be followed at threelevels of magnitude: macroscopic/microscopic/occult . Occult tumorspread can be traced with PCR-based molecular probes for specifictumor cellderived DNA and RNA sequences. Alternatively, occulttumor cells may be highlighted by immunocytochemistry albeit witha much lower detection rate. At present, only a few studies haveassessed occult local tumor propagation whereas numerousinvestigations have been devoted to the detection of regional andsystemic minimal residual disease (9). By identifying p53 mutationsin individual head and neck cancers, Brennan et al. showed thepresence of tumor cells at the surgical resection margins micro-scopically free of disease in 50% of the cases (10). Similar findingswere recently reported for nonsmall cell lung cancer (11).As many carcinomas of the uterine cervix harbor integrated HPV

16 and 18 sequences in their DNA, the sequence information of the5V and 3V viral-cellular junction loci can be used as the basis forindividual PCR assays to detect microscopically occult cervicalcancer (12). We have applied integrate-specific PCR in specimens ofanatomically precisely defined locations remote from the tumorborder in patients with cervical carcinoma treated surgically.Whereas a low background of cancer DNA was found in nearly allinvestigated pelvic and extrapelvic locations, most probably as aconsequence of hematogenous, lymphatic and i.p. tumor celldissemination, significantly more often tumor DNA was identifiedwithin an anatomic compartment comprised of the cervix,proximal vagina, and their posterolateral supply and supportstructures with a higher probability closer to the microscopictumor border than remote from it (13). This cervicovaginalcompartment can be regarded as the adult morphogenetic unitarising in the female individual from the distant segment of thebilateral paramesonephric (Mullerian) ducts under the influence ofthe HOXA-13 gene (14). From magnetic resonance imaging studies,

Table 1. Local recurrence rates (LRR) of selected tumor entities after surgical treatment with microscopically clear margins(R0) without adjuvant radiation

Tumor entity LRR (%) Remarks References

Head and neck carcinoma 4-50 (35, 10)Lung carcinoma 6-19 Regional recurrences included (36, 37)Breast carcinoma 22-39 After breast conserving surgery (38, 39)Gastric carcinoma 20-33 (40, 41)Rectal carcinoma 10-27 Regional recurrences included (42, 43)Cervical carcinoma 10-25 Regional recurrences included (44, 45)Vulvar carcinoma 16-19 (46, 47)Soft tissue sarcoma 8-36 (48)

Figure 1. Proposed pathomechanisms of local tumor spread andlocal relapse formation. Three adjacent anatomical compartments aredesignated C1, C2, C3. Both C2 and C3 compartments exhibitgenetically altered epithelia (GAE ) representing field cancerization dueto a common carcinogen. A, carcinoma (CA ) that developed from thegenetically altered epithelium of C2. Carcinoma spreads locally byproliferation, collective and single cell migration within the developmentallydefined compartment of its origin (C2), which provides the adequate positionalinformation. Surgical-wide resection erased the tumor with microscopicallyclear (R0) margins (B ). Minimal residual cancer interacting with the surgicalwound can lead to a scar recurrence (SR ). Progression from geneticallyaltered epithelium remote from the site of the resection can result in an in siturecurrence (ISR ) in C2 and in another primary cancer (P) in C3. At theintersection of the surgical field with the genetically altered epithelia in theresidual C2 and in C3 all three pathomechanisms are possible (X). Wideresection unnecessarily removed parts of C1.

Cancer Research

Cancer Res 2005; 65: (8). April 15, 2005 2998 www.aacrjournals.org











the histopathologic investigation of surgical specimens andmolecular mapping local cervical cancer spread can be locatedwithin the cervicovaginal compartment over a long time inmalignant progression. Only late in that process, the uterinecorpus and the adjacent visceral compartments such as the bladdercompartment anteriorly and rectum compartment posteriorly maybe invaded by the tumor, although these tissues are just a fewmillimeters apart from the site of tumor origin and no fascia orcondensed fibrous tissue layer separates them. The observationthat tumor cells migrate and proliferate within a developmentallydefined tissue or organ compartment is the same in vulvar,endometrium, ovarian, and colorectal carcinoma and could bea general principle. What makes local tumor spread remainconfined to the morphogenetic unit for a relatively long periodof time?Local tumor spread takes place by the collective migration of

cancer cells or by migrating individual tumor cells (15). There isevidence that individualmigrating cancer cells separate from the cellcollective by epithelial-mesenchymal transition (EMT), a highlyconserved process of morphogenesis in multicellular organisms(15, 16). The loss of the epithelial phenotype in the cancer cell often isa consequence of E-cadherin lack of function that may result frommutation or more frequently from epigenetic down-regulationthrough transcription repression or promoter hypermethylation(16). Several transcriptional repressors of E-cadherin such as snail,slug, and snip 1 have been found to be expressed with morphoge-netic processes in the embryo as well as in highly invasive cancer celllines. An important upstream signaling mechanism is the binding ofhepatocyte growth factor (scatter factor) to the Met receptor.Overexpression of Met in cancer cells can be induced by hypoxicconditions and may represent a clinically significant event inhypoxia-associated tumor progression and invasion (17, 18).Similar signaling mechanisms resulting in E-cadherin down-regulation and EMT are activated by the binding of semaphorinsto their receptors (plexins) both of which are structurally related toMet (19). Likewise, binding of several other structurally unrelatedgrowth factors to their ligands including insulin-like growth factor,epithelial cell growth factor, and the ErbB family, fibroblast growthfactors and transforming growth factor h (TGFh) may induce EMT.These cascades also orchestrate other molecular mechanismsessential for cell migration, particularly the synthesis, affinity, andavidity of integrins and the activity of matrix metalloproteinases(MMP). Integrins form the contacts to the extracellular matrix(ECM) and MMPs focally degrade and remodel ECM during cellmigration. These processes are interrelated by the integrins abilityto mediate cellular inside-out and outside-in signaling and toactivate and localize MMPs (20).We hypothesize that migrating tumor cells, either individual or as

members of a moving collective, respect positional information ofthe local environment until late in malignant progression. Theirmigration, survival, and proliferation is actively guided by tissue-specific contacts and signals rather than following paths of lowmechanical resistance. Positional information is generated by thegene expression of the local fibroblasts, blood, and lymphaticvessels and is presented through the corresponding ECM andvarious molecular signals (21, 22). Homologous positional informa-tion is retained in adult anatomic structures derived froma common precursor tissue (anlage) during embryologic and fetaldevelopment defining a morphogenetic unit. The integrin expres-sion on the surface of the migrating tumor cells may fit to thecompartment-specific ECM ligands preventing integrin-mediated

apoptosis (23). Tumor cell migration initiated within the tumormicroenvironment by the various molecular mechanisms describedabove may be maintained beyond the tumor border by compart-ment specific promigratory signals, such as local semaphorins andplexins (19). Only late in malignant progression when tumor cellshave gained migrational plasticity (15) and apoptotic resistance(24), these permissive effects of the compartment specificenvironment are no longer relevant for local tumor propagation.

Interaction between Minimal Residual Cancer andSurgical Wounds

The interaction of occult tumor cells and surgical wounds mayhappen locally at the site of tumor resection as well as remote fromit at the sites of surgical access to the body cavities orcompartments. If the migration of tumor cells detached from theprimary neoplasm is guided by positional information expressedwithin the developmentally defined morphogenetic unit, then theprobability of occult tumor cells being present within or close tothe local surgical wound created during the process of tumorresection should be much higher than the probability of theirpresence in remote wounds. Moreover, wound-associated activatedblood vessels within the compartment may favor a reentry ofcirculating tumor (stem) cells into the compartment furtherincreasing the probability of minimal residual disease within thelocal wounds (25). Even lymphatic and hematogenous tumor celldissemination as well as contamination will probably contribute toa higher load of minimal residual disease at the site of the localcompared with the remote surgically produced tissue injury due todilution effects. Consequently, recurrences in the local surgical scararea should be found much more often than remote scarrecurrences that certainly reflects the clinical experience. Recur-rences after tumor resection with microscopically free margins (R0)occur locally at the site of surgical resection in up to 50%, whereasrecurrences in the distant surgical scar are reported in only 1% to2% (26). Once exposed to the wound environment, neoplastic cellsreceive various stimuli and experience conditions favoring theformation of a tumor relapse. Table 2 gives a survey of proposedmechanisms of the interaction between minimal residual cancerand the surgical wound. Depending on the different phases of thehealing process cancer cells may be recruited, replicated, andselected at the site of the surgical wound (2731).Although the local immunomodulating effects of the surgical

wound during the different phases of healing may be complex, theestablished systemic immunosuppressive consequences of surgeryshould support the relapse formation. It can be assumed thatmost of the wound-associated tumor promoting effects are relatedto the wound volume, especially to the extent of ischemia within awound. Secondary healing wounds should therefore be moreprone to support scar recurrence formation than wounds healingper primam .

Therapeutic Implications

Following the arguments of the preceding chapters the currentprinciples of surgical oncology have to be revisited. A radical tumoroperation must remove not only the macroscopic and microscopictumor but also a maximum of the microscopically occult localcancer with a minimum of tissue trauma. The correspondingsurgical anatomy has to be deduced from embryologic and fetaldevelopment defining the morphogenetic units in the adult.Conventional radical tumor operations based on the Halsted

Local Tumor Spread and Recurrence










principles may not adequately address these criteria being tootraumatic due to unnecessary removal of tissue not infiltrated byoccult cancer on one hand and not sufficiently radical with respectto the eradication of occult local cancer on the other hand (Fig. 1).Both, total mesorectal excision for rectal cancer and total

mesometrial resection for cervical cancer are successful examplesof the new concept of radical operations based on surgical anatomyfrom the developmental perspective (32, 33). The surgical treatmentalternative for a radical tumor operation as defined above is a wideexcision , again with minimal trauma, supplemented by adjuvant(chemo)radiation in the presence of high risk factors. The treatmentyielding the best therapeutic index should be chosen for theindividual patient.With respect to the concepts introduced herein the design of

radiation fields for adjuvant radiotherapymay have to bemodified aswell. Complex geometric target volumes addressing the topographyof the morphogenetic units can be generated with the newtechnology of intensity modulated radiation therapy and tailoredfor the individual patient.In case of an insufficiently radical primary operation, serious

considerations should be given to the indication of reoperationswith the goal to resect the perifocal scar tissue (34) eventuallyalong with remnants of the corresponding morphogenetic unit.

Future Research

The investigation of gene expression profiles associated withthe positional identity of cells may open an exciting new field,molecular topographical anatomy. Do mesenchymal cells of amorphogenetic unit that represents the adult anatomic com-partment resulting from a certain anlage share a commontopographical differentiation fixed in a defined gene expressionprofile as a positional memory? Can it be discriminated fromthe transcriptional pattern of corresponding cells from adjacentcompartments? The transcriptome characterizing the topograph-ical identity of fibroblasts may be reduced to the HOX genecode according to the suggestion of Chang et al. (21). Based onthe HOX code of fibroblasts a topographical mapping of tissuesmay be deduced which might be important for the prediction oflocal tumor spread.The identification of those molecular substrates guiding tumor

cell migration within the permissive environment would be thenext step. As the permissive environment for local tumorpropagation seems to be topographically confined to themorphogenetic unit for a relatively long period in malignantprogression a response of cancer cells to positional informationpresented by the mesenchymal cells of this compartment can beexpected.

Table 2. Proposed mechanism of interaction between minimal residual cancer and the surgical wound at the site of tumorresection

Cancer Research










Whereasmuchwork has been devoted to the detection ofminimalresidual disease with regard to the prediction and prevention ofmetastases, research on local occult cancer spread has been sparse.With advanced biotechnology, molecular probes to highlightmicroscopically occult cancer are available now for a variety ofsolidmalignancies. Themain questions that should be addressed arethe following: What is the exact topographical distribution of occultcancer cells adjacent to the macroscopic tumor mass and its micro-scopic border with respect to the site of tumor origin, resectionmargins, and tissues remaining in situ? What is the prognosticsignificance of local minimal residual disease? Is molecular stagingof the resection margins appropriate for the indication of additionallocal treatment (reresection or adjuvant radiation)?Understanding the multifaceted interactions between surgical

wounds and minimal residual disease at the site of tumor resectionseems to be particularly rewarding as the wound is a topograph-ically and histologically defined environment suitable for variousstrategies of therapeutic intervention with the goal to erase aminimal tumor load. Of course, all strategies of wound-directedantitumor treatment have to respect side effects caused by theirinterference with healing in terms of restoration of tissue integrityand tensile strength.In summary, with this perspective we want to alert researchers

to the clinical importance of local relapse formation after

seemingly adequate surgical treatment (i.e., resection with clearmargins). In the light of recent clinical and molecular data, thepicture seems much more intricate than its mechanistic perceptionassuming that local tumor spread follows spaces of leastmechanical resistance and that local recurrences are simply dueto undetected macroscopic or microscopic tumor left behind aftersurgical treatment. We propose that local tumor spread is guidedby positional information presented molecularly in the develop-mentally defined anatomic compartment of the tumor origin for arelatively long period during malignant progression. Microscopi-cally occult minimal residual cancer recruited, expanded andselected by an ischemic surgical wound could be an important butpotentially avoidable cause of local recurrence. Research focusedon the mechanisms of local cancer spread and the interaction ofcancer cells with the wound environment may translate intosignificant clinical progress.

Acknowledgments

Received 10/27/2004; revised 1/17/2005; accepted 2/1/2005.Grant support: Fresenius-Kabi Deutschland GmbH (M. Hockel) and Else Kroner-

Fresenius-Foundation (N. Dornhofer).The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

References1. World Health Organisation. Steward BW, Kleihues P,editors. World Cancer Report. Lyon: IARC Press; 2003.

2. Huang E, Buchholz TA, Meric F, et al. Classifyinglocal disease recurrences after breast conservationtherapy based on location and histology. Cancer 2002;95:205967.

3. Vicini FA, Kestin L, Huang R, Martinez A. Does localrecurrence affect the rate of distant metastases andsurvival in patients with early-stage breast carcinomatreated with breast-conserving therapy? Cancer 2003;97:9109.

4. Rouzier R, Haddad B, Plantier F, Dubois P, Pelisse M,Paniel B-J. Local relapse in patients treated forsquamous cell vulvar carcinoma: incidence and prog-nostic value. Obstet Gynecol 2002;100:115967.

5. Slaughter DP, Southwick HW, Smejkal W. Fieldcancerization in oral stratified squamous epithelium.Cancer 1953;6:9638.

6. Garcia SB, Park HS, Novelli M, Wright NA. Fieldcancerization, clonality, and epithelial stem cells: thespread of mutated clones in epithelial sheets. J Pathol1999;187:6181.

7. Braakhuis BJM, Tabor MP, Kummer JA, Leemans CR,Brakenhoff RH. A genetic explanation of Slaughtersconcept of field cancerization: evidence and clinicalimplications. Cancer Res 2003;63:172730.

8. Partridge M, Li S-R, Pateromichelakis S, et al.Detection of minimal residual cancer to investigatewhy oral tumors recur despite seemingly adequatetreatment. Clin Cancer Res 2000;6:271825.

9. Pantel K, Cote RJ, Fodstad O. Detection and clinicalimportance of micrometastatic disease. J Natl CancerInst 1999;91:111324.

10. Brennan JA, Mao L, Hruban RH, et al. Molecularassessment of histopathological staging in squamous-cell carcinoma of the head and neck. N Engl J Med 1995;332:42935.

11. Jassem J, Jassem E, Jakobkiewicz-Banecka J, et al.P53 and K-ras mutations are frequent events inmicroscopically negative surgical margins frompatients with nonsmall cell lung carcinoma. Cancer2004;100:195160.

12. Luft M, Klaes R, Nees M, et al. Detection of integratedpapillomavirus sequences by ligation-mediated PCR

(DIPS-PCR) and molecular characterization in cervicalcancer cells. Int J Cancer 2001;92:917.

13. Einenkel J, Vinokurova S, Ziegert C, Horn L-C,von Knebel Doeberitz M, Hockel M. Occult localtumour spread in cervical cancer patients. Int J GynecolCancer 2004;14 Suppl 1:35.

14. Ma L, Benson GV, Lim H, Dey SK, Maas RL. AbdominalB (AbdB) Hoxa genes: regulation in adult uterus byestrogen and progesterone and repression in Mullerianduct by the synthetic estrogen diethylstibestrol (DES).Dev Biol 1998;197:14154.

15. Friedl P, Wolf K. Tumour-cell invasion and migration:diversity and escape mechanisms. Nat Rev Cancer 2003;3:36274.

16. Thiery JP. Epithelial-mesenchymal transitions intumour progression. Nat Rev Cancer 2002;2:44254.

17. Pennacchietti S, Michieli P, Galluzzo M, Mazzone M,Giordano S, Comoglio PM. Hypoxia promotes invasivegrowth by transcriptional activation of the metprotooncogene. Cancer Cell 2003;3:34761.

18. Hockel M, Schlenger K, Aral B, Mitze M, Schaffer U,Vaupel P. Association between tumor tissue hypoxiaand malignant progression in advanced cancer of theuterine cervix. Cancer Res 1996;56:450915.

19. Trusolino L, Comoglio PM. Scatter-factor andsemaphorin receptors: cell signalling for invasivegrowth. Nat Rev Cancer 2002;2:289300.

20. Hood JD, Cheresh DA. Role of integrins incell invasion and migration. Nat Rev Cancer 2002;2:91100.

21. Chang HY, Chi J-T, Dudoit S, et al. Diversity,topographic differentiation, and positional memory inhuman fibroblasts. PNAS 2002;99:1287782.

22. Cleaver O, Melton DA. Endothelial signaling duringdevelopment. Nat Med 2003;9:6618.

23. Stupack DG, Puente XS, Boutsaboualoy S, StorgardCM, Cheresh DA. Apoptosis of adherent cells byrecruitment of caspase-8 to unligated integrins. J CellBiol 2001;155:45970.

24. Graeber TG, Osmanian C, Jacks T, et al.Hypoxia-mediated selection of cells with diminishedapoptotic potential in solid tumors. Nature 1996;379:8891.

25. Ceradini DJ, Kulkarni AR, Callaghan MJ, et al.Article title: Progenitor cell trafficking is regulated byhypoxic gradients through HIF-1 induction of SDF-1.

Nat Med [online publication]. 2004 [cited 2004 July1-7]. Available from: http://www.nature.com/naturemedicine/.

26. Tannapfel A, Wittekind C. Definition of port-site andwound recurrences in cancer surgery. In: Reymond MA,Bonjer HJ, Kockerling F, editors. Port-site and woundrecurrences in cancer surgery. Berlin, Heidelberg:Springer; 2000. p. 15.

27. Balkwill F. Cancer and the chemokine network. NatRev Cancer 2004;4:54050.

28. Reid SE, Kaufman MW, Murthy S, Scanlon EF.Perioperative stimulation of residual cancer cellspromotes local and distant recurrence of breast cancer.J Am Coll Surg 1997;185:290306.

29. Tagliabue E, Agresti R, Carcangiu ML, et al. Role ofHER2 in wound-induced breast carcinoma prolifera-tion. Lancet 2003;362:52733.

30. Abramovitch R, Marikovsky M, Meir G, Neeman M.Stimulation of tumour angiogenesis by proximalwounds: spatial and temporal analysis by MRI. Br JCancer 1998;77:4407.

31. Balkwill F, Mantovani A. Inflammation and cancer:back to Virchow? Lancet 2001;357:53945.

32. Heald RJ, Husband EM, Ryall RDH. The mesorectumin rectal cancer surgery: the clue to pelvic recurrence?Br J Surg 1982;69:6136.

33. Hockel M, Horn L-C, Hentschel B, Hockel S,Naumann G. Total mesometrial resection: high reso-lution nerve-sparing radical hysterectomy based ondevelopmentally defined surgical anatomy. Int JGynecol Cancer 2003;13:791803.

34. Lewis JJ, Leung D, Espat J, Woodruff JM, BrennanMF. Effect of reresection in extremely soft tissuesarcoma. Ann Surg 2000;231:65563.

35. Slootweg PJ, Hordijk GJ, Schade Y, van Es RJJ,Koole R. Treatment failure and margin status in headand neck cancer. A critical view on the potentialvalue of molecular pathology*1. Oral Oncology 2002;38:5003.

36. Martini N, Bains MS, Burt ME, et al. Incidence oflocal recurrence and second primary tumors inresected stage I lung cancer. J Thorac Cardiovasc Surg1995;109:1209.

37. Harpole DH Jr, Herndon JE, 2nd, Wolfe WG,Iglehart JD, Marks JR. A prognostic model of recur-rence and death in stage I non-small cell lung cancer

Local Tumor Spread and Recurrence



http://cancerres.aacrjournals.org/

utilizing presentation, histopathology, and oncoproteinexpression. Cancer Res 1995;55:516.

38. Veronesi U, Marubini E, Mariani L, et al. Radiother-apy after breast-conserving surgery in small breastcarcinoma: long-term results of a radomized trial. AnnOncol 2001;12:9971003.

39. Fisher B, Anderson S, Bryant J, et al. Twenty-yearfollow-up of a randomized trial comparing totalmastectomy, lumpectomy, and lumpectomy plus irra-diation for the treatment of invasive breast cancer.N Engl J Med 2002;347:123341.

40. Cirera L, Balil A, Batiste-Alentorn E, et al. Random-ized clinical trial of adjuvant mitomycin plus tegafur inpatients with resected stage III gastric cancer. J ClinOncol 1999;17:38105.

41. Jeong JY, Kim YJ, Han JK, et al. Palliation ofanastomotic obstructions in recurrent gastric carcino-ma with the use of covered metallic stents: clinicalresults in 25 patients. Surgery 2004;135:1717.

42. Kockerling F, Reymond MA, Altendorf-Hofmann A,Dworak O, Hohenberger W. Influence of surgery onmetachronous distant metastases and survival in rectalcancer. J Clin Oncol 1998;16:3249.

43. Secco GB, Fardelli R, Rovida S, et al. Is intensivefollow-up really able to improve prognosis of patientswith local recurrence after curative surgery for rectalcancer? Ann Surg Oncol 2000;7:327.

44. Landoni F, Maneo A, Colombo A, et al. Randomisedstudy of radical surgery versus radiotherapy for stageIb-IIa cervical cancer. Lancet 1997;350:53540.

45. Kinney WK, Alvarez RD, Reid GC, et al. Value ofadjuvant whole-pelvis irradiation after wertheim hys-terectomy for early-stage squamous carcinoma of thecervix with pelvic nodal metastasis: a matched-controlstudy. Gynecol Oncol 1989;34:25862.

46. Piura B, Masotina A, Murdoch J, Lopes A, Morgan P,Monaghan J. Recurrent squamous cell carcinoma of thevulva: a study of 73 case. Gynecol Oncol 1993;48:18995.

47. Maggino T, Landoni F, Sartori E, et al. Patterns ofrecurrence in patients with squamous cell carcinomaof the vulva. A multicenter CTF study. Cancer 2000;89:11622.

48. Karakousis CP, Driscoll DL. Treatment and localcontrol of primary extremity soft tissue sarcomas. J SurgOncol 1999;71:15561.

Cancer Research



http://cancerres.aacrjournals.org/

2005;65:2997-3002. Cancer Res Michael Hckel and Nadja Dornhfer Locally after Microscopically Complete ResectionThe Hydra Phenomenon of Cancer: Why Tumors Recur

Updated version

http://cancerres.aacrjournals.org/content/65/8/2997

Access the most recent version of this article at:

Cited Articles

http://cancerres.aacrjournals.org/content/65/8/2997.full.html#ref-list-1

This article cites by 45 articles, 10 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/65/8/2997.full.html#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

SubscriptionsReprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

[email protected] at

To request permission to re-use all or part of this article, contact the AACR Publications


http://cancerres.aacrjournals.org/content/65/8/2997http://cancerres.aacrjournals.org/content/65/8/2997.full.html#ref-list-1http://cancerres.aacrjournals.org/content/65/8/2997.full.html#related-urlshttp://cancerres.aacrjournals.org/cgi/alertsmailto:[email protected]:[email protected]://cancerres.aacrjournals.org/

Documents

Mechanism of lcoal disease failure Cancer Res-2005-Höckel-2997-3002