Prognostic significance of mesenteric tumor nodules in patients with stage III colorectal cancer

Authors

  • Dorothy S. Lo MD,

    Corresponding author
    1. University of Toronto, Toronto, Ontario, Canada
    Current affiliation:
    1. St. Joseph's Health Centre, Toronto, Ontario, Canada
    • St Joseph's Health Centre, 30 The Queens Way, Toronto, ON M6R 1B5, Canada
    Search for more papers by this author
    • Fax: (416) 530-6656.

  • Aaron Pollett MD,

    1. University of Toronto, Toronto, Ontario, Canada
    2. Department of Pathology and Laboratory Medicine, Mount Sinai Hospital, Toronto, Ontario, Canada
    Search for more papers by this author
  • Lillian L. Siu MD,

    1. University of Toronto, Toronto, Ontario, Canada
    2. Department of Medical Oncology, Princess Margaret Hospital, Toronto, Ontario, Canada
    Search for more papers by this author
  • Steve Gallinger MD,

    1. University of Toronto, Toronto, Ontario, Canada
    2. Division of Surgical Oncology, Princess Margaret Hospital, Toronto, Ontario, Canada
    Search for more papers by this author
  • Ronald L. Burkes MD

    1. University of Toronto, Toronto, Ontario, Canada
    2. Department of Medical Oncology, Princess Margaret Hospital, Toronto, Ontario, Canada
    3. Department of Medical Oncology, Mount Sinai Hospital, Toronto, Ontario, Canada
    Search for more papers by this author

Abstract

BACKGROUND.

Tumor nodules are occasionally found in adjacent mesentery of colorectal cancer specimens and are felt to reflect a worse prognosis. The clinical significance of mesenteric tumor nodules was investigated.

METHODS.

A review of 786 patients with stage III colorectal cancer referred between 1995 and 1999 was undertaken. TNM staging was standardized by considering mesenteric nodules separately and not assigning them to T or N categories. Survival analyses were performed.

RESULTS.

Mesenteric tumor nodules were found in 116 (14.8%) patients: 48 (41.4%) with colon cancer and 68 (58.6%) rectal cancer. Mean age at surgery was 63 years. Adjuvant chemotherapy was given to 84.8% of colon cancer patients. Two (2.9%) rectal cancer patients received neoadjuvant chemoradiation, and 63 (92.6%) received adjuvant therapy (chemotherapy and/or radiation). In the cohort with mesenteric nodules, the median time to progression was 23.1 months, the median 5-year disease-free survival was 35%, and the median overall survival (OS) was 47.9 months, with 44% OS at 5 years. In the 19 (16.4%) patients with mesenteric nodules and no lymph nodes the 5-year OS was 60% (SEER stage II 5-year survival 82.5%), whereas in 97 patients who were lymph node-positive the 5-year OS was 40% (SEER 5-year survival stage IIIc 44.3%; stage IV 8.1%).

CONCLUSIONS.

In comparison to SEER survival data, the presence of mesenteric nodules appears to worsen the prognosis of any T/N0 disease to that of overall stage III disease. Mesenteric nodules with any T/N+ disease had prognosis similar to that of stage IIIC disease, but the prognosis was better than M1 disease. Cancer 2008. © 2007 American Cancer Society.

Colorectal cancer is the third most common cancer as well as the third most common cause of cancer death in both men and women, with an estimated 148,610 new cases diagnosed in the US in 2006.1 The staging of patients presenting with colorectal cancer is important for making treatment decisions and for estimating prognosis. However, in addition to the classic TNM (tumor, lymph node, metastasis) assessment of patient pathology, tumor nodules are occasionally found in the adjacent mesentery of cancer specimens. There is uncertainty about the clinical significance of these mesenteric nodules and how best to classify them pathologically. Clinically it has been felt that they bear a worse prognosis because they tend to be regarded as the transition between lymph node involvement and the development of metastatic disease. According to the TNM staging guidelines by the AJCC, these nodules are classified as part of T or N categories.2 More specifically, the recommendation is that if the nodule is smooth and round it should be considered nodal disease, and if it is spiculated and near a vessel, then it should be considered venous invasion (V1, V2) as part of tumor extension. In a prior study addressing this question, 544 patients who had resections for stage II or stage III colorectal adenocarcinoma at the Chikushi Hospital in Japan between July 1985 and 1995 were reviewed. It was found that tumor nodules were present in just over 17% of cases and they had the same prognostic significance as nodal disease.3 In another Japanese study,4 369 patients who had surgical resections for rectal cancer at the National Defense Medical College Hospital from 1980 to 1992 were reviewed. Cancer nodules that were not considered to be lymph node metastases were found in 30% of the total number of cases. However, looking at only the subset of the patients who had nonmetastatic disease, 22.7% had mesenteric nodules.

The objectives of our study were to investigate the clinical significance of mesenteric tumor nodules and to assess the impact of these nodules on survival relative to the nodal status in patients with resectable, nonmetastatic colorectal cancer.

MATERIALS AND METHODS

Ethics approval was obtained from the Princess Margaret Hospital Research Ethics Board. A list of 835 patients from the Princess Margaret Hospital Cancer Registry was generated using the search words: colon cancer, rectal cancer, stage III, and 1995–1999. We chose to look at patients with stage III disease because mesenteric nodules have been traditionally classified as stage III and thus could have been missed if we included only stage II patients. Patients with stage I or stage II disease and the presence of nodules would not have been captured in this study because “mesenteric nodules” was not a recognized search term and we could not select only patients with mesenteric nodules.

The charts of the 835 patients on this list were reviewed and patients with metastatic disease within 1 month of their surgery as demonstrated by staging radiology reports were excluded. We reviewed pathology reports of 786 patients with stage III colorectal cancer referred to Princess Margaret Hospital between 1995 and 1999. Those cases with pathology reports indicating mesenteric tumor nodules were highlighted. We standardized the TNM staging by assigning N status based on the number of definite lymph nodes reported. Mesenteric nodules were considered separately and not assigned to T or N categories. Mesenteric nodules were defined as tumor nodules without any evidence of lymph node tissue or lymph node architecture. All ambiguous pathology descriptions about tumor nodules were reviewed by a single pathologist (A.P.).

The charts of all patients with mesenteric tumor nodules were reviewed in detail. All data were collected by a single data abstractor (D.S.L.). All patients' paper charts were reviewed as were the electronic records, if available, as the electronic record only became available in 1999. Data on patient demographics, clinical history, details of pathology reporting, and clinical and radiological status of disease at each clinic visit was entered into a paper form. These data were then entered into an electronic database. Survival information was obtained through patients' medical records, from the Cancer Care Ontario registry, and by calling patients' family physicians. The date of disease progression was recorded as the date of either a radiological procedure that documented definite disease progression or as the date of a clinic appointment that a patient's medical oncologist felt there was disease progression. The opinion of the medical oncologist was used because it was felt that these patients would see their medical oncologists more often than physicians of other subspecialties during their treatment. Also, patients with colon cancer would not have routinely received radiation treatment, and patients with rectal cancer who had received radiation would most likely be followed by a medical oncologist as well. If both conditions were met, with radiological findings and clinical assessment indicating progression, then the earlier date would be used. Survival analyses were performed using Kaplan-Meier analyses.5 For patients lost to follow-up, the last date that they were seen (eg, date of clinic appointment, lab results, radiology results) was used, and then the time after this date was entered as unknown.

RESULTS

Of 786 cases reviewed, 116 patients or 14.8% were reported to have mesenteric nodules. In this cohort of 116 patients, 41.4% had colon cancer and 58.6% had rectal cancer. The male-to-female ratio was 1.2:1 with a mean age at the time of surgery of 63 years (Table 1). All tumors were adenocarcinomas, and the majority were moderately differentiated. The average tumor size was 4.3 cm. Resection margins were negative in all but 7 cases. These patients were not classified separately in this study as the numbers were small.

Table 1. Patient Characteristics
 No. (%)
Patients with mesenteric nodules116 (14.8)
Colon cancer48 (41.4)
Rectal cancer68 (58.6)
Females52 (44.8)
Males64 (55.2)
Mean age at surgery63 y

Of the 116 patients who had mesenteric nodules and were considered to have stage III colorectal cancer, 19 patients did not have any lymph node involvement. Two of these patients would otherwise be considered to have stage I disease and the other 17 patients stage II disease. The remaining 97 patients with mesenteric nodules had lymph node involvement and thus would be considered to have stage III disease. Of these patients, 61 patients had N1 disease, whereas the remaining 36 patients had N2 disease.

With respect to high-risk features, 6 (5.2%) cases had bowel perforation, 12 (10.3%) obstructive symptoms, 41 (35.3%) lymphovascular invasion, and 11 (9.5%) were T4 lesions. As high-risk features are mostly applicable to clinical decision-making in patients with stage II disease and only 5 of 19 patients with stage II disease had high-risk features in our study, further meaningful analyses could not be performed.

The majority of patients received treatment in addition to surgery. Of the patients who had colon cancer, 84.8% had adjuvant chemotherapy. Of the patients who had rectal cancer, 2.9% received neoadjuvant therapy (chemotherapy or radiation, or both) and 92.6% received adjuvant therapy (chemotherapy or radiation, or both). Twelve patients in the cohort of 116 patients had surgery alone. The reasons for not receiving adjuvant chemotherapy were mostly related to postoperative complications, other comorbidities, and patient preference (Table 2).

Table 2. Therapy
TreatmentNo. cases (%)
Adjuvant chemotherapy for colon cancer39 (84.8)
Neo-adjuvant therapy for rectal cancer2 (2.9)
Adjuvant therapy for rectal cancer63 (92.6)
Surgery alone12 (10.3)

The time to progression for all 116 patients was 23.1 months. Thirty-five percent of patients were disease-free at 5 years. The median survival duration was 47.9 months and the overall survival at 5 years was 44% (Table 3).

Table 3. Survival Data
  1. TTP indicates time to progression; DFS, disease-free survival; MST, median survival time; OS, overall survival.

TTP (mo)23.1
DFS (5 y)35%
MST (mo)47.9
OS (5 y)44%

The 5-year survival for the 19 patients with mesenteric nodules without lymph node involvement was 60%, whereas the 5-year survival for those 97 patients with lymph node involvement was only 40% (Table 4).

Table 4. Survival Data Compared With SEER Survival Data
StatusNo.5-Year OSSEER 5-y OS (6)
  1. OS indicates overall survival; LN, lymph node; MN, mesenteric nodules.

LN−, MN+1960%Stage II 82.5%
LN+, MN+9740%Stage III 59.5%
Stage IIIc 44.3%
Stage IV 8.1%

DISCUSSION

Mesenteric tumor nodules are not uncommon in patients with colorectal cancer. In this study the incidence of mesenteric nodules in stage III colorectal cancer was 14.8%, which is slightly lower than the 17.3% and 22.7% previously reported.3, 4 There has been no prior review of this pathologic feature from North American centers. Patient characteristics in this study are reflective of what is generally seen in clinical practice.

There has been uncertainty as to how to classify patients with mesenteric nodules in terms of staging. The AJCC recommends that if the mesenteric nodules are smooth and round, they should be considered nodal disease, and if they are spiculated and near a vessel then they should be considered an extension of the primary tumor and labeled as V1 or V2 for venous invasion.2 Traditionally, medical oncologists have considered them to represent at least stage III disease, even in the absence of lymph nodes, and most oncologists have felt they represent a poor prognosis more suggestive of metastatic disease. The results from Tateishi et al.'s study3 in 544 patients suggested that the 5-year survival rate for patients with either mesenteric nodules, metastatic lymph nodes, or both was worse than those without these pathologic features. Furthermore, graphical data illustrated no survival difference between those patients with lymph node involvement regardless of the presence or absence of mesenteric nodules. They concluded that the presence of a mesenteric nodule bears the same prognosis as having the presence of a metastatic lymph node. Ueno and Mochizuki4 examined several pathologic features including cancer nodules in the fat, which they termed extrabowel skipped cancer infiltration (ex), in 369 patients. In their review, the ex patterns included ‘scattering,’ ‘vessel invasion,’ ‘neural invasion,’ and ‘nodular.’ In our pathology department, aside from ‘nodular,’ the other pathologic features would probably have been as part of the T category in TMN staging. Thus, their study patients with ‘nodular’ ex patterns would be akin to our study population of interest. In their study the overall recurrence rate after curative resection was 52.8% in the ex+ group versus 24.0% in the ex− cases. Similarly, the survival rates for ex+ patients was significantly worse than for ex− patients. For patients without lymph node involvement, the presence of mesenteric tumor nodules decreased the 5-year survival rate from approximately (graphical data) 80% to 50% (P < .001). The authors concluded that these ex may be a significant prognostic predictor.

In this study we chose to compare survival data to Surveillance, Epidemiology, and End Results (SEER) data given that they represent 119,363 patients with colon adenocarcinoma during the time period of January 1, 1991, through December 31, 2000.6 The adjuvant treatment available and medical care during this time period should be comparable to what our patients received. We felt that the survival data of our retrospective series should be benchmarked against SEER instead of published literature data based on therapeutic clinical trials, because the latter may contain selected patient populations and may not be reflective of all cases at risk with the disease condition.

For the patients with mesenteric nodules without lymph node involvement, the 5-year survival of 60% was lower than the 82.5% seen for SEER stage II (84.7% for stage IIa and 72.2% for stage IIb). For patients with mesenteric nodules and lymph node involvement, the 5-year survival of 40% was lower than that seen in SEER overall stage III (83.4% for stage IIIa and 64.1% for stage IIIb) but was similar to that seen in SEER stage IIIc (44.3%.).6 However, the 44% 5-year survival for this group of patients in our study was greater than the 8.1% 5-year survival for SEER stage IV.6

Routine chemotherapy in the postoperative management of patients with stage II colon cancer is not recommended. According to the American Society of Clinical Oncology (ASCO) guidelines, adjuvant chemotherapy should be considered in patients with stage II colon cancer with high-risk features such as inadequately sampled nodes, T4 lesions, perforation, or poorly differentiated histology.7 Given that in the absence of lymph node involvement, patients with mesenteric nodules as seen in our study had a survival similar to the overall stage III disease SEER data, the presence of mesenteric nodes should be considered as a high-risk feature and these patients should be considered for adjuvant chemotherapy. Given the small numbers in our study, high-risk predictive features were not used in the regression analysis; however, future studies with greater numbers may allow evaluation to see if these features and mesenteric nodules are independently prognostic in stage II disease.

In patients with nodal involvement and mesenteric nodules the 5-year overall survival was found to be similar to that of stage IIIc disease (44.3%) when compared with SEER data.6 However, the prognosis was better than that seen in patients with distant metastatic disease. Thus, the presence of tumor nodules does not represent stage IV disease and therefore these patient should receive adjuvant chemotherapy with curative rather than palliative intent. Our findings are similar to those of Tateishi et al.3 published in 2005, confirming that mesenteric nodules have the same prognostic significance as having stage III disease. With respect to the AJCC recommendations for interpretation of mesenteric nodules, these results support the interpretation of mesenteric nodules as lymph nodes rather than as part of T (V1, V2) classification.

Clearly these results suffer from the usual limitations of a retrospective analysis from a single institution. Potentially confounding factors could not be taken into account; patients with mesenteric nodules could have had other poor risk features that were not captured. It is possible to perform a prospective study investigating the same question as in this study; however, the results would not be available until several years in the future.

This study was also limited by the lack of standardization of pathology reporting over the time period studied. Over that time period the number of total lymph nodes examined were inconsistently reported. There is a possibility due to lack of standardization of the number of lymph nodes examined that mesenteric tumor nodules could have been missed and not evaluated in the pathologic evaluation.

In interpreting the survival results it should be noted that we evaluated only patients who were classified with stage III disease. Patients with stage I or stage II disease and the presence of nodules would not have been evaluated. These patients would likely have had better survival results than the patients included in this study.

Also, the recent changes in pathology, chemotherapy, radiation, and surgery have altered the outcomes of patients with colorectal cancer. Pathology reporting has been standardized to examine at least 12 lymph nodes. Total mesorectal excision has become an optimal goal. New chemotherapy agents such as oxaliplatin have become standard since the study time period. Also, with adoption of neoadjuvant chemoradiation for rectal tumors, mesenteric tumor nodules may be treated and absent from surgical pathology reports.

The number of patients is small and a larger number of patients would be required to confirm these results. It would be interesting to compare our data with those of another large institution.

In conclusion, this study, in addition to others examining this question, provides data for continued evaluation of AJCC guidelines in classifying patients with mesenteric nodules.

Ancillary