Abstract
Purpose
NRG Oncology RTOG 9704 was the first adjuvant trial to validate the prognostic value of postresection CA19-9 levels for survival in patients with pancreatic carcinoma. The data resulting from this study also provide information about predictors of recurrence that may be used to tailor individualized management in this disease setting. This secondary analysis assessed the prognostic value of postresection CA19-9 and surgical margin status (SMS) in predicting patterns of disease recurrence.
Methods and materials
This multicenter cooperative trial included participants who were enrolled as patients at oncology treatment sites in the United States and Canada. The study included 451 patients analyzable for SMS, of whom 385 were eligible for postresection CA19-9 analysis. Postresection CA19-9 was analyzed at cut points of 90, 180, and continuously. Patterns of disease recurrence included local/regional recurrence (LRR) and distant failure (DF). Multivariable analyses included treatment, tumor size, and nodal status. To adjust for multiple comparisons, a P value of ≤ .01 was considered statistically significant and > .01 to ≤ .05 to be a trend.
Results
For CA19-9, 132 (34%) patients were Lewis antigen–negative (no CA19-9 expression), 200 (52%) had levels <90, and 220 (57%) had levels <180. A total of 188 patients (42%) had negative margins, 152 (34%) positive, and 111 (25%) unknown. On univariate analysis, CA19-9 cut at 90 was associated with increases in LRR (trend) and DF. Results were similar at the 180 cut point. SMS was not associated with an increase in LRR on univariate or multivariate analyses. On multivariable analysis, CA19-9 ≥ 90 was associated with increased LRR and DF. Results were similar at the 180 cut point.
Conclusions
In this prospective evaluation, postresection CA19-9 was a significant predictor of both LRR and DF, whereas SMS was not. These findings support consideration of adjuvant radiation therapy dose intensification in patients with elevated postresection CA19-9.
Summary.
In this large prospective clinical trial of adjuvant therapy for pancreatic carcinoma, the prognostic value of postresection CA19-9 level was identified as a significant predictor of both local/regional recurrence and distant failure, whereas surgical margin status was not. These findings challenge the commonly held belief that surgical margin status and elevated postresection CA19-9 levels are major predictors primarily of local/regional recurrence and distant failure, respectively. These findings could influence the prioritization and/or timing of adjuvant chemotherapy and/or radiation therapy.
Alt-text: Unlabelled box
Introduction
Adenocarcinoma of the pancreas remains one of the most aggressive solid tumors and the fourth most common cause of cancer death in the United States today.1 Patients who undergo potentially curative gross total tumor resection have the best chance for cure. Although some improvements in survival have been associated with the addition of chemotherapy and/or radiation therapy (RT), survival remains limited.2, 3, 4, 5, 6, 7 Despite multiple phase 3 adjuvant trials, analyses of the correlation of surgical margin status (SMS) with pattern of disease recurrence among patients for whom postresection CA19-9 levels are also known are lacking.2, 3, 4, 5, 6, 7, 8
A common presumption among patients being evaluated for adjuvant therapy, and often within the conduct of multidisciplinary tumor boards, is that an elevation in postresection CA19-9 levels reflects the eminent development of distant disease spread of pancreatic cancer, without consideration of the potential risk of concurrent local/regional disease recurrence. NRG Oncology RTOG 9704 was the first phase 3 adjuvant pancreatic cancer trial to prospectively validate the prognostic value of postresection CA19-9 levels for overall survival (OS), with values of >90 and >180 associated with worse OS.9 However, analysis of its predictability of pattern of disease recurrence, not limited to the first recurrence, in this setting of prospective evaluation and in associated contexts with SMS has not been performed and is the subject of this report.
Methods and materials
The study design of NRG Oncology RTOG 9704 has been previously reported in detail.6, 7 Protocol approval was received from the institutional review board at each study site, and informed consent was obtained from each patient prior to participation in the study. Patients with histologic proof of adenocarcinoma of the pancreas underwent potentially curative gross total resection of all disease. Specific protocol recommendations with regard to surgery and determination of surgical margin status were as follows: A standard Whipple or pylorus-preserving pancreaticoduodenectomy was preferred for lesions of the pancreatic head (neck/uncinate process). The margins of the resected tissue (left lateral, bile duct), superior, and inferior were to be marked with metallic clips (when possible). If the tumor was adherent to and resected from adjacent structures (eg, a major blood vessel), small vascular or titanium clips were to be used to mark the margins of adherence. Also, resected specimens were to be marked with suture at sites of adherence so pathologists could determine whether radial margins were free of disease. In addition, the operative note and pathology report of each patient was centrally reviewed prior to patient registration by the study's surgical oncology principal investigator (J.P.H.). Patients were stratified at randomization according to documentation of tumor status at surgical margins as stated on the official pathology report. Therefore, strata included “negative,” “positive,” or otherwise not mentioned/commented on (ie, “unknown”).
After stratification by SMS, nodal status, and tumor diameter, patients were randomly assigned to 1 of 2 treatment arms as depicted in eFigure 1; available as supplementary material online only at www.practical.radonc.org. All patients received adjuvant gemcitabine or 5-fluorouracil and chemoradiation therapy (CRT).
CA 19-9 testing
Lewis antigen expression is essential for expression of CA19-9; therefore, red cell phenotyping for Lewis A and B antigens was required for study eligibility and was obtained at each institution's laboratory. If patients were negative for both Lewis A and B antigens, they were considered CA19-9 nonexpressers. For patients who expressed either antigen, blood was drawn no more than 3 weeks before random assignment, or after random assignment but before the start of protocol treatment. Subsequently, serum was prepared, frozen at 20°C, and shipped (frozen) to the AAAA tissue bank, which was located at Latter-Day Saints Hospital. Centralized determination of CA19-9 was done using enzyme-linked immunosorbent assay GI-MA kits (Diagnostic Products Corporation, a Siemens Company; Gwynedd, United Kingdom). CA19-9 nonexpressing patients (Lewis antigen A and B negative) were assigned values of 0 because by definition, they did not have the ability to secrete CA19-9 into their serum.10
Statistical methods
The following baseline characteristics were dichotomized: primary tumor location (head vs everything else), pathologic T-stage (T1, T2 vs T3, T4), and American Joint Committee on Cancer stage (I, II vs III, IV). Race was categorized as white versus African American/other. Statistical comparisons to assess potential associations among baseline characteristics and 1) missing CA19-9 data, 2) CA19-9 levels, and 3) SMS were carried out using the χ2 or Fisher's exact test. CA19-9 baseline expression was analyzed and grouped in 2 different ways (Lewis antigen negative vs CA19-9 < 180 vs CA19-9 ≥ 180 and Lewis antigen negative vs CA19-9 < 90 vs CA19-9 ≥ 90) with <180 and <90 as the reference levels. The CA19-9 cut points of 180 and 90 were used as in the previously published analysis9 and based on previously published data from the Fox-Chase Cancer Center (180 cut point)11 as well as the previously published adjuvant chemotherapy trial CONKO-001 (90 cut point).3 CA19-9 was also analyzed as a continuous variable. SMS was analyzed as negative versus positive versus unknown (ie, no margin comment in pathology report; shown to have disease-free survival [DFS] and OS outcomes similar to negative-margin patients).6 This variable was broken into 2 dummy variables with a value of negative as the reference level.
OS and DFS were estimated univariately with the Kaplan-Meier method,12 and levels of CA19-9 and SMS were compared using the log-rank test. The first follow-up evaluation was required at 2 to 4 weeks after completion of chemoradiation and prior to the start of maintenance chemotherapy. Thereafter, follow-up examinations were required q 3 months for one year, then q 6 months for 2 years, then yearly and included use of abdominal computed tomography scans. Local relapse (LR) was defined as recurrence at the primary resection site; regional relapse (RR) was defined as recurrence in the regional lymph nodes associated with the primary resection site, local/regional relapse (LRR) combined both LR and RR, and all other disease relapses were defined as distant failure (DF). LR, RR, LRR, and DF were estimated by the cumulative incidence method,13 and levels of CA19-9 and SMS were compared using Gray's test. LR, RR, LRR, and DF failures were counted regardless of when they occurred relative to each other. Only death was considered a competing risk for LR, RR, LRR, and DF. Univariate and multivariable Cox proportional hazards models14 were used to identify the impact of CA19-9 and SMS on the OS, DFS, LR, RR, LRR, and DF. CA19-9 and SMS were forced into their respective models, and stepwise selection procedures were used to choose other variables using α = .05 level as the entry and exit criteria for the model building. The following variables were assessed in the models along with CA19-9: treatment arm, age, sex, race, tumor location, nodal involvement, tumor diameter, and SMS. The same variables were assessed in the models for SMS with inclusion of CA19-9. To adjust for multiple comparisons, a P value ≤ .01 was considered statistically significant and >.01 to ≤ .05 a trend.
Results
The study opened on July 20, 1998 and closed on July 26, 2002 with a total of 538 patients. Of the 538 patients entered, 451 were eligible and analyzable for SMS. Of the 87 cases excluded, 85 were ineligible as detailed in the original report6 and 2 withdrew their consent. Of the 451 eligible patients, 66 had no analyzable postresection CA19-9 data. These 66 patients were Lewis antigen–positive cases for whom tissue was sent to the AAAA tissue bank, per the protocol, but CA19-9 levels could not be determined. Therefore, the sample size for this analysis was 385. The CONSORT diagram is shown in Figure 1, Figure 1.
Figure 1.
U.S. intergroup NRG Oncology RTOG 9704 phase 3 CONSORT diagram.
The median follow-up times were the same for both analysis groups: 1.48 years for all patients and 6.98 years for surviving patients (n = 73). Table 1 shows the pretreatment characteristics for patients entered into this study by the postresection CA19-9 value cut point of 90. No significant difference in pretreatment characteristics was observed by CA19-9 cut point grouping of 90. Results were similar at the CA19-9 180 cut point (220 patients <180, 33 ≥ 180), except that patients with ≥180 were more likely to have tumors ≥3 cm (P = .048). Patients with CA19-9 levels ≥90 or ≥180 were not more likely to be associated with a positive SMS compared with patients with values <90 or <180.
Table 1.
Pretreatment characteristics by CA19-9 (cut point = 90; n = 385)
| Lewis antigen–negative (n = 132) | <90 (n = 200) | ≥90 (n = 53) | P-valuea | |
|---|---|---|---|---|
| Age (years) | ||||
| Median | 60 | 63 | 62 | |
| Range | 37-82 | 35-84 | 39-78 | |
| Sex | .45 | |||
| Male | 74 (56.1%) | 116 (58.0%) | 35 (66.0%) | |
| Female | 58 (43.9%) | 84 (42.0%) | 18 (34.0%) | |
| Race | .16 | |||
| White | 119 (90.2%) | 180 (90.0%) | 43 (81.1%) | |
| Other | 13 (9.8%) | 20 (10.0%) | 10 (18.9%) | |
| Primary location | .19 | |||
| Head | 117 (88.6%) | 176 (88.0%) | 42 (79.2%) | |
| Everything else | 15 (11.4%) | 24 (12.0%) | 11 (20.8%) | |
| Karnofsky performance score | .13 | |||
| 60-80 | 51 (38.6%) | 63 (31.5%) | 24 (45.3%) | |
| 90-100 | 81 (61.4%) | 137 (68.5%) | 29 (54.7%) | |
| T-stage | .13 | |||
| T1, T2 | 23 (17.4%) | 54 (27.0%) | 12 (22.6%) | |
| T3, T4 | 109 (82.6%) | 146 (73.0%) | 41 (77.4%) | |
| N-stage (surgical) | .31 | |||
| N0 | 37 (28.0%) | 72 (36.0%) | 18 (34.0%) | |
| N1 | 95 (72.0%) | 128 (64.0%) | 35 (66.0%) | |
| American Joint Committee on Cancer stage | .088 | |||
| I, II | 31 (23.5%) | 69 (34.5%) | 18 (34.0%) | |
| III, IV | 101 (76.5%) | 131 (65.5%) | 35 (66.0%) | |
| Largest tumor dimension of primary | .21 | |||
| <3 cm | 56 (42.4%) | 87 (43.5%) | 16 (30.2%) | |
| ≥3 cm | 76 (57.6%) | 113 (56.5%) | 37 (69.8%) | |
| Primary tumor status | .70 | |||
| Complete resection/negative margins | 50 (37.9%) | 80 (40.0%) | 24 (45.3%) | |
| Complete resection/positive margins | 51 (38.6%) | 65 (32.5%) | 17 (32.1%) | |
| Complete resection/unknown margins | 31 (23.5%) | 55 (27.5%) | 12 (22.6%) | |
| Prescription | .86 | |||
| Radiation therapy + 5-fluorouracil | 64 (48.5%) | 101 (50.5%) | 28 (52.8%) | |
| Radiation therapy + gemcitabine | 68 (51.5%) | 99 (49.5%) | 25 (47.2%) |
P-value from χ2 test.
Table 2 shows the pretreatment characteristics of patients entered into this study by the SMS. Patients with positive SMS were more likely to have head of pancreas tumors, Karnofsky performance scores of 60 to 80, T3/T4/N1, American Joint Committee on Cancer stage III/IV disease, and primary tumor diameters of ≥3 cm.
Table 2.
Pretreatment characteristics by surgical margin status (n = 385)
| Negative margins (n = 154) | Positive margins (n = 133) | Unknown margins (n = 98) | P-valuea | |
|---|---|---|---|---|
| Age (years) | ||||
| Median | 61 | 61 | 63 | |
| Range | 38-84 | 36-80 | 35-82 | |
| Sex | .24 | |||
| Male | 82 (53.2%) | 82 (61.7%) | 61 (62.2%) | |
| Female | 72 (46.8%) | 51 (38.3%) | 37 (37.8%) | |
| Race | .13 | |||
| White | 131 (85.1%) | 123 (92.5%) | 88 (89.8%) | |
| Other | 23 (14.9%) | 10 (7.5%) | 10 (10.2%) | |
| Primary location | .019 | |||
| Head | 125 (81.2%) | 122 (91.7%) | 88 (89.8%) | |
| Everything else | 29 (18.8%) | 11 (8.3%) | 10 (10.2%) | |
| Karnofsky performance score | .0047 | |||
| 60-80 | 49 (31.8%) | 62 (46.6%) | 27 (27.6%) | |
| 90-100 | 105 (68.2%) | 71 (53.4%) | 71 (72.4%) | |
| T-stage | < .0001 | |||
| T1, T2 | 55 (35.7%) | 9 (6.8%) | 25 (25.5%) | |
| T3, T4 | 99 (64.3%) | 124 (93.2%) | 73 (74.5%) | |
| N-stage (surgical) | .12 | |||
| N0 | 52 (33.8%) | 36 (27.1%) | 39 (39.8%) | |
| N1 | 102 (66.2%) | 97 (72.9%) | 59 (60.2%) | |
| American Joint Committee on Cancer stage | .025 | |||
| I, II | 50 (32.5%) | 30 (22.6%) | 38 (38.8%) | |
| III, IV | 104 (67.5%) | 103 (77.4%) | 60 (61.2%) | |
| Largest tumor dimension of primary | .0003 | |||
| <3 cm | 82 (53.2%) | 40 (30.1%) | 37 (37.8%) | |
| ≥3 cm | 72 (46.8%) | 93 (69.9%) | 61 (62.2%) | |
| CA19-9 | .70 | |||
| Lewis antigen-negative | 50 (37.9%) | 80 (40.0%) | 24 (45.3%) | |
| <90 | 51 (38.6%) | 65 (32.5%) | 17 (32.1%) | |
| ≥90 | 31 (23.5%) | 55 (27.5%) | 12 (22.6%) | |
| Prescription | .35 | |||
| Radiation therapy + 5-fluorouracil | 84 (54.5%) | 64 (48.1%) | 45 (45.9%) | |
| Radiation therapy + gemcitabine | 70 (45.5%) | 69 (51.9%) | 53 (54.1%) |
P-value from χ2 test.
eTable 1; available as supplementary material online only at www.practical.radonc.org shows the overall univariate analyses of disease recurrence for postresection CA19-9 value at the cut point of 90 and for SMS, with 4-year cumulative incidence reported. Postresection CA19-9 was associated with significant increases in both LRR (58% at ≥90 vs 44% at <90, P = .012; Fig 2A) and DF (89% at ≥90 vs 73% at <90, P < .0001; Fig 2B). Although these results did not meet the multiple comparisons adjusted P-value, a trend toward statistical significance for LRR was observed. In the gemcitabine treatment arm, this postresection CA19-9 association was significant for DF and not for LRR. In the 5-fluorouracil treatment arm, this postresection CA19-9 association was significant for both LRR and DF. These findings overall and by treatment arm were similar at the postresection CA19-9 180 cut point. SMS was not associated with significant differences in LRR or DF (Figs 2C and D).
Figure 2.
(A) Local–regional failure by CA19-9; (B) distant failure by CA19-9; (C) local-regional failure by surgical margin status; and (D) distant failure by surgical margin status.
Table 3 shows the multivariable modeling of postresection CA19-9 value at the cut point of 90 for associations with survival and pattern of disease recurrence. In addition to being a predictor of OS and DFS, as previously reported,9 postresection CA19-9 showed associations with LR, RR, LRR, and DF. Patients with CA19-9 ≥ 90 had a significant increased risk of LRR compared with patients with CA19-9 values of <90 (P < .0001; hazard ratio [HR]: 2.91; 95% confidence interval [CI], 1.90-4.46) and a significant increased risk of DF compared with patients with CA19-9 values of <90 (P < .0001; HR: 2.69; 95% CI, 1.92-3.77). Results were similar at the 180 cut point. eTable 2; available as supplementary material online only at www.practical.radonc.org shows the multivariable modeling of SMS for associations with survival and pattern of disease recurrence. On multivariable analysis, positive SMS was not associated with significant differences in LRR.
Table 3.
Stepwise multivariable Cox proportional hazards models: CA19-9 forced into modals (n = 385)
| Endpoint | Adjustment variables | Comparison | Adjusted HRa | 95% CI LL | 95% CI UL | P-value* |
|---|---|---|---|---|---|---|
| LR |
CA19-9 | Lewis antigen–negative | 0.90 | 0.61 | 1.32 | .57 |
| <90 | 1.00 | – | – | – | ||
| ≥90 | 2.59 | 1.59 | 4.21 | .0001 | ||
| Age | Continuous | 0.98 | 0.96 | 0.99 | .021 | |
| Surgical margin status | Negative | 1.00 | – | – | – | |
| Positive | 1.00 | 0.69 | 1.46 | .99 | ||
| Unknown | 0.55 | 0.34 | 0.87 | .012 | ||
| RR |
CA19-9 | Lewis antigen–negative | 1.19 | 0.72 | 1.95 | .50 |
| <90 | 1.00 | – | – | – | ||
| ≥90 | 4.06 | 2.24 | 7.33 | < .0001 | ||
| Nodal involvement | No | 1.00 | – | – | – | |
| Yes | 1.65 | 1.01 | 2.70 | .045 | ||
| LRR |
CA19-9 | Lewis antigen–negative | 0.96 | 0.68 | 1.36 | .82 |
| <90 | 1.00 | – | – | – | ||
| ≥90 | 2.91 | 1.90 | 4.46 | < .0001 | ||
| Age | Continuous | 0.98 | 0.97 | 0.99 | .031 | |
| Sex | Female | 1.00 | – | – | – | |
| Male | 1.38 | 1.01 | 1.90 | .045 | ||
| Surgical margin status | Negative | 1.00 | – | – | – | |
| Positive | 0.99 | 0.71 | 1.39 | .95 | ||
| Unknown | 0.63 | 0.42 | 0.95 | .025 | ||
| DF |
CA19-9 | Lewis antigen–negative | 0.98 | 0.75 | 1.27 | .86 |
| <90 | 1.00 | – | – | – | ||
| ≥90 | 2.69 | 1.92 | 3.77 | < .0001 | ||
| Surgical margin status | Negative | 1.00 | – | – | – | |
| Positive | 1.31 | 1.00 | 1.70 | .048 | ||
| Unknown | 0.84 | 0.62 | 1.13 | .25 | ||
| OS |
CA19-9 | Lewis antigen–negative | 1.27 | 0.99 | 1.62 | .06 |
| <90 | 1.00 | – | – | – | ||
| ≥90 | 3.42 | 2.47 | 4.73 | < .0001 | ||
| Race | White | 1.00 | – | – | – | |
| African-American/other | 1.53 | 1.10 | 2.14 | .012 | ||
| Nodal involvement | No vs yes | 1.51 | 1.18 | 1.93 | .0011 | |
| DFS | CA19-9 | Lewis antigen–negative | 1.15 | 0.91 | 1.45 | .26 |
| <90 | 1.00 | – | – | – | ||
| ≥90 | 2.82 | 2.05 | 3.87 | < .0001 | ||
CI, confidence interval; DF, distant failure; DFS, disease-free survival; HR, hazard ratio; LL, lower limit; LR, local relapse; OS, overall survival; LRR, local/regional recurrence; RR, regional relapse; UL, upper limit.
A hazard ratio of 1 indicates no difference between the 2 subgroups.
P-value from χ2 test using the Cox proportional hazards model.
Although 90 and 180 are well-established cut points for CA19-9, these analyses were also conducted using CA19-9 as a continuous variable, which produced similar results.
Discussion
The primary objectives of the phase 3 NRG Oncology RTOG 9704 trial included a prospective evaluation of the ability of postresectional CA19-9 to predict survival among adjuvantly treated patients who had undergone a potentially curative resection for adenocarcinoma of the pancreas and the pattern of disease recurrence after adjuvant therapy. There are 2 major findings from this analysis: Postresection CA19-9 predicts for pattern of disease recurrence after adjuvant therapy, inclusive of both LRR and DF, whereas SMS is only associated with DF. Although SMS and postresection CA19-9 levels have been described as prognostic factors for survival in patients with pancreatic carcinoma who are able to undergo surgical resection,9, 15, 16 the findings in this report challenge the commonly held beliefs that SMS and elevated postresection CA19-9 levels are the major predictors of primarily LRR and DF, respectively. These findings could greatly influence the prioritization and/or timing of adjuvant chemotherapy and/or RT.
NRG Oncology RTOG 9704 was the first phase 3 trial to perform a prospective analysis of postresection CA 19-9 levels in patients treated with adjuvant CRT. The trial demonstrated that in the postoperative setting, postresection CA 19-9 levels were the most important predictor of OS. Patients with CA19-9 of ≥180 had a median survival of 9 months, compared with 21 months for those with CA 19-9 < 180 (P < .0001; HR: 3.58). At the cut point of 90, as used for eligibility for the phase 3 adjuvant CONKO-0001 trial (only patients with values <90 were allowed in the trial), the median survival time for patients with CA 19-9 ≤ 90 was 23 months, whereas patients with CA 19-9 > 90 had a median survival of 10.4 months (P < .0001; HR: 3.34).9
These findings became the basis for stratification of patients within the successor, and currently active, YYYY adjuvant phase 3 trial. In this study, patients are stratified at the postresection CA19-9 cut point of 90, and patients with values >180 are ineligible. Although the primary analysis of NRG Oncology RTOG 9704 demonstrated that >70% of patients developed DF and approximately 30% LRR at the time of the first relapse of disease, the pattern of disease recurrence potentially associated with postresection CA19-9 levels had not been evaluated.6, 7, 9 Postresection elevation in CA19-9 is often presumed to be primarily a predictor of DF without consideration of the associated potential for LRR. Although the report of this analysis supports the association with DF, the findings have also demonstrated that such an elevation in CA19-9 levels is associated with LRR at a similar magnitude. In addition, patients with postresection CA19-9 levels ≥90 had an almost 3-fold increase in LRR compared with those with postresection CA19-9 levels of 90.
Postresection SMS has been reported in large institutional series to be a significant and independent predictor of survival.15, 16 However, this observation has not been routinely demonstrated in prospective phase 3 trials evaluating use of adjuvant therapy. In fact, in the 3 randomized trials that have demonstrated at least a suggested benefit in survival with the use of adjuvant therapy, SMS was not found to be an independent predictor of survival.3, 4, 5, 6, 7 In addition, evaluation of SMS on its association with pattern of disease recurrence in a large institutional series and in the setting of prospective adjuvant phase 3 trials has not been routinely reported.3, 4, 5, 6, 7, 8, 15, 16 Among patients with a postresection positive SMS, LRR is typically presumed to be a significant contributor to the pattern of disease recurrence.
Furthermore, the presence of positive SMS is often, in part, the basis for recommendation of adjuvant RT15, 16 and is used as the rationale for consideration of RT dose escalation.17, 18 This report challenges the utility of the use of postresection SMS as a predictor of LRR and as a basis for the use of RT or RT dose escalation in the setting of available postresection CA19-9 levels. Positive SMS has been previously demonstrated to not be an independent predictor of patient survival,6, 7 with NRG Oncology RTOG 9704 being the first phase 3 trial that required central RT quality assurance review.19 The current analysis demonstrates that in the setting of central RT quality assurance, positive SMS is also not an independent predictor of LRR. In addition, evaluation of SMS in the setting of available postresection CA19-9 levels as performed in this report demonstrates that postresection CA19-9 was the only independent predictor of both LRR and DF.
Conclusions
In this unique adjuvant setting of a prospective evaluation of postresection CA19-9 levels, CA19-9 has significant association with both LRR and DF that was not seen with SMS. These findings contradict the commonly held presumption that an elevation in postresection CA19-9 levels reflects the development of distant disease spread of pancreatic cancer without consideration of LRR. These findings support consideration of adjuvant RT dose intensification among patients with elevated postresection CA19-9.
Footnotes
Sources of support: This project was supported by grants U10CA21661 (RTOG-Ops-Stat), U10CA180868 (NRG Oncology Operations), U10CA180822 (NRG Oncology SDMC), and U10CA37422 (CCOP) from the National Cancer Institute (NCI) and Eli Lilly.
Conflicts of interest: Dr. Delouya reports personal fees from Elekta, Bayer, Astellas, and Janssen outside the submitted work. Dr. Suh reports personal fees from Varian Medical Systems and Philips as well as other fees from Elekta outside the submitted work. Ms. Winter reports the receipt of an NCI grant as a subcontract to ACR (her employer during the conduct of the study). All other authors have no conflicts of interest to declare.
Supplementary material for this article (https://doi.org/10.1016/j.adro.2018.01.003) can be found at www.practicalradonc.org.
Supplementary data
The following is the supplementary data to this article:
US Intergroup/NRG Oncology RTOG 9704 Phase III Study Schema.
eTables 1 and 2.
References
- 1.Siegel R., Ma J., Zou Z., Jemal A. Cancer statistics, 2014. CA Cancer J Clin. 2014;64:9–29. doi: 10.3322/caac.21208. [DOI] [PubMed] [Google Scholar]
- 2.Kalser M.H., Ellenberg S.S. Pancreatic cancer. Adjuvant combined radiation and chemotherapy following curative resection. Arch Surg. 1985;20:899–903. doi: 10.1001/archsurg.1985.01390320023003. [DOI] [PubMed] [Google Scholar]
- 3.Oettle H., Post S., Neuhaus P. Adjuvant chemotherapy with gemcitabine vs observation in patients undergoing curative-intent resection of pancreatic cancer: A randomized controlled trial. JAMA. 2007;297:267–277. doi: 10.1001/jama.297.3.267. [DOI] [PubMed] [Google Scholar]
- 4.Oettle H., Neuhaus P., Hochhaus A. Adjuvant chemotherapy with gemcitabine and long-term outcomes among patients with resected pancreatic cancer. JAMA. 2013;310:1473–1481. doi: 10.1001/jama.2013.279201. [DOI] [PubMed] [Google Scholar]
- 5.Neoptolemos J.P., Stocken D.D., Friess H. A randomized trial of chemoradiotherapy and chemotherapy after resection of pancreatic cancer. N Engl J Med. 2004;350:1200–1210. doi: 10.1056/NEJMoa032295. [DOI] [PubMed] [Google Scholar]
- 6.Regine W.F., Winter K., Abrams R.A. Fluorouracil vs gemcitabine chemotherapy before and after fluorouracil-based chemoradiation following resection of pancreatic adenocarcinoma: a randomized controlled trial. JAMA. 2008;299:1019–1026. doi: 10.1001/jama.299.9.1019. [DOI] [PubMed] [Google Scholar]
- 7.Regine W.F., Winter K., Abrams R. Fluorouracil-based chemoradiation with either gemcitabine or flourouracil chemotherapy following resection of pancreatic adenocarcinoma: 5-year analysis of the U.S. Intergroup/RTOG 9704 phase III trial. Ann Surg Oncol. 2011;18:1319–1326. doi: 10.1245/s10434-011-1630-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Klinkenbijl J.H., Jeekel J., Sahmoud T. Adjuvant radiotherapy and 5-fluorouracil after curative resection of cancer of the pancreas and periampullary region: Phase III trial of the EORTC Gastrointestinal Tract Cancer Cooperative Group. Ann Surg. 1999;230:776–782. doi: 10.1097/00000658-199912000-00006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Berger A.C., Garcia M., Hoffman J.P. Postresection CA 19-9 predicts overall survival in patients with pancreatic cancer treated with adjuvant chemoradiation: a prospective validation by RTOG 9704. J Clin Oncol. 2008;26:5918–5922. doi: 10.1200/JCO.2008.18.6288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tempero M.A., Uchida E., Takasaki H., Burnett D.A., Steplewski Z., Pour P.M. Relationship of carbohydrate antigen 19-9 and Lewis antigens in pancreatic cancer. Cancer Res. 1987;47:5501–5503. [PubMed] [Google Scholar]
- 11.Montgomery R.C., Hoffman J.P., Riley L.B. Prediction of recurrence and survival by post-resection CA 19-9 values in patients with adenocarcinoma of the pancreas. Ann Surg Oncol. 1997;4:551–556. doi: 10.1007/BF02305535. [DOI] [PubMed] [Google Scholar]
- 12.Kaplan E.L., Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457–481. [Google Scholar]
- 13.Kalbfleish J.D. John Wiley & Sons; New York: 1980. The Statistical Analysis of Failure Time Data. [Google Scholar]
- 14.Cox D.R. Regression models and life-tables. J Roy Statist Soc. 1972;34:187–202. [Google Scholar]
- 15.Herman J.M., Swartz M.J., Hsu C.C. Analysis of fluorouracil-based adjuvant chemotherapy and radiation after pancreaticoduodenectomy for ductal adenocarcinoma of the pancreas: Results of a large, prospectively collected database at the Johns Hopkins Hospital. J Clin Oncol. 2008;26:3503–3510. doi: 10.1200/JCO.2007.15.8469. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chang D.K., Johns A.L., Merrett N.D. Margin clearance and outcome in resected pancreatic cancer. J Clin Oncol. 2009;27:2856–2862. doi: 10.1200/JCO.2008.20.5104. [DOI] [PubMed] [Google Scholar]
- 17.Willett C.G., Lewandrowski K., Warshaw A.L., Efird J., Compton C.C. Resesection margins in carcinoma of the head of the pancreas implications for radiation therapy. Ann Surg. 1993;217:144–148. doi: 10.1097/00000658-199302000-00008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rwigema J.C., Heron D.E., Parikh S.D. Adjuvant stereotactic body radiotherapy for resected pancreatic adenocarcinoma with close or positive margins. J Gastrointest Cancer. 2012;43:70–76. doi: 10.1007/s12029-010-9203-7. [DOI] [PubMed] [Google Scholar]
- 19.Abrams R.A., Winter K.A., Regine W.F. Failure to adhere to protocol specified radiation therapy guidelines was associated with decreased survival in RTOG 9704––a phase iii trial of adjuvant chemotherapy and chemoradiotherapy for patients with resected adenocarcinoma of the pancreas. Int J Radiat Oncol Biol Phys. 2012;82:809–816. doi: 10.1016/j.ijrobp.2010.11.039. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
US Intergroup/NRG Oncology RTOG 9704 Phase III Study Schema.
eTables 1 and 2.