Radiation Oncology/Esophagus/Review

Esophageal Cancer Review Epidemiology 2009 U.S. Data: incidence ~16,500; mortality: ~14,500 (88% death rate) High prevalence worldwide: Asia, southern/eastern Africa, southern France Risk factors: age, male gender, Plummer-Vinson, caustic injury Squamous cell: smoking, tobacco; risk decreases with cessation AdenoCA: GERD, Barrett's (40x), smoking, tobacco; risk does not decrease with cessation Clinical Presentation & Work Up Clinical presentation: dysphagia, weight loss, hiccups (phrenic nerve), reflux (suggests EGJ tumor), satiety (suggests infiltration into stomach wall) EGD: distance from incisors, extent of lesion/mucosal changes CT scan: frequently not helpful, since Barretts shows wall thickening; miss 50% nodes Endoscopic ultrasound: T-stage, LN evaluation/biopsy PET/CT Bronchoscopy: if above carina, to rule out fistulas Surgery Surgically resectable: Cervical esophagus: typically not treated surgically for larynx preservation, need >5 cm from cricopharyngeus Only 30-40% present with resectable disease, only 50% of these undergo curative (R0) resection Technique: Transthoracic (Ivor Lewis) vs transhiatal Lymph node dissection: based on SEER data, survival better if >10 LN removed Surgery alone: 5-year OS 20-25%; median OS ~1.5 years Local control? Surgery vs RT: 2 older trials showed survival benefit to surgery; recent Chinese trial no difference in 5-year OS (~35%) Pre-op RT: 6 older RCTs. No survival benefit. Meta-analysis using individual patient data showed no conclusive survival benefit for RT Pre-op chemo: 9 RCTs. Two (MRC and Netherlands) showed significant survival benefit. Meta-analysis suggests 2-year 7% OS benefit and 5-year 4% benefit Pre-op chemo-RT: 10 RCTs. Two (Walsh, Tepper) showed significant survival benefit. Meta-analysis showed 2-year 13% OS benefit, which persisted even after excluding Walsh trial Dublin (Walsh): median OS 16 months vs 11 months (SS); 3-year OS 32% vs 6% (SS). Critique poor outcome in surgery only arm CALGB 9781 (Teper): median OS 4.5 years vs 1.8 years (SS); 5-year OS 39% vs 16%. pCR 40%; Grade 3+ toxicity 40% Post-op chemo: 3 RCTs. Meta-analysis showed no significant benefit Post-op chemo-RT: Gastric/EGJ trial (INT 0116, MacDonald): 3-year OS benefit 41% vs 50% (SS); 3-year local relapse 29% vs 19%; Grade 4 toxicity 32% Pre-op chemo-RT + Surgery vs Chemo-RT alone: 2 RCTs, majority squamous cell. Non-inferiority 10-15%, both trials negative for OS difference. Germany (Stahl): Randomized upfront, no lower esophagus. 2-year OS 40% vs 35% (NS). 2-year local control 64% vs 41% (SS). Periop mortality 13% vs 4% French FFCD-9102 (Bedenne): Only responders to CRT randomized. 2-year OS 34% vs 40% (NS). 2-year local control 66% vs 57% (SS). Periop mortality 12% vs 0% Concurrent chemotherapy CALGB 9781: 5-FU 1000 mg/m2 + cisplatin 100 mg/m2 RTOG 0436 (ongoing): paclitaxel 50 mg/m2 + cisplatin 25 mg/m2 QW German (Stahl): cisplatin + etoposide Conclusions: If surgery will be done, preop chemo-RT improves 2-year survival In patients with SCC at risk for surgical morbidity/mortality, primary chemo-RT may be comparable/superior NCCN guidelines (v1.2010) T1a: esophagectomy or EMR T1b: esophagectomy for non-cervical, chemo-RT for cervical primaries T2-T4N0-N1: pre-op chemo (adenoCA only), pre-op chemo-RT, or definitive chemo-RT T2-T4N0-N1: if surgery done first, post-op chemo-RT (MacDonald) for distal/GEJ tumors or R1/R2 resections Non-surgical approaches RT vs Chemo-RT RTOG 8501 (Herskovic) Concurrent chemo: cisplatin 75 mg/m2 + 5-FU 1000 mg/m2 (weeks 1, 5, 8, 11) Survial benefit: 3-year OS 10% vs 40% (SS); 5-year OS 0% vs 26% (SS); median OS 9 months vs 14 months (SS) Locoregional control benefit: 53% vs 38% (SS) Worse toxicity: Acute Grade 4-5 2% vs 10% (including 2% deaths) Chemo-RT: Dose escalation RTOG 9207 (Gaspar) brachytherapy boost EBRT 50/25 + HDR 15/3 or LDR 20/1 with concurrent cisplatin 75 mg/m2 + 5-FU 1000 mg/m2 Grade 4 toxicity 24% (fistula 18%), deaths 10% RTOG 9405 (Minsky) 3DCRT boost EBRT 50.4 Gy vs 64.8 Gy with concurrent cisplatin 75 mg/m2 + 5-FU 1000 mg/m2 No difference in survival, 2-year 40% vs 31% No difference in local failure: 56% vs 52% Toxicity: 11 deaths in high-dose arm Chemo-RT: Different chemo regimens RTOG 0113: Induction chemo -> chemo-RT comparing cisplatin/paclitaxel vs 5-FU/paclitaxel Neither arm successful Local control: ~50% despite dose escalation or chemo changes Survival: median OS ~1.5 years, 5-year OS ~25%