Intra-Abdominal Patterns of Disease Dissemination in Colorectal Cancer Identified Using Radioimmunoguided Surgery Mark W. Arnold, M.D.,* Charles L. Hitchcock, M.D., Ph.D.,t Donn C. Young, Ph.D.,:~ William E. Burak, Jr., M.D.,~ David J. Bertsch, M.D.,~ Edward W. Martin, Jr., M.D.,~

From the *Division of General Surgery, Department of Surgery, ~Department of Pathology, ~Department of Biostatistics, Comprehensive Cancer Center, 5Division of Surgical Oncology, Department of Surgery, The Arthur G. James Cancer Hospital and Research Institute, The Ohio State University, Columbus, Ohio

PURPOSE: Patterns of metastatic spread are difficult to de- termine with routine postoperative follow-up. This study was undertaken to evaluate two selected populations of colorectal cancer patients injected and screened with anti- tumor antibody. METHODS: Eighty-six evaluable patients with colorectal cancer underwent exploratory laparotomy with both traditional surgical exploration and radioimmu- noguided surgery (RIGS) following injection of lzSI-labeled CC49 monoclonal antibody. RIGS-positive tissue detectable with a handheld gamma-detecting probe was defined as tissue involved with the disease process. Comparisons were made between extent of disease using traditional explora- tion and extent using RIGS. RESULTS: In 41 patients with primary disease, traditional exploration detected 45 sites of disease (1.1 sites/patient) compared with 153 RIGS-positive sites (3.7 sites/patient). In 45 patients with recurrent dis- ease, traditional exploration found 116 sites (2.6 sites/pa- tient) vs. 184 RIGS-positive sites (4.1 sites/patient). Involve- ment by selected anatomic sites is shown below.

Primary (n = 41) Recurrent (n = 45) Site Traditional RIGS Traditional RIGS

(%) (%) (%) (%)

Liver 9 (22) 9 (22) 29 (64) 29 (64) Gastrohepatic 4 (10) 28 (68) 8 (18) 30 (67)

ligament Celiac 2 (5) 20 (49) 5 (11) 22 (49) Retroperitoneal 2 (5) 14 (34) 4 (9) 5 (11) Small bowel 3 (7) 15 (37) 11 (24) 13 (29)

mesentery Pelvis 6 (15) 15 (37) 19 (42) 22 {49)

RIGS = radioimmunoguided surgery.

CONCLUSION: RIGS detected more tissue involved in dis- ease process for all sites in both primary and recurrent disease except liver metastases. Areas with highest propor-

Read at the meeting of The American Society of Colon and Rectal Surgeons, Montreal, Quebec, Canada, May 7 to 12, 1995. Address reprint requests to Dr. Arnold: The Ohio State University Hospitals, 410 West 10th Avenue, Doan N-717, Columbus, Ohio 43210.

509

tion of RIGS-positive tissue, the gastrohepatic ligament and celiac nodes, are rarely resected and are not pathologically examined. Positive RIGS localization of tumor antigen in these areas suggests more extensive dissemination of dis- ease process. [Key words: Radioimmunoguided surgery (RIGS); Metastatic spread; Lymph nodes; Tumor antigens]

Arnold MW, Hitchcock CL, Young DC, Burak WE Jr, Bertsch DJ, Martin EW Jr. Intra-abdominal patterns of disease dis- semination in colorectal cancer identified using radioimmu- noguided surgery. Dis Colon Rectum 1996;39:509-513.

D e tec t ion of ex ten t a n d pa t t e rn o f d i s semina t ion

o f co lorec ta l cancer has b e e n l imi ted b y the

inabi l i ty to de tec t micrometas ta t ic disease . Most s tud-

ies have l o o k e d at the p r o b l e m re t rospect ively , t rying

to ident i fy metas ta t ic loca t ion and recur rence in se-

l ec ted p o p u l a t i o n s o f colorecta l cancer pat ients . Vari-

a t ion in t rea tment and fo l low-up, however , m a k e s it

difficult to in terpre t actual s e que nc ing a n d pa t te rns o f

metas ta t ic disease. For example , pa t ien ts fo l lowed b y

c o m p u t e d t o m o g r a p h y (CT) scan m a y p re sen t differ-

ent ly than pat ients fo l l owed wi th serial ca rc inoembry -

onic an t igen (CEA) evaluat ions . In addi t ion, tests for

recur rence , i.e., phys ica l examina t ion , CEA, CT scan,

a n d c o lonosc opy , are all l imi ted in the abi l i ty to de tec t

d i sease or m a y highl ight on ly se lec ted areas of con-

cern. An equa l ly l imit ing factor is that each eva lua t ion

and p rocedure , even if the pa t i en t is f o l l o w e d closely,

r ep resen t s a re t rospec t ive l ook at on ly a s ingle po in t

in time. A n e w a n d en larg ing les ion in the liver, for

example , m a y no t be i m a g e d o n CT scan until late in

the d isease , or a c o m p l e x pa t te rn o f a n t e c e d e n t lym-

pha t ic s p r e a d m a y b e u n d e t e c t e d or ove r looked . To

o v e r c o m e s o m e of these l imitat ions a n d be t te r def ine

loca t ion a n d pa t te rns of cancer spread , w e c o m p a r e d

two p o p u l a t i o n s o f co lorec ta l cancer pa t ien ts (pr ima-

510 ARNOLD E T A L

ry and recurrent) who underwent operations using a new technology, radioimmunoguided surgery (RIGS).

Dis Colon Rectum, May 1996

METHODS

NineW-nine patients with primary or recurrent co- lon and rectal cancer were enrolled at The Ohio State Hospitals in a study using the RIGS system from July 1990 to April 1992 under National Cancer Institute's Cancer Therapy Evaluation Program protocol TA90- 0038. All patients signed a written consent approved by Ohio State Hospital's Human Subject's Review Board. Patients with extra-abdominal disease were excluded. This protocol was described in detail in a previous publication. 1 Briefly, eighw-six of these pa- tients (87 percent) had positive antibody localization at surgery and were considered evaluable for the study. Forty-one patients underwent surgery for a primary colon or rectal lesion, and 45 patients under- went surgery for recurrent tumors because of rising serum CEA levels, changing CT scan, or other physical evidence of recurrent disease.

M1 patients were injected with murine monoclonal antibody CC49 (0.1-10 mg) radiolabeled with iodine 125 (~z5I) (2 mCi) approximately three weeks before surgery. At surgery a detailed exploration and evalu- ation of the abdomen were performed both tradition- ally and using the handheld gamma-detecting probe (Neoprobe | 1000, Neoprobe, Dublin, OH). RIGS-pos- itive tissue was defined as tissue with probe counts three standard deviations above background. 2 Lower counts were considered negative.

Location of all RIGS-positive tissues was recorded by site and by zone. Zones of the abdomen were established as a convenient grouping of sites for eval- uation of spread of disease (Fig. 1). RIGS-positive tissue was biopsied whenever practical and safe, and all tissues removed from the patient were confirmed as RIGS-positive by e x vivo counts taken by the gam- ma-detecting probe. For purposes of this study, RIGS- positive tissue was not compared with results of con- ventional pathology, i.e., hematoxylin and eosin (H&E) staining of removed tissues. Rather, manner and location of sites found by traditional surgical means were compared with those found in v ivo by the RIGS system, and those results were compared between primary and recurrent patients.

RESULTS

In 41 primary colorectal cancer patients, 12 right or transverse colon lesions, 13 left colon lesions and 16

Figure 1. Location of radioimmunoguided surgery-posi- tive tissue found was recorded by site and by zone. Zones were established as a convenient grouping of sites for evaluation of spread of disease.

rectal lesions were found. Six (50 percen0 right colon lesions, six (46 percent) left colon lesions, and eight (50 percent) rectal lesions had regional metastatic spread as determined by H&E-positive nodes. Loca- tions of extraregional disease found by both surgical exploration and by RIGS exploration are shown in Figure 2. A total of 45 extraregional sites were found without RIGS and 153 extraregional sites were found by combined traditional and RIGS evaluation (240 percent increase). Number of sites and location by zone are given in Table 1. Average number of sites per patient for right colon, left colon, and rectal le- sions were 4.1, 3, and 4.1, respectively. More sites were found with RIGS in Zones II, III, and IV.

Of 45 colorectal cancer patients with recurrent tu- mors, 15 patients had a right or transverse colon primary tumor, 14 patients had a left colon primary tumor, and 16 patients had a rectal primary tumor (Fig. 3). Surgical exploration without RIGS identified 116 sites, whereas surgical exploration with RIGS identified 184 sites (59 percent increase). Average number of recurrent sites per patient with right, left, and rectal primary lesions was

Vol. 39, No. 5

Without RIGS With RIGS

45 153 Sites Sites

N =41

Figure 2. Locations of extraregional disease found both by surgical exploration and by radioimmunoguided sur- gery (RIGS) exploration in primary colorectal cancer pa- tients.

4.7, 3.7, and 3.9, respectively. Recurrent sites by location and zone are shown in Table 1. More sites were found with RIGS in Zone II.

DISCUSSION

Ideal study of metastatic disease would follow in- dividual patients prospectively after primary tumor resection to determine where and when secondary tumor sites present. Unfortunately, tests we use for follow-up, for all their sophistication, are highly fo- cused and detect only relatively large amounts of tumor. Patients undergo reoperation only infrequent- ly; therefore, tissue is not usually available for histo- logic analysis. Thus, for the most part, studies of metastatic sites are done retrospectively to see when recurrence was ultimately detected.

This study used RIGS in conjunction with the CC49 monoclonal antibody to study two distinct popula- tions of patients with colorectal carcinoma. Each tu- mor was evaluated by histopathology and by biologic expression of the TAG-72 antigen as detected by the RIGS system. Ability to determine distribution of RiGS-identified tumor antigens allowed us to charac- terize each individual's tumor by its gross location, but more importantly, by pathways of tumor antigens.

One of the most striking results was association of disease found in Zone II (gastrohepatic ligament, ce- liac axis, and retropancreatic aorta) with development of liver metastasis. For example, ratio of evident me- tastasis in liver to TAG localization in the gastrohe- patic ligament in primary patient population was 22:68 percent. In recurrent group, ratio of liver dis- ease to gastrohepatic ligament TAG localization was 64:67 percent, almost 1:1. Interestingly, percentage of RIGS-identified TAG activity in gastrohepatic ligament in both primary and metastatic populations was

RIGS LOCATES CANCER PATTERNS 511

nearly identical, 68 vs. 67 percent, respectively. Equally striking were identical RIGS findings in celiac axis (49 percent) of both primary and recurrent pa- tients. Likewise in Zone III, there was evidence of tumor antigen in both patient populations, primary and recurrent, in the retroperitoneum and small bowel mesentery and in Zone IV in the pelvic and iliac areas. Also of note, there appeared to be no difference in pattern of tumor antigen spread based on primary tumor location.

Similarity of findings in lymph node-bearing re- gions in both primary and recurrent patients strongly suggests that these areas are already processing tumor antigen at time of the first resection. Likewise, close correlation between disease in gastrohepatic ligament and liver suggests that RIGS-positive tissue in gastro- hepatic ligament at primary surgery may harbor future hepatic metastatic diseasel

Past studies have not looked as closely at the exten- sive network of intra-abdominal lymphatic zones be- cause technology to do so was not available. Gross metastatic deposits such as those typically found in the liver are easy to identify, but identifying disease in ex- traregional lymph nodes is much harder. Simply sam- piing extraregional lymph nodes for H&E histologic evaluation is an ineffective prognostic indicator. Galan- diuk and coworkers 3 undertook a detailed review of recurrent disease in 818 Dukes B2 and C colon cancer patients and found liver involvement in 33 percent of patients but found retroperitoneal and peripheral lymph node disease in only 10 and 4 percent, respectively. Certainly in a population of patients with known re- gional lymph node spread, i.e., Dukes C patients, one would expect lymphatic pathways to play a prominent role in progression of disease, even though they may not harbor metastatic sites, p e r se. Using RIGS, we de- tected evidence of a complex pattern of lymphatic spread of tumor antigens that is similar in both primary and recurrent patients.

A major concern in past RIGS studies has focused on determining importance of RIGS-positive lymph nodes. Frequently, routine pathologic evaluation with light microscopy fails to demonstrate tumor ceils in these nodes. New clinical studies indicate that these lymph nodes represent evolution of cancer process. Recent 30-month to 54-month survival data from pri- mary patients demonstrate that patients who have residual, unresectable, in vivo RIGS-positive tissue remaining at close of surgery almost always die. This outcome is independent of traditional pathologic stage. 4 Another recent study of patients with recurrent

512 ARNOLD ET AL

Table 1. Metastatic Cancer Sites Found at Surgery

Dis Colon Rectum, May 1996

RIGS Zone/Site Primary Colorectal Cancer (n = 41)

Without RIGS (%) With RIGS (%)

Recurrent Colorectal Cancer (n = 45)

Without RIGS (%) With RIGS (%)

Zone I Liver 9 (22) 9 (22) 29 (64) 29 (64) Diaphragm 4 (10) 5 (12) 11 (24) 13 (29)

Zone II Gastrohepatic ligament 4 (10) 28 (68) 8 (18) 30 (67) Celiac 2 (5) 20 (49) 5 (11) 22 (49) Retropancreatic 0 (0) 8 (20) 0 (0) 12 (27)

Zone Ill Perirenal 0 (0) 6 (15) 3 (7) 3 (7) Retroperitoneal 2 (5) 14 (34) 4 (9) 5 (11) Small bowel mesentery 3 (7) 15 (37) 11 (24) 13 (29) Abdominal wall 9 (22) 7 (17) 11 (24) 10 (22) Omental 3 (7) 7 (17) 3 (7) 6 (13) Anastomotic recurrence NA NA 3 (7) 3 (7)

Zone IV Pelvis 6 (15) 15 (37) 19 (42) 22 (49) Iliac 2 (5) 12 (29) 8 (18) 14 (31) Sacral prominence 1 (2) 7 (17) 1 (2) 2 (4)

Totals 45 (100) 153 (100) 116 (100) 184 (100) RIGS = radioimmunoguided surgery; NA -- not applicable.

Without RIGS With RIGS

116 184 Sites Sites

N = 4 5

Figure 3. Locations of extraregional disease found both by surgical exploration and by radioimmunoguided sur- gery (RIGS) exploration in recurrent colorectal cancer patients.

or advanced disease demonstrated that RIGS informa- tion led to a lower resectability rate, sparing some

patients extensive surgery, and to a better selection of resectable patients. 5 Based on a minimum of 28 months follow-up, survival was 22 percent, in contrast to the expected rate of 10 percent in these patients. 6 FurthetTnore, there is increasing evidence that bio- logic behavior of cancer ceils is a highly organized nonrandom selection involving both host (soil) and tumor (seed) factors. 7 Pathogenesis of metastatic cells depends on many homeostatic mechanisms, which are newly under investigation. There is an ever grow- ing list of cytokines, lymphokines, angiogenesis,

growth factors, angiogenic, and genetic factors that influence trafficking of metastatic tumor cells. To this

list we add RIGS with CC49 for i n v i vo identification of TAG antigen, which provides valuable information on

trafficking status of colorectal carcinoma.

C O N C L U S I O N

Lymphatic pathways play an important role in char- acterizing progression of metastatic disease in colon

and rectal cancer. RIGS with CC49 appears to target similar groups of lymph nodes in both primary and metastatic patients, indicating that these tissues are involved in processing minor antigens much earlier than we have been able to determine before. Finding

RIGS-positive tissue in primary patients may help identify those patients, prospectively, who ultimately

clinically recur and may be useful in directing future

adjuvant therapy.

REFERENCES

1. Arnold MW, Schneebaum S, Berens A, et al. Intraoper- ative detection of colorectal cancer with radioimmu- noguided surgery and CC49, a second-generation monoclonal antibody. Ann Surg 1992;216:627-32.

2. Thurston MO. Development of the gamma-detecting probe for Radioimmunoguided surgery. In: Martin EW

Vol. 39, No. 5

Jr, ed. Radioimmunoguided surgery (RIGS) in the de- tection and treatment of colorectal cancer. Austin: RG Landes, 1994:41-65.

3. Galandiuk S, Wieand HS, Moertel CG, et al. Patterns of recurrence after curative resection of carcinoma of the colon and rectum. Surg Gynecol Obstet 1992;174: 27-32.

4. Arnold MW, Young DC, Hitchcock CL, et al. Radioim- munoguided surgery in primary colorectal carcinoma: an intraoperative prognostic tool and adjuvant to tradi- tional staging. Am J Surg (in press).

5. Bertsch DJ, Young DC, Arnold MW, et al. Radioim-

RIGS LOCATES CANCER PATTERNS 513

.

7.

munoguided surgery (RIGS) improves survival for recurrent colorectal cancer patients. Surgery (in press). Cohen Am, Minski BD, Schiloky RL. Colon cancer. In: DeVita VT Jr, Hellman S, Rosenberg SA, eds. Cancer: principles & practice of oncology. 4th ed. Philadelphia: JB Lippincott, 1993:929-77. Fidler IJ, Hart IR. Principles of cancer biology: biology of cancer metastasis. In: DeVita VT Jr, Hellman S, Rosenberg SA, eds. Cancer: principles & practice of oncology. 1st ed. Philadelphia: JB Lippincott, 1982: 80-92.

NOTICE

T h e Editorial Off ice has m o v e d . P lease m a k e a n o t e o f the n e w add re s s

Diseases o f the Colon & R e c t u m Editorial Off ice

201 1st A v e n u e S.W.

Suite 321

Roches te r , MN 55902