The treatment of esophageal cancer with curative intent remains highly controversial, with advocates of surgery alone, chemoradiotherapy alone, surgery with adjuvant therapy (including neoadjuvant and postoperative), and trimodality therapy each contributing prospective randomized controlled trials (PRCTs) to the body of scientific publications between 2000 and 2008. Any improvements in survival have been small in absolute percentage terms, and as such PRCTs published over the last decade have met the same primary obstacle encountered by the studies from the two prior decades, namely lack of power to detect small differences in outcome. Variations in staging methods, surgical technique, radiotherapy technique, and chemotherapy regime have in turn been the subject of PRCTs over the last nine years. In many cases primary end points have not been survival but rather rates of complication or response. As well as giving an overview of PRCTs, this article collates the level Ia evidence published to date.
The treatment of esophageal cancer with curative intent remains highly controversial, with advocates of surgery alone, chemoradiotherapy (CRT) alone, surgery with adjuvant therapy (including neoadjuvant and postoperative), and trimodality therapy each contributing prospective randomized controlled trials (PRCTs) to the body of scientific publications between 2000 and 2008. Any improvements in survival have been small in absolute percentage terms, and as such PRCTs published over the last decade have met the same primary obstacle encountered by the studies from the two decades prior and reviewed in the previous summary article, namely lack of power to detect small differences in outcome. Many of the best trials in attempting to address this issue by careful power calculations have been thwarted by lack of accrual due largely to patient and physician preference for a particular treatment modality.
Attempts have been made to address this deficiency by the publication of 8 meta-analyses ; these in turn are limited by: the heterogeneity of patients included with respect to clinical stage, methods used to assess clinical stage (ranging from upper gastrointestinal endoscopy [UGIE], abdominal ultrasound [US], and chest radiography (CXR) alone, through to high-resolution computed tomography [CT], positron emission tomography [PET]-CT, endoscopic ultrasound [EUS], staging laparoscopy, and thoracoscopy), performance status of patients, and histology of tumor; the heterogeneity of treatment with respect to anesthetic and perioperative care, surgery (including operative approach, conduit used and technique of reconstruction), radiotherapy (RT) including planning (hardware, software, and extent of treatment) and delivery (hardware, software, fractionation, and total dose), chemotherapy varying in agent, dose, and number and timing (induction, postoperative, or both) of cycles, and CRT varying with respect to the factors already mentioned, as well as whether administered concurrently or sequentially; and publication bias .
These limitations remain despite sophisticated statistical methods employed to address said limitations. These deficiencies are seemingly intuitively apparent to clinicians working in the area and make the conclusions of the meta-analyses (which themselves vary) difficult to confidently and broadly apply. In any case, meta-analyses do not represent level Ia evidence and are not the subject of this review.
These variations in staging methods, surgical technique, RT technique, and chemotherapy regime have in turn been the subject of PRCTs over the last nine years. In many cases primary end points have not been survival but rather rates of complication or response.
Curative intent: surgery versus chemoradiotherapy
The only PRCT published in the last nine years comparing surgery alone to definitive CRT was conducted in Hong Kong. This multicenter trial enrolled biopsy-proven squamous cell carcinoma (SCC) of the mid and lower esophagus with preoperative stage T 1-3 , N 0-1 , M 0-1a , and had 80% power to discern a doubling in median survival from 15 to 30 months. Such an expectation was perhaps overly optimistic, and despite near equivalence in two-year survival (58% vs 55%, P = .45) and a nonsignificant risk ratio (0.89 confidence interval [CI] 0.37–2.17), a smaller but clinically relevant improvement in overall survival (OS) has not been discounted. The secondary end point of median hospital stay was longer in the CRT group (41 vs 27 days, P = .022). Despite stated analysis on an intention-to-treat basis, one patient randomized to esophagectomy was removed from analysis after failing to undergo planned treatment. This trial has much to commend it: extensive staging of patients with modalities including EUS, CT, and neck US; modern RT techniques including concurrent CRT and 3D conformal delivery; and inclusion of patients with clinically advanced disease (ie, M 1a , a group often included in studies of CRT but excluded from surgical trials); however, it remains underpowered to confirm a clinically relevant superiority of one treatment over the other.
Curative intent: surgery alone versus neoadjuvant therapy
If from the aforementioned PRCT one accepts surgery as a standard of care, it is reasonable to include it as the control arm of further PRCTs of adjuvant therapy (including neoadjuvant and postoperative). Ten such trials employing neoadjuvant therapy have been published in the last nine years, including two previously reported trials with updated follow-up ( Table 1 ).
| Authors (Institution, Region) | Accrual | Power Calculation | n | Induction Regime | Survival, Induction & S vs S | Significance | Hazard Ratio |
|---|---|---|---|---|---|---|---|
| Walsh et al (single, Ireland) | 1990–1995 | 80%, 2 y from 23% to 43% required 190 pts | 113 | C, 5-FU (×2) 2.5 Gy × 16 (40 Gy) | Median: 17 vs 12 mo; 5 y: 32% vs 6% | 0.002 | |
| Urba et al (single, North America) | 1989–1994 | 80%, Median from 1 to 2.2 y | 100 | C, 5-FU, V (×2) 1.5 Gy BD ×15 (45 Gy) | Median: 16.9 vs 17.6 mo | NS | 0.73 |
| Burmeister et al AGITG (multi, Australasia) | 1994–2000 | 80%, 3 y PFS from 20% to 35% | 256 | C, 5-FU (×1) 15 Gy ×1.3 (35 Gy) | Median: 16 vs 12 b mo | NS | 0.82 |
| Tepper et al CALGB 9781 (multi, North America) | 1997–2000 | 90%, 5 y from 20% to 32% required 500 pts | 56 | C, 5-FU (×2) 1.8 Gy ×28 (50.4 Gy) | Median: 54 vs 21 mo; 5 y: 39% vs 16% | 0.002 | |
| Natsugoe et al (single, Japan) | 1997–2001 | None reported | 45 | C, 5-FU (continuous) 2 Gy ×20 (40 Gy) | 5 y: 57% vs 41% | NS | |
| Lee et al (single, South Korea) | 1999–2002 | 80%, 2 y from 20% to 30% | 101 | C, 5-FU (×2, S, ± ×3) 1.2 Gy BD ×19 (45.6 Gy) | Median: 27.3 vs 28.2 mo | NS | |
| Kelsen et al Intergroup 113 (multi, North America) | 1990–1995 | 90%, 38% improvement in median | 443 | C, 5-FU (×3, S ± ×2) | Median: 14.9 vs 16.1 mo | NS | |
| Ancona et al (single, Italy) | 1992–1997 | 80%, 2 y improved 20% required 240 pts | 94 | C, 5-FU (×2–3) | Median: 25 vs 24 mo; 5 y: 34% vs 22% | NS | |
| Girling et al MRC (multi, Europe) | 1992–1998 | 90%, 2 y from 20% to 30% | 802 | C, 5-FU (×2) | Median: 16.8 vs 13.3 mo | 0.004 | 0.79 |
| Cunningham et al “MAGIC”, MRC (multi, Europe) | 1994–2002 | 90%, 5 y from 23% to 38% | 503 a | C, 5-FU, E (×3, S, ×3) | 5 y: 36% vs 23% | 0.009 | 0.75 |
a 131 patients esophageal or gastroesophageal junction.
Of the six PRCTs employing neoadjuvant CRT, two showed improvement in OS and four did not. All trials used concurrent CRT, with total doses ranging from 35 to 50.4 Gy. Chemotherapy consisted of cisplatin at total doses of 60 to 100 mg/m 2 per cycle and 5-flourouracil (5-FU).
Of the four PRCTs employing neoadjuvant chemotherapy, two showed improvement in OS and two did not. Three of four trials employed cisplatin and 5-FU, whereas one employed epirubicin, cisplatin, and 5-FU.
Of note, each of the negative trials either did not report a power calculation, failed to accrue the calculated sample size, or had only 80% power to detect 50% or more improvement in survival. In this context the implication from these non significant trials is that there remains a 20% chance that at least a 50% improvement in survival exists but was not detected, or indeed that there is an even greater chance of a real difference in survival of less than 50% being present but undetected.
The notable exception to this implication is the North American Intergroup trial (INT113) which, despite power of 90% to detect a 38% increase in median survival, showed no difference in survival. The integration of the results of this trial to clinical practice has remained problematic for advocates of neoadjuvant chemotherapy alone. The larger MRC trial (OEO2) and INT113 are often compared as the accrual of demographically similar patients with seemingly similar treatment regimes resulting in disparate outcomes. Some differences between the two trials may explain the conflicting results. The OEO2 stratified patients, according to treatment center and surgeon, had a shorter time from randomization to surgery (63 vs 93 days), had a higher percentage of patients in the induction arm proceed to surgery (92% vs 80%), and used a less toxic cisplatin regimen (160 vs 450 mg/m 2 ). This last point is perhaps most salient. The inability of North American patients to tolerate the higher doses of cisplatin (71% completed planned preoperative chemotherapy compared with 90%, and only 38% completed planned postoperative treatment whereas OEO2 did not include postoperative chemotherapy) may imply that any benefit in terms of eradication of micrometastases was counterbalanced by side effects and toxicities resulting in no net change in survival. Finally, the negative result of INT113 may simply represent the 10% chance of a trial with 90% power not detecting a real difference in survival.
The observed magnitude of improvement in survival with CRT has been similar to that of chemotherapy alone, with median survival after neoadjuvant CRT improving from 12 to 17 months and after neoadjuvant chemotherapy improving from 13 to 17 months.
Curative intent: surgery alone versus neoadjuvant therapy
If from the aforementioned PRCT one accepts surgery as a standard of care, it is reasonable to include it as the control arm of further PRCTs of adjuvant therapy (including neoadjuvant and postoperative). Ten such trials employing neoadjuvant therapy have been published in the last nine years, including two previously reported trials with updated follow-up ( Table 1 ).
| Authors (Institution, Region) | Accrual | Power Calculation | n | Induction Regime | Survival, Induction & S vs S | Significance | Hazard Ratio |
|---|---|---|---|---|---|---|---|
| Walsh et al (single, Ireland) | 1990–1995 | 80%, 2 y from 23% to 43% required 190 pts | 113 | C, 5-FU (×2) 2.5 Gy × 16 (40 Gy) | Median: 17 vs 12 mo; 5 y: 32% vs 6% | 0.002 | |
| Urba et al (single, North America) | 1989–1994 | 80%, Median from 1 to 2.2 y | 100 | C, 5-FU, V (×2) 1.5 Gy BD ×15 (45 Gy) | Median: 16.9 vs 17.6 mo | NS | 0.73 |
| Burmeister et al AGITG (multi, Australasia) | 1994–2000 | 80%, 3 y PFS from 20% to 35% | 256 | C, 5-FU (×1) 15 Gy ×1.3 (35 Gy) | Median: 16 vs 12 b mo | NS | 0.82 |
| Tepper et al CALGB 9781 (multi, North America) | 1997–2000 | 90%, 5 y from 20% to 32% required 500 pts | 56 | C, 5-FU (×2) 1.8 Gy ×28 (50.4 Gy) | Median: 54 vs 21 mo; 5 y: 39% vs 16% | 0.002 | |
| Natsugoe et al (single, Japan) | 1997–2001 | None reported | 45 | C, 5-FU (continuous) 2 Gy ×20 (40 Gy) | 5 y: 57% vs 41% | NS | |
| Lee et al (single, South Korea) | 1999–2002 | 80%, 2 y from 20% to 30% | 101 | C, 5-FU (×2, S, ± ×3) 1.2 Gy BD ×19 (45.6 Gy) | Median: 27.3 vs 28.2 mo | NS | |
| Kelsen et al Intergroup 113 (multi, North America) | 1990–1995 | 90%, 38% improvement in median | 443 | C, 5-FU (×3, S ± ×2) | Median: 14.9 vs 16.1 mo | NS | |
| Ancona et al (single, Italy) | 1992–1997 | 80%, 2 y improved 20% required 240 pts | 94 | C, 5-FU (×2–3) | Median: 25 vs 24 mo; 5 y: 34% vs 22% | NS | |
| Girling et al MRC (multi, Europe) | 1992–1998 | 90%, 2 y from 20% to 30% | 802 | C, 5-FU (×2) | Median: 16.8 vs 13.3 mo | 0.004 | 0.79 |
| Cunningham et al “MAGIC”, MRC (multi, Europe) | 1994–2002 | 90%, 5 y from 23% to 38% | 503 a | C, 5-FU, E (×3, S, ×3) | 5 y: 36% vs 23% | 0.009 | 0.75 |
a 131 patients esophageal or gastroesophageal junction.
Of the six PRCTs employing neoadjuvant CRT, two showed improvement in OS and four did not. All trials used concurrent CRT, with total doses ranging from 35 to 50.4 Gy. Chemotherapy consisted of cisplatin at total doses of 60 to 100 mg/m 2 per cycle and 5-flourouracil (5-FU).
Of the four PRCTs employing neoadjuvant chemotherapy, two showed improvement in OS and two did not. Three of four trials employed cisplatin and 5-FU, whereas one employed epirubicin, cisplatin, and 5-FU.
Of note, each of the negative trials either did not report a power calculation, failed to accrue the calculated sample size, or had only 80% power to detect 50% or more improvement in survival. In this context the implication from these non significant trials is that there remains a 20% chance that at least a 50% improvement in survival exists but was not detected, or indeed that there is an even greater chance of a real difference in survival of less than 50% being present but undetected.
The notable exception to this implication is the North American Intergroup trial (INT113) which, despite power of 90% to detect a 38% increase in median survival, showed no difference in survival. The integration of the results of this trial to clinical practice has remained problematic for advocates of neoadjuvant chemotherapy alone. The larger MRC trial (OEO2) and INT113 are often compared as the accrual of demographically similar patients with seemingly similar treatment regimes resulting in disparate outcomes. Some differences between the two trials may explain the conflicting results. The OEO2 stratified patients, according to treatment center and surgeon, had a shorter time from randomization to surgery (63 vs 93 days), had a higher percentage of patients in the induction arm proceed to surgery (92% vs 80%), and used a less toxic cisplatin regimen (160 vs 450 mg/m 2 ). This last point is perhaps most salient. The inability of North American patients to tolerate the higher doses of cisplatin (71% completed planned preoperative chemotherapy compared with 90%, and only 38% completed planned postoperative treatment whereas OEO2 did not include postoperative chemotherapy) may imply that any benefit in terms of eradication of micrometastases was counterbalanced by side effects and toxicities resulting in no net change in survival. Finally, the negative result of INT113 may simply represent the 10% chance of a trial with 90% power not detecting a real difference in survival.
The observed magnitude of improvement in survival with CRT has been similar to that of chemotherapy alone, with median survival after neoadjuvant CRT improving from 12 to 17 months and after neoadjuvant chemotherapy improving from 13 to 17 months.
Curative intent: surgery alone versus adjuvant therapy
Four PRCTs have addressed the use of adjuvant therapy with curative intent after resection.
Adjuvant RT was tested at multiple institutions in China between 1986 and 1997 in 495 patients with SCC of the thoracic esophagus more than 4 cm in length randomized to 60 Gy external beam RT versus surgery alone. Five-year survival was 41.3% in the treatment group compared with 31.7% in the surgery alone group ( P = .45). Subgroup analysis showed 5-year survival to be better after adjuvant RT in stage III patients (35.1 vs 13.1%, P = .0027). The impact of this trial has been limited, as no power calculation was undertaken, groups were not well balanced with respect to sex and nodal status, and intention-to-treat principles were not applied.
Adjuvant chemotherapy was tested at multiple institutions in Japan by the Japanese Clinical Oncology Group (JCOG) between 1992 and 1997 in 242 patients with completely resected SCC randomized to two cycles of cisplatin and 5-FU versus surgery alone. Patients were stratified according to institution and pathologic nodal status. The primary end point of disease-free survival (DFS) favored the adjuvant treatment arm (55 vs 45%, P = .037). In subgroup analysis 5-year DFS in pN 0 patients showed no difference (70 vs 76%, P = .4), but was enhanced in pN 1 patients (52 vs 38%, P = .04). Of note, no significant difference was seen in OS.
Adjuvant chemotherapy was compared with adjuvant concurrent CRT at a single institution in Japan between 1991 and 2000 in 45 patients with completely resected SCC and pathologic stage I b to III tumors randomized to adjuvant CRT (cisplatin 50 mg/m 2 days 1, 15, 5-FU 300 mg/m 2 days 1–35, and concurrent external beam RT 45–50 Gy) versus chemotherapy alone (cisplatin 50 mg/m 2 days 1, 15, and 5-FU 300 mg/m 2 days 1–35). Patients with postoperative complication were not eligible. No power calculation was reported and no difference was found in median survival (31 vs 28 months, not significant).
Adjuvant CRT (leucovorin 20 mg/m 2 and 5-FU 450 mg/m 2 days 1–5, 28–31, 58–60, 90–94, 120–124, plus external beam RT of 45 Gy in divided doses 5 times per week days 28–60) was compared with surgery alone at multiple institutions in North America between 1991 and 1998 in 556 patients with completely resected adenocarcinoma (AC) of the stomach or gastroesophageal junction, and including approximately 20% with tumors of the cardia. Median OS was improved in patients receiving adjuvant therapy (36 vs 27 months, P = .005). Hazard ratio (HR) of death was 1.35 (CI = 1.09–1.66).
In summary, RT alone or in addition to chemotherapy did not improve primary end points. Adjuvant chemotherapy alone and CRT showed improvement compared with surgery alone.
Curative intent: definitive chemoradiotherapy versus trimodality therapy
If from the aforementioned PRCT one accepts definitive CRT as a standard of care, it is reasonable to include it as the control arm of further PRCTs of trimodality therapy. Two such trials have been undertaken.
The first trial to be published is perhaps the most significant to be discussed in the context of this review. Stahl and colleagues randomized 172 patients from 11 institutions in Germany with SCC of the upper and mid esophagus to trimodality therapy versus definitive CRT. After induction chemotherapy with cisplatin, 5-FU, etoposide, and leucovorin patients were randomized to either concurrent CRT with cisplatin, etoposide, and 40 Gy RT then transthoracic esophagectomy or definitive concurrent CRT with cisplatin, etoposide, and 60 to 65 Gy RT. The patients were staged with EUS (though the details are not reported) and CT scan, but not PET. T 3-4 (excluding airway invasion) and N 1-0 patients were stratified by: center, TNM, completeness of EUS, sex, and weight loss.
Of interest, the trial was designed to assess equivalence of treatments; as such a reduction in two-year survival of greater than 15% in the definitive CRT arm from the expected 35% in the surgery arm would have resulted in a negative trial. In fact two-year survival was 39.9% in the trimodality arm and 35.4% in the definitive CRT arm ( P = .007). Thus the trial was positive, implying there is only a 5% chance of a greater than 15% absolute difference in two-year OS between the two groups. Of note, if the more usual design attempting to discern a survival advantage had been employed one would be discussing an adequately powered (80%) trial that was negative, that is, had failed to show improvement in two-year survival from 20% to 35%, and likely questioning the overly optimistic expectation of a 75% improvement in survival.
When treatments are equivalent in terms of OS, attention logically turns to other measures of outcome. Secondary end points included progression-free survival (PFS), which at two years was 64.3% versus 40.7% ( P = .003) favoring the trimodality arm; and treatment-related deaths, which clearly favored the definitive CRT arm: 11 of 86 (12.8%) versus three of 85 (3.5%) ( P = .03) including two deaths due to complications of induction chemotherapy in each arm. Finally, a regression analysis revealed response to initial induction chemotherapy to be the only significant predictor of survival.
In summary, there was no difference in OS, progression was less common, and treatment-related death more common in the trimodality arm, and response to induction chemotherapy (regardless of other treatments) was the only significant predictor of survival.
A second European multi-institution trial undertaken in France has resulted in two publications. The first article addressed quality of life (which was a secondary end point of the trial and not the focus of a power analysis), and found a short-term reduction in the trimodality arm but no difference between arms beyond six months. The second reported the trial in detail, and is notable (particularly after the findings of the regression analysis in the first trimodality trial) in that only patients with esophageal SCC responding to induction therapy (259/444) were subsequently randomized to surgery versus continued CRT. Sequential or concurrent CRT was allowed in both arms and surgical approach was not mandated. The trial was again designed to assess equivalence and (despite planned accrual of 500 patients to provide 80% power to confirm <10% difference in OS between the arms) was ceased after interim analysis showed no chance of rejecting the hypothesis of equivalence with further enrollments. Two-year survival was 34% in the trimodality arm and 40% in the definitive CRT arm ( P = .03), implying equivalence of the 2 treatments.
Secondary end points included locoregional relapse (HR = 1.63, CI 1.04–2.55, P = .03) and treatment for dysphagia (24% vs 46%, P = .001), which both favored the trimodality arm, whereas 3-month mortality (9.3% vs 0.8%, P = .002) and length of stay (68 vs 52 days, P = .02) favored the definitive CRT arm.
Curative intent: surgical technique
Two PRCTs of surgical approach comparing esophageal resection with and without thoracotomy have been reported. The first, undertaken at two centers in the Netherlands and accruing between 1994 and 2000, compared transhiatal esophagectomy to transthoracic esophagectomy in 220 patients with T 1-3 , N 0-1 , M 0 AC of the mid esophagus to gastric cardia. The trial was powered to detect a 50% improvement in median survival from 14 to 22 months in the transthoracic arm. There was no such improvement, with median survival in the transhiatal compared with transthoracic arm 1.8 versus 2.0 years ( P = .38), 5-year survival 29% versus 38% (CI −3% to 23%), and median DFS 1.4 versus 1.7 years ( P = .15). Of note, transhiatal esophagectomy was associated with significantly less blood loss, pulmonary complication, chyle leak, and ventilated days. Despite this higher rate of complications there was no significant increase in perioperative mortality in the transthoracic arm. Thus this trial implies there is less than a 50% improvement in survival and no increased perioperative mortality when transthoracic esophagectomy is undertaken, despite an increase in perioperative morbidity. A second publication was generated based on this trial, and reported actual five-year survival of 34% versus 36% ( P = .71). Subgroup analysis of patients found to have 1 to 8 malignant lymph nodes had OS of 19% versus 39% ( P = .05) and DFS of 23% versus 64% ( P = .02). This retrospective analysis, while of interest, was based on pathologic results and excluded patients undergoing exploratory surgery (though R1 resections were included). As such, intention-to-treat principles were not adhered to, postoperative stage migration is likely, and the report at best represents a level “1c” PRCT.
The second trial, undertaken at 27 Japanese hospitals under the auspices of the JCOG and accruing between 1995 and 2003, compared left thoracoabdominal to transhiatal resection in 167 patients with T 2-4 , M 0 gastric AC with 3 cm or less esophageal involvement. The trial was designed with 80% power to detect a 10.5% improvement in five-year survival with thoracoabdominal resection, and initially planned enrollment of 302 patients over 4 years. After eight years and accrual of 167 patients, interim statistical analysis showed less than a 3.65% chance of proving thoracoabdominal resection superior with completion of accrual, thus the trial was closed. The investigators conclude that thoracoabdominal resection does not offer a survival advantage compared with transhiatal resection in Siewert II and III tumors.
Three single-institution PRCTs have compared hand-sewn to stapled anastomosis after esophagectomy. Rate of anastomotic leak, perioperative morbidity and mortality, dysphagia, anastomotic diameter, stricture rate, and quality of life were not different between the two approaches.
Further variations of reconstructive technique have been addressed by three PRCTs. Gupta and colleagues found a lower leak rate of 4.3% versus 20.8% ( P = .03) and stricture rate of 8.5% versus 29.2% ( P = .02) after a “novel” hand-sewn esophagogastric anastomosis compared with a standard hand-sewn anastomosis. Bhat and colleagues found anastomotic leak dramatically reduced after omental wrap of the esophagogastric anastomosis versus standard anastomosis (3.09% versus 14.43%, P = .005). Tabira and colleagues found no difference in anastomotic leak or postoperative nutritional status at 6 and 12 months after use of a slender gastric tube for reconstruction after esophagectomy when compared with a more generous gastric tube.
Shackcloth and colleagues completed a well-planned and executed study addressing the most appropriate use of nasogastric tubes (NGT) in the first 48 hours post esophagectomy. Thirty-four patients were randomized to NGT with continuous sump suction, single-lumen NGT with 4-hourly aspirations or no NGT. The patients receiving continuous suction via the sump system spent significantly less time with pH less than 5.5 than either of the other two groups (4.3% vs 39.7% vs 40.3% [ P = .007]). Patients randomized to no NGT had significantly more pulmonary complications, 7 of 12 versus 4 of 22 ( P = .02), and required an NGT to be inserted in 7 of 12 cases. This simple study argues strongly for the use of sump NGT in all patients in the immediate perioperative period.
Method of endoscopic resectional technique, extent of lymph node resection, thoracotomy versus thoracoscopy, antibiotic bowel decontamination, and PGE1 infusion have been reported in “1c” trials without difference in primary end points.
Related posts:
Randomized Clinical Trials in Soft Tissue Sarcoma
Randomized Clinical Trials in Melanoma
Randomized Clinical Trials in Breast Cancer
An Update on Randomized Clinical Trials in Advanced and Metastatic Colorectal Carcinoma
Randomized Clinical Trials in Colon Cancer
Randomized Clinical Trials in Rectal and Anal Cancers
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
