Familial Cancer Syndromes
Fig. 14-3). Linkage studies have successfully identified the majority of familial cancer syndromes through DNA analysis of large families with affected individuals, linking disease phenotypes or cancer subtypes to regions of the genome, with candidate genes subsequently sequenced for a causative mutation. These linkage studies have identified the rare, highly penetrant cancers of hereditary cancer syndromes that account for about 5% to 10% of reported cancers (Table 14-2).12 The more common, but lower-penetrant variants of genetic-, nonsyndrome-related cancers are typically found through large, genome-wide analysis of unrelated persons with common cancers. In genome-wide association studies (GWAS) the entire genome is sequenced and single alterations, or single-nucleotide polymorphisms (SNPs), are used to delineate increased risks of cancer related to genetic risks.13The discovery of inherited mutations of genes associated with an increased risk of cancer provides important opportunities for early detection and prevention of common and rare forms of human malignancies. Genetically related cancer is a spectrum, ranging from common variants with low penetrance to rare variants with either moderate or high penetrance (
In both high and low penetrant genetically related cancers a high clinical suspicion is necessary for diagnosis and to formulate appropriate screening and treatment plans. To aid the clinician, tools have been developed to obtain an accurate family history, thereby assessing the risk of genetically related cancers (Table 14-3).14 If the clinician has a reasonable suspicion that a patient may be either at risk for or has a genetically related cancer, further follow-up with a dedicated genetic counselor is recommend.15 We herein summarize the five most prevalent familial cancer syndromes with regard to genetic mechanism and diagnosis.
Hereditary Nonpolyposis Colon Cancer
Hereditary nonpolyposis colon cancer (HNPCC), previously known as Lynch syndrome, is the most prevalent familial cancer syndrome and is responsible for 2% to 5% of all colorectal cancer.16 An autosomal dominant syndrome, HNPCC is caused by a germline mutation in one of six genes: MLH1, MSH2, MSH3, MSH6, PMS1, or PMS2.17–21 All six of the genes function in DNA mismatch repair and between 45% and 70% of HNPCC families have mutations in the most common gene variants: MLH1, MSH2, or MSH6. Patients with mutations in mismatch repair genes exhibit microsatellite instability, with errors in replication of highly repetitive sequences unable to be repaired, resulting in alterations of the length of the total repeat sequence.
Figure 14-1. Trends in age-adjusted cancer incidence and death by gender, United States, 1975 to 2011. (From Siegel RL, Miller KD, Jemal A. Cancer statistics, 2015. Ca Cancer J Clin 2015;65:5–29.)
Figure 14-2. Trends in age-adjusted cancer incidence by gender and site, United States, 1975 to 2011. (From Siegel RL, Miller KD, Jemal A. Cancer statistics, 2015. Ca Cancer J Clin 2015;65:5–29.)
Figure 14-3. Genetic architecture by cancer risk.
Clinically, HNPCC is associated with right-sided colonic tumors with histopathology demonstrating poorly differentiated adenocarcinoma and signet ring features. Patients with HNPCC have a 50% to 80% lifetime risk of colorectal cancer, with a median age of diagnosis in the mid-40s. An increased risk of endometrial cancer, ovarian cancer, stomach cancer, small intestine cancer, ureteral cancer, and kidney cancer is also seen in HNPCC kindreds.22–24
Currently, testing for HNPCC is recommended for all newly diagnosed cases of colorectal cancer that fulfill the revised Bethesda guidelines, in families that meet the Amsterdam II criteria, in patients with endometrial cancer diagnosed before age 50, or in families with known HNPCC (Table 14-4).25,26
Table 14-2 Familial Cancer Syndromes
Table 14-3 Clinical Aid Suggesting the Presence of a Hereditary Cancer Disposition
Table 14-4 Amsterdam II and Revised Bethesda Guidelines for the Testing of Hereditary Nonpolyposis Colorectal Cancer
Familial Adenomatous Polyposis
Familial adenomatous polyposis (FAP) is a highly penetrant, autosomal dominant syndrome with an incidence between 1 in 5,000 and 1 in 10,000 and is responsible for approximately 1% of all colon cancer cases.27 Arising from a mutation in the antigen-presenting cell (APC) gene on chromosome 5q, nearly 75% of cases are due to familial germline mutations with the remainder secondary to first-generation de novo mutations.28
Clinically, FAP manifests with hundreds to thousands of adenomatous polyps at a young age with a resultant risk of colon cancer of 90% by age 45.29 In addition to colonic manifestations, duodenal and gastric polyps are also prevalent with a lifetime risk of duodenal cancer ranging from 5% to 12%, typically periampullary in location. Benign, extraintestinal manifestation of FAP includes desmoid tumors, mesenteric fibrosis, epidermoid cysts, osteomas, congenital retinal pigment epithelium, and dental anomalies and typically accompanying adenomatous polyp formation.29
Hereditary Breast and Ovarian Syndrome
Hereditary breast and ovarian syndrome (HBOC) occurs at an incidence rate of 1 in 500 or 1,000 and is inherited in an autosomal dominant fashion with a penetrance of nearly 85%.30,31 Germline mutations in BRCA1 (chromosome 17q) or BRCA2 (chromosome 13q) are responsible for approximately 60% of HBOC, with mutations in ATM, NBM, BRIP1, CHEK2, TP53, PTEN, and RAD51C responsible for the remaining cases.32,33 BRCA1 and BRCA2 are DNA damage repair genes and more than 1,000 variants have been identified with founder BRCA mutations documented in genetically isolated populations.34 In the United States, the most common founder mutations occur in Ashkenazi Jewish lineage with nearly 1 in 40 carrying the common BRCA1 or BRCA2 mutations.34 Genetic linkage studies of families with HBOC syndrome demonstrate that the lifetime risk of developing breast and ovarian cancers by age 70 is 56% and 17%, respectively.30,32
Familial Gastric Cancer
Hereditary diffuse gastric cancer (HDGC) is autosomal dominant in nature and is characterized by diffuse gastric cancer and lobular breast cancer with a nearly uniform penetrance and an 80% lifetime risk of gastric cancer.35 The average age of developing hereditary gastric cancer is 38 years. Thirty to 40% of families with HDGC have germline mutations in CDH1 at chromosome 16q, a gene producing the E-cadherin protein necessary for cell proliferation and adhesion within the stomach mucosa.36 Currently, the diagnosis of HDGC is suspected if a person or family meets any of the below criteria37:
At least two cases of stomach cancer in a family, with at least one being diffuse gastric cancer diagnosed before age 50
At least three cases of stomach cancer at any age
Diagnosis of diffuse gastric cancer before the age of 45
Diagnosis with both diffuse gastric cancer and lobular breast cancer
von Hippel–Lindau Disease
von Hippel–Lindau Disease (VHL) is an autosomal dominant syndrome characterized by the formation of hemangioblastomas of the brain, spinal cord, and retina.38 In addition, individuals with VHL disease may develop renal or pancreatic cysts and are at an increased risk for clear cell renal cell carcinoma, pheochromocytomas, and pancreatic neuroendocrine tumors. Arising from a germline mutation of the VHL gene on chromosome 3p, the protein product degrading hydroxylated hypoxia inducible factor-1α, nearly 80% of cases are familial.39 Unlike most autosomal dominant disorders, VHL disease subscribes to the Knudson “two-hit hypothesis” with mutation of both VHL alleles necessary for tumor or cyst formation.38
Environmental Risk Factors
Although the cause of most human cancers remains unidentified, the causal association between carcinogenesis and exposure to environmental risk factors is significant. The first known documented link between cancer and exposure to a carcinogen was documented in the treatises of the English surgeon Percivall Pott.40 In 1775, Pott documented the high incidence of scrotal cancer among chimney sweeps secondary to soot lodging in scrotal skin. Subsequent preventive measures aimed at increasing bathing and use of protective clothing led to a dramatic decline in the incidence of scrotal cancer. It was not until the 20th century that the active carcinogens in soot were shown to be polycyclic aromatic hydrocarbons.41 Table 14-5 lists the most common carcinogens associated with human cancers.
Table 14-5 Preventable Exposures Associated with Human Solid Organ Cancers
Tobacco and Cancer
Tobacco use is associated with more cancer-related deaths worldwide than any other environmental risk factor. There are now 19 cancers for which causal evidence exists between cancer formation and cigarette smoking: lung, oral cavity, nasopharynx, nasal cavity, larynx, esophagus, liver, stomach, colorectum, pancreas, kidney, ureter, urinary bladder, cervix, ovary, and bone marrow. Three cancers have known causality with smokeless tobacco use: oral cavity, esophagus, and pancreas.42 Table 14-6 details the relative risks by cancer site of tobacco smoking.
The causative carcinogens in tobacco products are broad and heterogeneous. In fact, the International Agency for Research on Cancer (IARC) working group has identified 72 carcinogens contained in cigarette smoke linked to cancer development.43 The underlying pathophysiology for smoking-related cancer is organ specific but likely dependent on the oxidative damage caused by the carcinogens producing free oxygen radicals and redox cycling.44
Viral Risk Factors
Nearly 15% of cancers or 1.3 million cancer cases worldwide can be attributed to viral infection.45 The search for human tumor viruses is accentuated by a long appreciation of viral-linked cancers in birds and rodents. Peyton Rous in 1911 demonstrated spindle cell sarcomas could be readily transmitted from diseased to healthy chickens using tumor cell infiltrates.46 This observation led to the identification of the Rous sarcoma virus (RSV), a member of the Retroviridae family, as causative agent responsible for Rous’ original observation decades earlier.46
The lack of convincing animal models confirming epidemiologic studies coupled with the lack of oncogene expression by human tumor–associated viruses has delayed recognition of virus-induced human cancers. Using both epidemiology and molecular biology analysis, six viruses are established as causative agents of cancer (Table 14-7). Hepatitis B and C viruses (HBV and HCV), HPV, and the human herpes virus 8 (HHV-8) (Kaposi sarcoma-associated herpes virus) will be discussed in further detail.
Table 14-6 Cancer Sites Associated with Tobacco Smoking by Relative Risk According to the International Agency for Research on Cancer Working Groups
Table 14-7 Human Tumor–Associated Viruses
Hepatitis B and Hepatitis C Viruses
HBV and HCV are hepatotropic viruses causing both acute and chronic viral hepatitis and are responsible for nearly 90% of hepatocellular carcinoma (HCC) cases diagnosed worldwide.47 HBV, a member of the Hepadnaviridae family, is a circular double-stranded DNA that incorporates into the cellular DNA of hepatocytes at random sites. Although the exact mechanism for induction of HBV-related HCC remains unknown due to the long latency period between viral infection and cancer appearance it is postulated to occur through either (1) induction of chronic liver cell injury leading to random chromosomal and genetic injury or (2) expression of X protein from the incorporated HBV activating the mitogen-activated protein kinase, c-jun N-terminal kinase, protein kinase C, phosphatidylinositol 3-kinase, protein kinase B, and JAK/STAT pathways.48–51
HCV is a single-stranded RNA virus from the Flaviviridae family, unique as the only human tumor virus utilizing RNA as genetic material. Responsible for the vast majority of HCC cases within the United States, HCV has six known genotypes and has a latency period from infection to HCC formation typically more than two decades. The actual pathophysiologic mechanism for HCV-related HCC development is unknown but like HBV-related HCC is thought to occur secondary to creating a state of chronic inflammation and regeneration within the liver parenchyma at the hepatocyte level.52,53
With the development of a HBV vaccine in the late 1970s the incidence of HBV-related HCC has decreased worldwide. Although there is no commercially available HCV vaccines, since 2010 there are orally bioavailable drugs that dramatically decrease the HCV viral load. However, it is unclear whether HCV treatment will impact HCC incidence rates as many patients have already progressed to cirrhosis prior to actual HCV diagnosis.
Human Papilloma Virus
HPV is a nonenveloped double-stranded DNA tumor virus that is responsible for a variety of epithelial disorders ranging from benign to malignant. Although consisting of over 100 known variants, HPV-16 and HPV-18 are designated high risk for tumor induction and are responsible for over 70% of known cervical and anal cancers.54,55 The molecular pathogenesis of HPV-related cancers is better understood than other viral-associated cancers and occurs following HPV infection of keratinocytes at the basal epithelial layer of the stratified squamous epithelium. Upon HPV integration into host DNA, two viral genes, E6 and E7, are expressed and their products bind to a p53 tumor suppressor gene (E6) and/or retinoblastoma suppressor protein (E7) to deregulate cell growth and inhibit apoptosis leading to the accumulation of mutations and the development of cancers following a length latency period.56,57 The recent development of quadrivalent and bivalent HPV vaccines is expected to have a significant impact on the development of HPV and subsequent cervical and anal cancer incidence.58
Human Herpes Virus 8
HHV-8, also known as Kaposi sarcoma-associated herpes virus, is a double-stranded DNA virus from the Herpesviridae family associated with the development of Kaposi sarcoma, primary effusion lymphoma, and multicentric Castleman disease. HHV-8 has a latent and lytic phase persisting in B-lymphocytes. During the latent phase HHV-8 evades the host immune system and has only minimal expression of gene products. However, in the lytic phase innumerable viral proteins are produced including viral G protein–coupled receptor, K1, v-cyclin, and v-Bcl-2 that modulates cell growth, induces vascular endothelial growth factor (VEGF) expression, and inhibits apoptosis, respectively, leading to induction of cancer.59,60
Dietary Risk Factors and Obesity
The most important impact of diet on the risk of cancer is mediated through body weight. The IARC working group on weight control and physical activity estimates in developed countries a body mass index over 25 kg/m2 accounts for approximately 39% of endometrial, 25% of kidney, 11% of colon, 9% of postmenopausal breast cancer, and 5% of total cancer incidence.66
The mechanisms linking obesity and cancer development are largely unknown but likely multifactorial. An example of a causal association between cancer and obesity is in postmenopausal breast cancer. A weight gain of 10 kg or more is associated with a significant increase in breast cancer incidence among women who have never used hormone replacement therapy.67 This is likely secondary to large increases in endogenous estrogen levels, also leading to increased incidence of endometrial cancer.68 The mechanism for obesity-related nonbreast or nonendometrial cancers remains unclear and is under clinical investigation.
Table 14-8 Human Cancers Associated with Dietary Factors and Nutrition
Control of the cell division cycle is central for governing when the cell should progress to DNA synthesis and proliferation versus growth arrest, DNA repair, or apoptosis. Cell division proceeds through a well-defined series of stages with tightly regulated and balanced processes dependent on oncogene and tumor suppressor gene expression (Fig. 14-4). When cells leave quiescence (G0), they enter a first gap phase (G1) where an explosion of growth factors and macromolecules is transcribed and translated allowing cells to divide but not lose overall size. Toward the end of G1, cells reach a restriction point governed by cell cycle checkpoint genes where thereafter they are now committed to division. It is at this restriction point where DNA repair or programmed cell death (apoptosis) occurs and where phosphorylation of the tumor suppressor product, RB protein, allows entry into the S phase. During the S phase, DNA is synthesized and progression to second gap phase (G2) follows. Within the middle of the G2 phase yet another restriction point or cell cycle checkpoint occurs prior to cell entry into mitosis (M), the actual cell division phase.69
Proto-oncogenes code for proteins that send signals to the cell nucleus promoting cell division, especially at the G1–S and G2–M transition points. Oncogenes are altered versions of proto-oncogenes also coding for signaling proteins but in a continuous fashion leading to incontrollable cell division and thus, tumor development. The conversion of proto-oncogenes into oncogenes occurs through three basic methods: (1) a mutation within a proto-oncogene producing an increase in protein activity (seen in the conversion of the Ras proto-oncogene), (2) an increase in the amount of protein within the cell resulting in amplified expression (i.e., c-MYC proto-oncogene), and (3) a chromosomal translocation where fusion proteins are produced or protein expression is altered (i.e., Philadelphia chromosome, BCR/ABL).70
Tumor suppressor genes also play a crucial role in the cell division cycle serving as a brake for division in response to DNA damage. The tumor suppressor genes, p53 and RB, play a critical role in maintaining the checkpoint at the G1–S transition point.71,72 Homozygous loss of p53 is found in 65% of colon cancers, 30% to 50% of breast cancers, and 50% of lung cancers. In addition, p53 germline mutations are associated with Li–Fraumeni syndrome, an autosomal dominant disorder associated with sarcomas, breast cancer, leukemia, and adrenal gland cancers.73
The fundamental tenet of the immune surveillance hypothesis, postulated by Thomas and Burnet nearly five decades ago, is that the underlying function of the immune system is to survey the human body, recognize and then eliminate tumors based on tumor antigen expression.74,75 Well documented in nonhuman animal models, the role of the immune system as a surveillance and treatment response in human malignancies is debatable and is best supported by tangential clinical evidence. In humans, cancer incidence rates correlate to advancing age presumably due to increased cell division and error rates in chromosomal replication over time that the immune system simply cannot overcome.
A corollary to the immune surveillance hypothesis is that immunodeficient individuals have an increased rate of cancer development.76 Epidemiologic studies of patients with heritable immunodeficiencies demonstrate mixed results for this hypothesis. Incidence rates of traditional noncommon cancers including Kaposi sarcoma and lymphoblastic lymphoma have demonstrated increased frequency in heritable immunodeficiencies.77 However, common epithelial-based cancers, including lung, colorectal, and breast, have similar incidence rates as the general population.77 The initial epidemiologic studies were conducted at a time when patients with heritable immunodeficient disorders rarely lived past 30 years of age, so it is difficult to ascertain whether subtle changes in incidence seen in the more common epithelial cancers are more obvious as patients age. A similar phenomenon is seen in acquired immunodeficiencies, including acquired immunodeficient syndrome (AIDS), where uncommon cancers such as Kaposi sarcoma and non-Hodgkin lymphoma and not epithelial-based tumors remain the most common AIDS-related cancers.78 As HIV-positive individuals are living longer due to more efficacious antiretroviral treatments, common epithelial cancers as well as cancers secondary to viral coinfectivity, with hepatitis viruses and/or HPV have increased in incidence.79,80
Figure 14-4. Mammalian cell cycle. Rb, a tumor suppressor gene, is complexed to E2F and is thereby unable to enter from G1 to S phases. Upon signaling from the checkpoint genes, CDK4, 6, and cyclin D, Rb is phosphorylated allowing passage into S phase.
Another cohort, with an acquired immunodeficiency, demonstrating increased incidence rates of common and noncommon cancers are patients receiving immunosuppressive drugs following transplantation. A recent study examining cancer incidence in patients following heart and/or lung transplantation receiving immunosuppression therapy demonstrated a sevenfold increased incidence of cancer compared to the general population, with leukemia and lymphoma being the most prevalent.81 Additionally, nearly a 200-fold increase in basal and squamous cell carcinomas is demonstrated following renal transplantation, secondary to immunosuppression and UV radiation damage.82
Tumors differ fundamentally from their normal cell counterparts through the production of tumor antigens, thereby allowing the immune system to differentiate self versus nonself. Tumor antigens are generated by the tumor cell and have historically been broken into two groups, tumor-specific antigens and tumor-associated antigens.76 Tumor-specific antigens are molecules, typically proteins produced by the tumor secondary to a mutation of a proto-oncogene or tumor suppressor gene. Tumor-associated antigens are proteins produced following gene mutations unrelated to tumor formation. Due to lack of specificity between the two historical terms, tumor antigens are more accurately classified according to mutation of normal self-proteins, overexpression or aberrant expression of normal self-proteins, glycoproteins, or glycolipids, and formation of protein products of oncogenes or oncoviruses.76
The innate immune system response to tumors can be divided into two general compartments: a set of barriers and cell activity.83 The barriers are comprised of mechanical factors including the skin and mucous membrane, chemical factors including high gastric acidity, and biologic barriers including commensal microbes. Natural killer cells, macrophages, and dendritic cells comprise the cellular component of the innate immune system response.
Natural killer cells constitute the primary innate immune cell type responsible for killing nonmajor histocompatibility complex (MHC) expressing cancer cells. Many tumors lose MHC class I molecules during malignant transformation while continuing to express ligands such as MICA and MICB that constitutively bind NK receptors, including NKG2D, inducing tumor cell apoptosis through perforin and granzyme release.84 Macrophages can efficiently eliminate apoptotic tumor cells through the release of lysosomal enzymes, reactive oxygen intermediaries, and nitric oxide. Dendritic cells link the innate immune system to the adaptive immune system. Residing in specific tissues according to dendritic cell lineage, dendritic cells phagocytize tumor antigens and carry the proteins to draining lymph tissue presenting to tumor-specific T cells. Depending on the state of dendritic cell maturity caused by costimulatory molecule signaling, either a state of anergy/tolerization (immature) or activation (mature) of tumor-specific T cells and the adaptive immune system response occur.76
Evidence supporting activation of the adaptive immune system response in mouse tumor cell models is well documented. However, its role in human malignancies is not as clearly defined due to the heterogeneity of the tumor microenvironment and the obvious lack of immunodeficient human models. Nevertheless, compelling clinical data supports the hypothesis that an adaptive immune system response serves as a mechanism for tumor cell death in humans. Tumor infiltrating lymphocytes (TILs) including CD8+ T cells, natural killer cells, or natural killer T cells have been associated with improved prognosis for a number of different tumor types.85–88 The initial association between favorable prognosis and TILs was first observed in melanoma patients where it was reported that patients with higher levels of CD8+ T cell tumor infiltration survived longer than patients with tumors containing lower numbers.85 The mechanism of TILs and especially CD8+ T-cell–mediated tumor elimination requires tumor antigen presentation by APCs, typically dendritic cells. Activation of CD8+ T cells occurs following binding of MHC class I molecules expressed by APCs and costimulatory molecule expression. The most widely studied costimulatory molecule pathway is the B7-CD28 interaction with B7 expressed on APCs and CD28 on CD8+ T cells. Following CD8+ T-cell activation, intracellular signaling activation of the NF-AT, NF-kb, and AP1 pathways leads to further CD8+ T-cell activation and promotion of tumor cell death.76
Although the original immune surveillance hypothesis is critical to understanding cancer cell eradication it is likely that tumor cell tolerance and evasion play a much more important role in tumor cell growth, thus offering opportunities for development of novel immunotherapies. The most successful molecules to be targeted in clinical cancer immunotherapy are the immune checkpoint receptors, cytotoxic T-lymphocyte–associated antigen 4 (CTLA4), and programmed cell death protein 1 (PD1). Both PD1 and CTLA4 are inhibitory receptors that regulate immune responses at different levels and via different mechanisms (Fig. 14-5).89
CTLA4, the first immune checkpoint receptor to be clinically targeted, is expressed exclusively on T cells, where it primarily regulates the amplitude of the early stages of T-cell activation. Although expressed on activated CD8+ T cells, the major physiologic role of CTLA4 appears to be through effects on CD4+ regulatory (Treg) and helper cells. Thus, blockade of CTLA4 appears to switch the tumor microenvironment from immunosuppressive to immunoreactive.90,91 An antihuman CTLA4 antibody, ipilimumab, was shown in a cohort of patients with advanced melanoma to induce an objective response rate in tumors previously treated with IL-2. Although there is significant immune-related toxicity involving skin, liver, or colon, ipilimumab has gained FDA approval for the treatment of advanced melanoma.92
In contrast to CTLA4, the major role of PD1 is to limit the activity of T cells in peripheral tissues at the time of T-cell activation to tumor antigen presentation. Similar to CTLA4, PD1 is highly expressed on Treg cells acting as a suppressor mechanism to effector T cells (CD8+ T cells).93 Currently, two anti-PD1 inhibitors, pembrolizumab and nivolumab, have demonstrated efficacy in advanced melanoma patients with disease progression following ipilimumab treatment.94,95
The Role of the Surgeon in Cancer Management
For many patients, surgeons are the entry point into the healthcare system when managing cancer and cancer-related illness. For many surgeons, palliative or supportive care for patients with incurable malignancy comprises a significant portion of their practice. Surgeons are involved in every phase of cancer care from diagnosis to palliation. Many of the concepts covered in this chapter are covered on a disease-specific basis elsewhere in the textbook. Understanding the concepts underpinning many of the treatment approaches can help surgeons understand the rationale for current practice, and can also potentiate the development of new therapeutic strategies.
Figure 14-5. A: Upon antigen expression to the T cell receptor on memory or naïve T cells and subsequent stimulation, CTLA4 is transported to the T cell surface dampening further T cell activation. B: The activity of PD1 is further downstream in the inflammation process. PD1 is induced by activated T cells within peripheral tissues and signals to dampen further effector T-cell activation-limiting inflammatory cascade. (Adapted from Pardoll DM. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer 2012;22:252–264.)
Although a significant portion of oncologic surgery practice focuses on curative resection of disease, the intent for many surgical situations does not include cure. Recognizing the goals of therapy prior to initiating a treatment course can help surgeons maintain a patient-centered approach with appropriate patient preoperative counseling and consent.
Diagnosis and Staging
The role of surgical biopsy to confirm the diagnosis of cancer prior to definitive therapy has been diminished recently by the advances of image-guided and endoscopic biopsy techniques. These techniques allow high levels of diagnostic accuracy with reduced patient discomfort, expense, and use of hospital resources. Procedures such as punch biopsy for suspicious cutaneous lesions or excisional lymph node biopsy for characterization of lymphoma still remain a common diagnostic procedure for many surgeons.
A key aspect of these biopsies is a clear communication with the responsible pathologist regarding specialized tissue handling or preservation prior to initiation of the procedure. Specimen marking and orientation are critical for subsequent resection needs. Incorrect preservation of the biopsy specimen may obviate the possibility for flow cytometry when attempting to characterize lymph node suspicious for lymphoma. Intraoperatively, surgeons need to plan for a subsequent definitive resection when appropriate. Forethought in the selection of an incision or the extent of tissue plane dissection and disruption can make a subsequent definitive resection less morbid and less complex for the patient and surgeon. For instance, on occasion, extremity soft tissue tumors may require excisional biopsy. A longitudinal incision in the orientation of the affected limb can minimize the size of the incision during the subsequent definitive resection.
Surgeons may also play a key role in the staging of cancers. As entry points into the healthcare system for patients dealing with cancer, surgeons are often responsible for distilling the physical examination, radiologic, endoscopic, and surgical findings into a clinical stage assignation. For cancers, such as pancreatic adenocarcinoma, gastric adenocarcinoma, and ovarian cancer, the operative assessment of the peritoneum may identify radiographical occult carcinomatosis 10% to 25% of the time.96–99 When performed laparoscopically in well-selected patients, this minor procedure can avoid the patient impact of a nontherapeutic laparotomy in the setting of metastatic, incurable cancer. In addition, lymphadenectomy, which will be discussed more fully elsewhere, serves an important prognostic role by allowing for complete staging of the nodal basins. The status of the nodal basins not only has important prognostic impact, but also defines the role of adjuvant therapy for many different cancer types.
The ability to completely resect all viable tumors requires consideration of technical, oncologic, and functional resectability. A small number of tumor types are associated with surgical survival benefit in the setting of incomplete resection. Patients with cancers such as ovarian cancer and mucinous appendiceal cancer may benefit from cytoreduction with the possible addition of intraperitoneal chemotherapy.100–104 Patients with life-limiting hormonal symptoms related to metastatic neuroendocrine tumors may gain survival benefit from cytoreduction and the resulting decrease in circulating hormones, such as insulin and somatostatin.105,106 However, for the vast majority of cancers, survival benefit of resection is only associated with complete resection.107–110 For that reason, a plan for resection should include extirpation of all viable tumor for patients with a treatment goal of cure.
The goals for curative resection should include not only complete removal of tumor, but also preservation of adequate patient function and the possibility of prolonged disease-free survival (DFS). These functional and oncologic aspects of resectability are the metrics by which any planned resection can be judged successful. For instance, the role of resection of hepatic colorectal metastases or hepatocellular cancer is defined to a great degree by the amount of healthy residual liver and not necessarily the volume of liver tumors.111–113 Conversely, even small-volume hepatic metastasis in the setting of pancreas adenocarcinoma precludes resection for curative intent.114 Resectability for patients with colorectal hepatic metastases relies on functional preservation; resection for patients with pancreas cancer metastases is limited by oncologic outcomes.
Symptoms related to the progression of malignant disease are a common problem faced by surgeons. A significant proportion of a surgeon’s oncologic practice can be dedicated to providing palliation for patients with incurable conditions. This need is not limited to patients with malignancy; however, the principles associated with surgical palliation of cancer-related symptoms are a sound model for discussion regarding surgical palliation in general.
The primary concept driving decision-making for patients with palliative needs is the critical starting point of goal assessment. In order to deliver the most individualized, risk-appropriate, effective palliation to any given patient, surgeons need to ascertain the goals of therapy for each individual patient at that given time. This assessment transcends diagnosis-based treatment algorithms or pathways. Patients with gastric outlet obstruction related to metastatic gastric cancer may seem like a homogeneous group; however, the primary therapeutic goals of individual patients may vary dramatically. Treatment goals in that setting may include resolution of nausea, eating independently, long-term maintenance of nutrition, or improving performance status sufficiently to enter home hospice. These variable treatment goals can lead to variable treatment strategies, such as placement of a nasogastric decompression tube, a percutaneous gastrostomy tube, gastrojejunostomy, and/or antiemetic medications.
A second factor particularly relevant to palliative surgery for patients with cancer is the increased risk for many procedures in a population with incurable malignancy. Many patients eligible for palliative operations or procedures have been debilitated by malnutrition, physical deconditioning, prolonged hospitalization, or cytotoxic chemotherapy.115,116 These factors, as well as the presence of metastatic disease, have all been associated with poor short-term outcomes following operations. Understanding the increased risks of even simple operations for this patient population can help the surgeon guide the patient and family conversations regarding the risks and benefits of a procedure. Making a decision with a patient and family regarding strategies for surgical palliation requires both an individualized assessment of the needs of those involved as well as a generalized awareness of the risks for that procedure based on associated risk factors.
Extent of Resection
For malignancies such as melanoma, breast cancer, or sarcoma, the requirement for radical resection has decreased significantly over the past several decades.117–121 Conversely, for diseases such as pancreas cancer and liver tumors, the frequency and safety of radical resections have increased dramatically over that same time period.122 A key point of judgment for surgeons involved in cancer operations is not only deciding when to operate, but also how extensive that operation should be.
An important distinction must be made between surgical and pathologic margins. Surgical margins are the planned lines or planes of resection around a grossly visible tumor. The aim of this strategy is not to ensure a wide swath of healthy tissue around the tumor specimen, but to ensure a final negative pathologic margin. Pathologic margins are the histologically assessed borders of uninvolved tissue around the microscopic tumor. Tumors with infiltrative behavior may extend up to the pathologically assessed margin even though the grossly assessed surgical margin appears to be uninvolved. An example of this discrepancy occurs in the management of melanoma. Resection with a 1-cm surgical margin is generally considered adequate for a nonmetastatic extremity lesion with a tumor thickness of 1 to 2 mm. If the final pathologic margin is only 0.3 mm, this is considered a negative, and adequate, pathologic margin.
Table 14-9 Terms Typically Used to Describe Pathologic Margins for Cancer Resections
The terms typically used to describe pathologic margins for cancer resections are defined below (Table 14-9). An R0 margin is typically considered negative and an R2 margin is considered grossly positive. For many, but not all, types of cancer, the presence of an R1 margin is associated with increased risk of recurrence. Management of a positive pathologic margin is dependent on technical and oncologic factors associated with that particular tumor. For instance, some early-stage cutaneous malignancies such as dermatofibrosarcoma protuberans or basal cell cancer are characterized by limited risk of distant disease dissemination and a high risk of local recurrence in the setting of positive pathologic resection margins. In a location where additional local excision may not be limited by adjacent structures, an effort for reexcision of a positive margin may be reasonable. Alternatively, the presence of cancer at the superior mesenteric artery (SMA) resection margin at the time of pancreaticoduodenectomy does not lead to an additional operative attempt to achieve a negative final margin for patients with pancreas adenocarcinoma. The hesitation to attempt reresection is driven not only by the technical challenges of attempting to achieve a broader margin along the SMA, but also by the very high risk of distant recurrence for those patients and the relatively lower impact of local recurrence in this setting.123–125
The removal of regional lymph nodes at the time of resection of the primary tumor has been the focus of significant investigation and controversy throughout the contemporary history of surgical management of cancer. The surgical principles attributed to William Halsted regarding the surgeon’s role in interrupting the progression of cancer from the primary tumor through lymphatic channels to regional lymph nodes and then to distant sites form the traditional rationale for regional lymphadenectomy.126,127 However, more contemporary studies such as MSLT-1, the Dutch Gastric Cancer Lymphadenectomy, and ACOSOG Z0011 clinical studies have raised questions regarding the survival benefit of lymphadenectomy for patients with melanoma, gastric cancer, and breast cancer, respectively.128–131
The potential prognostic benefits of lymphadenectomy undoubtedly are more germane than therapeutic advantages. For almost all types of malignancy, the presence of cancer within regional lymph nodes portends a worse prognosis for patients, as a harbinger of systemic cancer spread. The accuracy of the assessment of the regional lymph node basin can be increased with either focused evaluation of the most at-risk lymph nodes (i.e., sentinel lymph node biopsy) or thorough sampling of a large number of nodes (i.e., lymphadenectomy). Accuracy of cancer staging, for example in colorectal cancer, is oftentimes increased through the sampling of a sufficient number of lymph nodes. The threshold for quality assurance through adequate lymph node sampling has been established for surgeons who perform colorectal cancer resections, however, that well-defined performance metric is not common to all types of resection and all types of cancer.132–135
For patients with bulky involved regional lymph nodes, the benefits of node removal are not likely prognostic. In selected patients, removal of these lymph nodes may provide some palliation from local symptoms. The role of “prophylactic palliation” or removal of lymph nodes, which may develop and cause local problems is less clear. For example, in melanoma or rectal cancer, eliminating all of the regional lymph nodes in the setting of positive microscopic nodal burden decreases the potential of future development of bulky, symptomatic nodes.128,136–138 The therapeutic benefit of avoiding local recurrence in a subset of the operative population needs to be balanced against the potential adverse effects of lymphadenectomy. For many nodal basin sites, the adverse impact of interrupting lymphatic flow can have significant deleterious effects on patients following lymphadenectomy, including lymphedema, paresthesias, and even development of secondary malignancies, such as angiosarcoma (Stewart–Treves syndrome).139–142
In general, lymphadenectomy is an important aspect of surgical cancer care when any of the following criteria are met: (1) removal of regional lymph nodes will provide important information to either guide adjuvant therapy decisions or provide prognostic clarity (e.g., colon cancer, gastric cancer); (2) lymphadenectomy can eliminate a predictably involved site of disease involvement which can lead to local regional complications in the setting of nodal recurrence (e.g., rectal cancer, esophageal cancer); (3) removal of at-risk lymph nodes is a minor adjunct of the procedure for primary tumor removal (e.g., pancreatectomy). The extent of our operative management for cancer is likely to change as we learn more about the therapeutic benefit of regional nodal removal. Perhaps more strikingly, the prognostic information gained by evaluating a nodal basin is likely to diminish as our ability to profile tumors and their behavior becomes more dependent on genomic, proteomic, and other expression profiling.
Resection of Metastatic Disease
Outside the palliative setting, the indications for resection of metastatic disease are relatively narrow. Despite the growing awareness and increased frequency of resection for metastatic colon cancer, endocrine tumors, and gastrointestinal stromal tumors, the vast majority of patients with metastatic disease are not eligible for metastasectomy. Although it is important to recognize the diverse spectrum of presentations for malignant disease and to be wary of absolute contraindications against surgical management of metastases, several general principles do apply to the overwhelming majority of patients with metastatic disease.
Complete elimination of all sites of disease is generally associated with dramatically improved survival compared to incomplete resection. In addition, the outcomes for patients with incompletely resected disease are frequently comparable to outcomes for patients who underwent no resection at all.108–110,143–145 For these reasons, resection or ablation of all sites of disease is a common requirement when making a decision to proceed with resection of metastatic disease. This principle applies to patients with primary tumors in place as well; therefore, patients with unresectable primary tumors are generally not eligible for curative intent resection of their metastatic sites.
The behavior or biology of an individual’s cancer may suggest a less aggressive phenotype and warrant consideration of metastasectomy. Oftentimes characteristics of favorable cancer biology are associated with improved survival outcomes following resection. A prolonged period of progression-free survival (PFS) or DFS is commonly associated with improved outcomes following resection of metastatic disease.143–145 Progression of metastatic sites while receiving first-line systemic chemotherapy is often a harbinger of early systemic failure among patients undergoing resection of metastatic disease.146,147 As systemic chemotherapy has grown more effective over the last decades, the impact on tumor progression and control of systemic disease has led to increased opportunities for resection of metastatic sites. This concept has been most plainly demonstrated in the context of hepatic colorectal metastases. The frequency of hepatectomy in that setting has increased in harmony with the increased effectiveness and availability of both cytotoxic and targeted chemotherapy agents.122 Surgical management of metastatic disease may increase in frequency for other diseases as we develop more effective systemic therapy options.
Cancer-Related Surgical Outcomes
The short-term outcomes for surgical patients commonly include metrics, such as complication rates, length of stay, mortality rate, and overall survival (OS). These metrics are certainly important measures of quality and efficacy for patients undergoing operations for cancer. Similar to other surgical specialties, cancer-related surgery also carries specific outcome measures, which are particularly relevant to these clinical scenarios. Understanding these metrics is critical to balancing risk and benefit for individual patients as well as for evaluating the role of new therapeutic strategies and approaches.
Many surgical specialties are directed at the eradication of a tangible, anatomically identifiable lesion. Peripheral and coronary vascular bypass, gastroesophageal fundoplication, and parathyroidectomy are all measured by how effectively the treatment eliminates the arterial plaque, gastroesophageal reflux, or hyperparathyroidism. Early recurrence of the condition suggests a failure of therapy or poor patient selection for the procedure. Disease recurrence is also a critical measure of the effectiveness of surgical therapy for cancer patients.
A prerequisite for disease recurrence as a measure of outcome is the initial eradication of all visible cancer. Recurrence can be measured by development of symptoms, physical examination, radiographic evaluation, biochemical evaluation, or operative exploration. Unique to the practice of surgery for cancer patients is the distinction between locoregional and distant disease recurrence. Locoregional recurrence is typically defined as recurrence in the resection bed or region of the draining regional lymph nodes. Distant recurrence is typically defined as occurrence of malignant cells outside of the initial primary tumor and lymphadenectomy resection field.
Although resection is traditionally considered a means to achieve locoregional control, and adjuvant systemic therapy is employed to control distant disease recurrence, a decision to proceed with an operation should consider how the operation impacts both locoregional and distant recurrence. A well-performed, radical, margin-negative operation can only lead to prolonged DFS if it is performed with appropriate patient selection. This distinction is the underpinning behind the approach for patients with small-volume metastatic disease in the setting of pancreatic adenocarcinoma, gastric cancer, or biliary tract cancers. Many patients, when faced with this scenario, will ask their surgeon, “Why can’t you just take out all of it?” Patients with easily resectable, distant disease are poor candidates for resection not because they cannot be rendered free of all visible disease but rather because the findings of metastases are highly predictive of early distant disease recurrence even if local control can be achieved. The trade-off for a short period of DFS after a radical operation is the chance at a more prolonged PFS or OS with the immediate initiation of effective systemic chemotherapy.
For patients who cannot have all visible tumor resected due to locally advanced disease or metastases, PFS is an important measure for subsequent therapy. The propensity for a treatment to control, but perhaps not eliminate, a measurable volume of tumor is the critical measure for defining the effectiveness of that therapy. For many types of cancer, PFS is a reliable surrogate for OS.148,149 Typically, when patients experience disease progression while receiving a particular regimen of systemic therapy, the treating physician will consider initiating the next line of systemic treatment.
Radiographic progression of measurable lesions is measured by a well-defined set of criteria referred to as RECIST or modified RECIST criteria. Response Evaluation Criteria in Solid Tumors (RECIST) was developed in 2000 and subsequently modified in 2009 as a system for measuring tumor response or progression.150 These criteria are commonly employed in clinical trials as a standardized way to determine when a tumor has progressed on therapy. A simplified description of the criteria is included in Table 14-10. This strategy requires the upfront identification of index lesions, which are measurable radiographically. RECIST has proven to be a useful measure of potential treatment benefit in phase II clinical trials. Although useful in clinical trials, stringent application of RECIST measures of progression is not as readily used outside trial settings. Even small changes in the one-dimensional measurements or characteristics of measurable lesions may be considered evidence for clinical progression and justification for a change in treatment course.
Return to Intended Oncologic Treatment
A novel measure for the quality of surgical management of cancer patients is return to intended oncologic treatment (RIOT).151 This metric considers the place that resection holds in the multidisciplinary care of a cancer patient. Although surgical resection is required for a curative intent pathway for the majority of solid tumors, important oncologic adjuncts such as adjuvant chemotherapy, radiotherapy, and maintenance hormonal therapy can reduce the chances of recurrence and prolong survival for many patients. Delays or disqualifications due to operative adverse events may negatively impact the overall outcomes of patients. An important element in the assessment of the quality of cancer care is the ability for a health system to deliver all of the treatment modalities demonstrating benefit for a particular malignant condition. Recognizing that the definitive resection of a cancer is part of a larger treatment strategy is one of the important considerations for assessing surgical outcomes for patients with cancer.
Table 14-10 Common Criteria Used to Determine Therapy-Related Progression