Colorectal LLM Wiki

Browse
Home / Procedures / Complete Mesocolic Excision with D3 Lymphadenectomy
Procedure created 2026-08-25 · updated 2026-08-25
colorectalcolon-cancerCMElymphadenectomyminimally-invasivetextbookreference

Complete Mesocolic Excision with D3 Lymphadenectomy

A deep dive into complete mesocolic excision (CME) with central vascular ligation (CVL) and Japanese-style extended (D3) lymphadenectomy for colon cancer — the wiki's first dedicated treatment of this topic, cross-linked from Colon Cancer Resection's brief existing CME section (which this page substantially extends rather than duplicates). Content below is from a textbook reference chapter (ASCRS Ch37 - Minimally Invasive Complete Mesocolic Excision with Extended Lymphadenectomy for Colon Cancer), not primary literature.

Terminology: CME + CVL = D3 dissection

Hohenberger et al.'s 2009 seminal publication popularized "complete mesocolic excision" (CME), building on Bill Heald's TME concept (mesorectal dissection along embryologic planes, popularized for rectal cancer) and earlier work by Jamieson (1909) and Enker (1979). CME comprises: (1) sharp en bloc dissection of the colon and mesocolon with complete removal of the mesenteric envelope along the embryologic plane between the mesocolic and parietal fascia; (2) extirpation of all lymphovascular tissue from the pericolic area to the lymph nodes at the base of the feeding arteries; (3) central vascular ligation (CVL) — high vascular tie at the most proximal extent of the feeding arteries and veins; and (4) adequate bowel resection length to ensure adequate lymph node dissection.

Modern CME practice has evolved from the original Kocher-maneuver-based right colectomy description toward a modified approach matching the Japanese Society for Cancer of the Colon and Rectum (JSCCR) definition of D3 lymphadenectomy, in which only anterior dissection of the SMV is required (not the more extensive dissection the original CME description implied). Because "CME + CVL" as loosely used in Western literature is a misnomer with variable, inconsistently-applied extent of lymphadenectomy — hampering comparison of surgical technique and specimen quality across studies — this page (following the source chapter) uses "CME + CVL" to mean CME performed with radical lymphadenectomy corresponding to JSCCR D3 lymph node dissection (CME + D3).

The difference between "conventional" and CME + D2 dissection is not analogous to left vs. right colectomy: many surgeons already routinely ligate the IMA at the aorta during proctectomy for rectal cancer (functionally CME + D2 for that vessel), so CME + D3 for left-sided colon tumors (requiring exposure of the aortic wall for additional nodal tissue at the base of the IMA) is not a dramatic step up in complexity. The same cannot be said for right-sided tumors — safe exposure/dissection of SMV tributaries and the trunk of Henle, with added risk of injury and a long learning curve, has made surgeons more reluctant to adopt this approach for right and transverse colectomy, especially absent high-level evidence of long-term oncologic benefit relative to standard CME + D2.

Lymph node classification (D1/D2/D3/D4)

Per the Japanese classification of colonic regional lymph nodes (Table 37.1, Fig. 37.1–37.2):

  • D1: complete dissection of epicolic/paracolic lymph nodes along the marginal artery in the relevant colon segments, with no or incomplete dissection along the tumor-supplying arteries.
  • D2: complete D1 dissection plus intermediate lymph nodes along the tumor-supplying arteries (ileocolic, right colic, middle colic, left colic, sigmoid, or IMA — from the origin of the last sigmoid artery to the origin of the left colic artery). For right-sided tumors, D2 includes ligation of the ileocolic artery (ICA) at its root, at the level of the lower border of the duodenum; if the right branch of the middle colic (or main middle colic) artery is ligated as part of extended right colectomy (ascending colon/hepatic flexure tumors), high ligation can be performed without exposing the SMV. For left-sided tumors, D2 includes ligation of the IMA just proximal to the left colic artery origin.
  • D3: complete D1-D2 dissection plus central lymph nodes — for left-sided tumors (including mid-transverse tumors), nodes along the IMA between the aorta and the left colic artery; for right-sided tumors, nodes along the SMV and lateral to the SMA. For right-sided tumors, the anterior surface of the SMV is exposed (SMA dissection not required), vessels transected at their origin with removal of all nodal tissue along the SMV and lateral to the SMA (nodal groups 203 and 213) — nodal group 223 also removed for extended right colectomy. For left-sided tumors, the IMA is divided at its base from the aorta, with removal of all nodal tissue between the aorta and the left colic artery (nodal group 253) — nodal group 223 also removed with extended left colectomy.
  • D4: complete D1-D3 dissection plus nodes along the aorta/IVC or SMA/SMV central to the origin of the middle colic artery.
  • Alternative JSCCR n(#)(+) location-based classification: n(1)(+) = LN metastasis in the D1 area but within 5cm proximal/distal of the tumor edges; n(2)(+) = LN metastasis in D1 area >5cm from the tumor, or in the D2 area; n(3)(+) = LN metastasis in the D3 area; n(4)(+) = LN metastasis in the D4 area (considered distant metastasis).

Decision-making: which patients get D2 vs. D3

The scope of lymph node dissection is based on preoperative clinical staging and/or intraoperative assessment of suspected lymph node metastasis. In Japan, endoscopic assessment of depth of tumor invasion (chromoendoscopy, narrow-band imaging, occasionally EUS) plays an important role in clinical T staging, alongside CT/MRI.

  • pTis tumors completely removed endoscopically: lymph node dissection not typically required; if bowel resection is elected, D1 dissection is performed.
  • pT1 tumors removed endoscopically: D2 dissection recommended, given ~10% risk of D1/D2 LN metastasis and 1.9% risk of D2 node metastasis (JSCCR registry data).
  • cT2 cancer: at least D2 recommended; D3 can be considered given a 0.7-1% risk of central/main node metastasis and the limited accuracy of imaging-based preoperative cT staging.
  • cT3/T4 tumors: D3 dissection is the standard of care. In the 2000-2004 JSCCR registry (25,617 patients treated for colorectal cancer), lymph node metastasis around the origin of the feeding vessel (main/central/D3 node) was found in 0.9%, 2.3%, and 5.7-6.7% for pT2, pT3, and pT4a/pT4b tumors respectively.

Outside centers with expertise in endoscopic depth-of-invasion assessment, recommendations based on standard preoperative radiology-based staging are difficult to apply clinically, since this can only be determined with certainty postoperatively after histologic evaluation of the resected specimen — the exception being patients with superficial tumors undergoing complete endoscopic removal before colectomy.

Adoption of D3 dissection among Japanese surgeons increased from 58.4% (2001) to 75% (2010), per photodocumentation of D3 specimens in the JSCCR tumor registry.

Quality indicators of CME resections

Analogous to proctectomy specimen grading (Quirke grading), CME resection quality indicators include: (1) documentation of mesocolic plane dissection grade (muscularis propria plane [poor] / intramesocolic plane [intermediate] / mesocolic plane [good]); (2) number of lymph nodes harvested (which may include JSCCR nodal-station classification); (3) specimen morphometry — distance from vascular tie to tumor, distance from vascular tie to bowel wall, area of mesentery resected (Fig. 37.3); (4) large/small bowel length. Grading is documented via high-resolution photographs of the unfixed, unopened specimen (front and back, mesentery laid flat without stretching, alongside a ruler) — standard Japanese practice, increasingly adopted since the 2005 JSCCR guidelines recommending D3 dissection.

Comparative specimen morphometry (review of photographs from non-CME England, CME+CVL Germany, and D3-dissection Japan specimens): no significant differences in tumor location, pT, or pN stage between groups. Length of resected bowel and area of mesentery were significantly greater in NCME and CME+CVL specimens than D3. Distance from vascular tie to bowel wall was similar between CME+CVL and D3 specimens; both longer than NCME. Rates of mesocolic-plane surgery and lymph node count were highest in CME+CVL, followed by D3 — reflecting JSCCR guidelines that bowel resection should extend 5cm beyond the feeding vessel toward lymph flow and 10cm away from the tumor in the opposite direction, whereas longitudinal tumor spread >10cm beyond the tumor is rare in Japanese studies (0% left-sided, 1-4% right-sided) — in contrast to CME's principle of removing the next feeding vessel beyond the 10cm bowel margin to ensure adequate lymphovascular clearance. These length differences may also partly reflect differing body habitus between populations.

There is currently no consensus on which quality indicators are most critical when evaluating/ comparing CME resection quality; lymph node count alone is a biased surrogate (impacted by patient and pathologic-processing variability). West and Hohenberger (2009) demonstrated that CME with central ligation removed significantly more tissue than NCME resections, in addition to increasing lymph node yield, using the mesocolic-plane-grading + specimen-morphometry framework above. Contemporary trials evaluating CME + D2/D3 outcomes have incorporated CME quality-indicator assessment to validate completion of CME + D3 by participating surgeons — e.g. the JCOG0404 RCT used central peer review of photodocumentation of the completed D3 dissection surgical field and specimen, highlighting surgical margins, for quality control.

Anatomical considerations for D3 dissection

Surgeons adopting CME + CVL/D3 must thoroughly understand common and less-common variations in central vascular anatomy, especially around the SMV and SMA, to safely complete D3 dissection, lower the risk of inadvertent ligation of incorrect vessels, and avoid bleeding from direct vascular injury or inadvertent traction injury to venous tributaries.

  • Gastrocolic trunk of Henle (GCT) / gastro-pancreatico-colic trunk (GPCT): the confluence of the right gastroepiploic vein (RGEV), superior (accessory) right colic vein (SRCV), and anterior superior pancreaticoduodenal vein (ASPDV) — present in 37.8-89% of cases, with several described subtype classifications based on SRCV presence/location or number of right colic veins (RCV) merging with the trunk. Most common configuration: a single colic vein joins the RGEV and ASPDV (55.8%). In 20.7-21.6% of cases, there is no common trunk between the colic and gastroepiploic vein.
  • SMV tributaries: 1, 2, or 3 middle colic veins found in 72.1-74.1%, 22.4-26.1%, and 1.8-3.5% of patients respectively; the MCV drains directly into the SMV in 86-89%. An SRCV is identified in 28.8-83.6% of cases; an RCV is identified in only 19-24.3% but always drains directly into the SMV. A single ICV (ileocolic vein) is found in 100% of cases, draining directly into the SMV in 92.8-100% (or alternatively into the GCT or jejunal trunk). In rare cases the ICV may not travel with the ICA and can enter the SMV up to 3cm cranial to the base of the ICA.
  • Arterial anatomy relative to the SMV: a single ileocolic artery originates from the SMA in 100% of cases and crosses the SMV either anteriorly or posteriorly. A right colic artery (RCA) is identified in only 10-63% of cases; when present, it originates from the SMA in 33% or alternatively from the ICA/MCA, crossing the SMV anteriorly in 87-97%. The SMA typically courses just left and posterior to the SMV. The MCA arises from the SMA in 100% of cases and is a single artery in 88% (double in 11.7%; three MCAs exceedingly rare). During SMV dissection, critical variations to identify: whether the ICA crosses anteriorly or posteriorly to the SMV, and where the ICV drains into the SMV (which could be a few centimeters cranial to the base of the ICA); early identification of the ASPDV, SRCV, and RCV (the first venous tributaries of the Henle trunk) as dissection proceeds to separate the right mesocolon from the head of the pancreas reduces the risk of inadvertent venous tearing from overzealous traction.
  • IMA anatomy is more consistent: splits into the left colic artery and superior rectal (hemorrhoidal) artery 2-3cm from its aortic origin. No universally agreed nomenclature exists regarding the IMA and its branches — some surgeons assert the IMA "becomes" the superior rectal artery after crossing the common iliac artery, once it has given off some sigmoidal branches.

Techniques of MIS CME + D3

Right-sided and transverse colon cancer

  • Medial-to-lateral approach: similar to conventional CME + D2, but after the ileocolic vessels are ligated at the root of the SMV/SMA, dissection continues cephalad anterior to the SMA/SMV toward the right colic vessels, GCT, and middle colic vessels. Right colic vessels (if present) are dissected/transected at their base; the GCT is carefully dissected with preservation of the ASPDV and RGEV; RCV/SRCV transected. Middle colic vessels dissected at their base from the SMA/SMV. Extent of colonic mobilization and middle-colic-vessel division follows tumor location: for cecal/ ascending-colon tumors, the main and left branch of the middle colic artery are spared, with only the right branch divided; for hepatic-flexure/transverse-colon tumors, the middle colic vessels are usually divided at their base. Arterial flow from the IMA retrograde to the transverse colon should be assessed before dividing the main trunk of the MCA, especially in older/atherosclerotic patients. For distal-transverse tumors, left colectomy is performed with sparing of the right MCA branch and IMV ligation with splenic-flexure takedown.
  • "SMV-first" approach (popularized for robotic CME + D3): the base of the mesocolon is exposed, transverse colon retracted superiorly, an incision made inferior to the ileocolic vessels extending proximally until the SMV is identified; all lymphatic/areolar tissue anterior to the SMV surface is dissected en bloc heading cephalad, with SMV tributaries sequentially divided starting with the ICV, then ICA ligation at the base of the SMA. Cephalad dissection continues along the SMV until the middle colic vessels are identified; a right colic artery may be identified/divided, followed by ligation of the MCA or its branches based on tumor location. The GCT is then identified and dissected along the right anterolateral aspect of the SMV.
  • Retrofascial uncinate-first approach (Stefan Benz): based on an open-book model to understand vascular and meso-/fascial anatomy — the different planes (mesogastric, mesocolic, ileocolic) are depicted as pages of a booklet with the GCT as the vertical axis and the pancreas as the book's back. The avascular plane between the "pages" must be dissected according to CME principles and prior to dividing vessels, to avoid vascular injury. A 9-step protocol with associated "critical views" guides completion: 1. Dissection starts at the duodeno-jejunal flexure; the peritoneum is incised at the inferior/ posterior border of the visible duodenum; medial-to-lateral retromesenteric dissection extended caudally along the ascending mesocolon; right ureter/gonadal vessels separated from the mesocolon; hepatic flexure reached cranially; Toldt's fascia incised along the lower border of the duodenum, separating the uncinate process's anterior surface from the mesocolon (Critical View 1: third portion of the duodenum and uncinate process visible). 2. Small bowel repositioned left, ileocolic vessels placed on tension; a V-shaped configuration between the ileocolic and superior mesenteric vessels identified (Critical View 2: incision point highlighted after tensioning the ileocolic pedicle). 3. Peritoneum incised ~3cm distal to the ileocolic root, between ileocolic and SMV; dorsal dissection plane entered, pancreatic head/duodenum identified; dissection extended along the SMV (180-270° of its circumference); ICV divided after 360° dissection at its SMV confluence (Critical View 3). 4. ICA identified crossing anterior or posterior to the SMV, dissected 360° at least at the SMV level; SMA identified, lymphatic tissue on its right side resected; after ICA division, SMV dissection continued toward the GCT (Critical View 4). 5. Mesogastric "page" separated from the mesocolic "page" — stomach retracted cranially, gastrocolic ligament incised ~5cm to the left of midline (preserving the gastro-omental arcade, since the lesser sac is usually obliterated on the right side near the middle colic vessel); lesser sac entered, dorsal stomach visualized (Critical View 5). 6. Gastrocolic ligament divided medial-to-lateral; omental vessels transected; avascular plane between transverse mesocolon and dorsal mesogastrium (containing gastroepiploic vessels and infrapyloric nodes) dissected — crosses the second duodenal portion, followed laterally toward the hepatic flexure, merging with step 1 (Critical View 6: lesser sac dissected, mesogastrium/ mesocolon sulcus visible, SRCV/RGEV merging visible at the sulcus base — excessive tension here is a described bleeding point). 7. Mesocolon retracted cranially, transverse mesocolon incised anterior to the SMV; transverse colon divided stepwise cranial-to-caudal, dissecting the origin of the MCA's right branch (usually found at the base); right branch mobilized 360° preserving left-branch integrity, then divided; middle colic vein divided only if it impedes lymphadenectomy (Critical View 7). 8. Anterior surface of the trunk of Henle exposed; all veins merging with it from the mesocolon divided (SRCV/RCV); RGEV and ASPDV identified and preserved. Central dissection complete (Critical View 8: completed CME with central ligation). 9. Lateral attachments of the right colon divided; specimen exteriorized (extracorporeal anastomosis) or colon/terminal ileum divided in situ (intracorporeal anastomosis), with care to preserve ileocolic lymphatic drainage.
  • Lateral-to-medial approach: preferred if the base of the ileocolic pedicle is difficult to identify. White line of Toldt incised along the right gutter; embryological plane between right mesocolon and Gerota's fascia dissected sharply (avoiding right ureter/kidney/gonadal vessel injury), extended medially toward the duodenum. Ileocolic pedicle base exposed/dissected with SMV lymphatic-tissue dissection; ileocolic pedicle divided at its base, sharp dissection continued along the SMV to the root of the middle colic vessels; after dissecting all lymphatic tissue at the base of the middle colic pedicle, each branch skeletonized with arterial/venous ligation based on tumor location. Transverse colon retracted caudally; gastrocolic ligament divided along the right gastroepiploic vessels.
  • Cranial-to-caudal approach (increasingly adopted in Asia): early access to/exposure of the pancreas, GCT, and SMV. Omental bursa opened, GCT exposed using the right gastroepiploic vessel and SRCV as landmarks; after dividing the SRCV, SMV and middle colic vein identified, the latter divided at its root. Complete lymphatic tissue dissection along the SMV proceeds cranial-to-caudal; additionally, the middle colic artery is dissected and its right branch divided at its base; transverse colon pulled ventrally with lymphatic-tissue dissection along the SMV continuing in a cranial-caudal approach. Ileocolic and right colic vessels dissected/divided at their origin; ascending/transverse mesocolon dissected off Gerota's fascia, retroperitoneum, pancreatic head, and duodenum using a medial approach.

Left-sided colon cancer with D3 dissection

Medial-to-lateral approach to the sigmoid mesentery preferred: medial aspect of sigmoid mesocolon incised just above the sacral promontory, medial-to-lateral mobilization carried sharply through the avascular plane (as in standard proctectomy medial-to-lateral approach). IMA elevated/skeletonized at its base, with dissection of all lymphatic/areolar tissue at its root — care to avoid superior hypogastric plexus, left ureter, and gonadal vessel injury during medial-to-lateral mesocolic dissection. Following D3 dissection, tumor-specific CME principles govern extent of colonic mobilization and division of the IMA or left colic artery based on tumor location: for proximal left-colon tumors, the left colic artery may be divided at its origin, sparing the root of the IMA/ superior rectal artery if anastomosis to the distal sigmoid is contemplated. If the entire sigmoid is to be resected (mid-left-sided, sigmoid, rectosigmoid tumors), the IMA is ligated at its origin on the aorta. In either case, the IMV is typically divided just inferior to the pancreatic border, allowing inferior rotation of the proximal colonic conduit when combined with complete splenic-flexure mobilization. The ligament of Treitz is the landmark for identifying the IMV — care must be taken to ensure no arterial branch runs in the tissues adjacent to the IMV before dividing it, since the dominant arterial flow from the left branch of the middle colic artery occasionally runs at the root of the mesentery here; inadvertent division could compromise arterial flow to the proximal colonic conduit. Left mesocolon dissected away from Gerota's fascia, lesser sac entered, root of transverse mesocolon freed from the inferior border of the pancreas. Splenic flexure can also be mobilized via the medial-to-lateral approach. Once completed, lateral attachments of the left colon are taken down, as is the splenocolic ligament. Omentectomy completed just below the gastroepiploic vessels with preservation of the left gastroepiploic vessels, unless infiltrated by tumor.

Short-term outcomes of CME

A 2019 systematic review found two of three eligible studies demonstrating longer operative time with CME vs. non-CME (NCME) for right and left colon cancer; pooled perioperative morbidity across 6 studies was 22.5% (CME) vs. 19.6% (NCME) — not significant.

  • Danish Colorectal Cancer Group (DCCG): 529 patients undergoing CME/D3 for stage I-III colon cancer (2008-2013, single institution, early implementation period). Laparoscopic approach attempted in 48.8% with a 19.9% conversion rate — the only study documenting a higher rate of intraoperative organ injury with CME vs. a non-CME historical cohort (9.1% vs. 3.6%), with injuries including splenic and SMV injuries.
  • JCOG0404 RCT (Japan): 521 open vs. 527 laparoscopic CME + D3 for stage II/III right- and left-sided cancer — conversion rate 5.4%, longer OR time, and a higher (not statistically significant) rate of intraoperative organ injury in the laparoscopic CME group (3.6% vs. 1.7%). Injuries during laparoscopic (not open) surgery included duodenal, arterial, and gastrocolic-trunk injuries.
  • Freund et al.: 1.6% vascular injury rate across 159 open and 145 CME cases for right colon cancer over a 10-year period — injury mechanism attributed to misperception/lack of understanding of venous/arterial branch variations at the root of the mesentery, and avulsion of the middle colic vein from excessive traction.
  • Other CME-related risks: organ injury, chyle leak from extensive nodal dissection, severe diarrhea from injury to the superior mesenteric plexus, and genitourinary dysfunction during left-sided dissection near the IMA (inferior mesenteric/superior hypogastric plexus injury). No additional procedural risks documented specifically for laparoscopic/robotic-assisted CME performed by experienced high-volume surgeons.
  • Learning curve: limited data. Melich et al. described excellent lymph node harvest, a low major- complication rate (3.6%), and no conversions during an initial 81-laparoscopic-CME-case series; OR time decreased from 250 to under 200 minutes after 81 cases completed over 3 years.

Laparoscopic approach for CME + D3

Technically challenging (vascular dissection/skeletonization at the root of the mesentery); increased operative time, vascular-injury risk, and long learning curve. Conversion rates range 5-22%. When performed by trained/experienced surgeons, laparoscopic CME + D3 is associated with equivalent OR time, morbidity, and oncologic outcomes relative to open — the MIS approach may improve short-term outcomes (incision-related postoperative morbidity, bowel function recovery, length of stay). Relative contraindications mirror conventional cancer surgery (emergency setting, hemodynamic instability, inability to tolerate pneumoperitoneum, prior extensive abdominal surgery, T4 disease anticipated to need difficult en bloc resection, large/bulky tumors).

  • Danish DCCG dataset: laparoscopic CME adoption increased to 49% (2008-2013), with conversion rates dropping slightly (22% → 19.9%) over the study period. For right colectomy with CME specifically, laparoscopic adoption was 33% with a 21% conversion rate, reflecting difficulties unique to D3 dissection near the SMV, often necessitating an open approach for adequate vascular exposure.
  • Systematic review/meta-analysis (1 RCT + 7 case-control trials comparing laparoscopic with open CME + central lymphadenectomy): despite heterogeneity, a trend toward longer OR time and shorter length of stay with laparoscopic approach; laparoscopic CME associated with significantly lower wound infection rate; no differences in lymph node yield, local/distant recurrence, or OS.
  • Large multicenter RCT (521 open vs. 527 laparoscopic CME + D3, stage II/III right- and left-sided cancer, 2004-2009, 30 Japanese centers, surgeons with ≥30 open and ≥30 laparoscopic CME + D3 procedures previously completed): conversion rate 5.4%; longer OR time and a higher (not statistically significant) intraoperative-organ-injury rate with laparoscopic CME (3.6% vs. 1.7%); injuries during laparoscopic surgery (not open) included duodenal, arterial, and gastrocolic-trunk injuries. Laparoscopic approach associated with lower blood loss, faster resolution of ileus, and shorter hospital stay. Postoperative morbidity was also lower with laparoscopic vs. open. 86% of stage III patients in each group received adjuvant chemotherapy, with no differences in 5-year OS (90.4% vs. 91.8%, open vs. lap) or DFS (80% vs. 79%). Central review of procedural videos/ intraoperative photographs used for quality control. Demonstrated that in the hands of highly experienced surgeons, laparoscopic CME + D3 dissection is safe for stage II/III colon cancer, with improved short-term outcomes and equivalent oncologic outcomes vs. open.
  • COLD trial (multicenter RCT comparing D2 vs. D3 dissection): first 100 patients accrued showed 79-84% of procedures completed laparoscopically, no differences in OR time, blood loss, conversion rates (12% vs. 13%), or 30-day Clavien-Dindo grade I-IV complications (47% vs. 48%) between D2 and D3 groups — supporting the notion that experienced D2 surgeons can transition to D3 dissection with excellent results. There were three injuries during D3 dissection (left gonadal artery transection, two small bowel injuries requiring conversion for the latter). Participating surgeons required a minimum of 20 D2 and 20 D3 lymphadenectomy cases with submitted unedited technique-assessment videos as a study-participation requirement.
  • Sammour et al. (USA): institutional experience with laparoscopic/robotic CME + D3 in 141 patients vs. 56 CME + D2 cases. No significant difference in short-term outcomes except a higher median lymph node count with CME + D3 (31 vs. 27), though the number of positive nodes did not differ. Not possible to determine incidence of centrally positive nodes since nodal stations aren't routinely recorded on pathology in the USA, nor was level of vascular ligation (based on mesenteric/pedicle length between vascular tie and tumor/bowel wall) recorded.

Long-term oncologic outcomes of CME with D3

A 2019 pooled analysis of eligible studies comparing long-term outcomes between CME and NCME resections: 5 studies showed no differences in 3- or 5-year OS; 1 study showed lower 5-year cancer- specific survival (CSS) in CME + D3 vs. CME + D1/D2. 4 studies demonstrated higher DFS or CSS with CME — interpretation limited by historical biases, study-design heterogeneity, extent of lymphadenectomy variability (D1/D2/D3/CVL), and inconsistent pathologic assessment of CME specimens. Overall, the pooled evidence did not consistently support CME's oncologic superiority.

  • Hohenberger et al.'s 2009 seminal article: 1329 consecutive stage I-III CRC patients from the ERCRC registry — standardized CME with central ligation adoption resulted in a decrease in 5-year locoregional recurrence (6.5% → 3.6%) and improvement in 5-year cancer-related survival (82.1% → 89.1%) between the 1978-1984 and 1995-2002 time periods; only 5.6% of patients received adjuvant chemotherapy in that timeframe.
  • Updated 1978-2014 ERCRC registry (same authors), contrasting pre-CME, CME-development, implementation, and CME-practice periods: rise in ≥12-lymph-node-harvest achievement from 84.8% to 100%, and R0 resection rate from 97% to 100%. Among stage III patients, significant reduction in 5-year locoregional recurrence (14.8% → 4.1%) and improvement in 5-year CSS (61.7% → 80.9%) and 5-year OS (53.1% → 68.6%) reported between 1978 and 2009, against a background of increased adjuvant-chemotherapy use (0% → 79%) between 1978 and 2014 — this oncologic benefit was likely confounded by historical changes in adjuvant therapy.
  • Danish DCCG population-based study: 2008-2011 dataset, 364 CME performed at one specialized center vs. 1031 non-CME at 3 other centers, demonstrated higher 4-year DFS in the CME group, especially for stage I/II. 2008-2013 dataset for right colectomy specifically: 256 CME (same specialized center) vs. 813 non-CME (3 other centers) — 5.2-year recurrence rate for stage I-III colon cancer significantly lower in CME vs. non-CME (9.7% vs. 17.9%).
  • Cohort study: 189 consecutive stage I-II colon cancer patients, CME + D3 at one hospital vs. CME + D2 at 3 other institutions (2007-2008), demonstrated significant improvement of 3-year OS (88.1% vs. 79%) and DFS (82.1% vs. 74.3%) in the CME + D3 group.
  • Smaller institutional study: 45 CME vs. 58 historical non-CME for right colon cancer — CME associated with significantly lower local recurrence rate (0% vs. 20.6%) and higher 5-year CSS in stage III patients, despite similar adjuvant-chemotherapy rates (47% vs. 43%, CME vs. NCME).
  • Much of the criticism of these studies relates to reliance on historical cohorts as controls (introducing significant bias), inexplicably high rates of local/locoregional recurrence in some historical NCME cohorts (14.8% and 20.6%), and inconsistent reporting of CME quality indicators (other than lymph node count) — as a result, the extent/quality of CME and D2-vs-D3 dissection could not always be validated.
  • More recent interim analysis (220 open CME + D3 vs. 110 open non-CME for right/left colon cancer, prospective non-randomized double-blind trial): improved 3-year local recurrence-free survival in stage II/III cases, but no differences in OS or DFS.
  • Ongoing trials: the COLD trial (multicenter RCT, mostly laparoscopic D2 vs. D3 dissection) will provide long-term outcomes — the first 100 patients accrued showed equivalent lymph node harvest between groups. Three RCTs currently underway in China evaluate long-term outcomes of laparoscopic CME + D2 vs. CME + D3, for right colon cancer (RELARC, SLRC trials) and right- and left-sided cancer (LCME trial).

Central node metastasis: oncologic implications

Locoregional lymph node metastasis pattern in colon cancer remains poorly understood, with two opposing theories: the Halsted model (stepwise progression from primary tumor to pericolic, intermediate, then central nodes, ultimately to distant organs) and the Fisher model (metastatic progression is non-linear and can bypass proximal nodal basins altogether). Under the Fisher model, removal of all lymphatic tissue at all three nodal stages is unlikely to impact oncologic outcomes; under the Halsted model, systematic removal of central lymph nodes in stage III disease during D3 dissection — including potential skip metastases — may improve staging and local disease control and impact long-term oncologic outcomes, independent of adjuvant treatment.

  • Incidence of D3+ nodal metastases increases with tumor stage. In stage III disease, apical/central node metastasis has been reported as high as 19.7% (244 stage III right colon cancers treated with CME + D3, with a skip-metastasis rate of 19.8%). Among 1355 stage III colon cancers treated with D3 dissection across 71 Japanese centers (minimum 5-year follow-up from the 2000-2002 JSCCR registry), the rate of D3+ nodes was 8.3%.
  • Review of the 1999-2007 JSCCR database (4034 stage III tumors treated with D3 dissection): significant differences in central node metastasis pattern between right- and left-sided colon cancer, and D3+ metastasis rate significantly higher in right-sided tumors (8.5% vs. 3.7%), as were skipped node metastases (13.7% vs. 9%). 5-year DFS was similar between right- and left-sided cancer, but 5-year unadjusted OS was significantly worse for right-sided cancer (77.4% vs. 80.9%). By multivariate analysis, D3+ status was associated with worse OS in left-sided cancer but neither DFS nor OS were affected by lymph node metastasis pattern in right-sided cancer — suggesting D3 status may have greater prognostic significance, and D3 dissection a greater impact on survival, in stage III left-sided colonic cancer.
  • Prognostic significance of central node metastasis: in a JSCCR registry of 1355 stage III colon cancers, incidence of central node metastases in D3+ specimens was 13.8% in stage III disease, with a 2.7% rate of skip metastasis. After adjusting for tumor/node classification and other prognostic factors, central node metastasis was independently associated with cancer-specific death (HR 2.29; 95% CI 1.49-3.52). The presence of metastatic central lymph nodes at the origin of the IMA has similarly been shown to correlate with advanced tumor stage and worse DFS despite adjuvant treatment. Some authors argue these findings support high ligation of the IMA at the aorta with CME and D3 dissection to reduce the risk of residual unsuspected metastatic IMA lymph nodes and optimize nodal staging.
  • Overall, these results support the recommendation of extended right colectomy with CME + D2 dissection at a minimum, with consideration for CME + D3 dissection for suspected clinical stage III cancer based on the 3.7-19.7% risk of central node metastasis and the potential prognostic significance of this finding in stage III disease — taking into account the shortcomings of preoperative staging by CT scan in predicting nodal involvement. It has been argued that CME + D3 dissection may help upstage tumors and improve local control by reducing the risk of residual metastatic/unsuspected micrometastatic lymph nodes, and may improve survival relative to D2 dissection alone, particularly in stage III colon cancer — however, there are no definitive data to support that contention. There is a limit to lymphadenectomy's benefit: at some distance from the tumor, finding involved lymph nodes is simply a marker for disseminated disease rather than local disease that can be surgically removed. To date, it has been difficult to quantify the potential benefit vs. risk of extending lymphadenectomy beyond the level of the commonly named vessels at the root of the mesentery.

Rationale for CME adoption

There is no universal agreement as to what constitutes "CME," "CME + CVL," or "conventional" dissection, which hampers outcome comparisons. It is nearly impossible to perform randomized studies of CME vs. conventional colectomy, since surgeons would be unlikely to substantially alter technique between patients (with the possible exception of level of nodal harvest at the root of the mesentery). Most published CME-vs-conventional studies use historical controls or compare outcomes between hospitals/geographic regions, introducing likely hidden confounders and substantial bias. Some studies retrospectively subgroup patients by stage, which lacks a clinical correlate given the difficulty of reliably differentiating stage III from stage II disease preoperatively.

In an ideal world, patients with true stage I and stage II disease should garner no benefit from lymphadenectomy — extended lymphadenectomy would pose excess risk in these patients and provide no benefit. "Improved" oncologic outcome from extended lymphadenectomy in patients with stage I/II disease would therefore reflect stage migration, or inherent inaccuracies in histologic staging (the magnitude of which may be co-linear with differences in basic surgical technique), further confounding the analysis.

Impact of CME on mesocolic dissection grade: colonic mesocolon resection along the embryologic mesocolic plane, with preservation of intact fascial layers, has been associated with good outcomes, analogous to Quirke's grading of TME dissection quality. West et al. demonstrated significantly greater mesentery/colon length, higher lymph node yield, and a higher proportion of mesocolic-plane- resection specimens in the CME group (review of 399 photographed/graded specimens: mesocolic dissection confirmed in 32% of specimens, associated with a 27% 5-year survival advantage in stage III colon cancer relative to dissection along the muscularis propria plane). A higher rate of mesocolic- resection-plane achievement was seen with left-sided CME than right-sided or transverse CME, again reflecting the higher complexity of vascular dissection for right-sided/transverse lesions. The Danish DCCG compared CME (adopted 2008) at one institution vs. NCME at three other hospitals: CME was an independent predictor of higher DFS relative to NCME, especially for stage I/II disease. In an ongoing prospective non-randomized trial comparing CME + D3 with NCME (specimen morphometry assessed via photodocumentation), CME was associated with improved 3-year local recurrence-free survival in stage II and III patients. Overall, current evidence supports adoption of tumor-specific mesocolic excision for colon cancer, similar to tumor-specific TME, with complete resection of the mesocolic envelope and lymphadenectomy beyond paracolic nodes (D2 or D3 dissection).

Impact of CME on lymph node harvest, central/skip/occult metastases: CME has consistently demonstrated an association with increased lymph node harvest, improving staging accuracy and prompting consideration of adjuvant chemotherapy. A systematic review of retrospective studies comparing CME vs. non-CME colectomy found lymph node harvest significantly higher in the CME group in 10 of 12 studies. CME with D3 dissection may also remove clinically suspicious nodes detected intraoperatively, plus lymph node micrometastases not otherwise detectable pre- or intraoperatively. Extended lymphadenectomy would permit removal of skip metastases (tumor metastases found in central lymph nodes in the absence of pericolic/intermediate lymph node involvement) — reported in 0-19.8% of stage III colon cancers.

Conclusions

Although published evidence supports completion of CME for colon cancer based on global improvement in surgical quality and prognostic/survival advantage from higher lymph node counts achieved, routine D3 lymphadenectomy for stage II/III colon cancers remains highly controversial, especially given the lack of direct and high-level evidence demonstrating long-term oncologic benefits. While in experienced hands CME + D3 has an acceptable risk profile — when applied to right colectomy in particular, it requires thorough knowledge of the anatomy of the superior mesenteric vessels and their tributaries. These are technically challenging procedures with added risk of organ and vascular injury during the early learning curve. Laparoscopic and robotic approaches for CME + D3 have been increasingly described, with acceptable conversion rates, improved short-term results, and non-inferior oncologic outcomes in experienced hands and at high-volume centers. Several ongoing prospective observational and randomized trials are evaluating the oncologic impact of CME + D2/D3 resection in stage II/III disease; unlike older trials, these incorporate video-based standardized assessment of CME quality and standardized pathologic assessment of quality indicators for CME specimens using photodocumentation.

Current evidence only supports selective use of CME + D3 for tumors with high risk of lymph node metastasis, based on preoperative imaging, tumor location, and/or clinical T stage — until further data emerge, CME + D2 dissection remains standard of care, with selective consideration of D3 lymphadenectomy based on preoperative detection of suspicious central lymph nodes in select patients.

Open items / gaps

  • JCOG0404, the COLD trial, RELARC, SLRC, LCME, the Danish DCCG datasets, the ERCRC registry, and the various retrospective CME-vs-NCME comparative studies are not independently ingested as primary literature — all captured only via this textbook chapter's narrative summary.
  • No consensus exists (per the chapter itself) on which CME quality indicators are most critical — this page presents the debate as unresolved, per wiki convention.
  • The Halsted-vs-Fisher lymphatic-spread-pattern debate is presented as genuinely unresolved biology, not adjudicated here.
  • Table 37.1 and Figs. 37.1-37.2's full Japanese lymph node station numbering (201, 202, 203, etc.) were not individually transcribed onto this page — refer back to ASCRS Ch37 - Minimally Invasive Complete Mesocolic Excision with Extended Lymphadenectomy for Colon Cancer if exact station numbers are needed clinically.