The Authors Reply
It is well supported that pathologic assessment of specimens has become more thorough with the adoption of fat-clearing techniques and methylene blue injection over manual dissection alone to meet the guidelines in pathologic reporting over recent decades.1 In addition, our data identified higher rates of inadequate LNH in low-income (19.5% vs 17.8%, p < 0.001) and low-education (17.2% vs 15.6%, p < 0.001) patients, and in those treated at Community programs compared with Academic/Research Programs (29.9% vs 17.1%, p < 0.001) and lower-volume centers compared with higher-volume centers (26.4% vs 21.1%, p < 0.001), indicating that inadequate LNH may be a proxy for a system of deficient oncologic care that surrounds the patient. Even despite standardization of the surgical, pathologic, and treatment guidelines expected of Commission on Cancer reporting hospitals over the past decade, 12.5% of specimens at the end of our study period yielded inadequate numbers of lymph nodes, suggesting that perhaps immunologic or other biologic alterations may contribute to inadequate LNH and may explain some molecular pathway for decreased cancer-specific survival by stage.2 In the absence of a clear causal relationship between LNH and the potential contributing factors listed above, we believe that causes for inadequate LNH are multifactorial. Regardless of the potential causes of inadequate LNH, its presence is a significant poor prognostic feature, and omission of adjuvant chemotherapy in this group is associated with decreased overall survival.
The authors agree that large national databases have significant shortcomings, especially in the details of procedures and chemotherapy regimens.3 However, we disagree with the assumption that conclusions drawn from large observational studies are essentially erroneous. One of the values of large, observational databases is the ability to observe treatment effects on a vast and generalizable scale. This is especially valuable when examining disease processes with infrequent negative outcomes, as in stage II colorectal cancer, where randomized clinical trials are limited by lack of resources, nongeneralizable samples, and underpowering.4 In these scenarios, well-designed, large observational studies offer the best possible evidence with which to guide the care of our patients. In this article, our central finding is that patients who have T3N0 cancer with inadequate LNH who undergo chemotherapy have improved survival compared with those that do not receive chemotherapy. From this, we conclude that T3N0 patients should be referred to medical oncology for treatment according to national guidelines. We find this conclusion to be far from erroneous and entirely appropriate for this high-risk group of patients.