Table 3. Chart review of 79 adults with esophageal cancer in relation to nutritional procedure received up to 10 days after the start of neoadjuvant therapy. Kaiser Permanente Northern California, 2010–2017.
| Characteristic | No procedure, N=20 | J-tube, N=39 | Stent, N=20 | P value |
|---|---|---|---|---|
| Anatomic location | ||||
| Upper | 0% | 10% | 5% | 0.49f |
| Middle | 25% | 18% | 5% | |
| Lower | 50% | 54% | 65% | |
| Gastroesophageal junction | 25% | 18% | 25% | |
| Adjuvant therapy | ||||
| Definitive | 25% | 28% | 20% | 0.79 |
| Preoperative | 75% | 72% | 80% | |
| Fistula at diagnosis | 0% | 0% | 5% | NE |
| Dysphagia at diagnosis | 75% | 100% | 100% | NE |
| Reason for intervention | <0.001f | |||
| Dysphagia | NA | 36%a | 95%b | |
| Anticipated weight loss | NA | 49% | 0% | |
| Actual weight loss | NA | 15%a | 0% | |
| Fistula | NA | 0 | 5%b | |
| Complication after intervention | ||||
| None vs. any | 95% | 26% | 15% | <0.001 |
| Pain | 5% | 10% | 35% | |
| Bleeding | NA | 0% | 25% | |
| Migration | NA | NA | 35% | |
| Fistula | NA | 3% | 5% | |
| Reflux | NA | 18% | 10% | |
| Pneumonia | NA | 3% | 5% | |
| Wound infection | NA | 26%c | 10% | |
| J-tube device failure | NA | 64%d | NA | |
| J-tube take-down for complications | NA | 13% | NA | |
| Stent adjustment or removal | NA | NA | 40% | |
| Intervention ≥11 days after start of neoadjuvant therapye | ||||
| None | 75% | 85% | 30% | <0.001f |
| Gastrostomy tube | 15% | 0% | 10% | |
| J-tube | 5% | 10% | 35% | |
| Stent | 5% | 5% | 0% | |
| Nasal gastric tube | 0% | 0% | 15% | |
| Total parenteral nutrition | 0% | 0% | 10% | |
| Final decision to go to surgery | ||||
| Yes | 65% | 51% | 40% | 0.28 |
| No, death | 0% | 5% | 10% | |
| No, disease not cleared | 5% | 3% | 0% | |
| No, metastatic disease | 10% | 13% | 25% | |
| No, patient not fit for surgery | 5% | 13% | 25% | |
| No, bowel perforation from J-tube | 0% | 3% | 0% | |
| No, patient preference | 0% | 0% | 5% | |
| No, surveillance | 15% | 10% | 0% | |
| No, tumor too proximal | 0% | 3% | 0% |
a, one patient had both dysphagia and actual weight loss and is counted twice; b, one patient had dysphagia and fistula and is counted twice; c, included 9 wound infections and one urinary tract infection; d, device failures included clogging, falling out, and plug and cap problems; e, for patients classified with no procedure at entry into follow-up, this was their first intervention. For patients classified with j-tube or stent at entry into follow-up, this was their second intervention; f, Fisher’s exact test was used in place of chi square because of small cell counts. NE, not estimated; NA, not applicable.