Back to 2016 Annual Meeting Posters

---

Impact of a 12 Lymph Node Harvest Quality Metric on Colon Cancer Outcomes
*Ramzi Amri, *Lawrence S. Blaszkowsky, *Anne M Dinaux, *Hiroko Kunitake, Liliana G Bordeianou, David L Berger
Harvard Medical School/Massachusetts General Hospital, Boston, MA

Objective: A lymph node (LN) harvest of 12 or more nodes has become a quality metric for adequacy of resection. A <12LN yield is considered a high-risk threshold for stage II patients. In theory, as institutions achieve ≥12LN resections, there should be a migration of patients previously felt to be stage II into stage III, as a greater LN harvest should capture otherwise missed positive nodes. This should then lead to outcome improvements for stage II cases, as it will include only “true” stage II patients.
Design: Retrospective review of prospectively maintained data repository.
Setting: Tertiary care center.
Patients: All operative AJCC stage I-III colon cancer patients (2004-2011;n=926).
Main outcome measures: Changes in ≥12LN rates, recurrence and mortality over two time intervals (2004-2007;2008-2011).
Results: Staging distributions did not change significantly, while the number of stage II patients with <12LN decreased drastically from 18.1% to 2.4% (P<0.001). Concurrently, recurrence rates remained stable (P=0.289), while survival improved (P=0.025).

	2004-2007 (n=461)	2008-2011 (n=465)	P-value
Stage I	27.7% (137)	26.9% (129)	0.82
Stage III	35.4% (175)	35.7% (171)	0.93
Stage II	30.2% (149)	34.2% (164)	0.17
Adjuvant chemotherapy	18.1%	18.3%	0.97
Screening diagnoses	19.5%	22.6%	0.5
≥12LN yield	81.9%	97.6%	<0.001
LN yield (median, IQR)	17 (13-22)	23 (18-30)	<0.001
3-yr recurrence*	10.6% ±2.7	16.5% ±3.0	0.289
5-yr recurrence*	14.2% ±3.1	8.1% ±3.2
3-yr survival*	82.9% ±3.2	86.8% ±2.8	0.025
5-yr survival*	71.7% ±3.9	81.4% ±3.3
* Kaplan-Meier estimates (percentage outcome ±standard error)

Conclusions: In a center that already had a relatively high success rate in harvesting ≥12LN, improving the rate to nearly 100% did not impact surgical outcomes. With nearly all patients above the ≥12LN threshold, it becomes less useful as a prognostic litmus test; implying that a node of harvest of ≥12LN reached its upper limit as a quality metric. In similar situations, other factors and quality metrics should be explored to estimate risk and resection quality in stage II colon cancer patients.

---

Back to 2016 Annual Meeting Posters