Matthew F. Kalady, MD

Cancer continues to be a principal component of general surgery practice. Various technological advances have led to improved screening, earlier diagnosis, and more accurate staging, all of which have enhanced patient care. Among the new technologies, accurate imaging remains an integral part of multidisciplinary cancer care. Perhaps more importantly, close communication between surgeons and radiologists maximizes the clinical information gained from radiological imaging.

Douglas S. Tyler, MD

The general surgeon treats numerous cancers, each of which dictates unique imaging requirements (Table 1, page 16). Extensive discussion of each cancer is beyond the scope of this review. This article focuses on the integration of radiological imaging and surgical care. From the surgical perspective, imaging augments patient care by providing information within four main contexts:

1. Providing diagnosis through image-guided interventional tissue biopsy;
2. Staging disease to determine the appropriate method of treatment;
3. Defining anatomic detail for intraoperative planning; and
4. Adding technical assistance for identifying disease intraoperatively.

Image-guided Biopsy

Occasionally, it is difficult to differentiate a benign process from malignant disease based on clinical history and the support of imaging studies. For example, a patient with a history of chronic pancreatitis who is found to have a mass in the head of the pancreas by CT presents a diagnostic dilemma. Imaging by CT may not accurately distinguish an inflammatory mass of chronic pancreatitis from a pancreatic neoplasm. Another common example involves pulmonary cavitary lesions. Infectious processes such as fungal disease or tuberculosis may be indistinguishable from a cavitary malignancy on CT scan. Making the correct diagnosis is paramount to further clinical management.
Tissue biopsy may provide cytologic or histologic samples to assist the surgeon in making a definitive diagnosis. The area of abnormality, however, is often not accessible by routine percutaneous biopsy. Using CT, ultrasound, or fluoroscopy-guided imaging, the interventional radiologist may obtain percutaneous biopsies with low risk of injury to vital adjacent structures such as blood vessels or viscous organs.

Table 1. Selected general surgery malignancies and associated commonly used imaging techniques.

Image-guided biopsies are most commonly employed by surgeons for the evaluation of breast abnormalities identified on mammography, particularly for those breast lesions that are nonpalpable on physical examination. Ultrasound guidance is also utilized to biopsy suspected malignant thyroid nodules or parathyroid nodules. Within the abdominal cavity, the introduction of endoscopic ultrasound has greatly improved the ability to obtain tissue diagnosis from periampullary lesions. In addition, CT-guided biopsy of liver lesions suspicious for metastatic malignancy may reduce the number of diagnostic exploratory laparotomies by providing a diagnosis. For example, if percutaneous biopsy reveals pancreatic carcinoma with hepatic metastasis, surgery is not indicated and the patient is spared an unnecessary operation.

Disease Staging

Once a diagnosis of cancer has been established, the disease must be accurately staged to determine the next step in clinical management. As a generalization, patients with advanced disease will not gain a survival benefit from resective surgery and may be more suitable for systemic therapy or palliative care. Furthermore, as more neoadjuvant protocols are being developed, enrollment of the appropriate patients depends on precise staging of both primary tumor and systemic involvement. Highly accurate noninvasive imaging modalities are preferable to surgical approaches such as a staging laparotomy to obtain this information.

Various imaging techniques have positively affected staging and treatment of several cancers. For example, the development of endoscopic rectal ultrasound has refined the treatment of rectal cancers. Endorectal ultrasonography can precisely identify the layers of the rectal wall and the corresponding level of tumor invasion.1,2 In addition, this technology may identify regional nodal enlargement and involvement of adjacent muscle or bone, so that the tumor may be more accurately staged. Tumor stage then determines the next appropriate management decision. For example, preoperative radiation and chemotherapy have been shown to benefit patients with tumors involving the full thickness of the abdominal wall (T3 disease). However, patients with tumors involving only partial thickness of the rectal wall (T2 disease) do not gain added benefit from neoadjuvant therapy compared to surgery alone.

The use of fluorine 18-labeled deoxyglucose (FDG) uptake positron emission tomography (PET) scanning to determine occult metastases is another example of how new imaging techniques have influenced patient care. PET appears most useful in staging sarcoma,3 colorectal tumors,4 melanoma,5 and lung cancer.6 By providing a total body image, PET allows for entire body staging and provides additional information about a given site of interest. Figure 1 displays FDG PET scan images of a patient with a right upper extremity sarcoma. Increased tracer uptake in the spine and hilar lymph nodes represents metastatic disease, defining a stage IV sarcoma. It is important to note that since some normal cells may also utilize FDG, areas of increased uptake on PET scan do need to be evaluated carefully to avoid false-positive conclusions. Newer imaging technologies that combine PET and CT imaging should help improve PET accuracy.

Operative Planning

Figure 1. Positron emission tomography on a patient with a primary right arm sarcoma demonstrated increased uptake in (A) the spine, suggesting vertebral metastasis (VM), and (B) hilar lymph nodes, indicating advanced stage sarcoma.

If the patient is an operative candidate based on staging criteria, the surgeon then focuses attention on the regional anatomic detail to determine an operative plan. Tumor involvement of adjacent organs or major blood vessels may render a tumor unresectable, or change the procedure to be performed. For example, complete resection of involved tissue provides the only chance for prolonged survival from pancreatic cancer. However, local invasion leading to encasement of the superior mesenteric or celiac arteries, regional nodal spread, or distant metastases are all contraindications to surgical resection. Abdominal CT accurately provides this information so that the surgeon may plan the appropriate procedure, or avoid a nontherapeutic laparotomy. Figure 2 provides an example of resectable (A) and non-resectable pancreatic cancer (B). The patient represented in panel A underwent a pancreaticoduodenectomy; the patient in panel B was deemed unresectable due to a thrombosed superior mesenteric vein secondary to tumor involvement and was spared an unnecessary laparotomy.

Figure 2. Abdominal CT of patients with pancreatic cancer. (A) Patient with fullness in the pancreatic head and dilated pancreatic duct had biopsy- proven adenocarcinoma, which was resected by pancreaticoduodenectomy. (B) Tumor involvement of the superior mesenteric vessels resulted in superior mesenteric vein thrombosis and thus unresectable disease. Imaging results spared this patient an unnecessary laparotomy.

For resectable tumors, precise imaging also helps to preoperatively define the necessary extent of resection. This detail is demonstrated by the preoperative planning of a sarcoma resection. For both extremity and visceral sarcomas, the surgical goal is complete tumor resection with histologically negative margins while preserving as much normal function as possible. Both CT and MRI provide detailed imaging to differentiate tumor from normal adjacent tissue such as muscle, bone, and nerves.7,8 Thus, tumor demarcation and fascial planes are more clearly defined to determine the necessary extent of surgical resection.

Intraoperative Assistance

Utilization of imaging modalities during surgery is often an integral part of staging procedures or identification of tissue for resection. Intraoperative sentinel lymph node (SLN) mapping and biopsy in melanoma demonstrate the importance of interaction between the surgeon and radiologist in the operating room. It has been shown that skin lymphatic drainage first flows to one or a few lymph nodes, the SLN(s), within a particular nodal basin. Thus, the pathological status of the SLN accurately predicts the pathological status of the entire nodal basin. In the last decade, physicians have developed and refined techniques to identify the SLN intraoperatively.

Figure 3. Lymphoscintigraphy of a patient with a left lower extremity melanoma revealed four sentinel lymph nodes in the inguinal nodal basin. Imaging findings directed the intraoperative dissection.

The technique of SLN mapping involves injection of radiolabeled sulfur colloid at the cutaneous melanoma site. The radioactive tracer follows the lymphatic drainage patterns and accumulates in the SLN(s). Areas of accumulation seen on preoperative lymphoscintigraphy determine the nodal basins to be explored. At the time of surgery, the surgeon uses a handheld gamma probe to identify radioactivity within the SLN to be resected. Using gamma probe detection of radioactivity in combination with visually tracing blue dye injected at the melanoma site during the surgery, the SLN is successfully identified in more than 90% of cases.9-11 The imaging and technology utilized for SLN mapping in melanoma are increasingly being applied to nodal mapping for patients with breast cancer and explored in other diseases such as colorectal cancer. Figure 3 shows a lymphoscintigram of a patient with a lower extremity melanoma in which four SLNs were identified. A small incision was made over the inguinal region and, with guidance of the handheld gamma probe, the SLN was localized and resected.

In summary, applications of cancer imaging are diverse and remain an invaluable adjunct to management of the surgical patient. As surgeons and radiologists continue to provide multidisciplinary care, newly developed imaging tests and techniques will direct patient care.?


Matthew F. Kalady, MD, is a postdoctoral research fellow and resident in general surgery at Duke University Medical Center, Durham, NC,

Douglas S. Tyler, MD, is an associate professor of surgery at Duke University Medical Center and chief of the Section of Surgical Oncology. He is also chief of general surgery at the Durham Veterans Affairs Medical Center.


  1. Kim H, Wong W. Role of endorectal ultrasound in the conservative management of rectal cancers. Semin Surg Oncol. 2000;19:358-366.
  2. Massari M, DeSimone M, Cioffi U, et al. Value and limits of endorectal ultrasound for preoperative staging of rectal carcinoma. Surg Laparosc Endosc. 1998;8:438-434.
  3. Schwarzbach M, Dimitrakopoulou-Strauss A, Willeke F, et al. Clinical value of [18-F] fluorodeoxyglucose positron emission tomography imaging in soft tissue sarcomas. Ann Surg. 2000;231:380-6.
  4. Hung G, Shiau Y, Tsai S, et al. Value of 18F-fluoro-2-deoxyglucose positron emission tomography in the evaluation of recurrent colorectal cancer. Anticancer Res. 2001;21(2B):1375-8.
  5. Tyler D, Onaitis M, Kherani A, et al. Positron emission tomography scanning in malignant melanoma. Cancer. 2000;89:1019-1025.
  6. Marom E, Erasmus J, Patz E. Lung cancer and positron emission tomography with fluorodeoxyglucose. Lung Cancer. 2000;28:187-202.
  7. Varma D. Imaging of soft-tissue sarcomas. Current Oncology Reports. 2000;2:487-490.
  8. Hughes T, Spillane A. Imaging of soft tissue tumours. Br J Surg. 2000;87:259-260.
  9. Morton D, Thompson J, Essner R, et al. Validation of the accuracy of intraoperative lymphatic mapping and sentinel lymphadenectomy for early-stage melanoma: a multicenter trial. Ann Surg. 1999;230:453-465.
  10. Kalady M, White D, Fields R, et al. Validation of delayed sentinel lymph node mapping for melanoma. Cancer Journal. In press.
  11. White D, Schuler F, Pruitt S, et al. Timing of sentinel lymph node mapping after lymphoscintigraphy. Surgery. 1999; 126(2):156-161.