By William B. Ershler, MD, Editor
Synopsis:In a retrospective review of several clinical trials for metastatic colorectal cancer conducted in Germany, a comprehensive examination of risk factors for mortality within 2 months of chemotherapy, white blood cell count and performance status were found to be significant. A combination of these two factors results in an index (60-day mortality score) with reasonable sensitivity and high specificity. Application of such an index may be useful in pre-treatment chemotherapy decision making.
Source:Giessen C, et al. Prognostic factors for 60-day mortality in first-line treatment of metastatic colorectal cancer (mCRC): Individual patient analysis of four randomized, controlled trials by the AIO colorectal cancer study group. Ann Oncol 2013;24:3051-3055.
The reported 60-day mortality in clinical trials for metastatic colorectal cancer (mCRC) is an established parameter for chemotherapy-related safety. However, despite being widely used, reporting of the 60-day mortality in clinical trials is not standardized.1 Identification of prognostic factors associated with 60-day mortality in patients with mCRC would be useful to identify patients at higher risk. In 2011, Giessen and colleagues identified white blood cell count (WBC) and lactate dehydrogenase (LDH) as prognostic factors for all-cause, 60-day mortality in an irinotecan-based trial for mCRC.2 However, a large analysis of prognostic factors associated with 60-day mortality has not been conducted to date.
To further define prognostic factors for early mortality, the Arbeitsgemeinschaft Internistische Onkologie (AIO) study group recently conducted a large, retrospective analysis on data collected from four randomized AIO mCRC trials. The first study (CAPOX/FUFOX) was a randomized, Phase 3 study that compared first-line treatment with capecitabine plus oxaliplatin (CAPOX) vs fluorouracil and leucovorin plus oxaliplatin (FUFOX) as first-line treatment of mCRC.3 The second study (AIO KRK 0104) was a randomized, Phase 2 study that compared first-line treatment with cetuximab plus capecitabine and irinotecan (CAPIRI) with cetuximab plus CAPOX.4 The third study (AIO KRK 0604) was a randomized, Phase 2 trial that compared treatment with CAPIRI vs CAPOX in combination with bevacizumab.5 The fourth study (FIRE-1) was a Phase 3 trial that compared treatment with either FUFIRI or mIROX.6 These studies were all conducted at multiple treatment sites throughout Germany between July 2000 and December 2006.
The 60-day mortality was defined as death within the first 60 days of initiation of chemotherapy treatment. In this retrospective analysis, the investigators reviewed medical records, trial databases, and adverse effect reports to investigate causes of death and such were categorized as: disease-related (tumor-progression), sepsis, severe gastrointestinal toxicity, thromboembolic events, cardiac, and bleeding events. Prognostic factors were determined using univariate and multivariate logistic regression models.
Two different cohorts were evaluated due to varying performance status inclusion criteria. Initial analysis was done on the 898 patients with ECOG performance status of 0, 1, or 2 who were enrolled in the AIO 0104 trial, the AIO 0604 trial, or the CAPOX/FUFOX trial. Among these participants, there were 33 deaths within the first 60 days of treatment (60-day mortality rate = 3.7%). Since the FIRE-1 trial required Karnofsky performance status ≥ 70%, a second cohort was analyzed that included patients enrolled in this study in addition to the subset of patients with ECOG performance status of 0 or 1 who were enrolled in the other studies. Of the 1302 patients assessed in this cohort, there were 44 deaths (60-day mortality rate = 3.4%).
The main cause of death within the first 60 days was disease related in both cohorts (18/33 [54.5%] in cohort 1 and 21/44 [47.7%] in cohort 2).
In the first cohort, the median age of the 33 patients who died within the first 60 days of treatment was slightly older than the median patient population (66 vs 64 years). Of the total number of enrolled subjects, the proportion of patients with ECOG 2 who died within the first 60 days was significantly greater than the proportion alive after 60 days (39% vs 4%, P < 0.001). Similarly, the proportion of patients with ECOG 0 who died within the first 60 days was significantly less than the proportion alive after 60 days (27% vs 54%, P < 0.001). ECOG 1 was comparable among the two groups. In addition, the proportion of patients with baseline WBC ≥ 8K/µL was significantly higher in patients who died within the first 60 days (67% vs 43%, P = 0.007).
Similar results were seen in the second cohort. The 44 patients who died within the first 60 days were older than those alive after 60 days (66.0 vs 63.6 years). The proportion of patients with ECOG 0 who died within the first 60 days was significantly less than the proportion alive after 60 days (43% vs 63%, P = 0.010). Additionally, the proportion of patients with baseline WBC ≥ 8.00/µL was significantly higher in patients who died within the first 60 days (66% vs 42%, P = 0.006, odds ratio 7.60). There were no statistically significant differences in LDH and alkaline phosphatase between the groups in either cohort.
Multivariate analysis of all patients revealed that both ECOG PS 2 and increased WBC were associated with an increased risk of death within the first 60 days. In the first cohort, odds ratio for ECOG PS 2 vs 0 was 12.92 (95% confidence interval [CI], 4.73-36.52; P < 0.001), and odds ratio for WBC was 6.28 (95% CI, 2.17-19.06; P < 0.001). In the second cohort, the odds ratio for WBC was 7.60 (95% CI, 3.11-18.60; P < 0.001). Other tumor and baseline patient characteristics were not associated with increased likelihood of early mortality.
This comprehensive analysis of factors that might predict early death after mCRC treatment yielded nothing surprising but provided evidence that may be useful for treatment decisions. Notably, the majority of deaths at 60 days was related to progressive malignancy. However, the risk was higher for those with poorer performance (PS 2 vs PS 0/1) and for those with WBC > 8K/µL. In the analysis, having both of these prognostic factors (i.e., AIO-60-Day-Mortality Score 2) yielded an odds ratio of 30.54 (10.58-88.20; P < 0.001). The calculated sensitivity for this combined score in predicting 60-day mortality was 18% and the specificity was 98%. The positive predictive value was 32%. Thus, calculating 60-day mortality risk by these readily available parameters (ECOG PS and WBC) can be used to identify high-risk patients who may require heightened surveillance after standard or experimental chemotherapy or, alternatively, a reduction in treatment intensity.