Combination therapy with gemcitabine and nanoparticle albumin-bound paclitaxel (nab-paclitaxel), known as GnP therapy, significantly prolongs the survival of pancreatic cancer patients compared with gemcitabine monotherapy. However, it may cause severe neutropenia, requiring discontinuation of treatment. This study aimed to clarify the risk factors for Grade 3/4 neutropenia during GnP therapy.
Clinical data of pancreatic cancer patients who underwent GnP therapy at the Cancer Institute Hospital of the Japanese Foundation for Cancer Research from December 2014 to December 2016 were retrospectively collected. The relationship of Grade 3/4 neutropenia onset to laboratory values and patient background factors was investigated by multivariate logistic regression analysis.
Clinical data of 222 patients were analyzed. Grade 3/4 neutropenia occurred in 118 patients (53.2%) in the first cycle of GnP therapy. Multivariate analysis identified low absolute neutrophil count (ANC), high total bilirubin (T-Bil), and low C-reactive protein (CRP) as risk factors for Grade 3/4 neutropenia. Age was not a risk factor. The incidence of neutropenia was 85.7% in patients with all three risk factors, but only 27.7% in patients with none of them.
Citation: Ito G, Kawakami K, Aoyama T, Yokokawa T, Nakamura M, Ozaka M, et al. (2021) Risk factors for severe neutropenia in pancreatic cancer patients treated with gemcitabine/nab-paclitaxel combination therapy. PLoS ONE 16(7): e0254726. https://doi.org/10.1371/journal.pone.0254726
Editor: Ming-Lung Yu, Kaohsiung Medical University Chung Ho Memorial Hospital, TAIWAN
Received: March 24, 2021; Accepted: July 1, 2021; Published: July 14, 2021
Copyright: © 2021 Ito et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data cannot be shared publicly because of the privacy and ethical restrictions. Data are available from the Ethics Committee of Cancer Institute Hospital (contact via e-mail: email@example.com) for researchers who meet the criteria for access to confidential data. Researchers need to be approved by conducting a research ethics review at the organization to which the researchers using the data.
Funding: This study was supported by Keio University Faculty of Pharmaceutical Sciences Research Fund.
Competing interests: The authors have declared that no competing interests exist.
Cancer chemotherapy may cause various adverse events, including nausea, hair loss, diarrhea, myelosuppression, liver damage and renal damage. Myelosuppression reduces neutrophils and increases the risk of developing a fatal infection. Therefore, neutropenia is a dose-limiting factor for many anticancer agents and may require a decrease in treatment intensity, such as dose reduction, postponement, or even discontinuation [1–3].
Risk factors for chemotherapy-induced neutropenia have been investigated in various dosing regimens and cancer types. A prospective observational study that analyzed data from 3,760 patients with solid tumor or malignant lymphoma identified high alkaline phosphatase (ALP), high total bilirubin (T-Bil), low glomerular filtration rate (GFR), low white blood cell (WBC) count, high relative dose intensity (RDI), and immunosuppressant administration as risk factors for severe neutropenia . Other reported risk factors include female gender, renal disorder, liver dysfunction, low absolute neutrophil count (ANC) at the start of treatment, and bone metastasis [5–10].
Combination therapy with gemcitabine and nanoparticle albumin-bound paclitaxel (nab-paclitaxel), known as GnP therapy, significantly prolongs the survival of pancreatic cancer patients compared with conventional therapies such as gemcitabine monotherapy, and is considered as a standard treatment for pancreatic cancer chemotherapy, together with FOLFIRINOX (combination of 5-fluorouracil, leucovorin, irinotecan, and oxaliplatin) [11, 12]. On the other hand, myelosuppression is frequently observed, and a domestic phase II study of GnP therapy in Japan found Grade 3 or higher neutropenia in 70.6% of treated patients and febrile neutropenia in 5.9% . Therefore, the target patients of GnP therapy are primarily reasonably strong patients in good general condition, and the therapy is generally considered unsuitable for elderly people. However, little is yet known about risk factors for severe neutropenia in GnP-treated patients. Recently, Makita et al. reported that low WBC and high alanine aminotransferase (ALT) levels may be risk factors for developing severe neutropenia in the first cycle of GnP therapy . However, theirs was a retrospective observational study of a small population of 52 cases, and further analysis in a larger population is desirable.
In the present study, we aimed to identify the risk factors for severe (Grade 3/4) neutropenia using clinical data collected from a larger population of pancreatic cancer patients who received 1–6 cycles of GnP therapy.
Study setting and population
Three hundred and two patients with unresectable pancreatic cancer who underwent standard GnP therapy were targeted in this study. Data were collected from the electronic medical records of patients seen from December 1, 2014 to December 31, 2016 at the Cancer Institute Hospital of the Japanese Foundation for Cancer Research. All data were analyzed anonymously and informed consent was waived due to the retrospective observational design of the study. The data included patient background factors and clinical laboratory parameters, including sex, age, weight, body surface area (BSA), Eastern Cooperative Oncology Group Performance Status (ECOG PS), concomitant drug(s), complications, gemcitabine dose, nab-paclitaxel dose, WBC, red blood cell (RBC), hemoglobin (Hgb), platelet (PLT), ANC, lymphocyte (Lymph), albumin (ALB), T-Bil, direct bilirubin (D-Bil), γ-glutamyltransferase (γ-GTP), alkaline phosphatase (ALP), lactate dehydrogenase (LDH), aspartate aminotransferase (AST), alanine aminotransferase (ALT), serum glucose (S-Glu), urea nitrogen (UN), serum creatinine (S-Cre), estimated GFR (eGFR), C-reactive protein (CRP), and carcinoembryonic antigen (CEA).
Standard GnP therapy was administered as follows. Patients received a 30-min intravenous infusion of nab-paclitaxel at a dose of 125 mg/m2, followed by a 30-min intravenous infusion of gemcitabine at a dose of 1,000 mg/m2, on days 1, 8, and 15 every 4 weeks.
The clinical laboratory data on day 1, 8, 15 of each cycle of GnP therapy were collected. Regarding the data on days 8 and 15, a one-day deviation from the scheduled test date was allowed, but patients with a shift of two days or more were excluded. The definition of neutropenia used here was ANC <1,000/mm3, which is Grade 3 or higher in the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. This study was approved by the ethics committees of the Cancer Institute Hospital (approval No. 2019–1040) and the Keio University Faculty of Pharmacy (approval No. 190613–4).
Outcome and predictor variables
The primary outcome of the study was onset of neutropenia of Grade 3 or higher. As explanatory variables for risk factor analysis, we selected sex, age, weight, BSA, ECOG PS, concomitant drug(s), complications, gemcitabine dose, nab-paclitaxel dose, and laboratory values (WBC, RBC, Hgb, PLT, ANC, Lymph, ALB, T-Bil, D-Bil, γ-GTP, ALP, LDH, AST, ALT, S-Glu, UN, S-Cre, eGFR, CRP) on day 1 of the first cycle (before administration). These risk factors were selected with reference to previous reports [4–10].
Univariate analyses (Mann-Whitney U test, χ2 test) comparing the Grade 3/4 neutropenia group and the non-neutropenia group were performed for each candidate variable described above. Multivariate logistic regression analysis was performed by the simultaneous forced entry method for the variables with p<0.1 in the univariate analyses, as well as factors considered likely to be relevant based on the medical/pharmacological findings. Quantitative variables were analyzed by receiver operating characteristic (ROC) curve analysis to find the cutoff value that maximizes the Youden index, and then were stratified above and below the cutoff value. Qualitative variables were converted into dummy variables. A p value less than 0.05 was considered statistically significant, and the odds ratio (OR) and 95% confidence interval (CI) were calculated for each factor. The number of explanatory variables input to the multivariate logistic regression analysis was set to one-tenth of the number of event occurrence samples .
For statistical analysis, statistical software IBM SPSS Statistics 25.0 (SPSS Inc., Chicago, IL, USA) was used.
Patient characteristics and incidence of neutropenia
Of the 302 patients for whom data were collected, 34 were excluded because of the absence of clinical laboratory data on Day 1, 8, or 15 of the first cycle. We also excluded 46 patients who were discontinued on Day 8 of the first cycle for reasons other than neutropenia Grade 3/4 (physical condition, peripheral neuropathy, anemia, neutropenia < Grade 3). Finally, 222 patients were analyzed in this study (Fig 1). The 222 patients analyzed in this study had a median age of 66 years (range 41–82). As summarized in Table 1, 15.3% were ≥ 75 years old, 117 were male, 79.3% were ECOG PS (0) and 20.7% were PS (1), 39.1% had liver metastasis, 30.6% had a history of chemotherapy. Of the 222 patients, 18 patients (8.1%) required starting dose reduction. The characteristics of these patients were compared with those of all patients, and they tended to include a higher proportion of patients aged > 75 years old (33.3% vs 15.3%), and those with previous chemotherapy history (55.5% vs 30.6%).
The incidence of Grade 3/4 neutropenia in patients who started each cycle (1 to 6 cycles) of standard GnP therapy is shown in Fig 2. Of the 222 patients, 118 developed Grade 3/4 neutropenia; 96 (81.4%) and 22 (18.6%) of 118 patients had Grade 3 and 4 neutropenia, respectively. Of the patients with Grade 3 neutropenia, the numbers of patients who had fever were as follows: 1 patient on day 8, 1 patient on day 15, and 3 patients on both days 8 and 15. Of the patients with Grade 4 neutropenia, the numbers of patients who had fever, that is febrile neutropenia, were as follows: 1 patient on day 15, and 3 patients on both days 8 and 15. GnP therapy was discontinued in all patients who developed G3/G4 neutropenia with or without fever. The incidences of Grade 3/4 neutropenia in the first cycle and in the second or a subsequent cycle of GnP therapy were 53.2% and 25%, respectively. It should be noted that patients who developed Grade 3/4 neutropenia and discontinued the drug in the first cycle were sometimes subsequently given a reduced dose regimen, or a schedule change to biweekly administration, or were switched to a different therapy.
Day 8&15 (Black): Neutropenia developed on Day 8 and the drug was discontinued, but Grade 3 or higher neutropenia continued until Day 15. Day 15 (Dark gray): Neutropenia was absent on Day 8 but developed on Day 15 and the drug was discontinued. Day 8 (Light gray): Neutropenia developed on Day 8 and the drug was discontinued, but it was reinstated on Day 15.
The incidence of neutropenia in the 18 patients who received a reduced starting dose was 56% (9 patients with Grade 3, 1 patient with Grade 4 neutropenia), which was almost the same as that of all subjects.
The results of univariate analysis using the Mann-Whitney U test for continuous variables and the Pearson’s chi-square test for qualitative variables are shown in Table 2. Liver metastases, WBC, PLT, ANC, T-Bil, γ-GTP, ALP, LDH, CRP and CEA showed values of p < 0.1. CRP values in the severe neutropenia and non-neutropenia groups were 0.43 ± 1.10 mg/dL and 0.89 ± 1.57 mg/dL, respectively.
Multivariate logistic analysis
Multiple logistic regression analysis was conducted by the simultaneous forced entry method for the items selected according to the following criteria and for the following reasons (Table 3).
- Patient background characteristics (sex, age) and factors with p < 0.1 in univariate analysis and < 10% missing values (liver metastasis, WBC, PLT, ANC, T-Bil, γ-GTP, ALP, CRP).
- For WBC/ANC (r = 0.9) and γ-GTP/ALP (r = 0.6), which showed high Spearman’s correlation coefficients, only ANC and ALP were used. For WBC/ANC, ANC was used because the objective variable was neutropenia ≥ Grade 3. For γ-GTP/ALP, ALP was used because of a low rate of missing values.
- ALT was included based on previous findings .
- BSA was included because GEM and nab-PTX doses were calculated based on BSA.
The number of explanatory variables in the multiple logistic regression analysis was selected to be 10 which is less than one-tenth of the number of event occurrence samples (neutropenia 118 and non-neutropenia 104 patients) . As factors with p < 0.05, we identified ANC < 3.03×103 /μL (OR: 4.806, 95% CI: 2.416–9.558, p = 0.000), T-Bil ≥ 0.6 mg/dL (OR: 1.964, 95% CI: 1.040–3.708, p = 0.037), and CRP < 0.13 mg/dL (OR: 2.607, 95% CI: 1.331–5.106, p = 0.005).
Comparison of incidence rates of neutropenia according to the number of risk factors
The incidence of Grade 3/4 neutropenia in the first cycle of GnP therapy was 53.2%. The incidence was 85.7% in patients with all three risk factors (ANC < 3.03×103 /μL, T-Bil ≥ 0.6 mg/dL, CRP < 0.13 mg/dL) and 27.7% in patients without any of these risk factors (Fig 3).
In the present study, we clarified the incidence and risk factors of severe neutropenia (> Grade 3) due to GnP therapy with or without prior treatment in Japanese patients with unresectable pancreatic cancer by the largest retrospective observational study so far.
Incidence rate of severe neutropenia
The incidence of neutropenia ≥ Grade 3 in the first cycle of GnP therapy in patients with pancreatic cancer was 53.2% in this study, whereas Braiteh et al . reported an incidence of 28% in patients with metastatic pancreatic cancer and an incidence of 4% in patients with a previous history of chemotherapy. However, Makita et al . reported an incidence of 61.5% (32/52) in unresectable pancreatic cancer patients in Japan, and Yano et al . found that the incidence of neutropenia due to docetaxel in clinical trials conducted in Asia was higher than in non-Asians. GnP therapy includes paclitaxel, which is a taxane-based anticancer drug similar to docetaxel. Therefore, the difference in the incidence of severe neutropenia between our study and that of Braiteh et al.  may reflect racial differences. This view is supported by the fact that the subjects in our study were similar to those of Makita et al.  in terms of country of origin, disease extent, and previous chemotherapy history, and the incidence of severe neutropenia in our study was close to that found in their study.
Risk factors for severe neutropenia
We identified ANC < 3.03 × 103 /μL, T-Bil ≥ 0.6 mg/dL, CRP < 0.13 mg/dL as risk factors for Grade 3/4 neutropenia in standard GnP therapy (CTCAE ver5.0). The reference ranges are ANC ≧ 2.0 × 103 /μL, T-Bil 0.4–1.5 mg/dL, and CRP 0.00–0.14 mg/dL , so the criterion values for increased risk fall within the reference ranges. Although T-Bil and CRP have not been previously reported as risk factors for Grade 3/4 neutropenia in GnP therapy, our results suggest that close monitoring of adverse events is important even in patients whose clinical laboratory values are within the reference ranges.
Low ANC level has been proposed to be a risk factor of neutropenia in cancer chemotherapy with various agents , not only GnP, so it could be a risk factor for neutropenia regardless of the regimen.
As regards T-Bil, Joerger et al. reported that this parameter is a good predictor of paclitaxel excretion and paclitaxel-induced myelosuppression . Thus, it seems likely that patients with high T-Bil levels in the present study would have had an increased blood concentration of paclitaxel due to decreased clearance, and this could have caused neutropenia.
CRP in neutropenia Grade 0–2 was higher than that in Grade 3/4 (Grade 0–2: 0.28 mg/dL, Grade 3/4: 0.10 mg/dL). Thus, low CRP level might be a risk factor, because ANC would remain high in patients with high CRP level due to increased inflammatory response, and ANC tended to be low in patients with low CRP level. On the other hand, Razzaghdoust et al. reported that high CRP (≥ 0.6 mg/dL) is a risk factor for severe neutropenia or febrile neutropenia in patients with breast cancer and gastrointestinal cancer . In their study, the CRP (mean ± SE) was 2.46 ± 0.06 mg/dL in patients with severe neutropenia or febrile neutropenia and 1.29 ± 0.17 mg/dL in patients without severe neutropenia or febrile neutropenia. In the present study, the CRP values in the severe neutropenia and non-neutropenia groups were 0.43 ± 1.10 mg/dL and 0.89 ± 1.57 mg/dL, respectively. Patients with PS = 2, who might have been predicted to have higher CRP values were not included in this analysis. A possible explanation for the apparent discrepancy is that pancreatic cancer patients with high CRP may not have been selected for GnP therapy, i.e., there could be selection bias in our study.
Since the incidence of pancreatic cancer increases with age, it is important to assess the suitability of chemotherapy regimens for the elderly. At present, GnP therapy generally tends not to be offered to elderly people (≥ 75 years) because of safety concerns. However, we did not find that age is a risk factor for neutropenia, although this study included a small proportion of elderly people (34 patients, 15.3%), who had either PS 0 (26 patients) or PS 1 (8 patients). Therefore, GnP therapy may be an option in elderly people (≥ 75 years) with good PS.
Strengths and limitations
This study is the largest retrospective observational study so far conducted to evaluate risk factors for severe neutropenia in patients treated with GnP therapy. This study design inevitably has various potential limitations, including selection bias, variations of blood sampling date/time, parameters evaluated, inclusion of patients whose dose regimen had to be altered, differences in patients’ characteristics, and so on. These limitations should be noted when interpreting the results of this study.
Our findings indicate that low ANC, high T-Bil, and low CRP are risk factors for severe neutropenia in patients receiving GnP therapy, even if the clinical laboratory test values lie within the normal reference ranges. The incidence of neutropenia was 85.7% in patients with all three risk factors, but only 27.7% in patients with none of them. Thus, careful monitoring of adverse events is needed in patients with these risk factors who are receiving GnP.
The authors thank the staff of the Cancer Institute Hospital, Japanese Foundation for Cancer Research for their cooperation in this study, and also thank Ryoichi Naito and Minori Teramae (Keio University) for assistance in reviewing the collected data.
- 1. Taplitz RA, Kennedy EB, Bow EJ, Crews J, Gleason C, Hawley DK, et al. Antimicrobial prophylaxis for adult patients with cancer-related immunosuppression: ASCO and IDSA clinical practice guideline update. J Clin Oncol. 2018;36(30):3043–3054. pmid:30179565
- 2. Lyman G, Abella E, Pettengell R. Risk factors for febrile neutropenia among patients with cancer receiving chemotherapy: A systematic review. Crit Rev Oncol Hematol. 2014;90(3):190–199. pmid:24434034
- 3. Pettengell R, Schwenkglenks M, Leonard R, Bosly A, Paridaens R, Constenla M, et al. Neutropenia occurrence and predictors of reduced chemotherapy delivery: Results from the INC-EU prospective observational European neutropenia study. Support Care Cancer. 2008;16(11):1299–1309. pmid:18351398
- 4. Lyman GH, Kuderer NM, Crawford J, Wolff DA, Culakova E, Poniewierski MS, et al. Predicting individual risk of neutropenic complications in patients receiving cancer chemotherapy. Cancer. 2011;117(9):1917–1927. pmid:21509769
- 5. Choi YW, Jeong SH, Ahn MS, Lee HW, Kang SY, Choi JH, et al. Patterns of neutropenia and risk factors for febrile neutropenia of diffuse large B-cell lymphoma patients treated with rituximab-CHOP. J Korean Med Sci. 2014;29(11):1493–1500. pmid:25408580
- 6. Watanabe H, Ikesue H, Oshiro M, Nagata K, Mishima K, Takada A, et al. Risk factors for predicting severe neutropenia induced by amrubicin in patients with advanced lung cancer. Chemotherapy. 2013;58(6):419–425. pmid:23295219
- 7. Shiota M, Yokomizo A, Takeuchi A, Kiyoshima K, Inokuchi J, Tatsugami K, et al. Risk factors for febrile neutropenia in patients receiving docetaxel chemotherapy for castration-resistant prostate cancer. Support Care Cancer. 2014;22(12):3219–3226. pmid:24996830
- 8. Mitani Y, Usami E, Kimura M, Nakao T, Okada K, Matsuoka T, et al. Risk factors for neutropenia with lenalidomide plus dexamethasone therapy for multiple myeloma. Pharmazie. 2016;71(6):349–351. pmid:27455556
- 9. Shigeta K, Kosaka T, Yazawa S, Yasumizu Y, Mizuno R, Nagata H, et al. Predictive factors for severe and febrile neutropenia during docetaxel chemotherapy for castration-resistant prostate cancer. Int J Clin Oncol. 2015;20(3):605–612. pmid:25196861
- 10. Pfeil AM, Vulsteke C, Paridaens R, Dieudonné AS, Pettengell R, Hatse S, et al. Multivariable regression analysis of febrile neutropenia occurrence in early breast cancer patients receiving chemotherapy assessing patient-related, chemotherapy-related and genetic risk factors. BMC Cancer. 2014;14(1). pmid:24641830
- 11. Von Hoff DD, Ervin T, Arena FP, Chiorean EG, Infante J, Moore M, et al. Increased survival in pancreatic cancer with nab-paclitaxel plus gemcitabine. N Engl J Med. 2013;369(18):1691–1703. pmid:24131140
- 12. Sasaki T, Kanata R, Yamada I, Matsuyama M, Ozaka M, Sasahira N. Improvement of treatment outcomes for metastatic pancreatic cancer: A real-world data analysis. In Vivo (Brooklyn). 2019;33(1):271–276. pmid:30587635
- 13. Ueno H, Ikeda M, Ueno M, Mizuno N, Ioka T, Omuro Y, et al. Phase I/II study of nab-paclitaxel plus gemcitabine for chemotherapy-naive Japanese patients with metastatic pancreatic cancer. Cancer Chemother Pharmacol. 2016;77(3):595–603. pmid:26842789
- 14. Makita T, Rikimaru R, Kanzaki H, Horiguchi S, Kato H, Tsutsumi K, et al. Distinct risk factor profiles of gemcitabine plus Nab-paclitaxel-induced neutropenia in patients with pancreatic cancer. Iryo Yakugaku (Japanese J Pharm Heal Care Sci. 2018;44(3):107–116.
- 15. Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49(12):1373–1379. pmid:8970487
- 16. Braiteh F, Patel MB, Parisi M, Ni Q, Park S, Faria C. Comparative effectiveness and resource utilization of nab-paclitaxel plus gemcitabine vs FOLFIRINOX or gemcitabine for the first-line treatment of metastatic pancreatic adenocarcinoma in a US community setting. Cancer Manag Res. 2017;9:141–148. pmid:28461766
- 17. Yano R, Konno A, Watanabe K, Tsukamoto H, Kayano Y, Ohnaka H, et al. Pharmacoethnicity of docetaxel-induced severe neutropenia: Integrated analysis of published phase II and III trials. Int J Clin Oncol. 2013;18(1):96–104. pmid:22095245
- 18. Japan Clinical Oncology group (JCOG). Common Terminology Criteria for Adverse Events (CTCAE) v5.0 edited by Japanese Committee for Clinical Laboratory Standards (JCCLS) (http://www.jcog.jp/doctor/tool/JCOG_kyouyoukijunchi_50.pdf) December 21,2020.
- 19. Nomura H, Tsuji D, Demachi K, Mochizuki N, Matsuzawa H, Yano T, et al. ABCB1 and ABCC2 genetic polymorphism as risk factors for neutropenia in esophageal cancer patients treated with docetaxel, cisplatin, and 5-fluorouracil chemotherapy. Cancer Chemother Pharmacol. 2020;86(2):315–324. pmid:32748110
- 20. Joerger M, Huitema AD, Huizing MT, Willemse PH, de Graeff A, Rosing H, et al. Safety and pharmacology of paclitaxel in patients with impaired liver function: A population pharmacokinetic-pharmacodynamic study. Br J Clin Pharmacol. 2007;64(5):622–633. pmid:17935602
- 21. Razzaghdoust A, Mofid B, Moghadam M. Development of a simplified multivariable model to predict neutropenic complications in cancer patients undergoing chemotherapy. Support Care Cancer. 2018;26(11):3691–3699. pmid:29736867