To the Editor
Patients with acute leukemia require long-term central venous catheter (CVC), which provides rapid and reliable access for the administration of cytotoxic agents (mostly, cytarabine, fludarabine, methotrexate and/or anthracycline-based regimens, all strongly phlebitogenic due to their acidic [< 5] or alkaline [≥ 9] pH), antimicrobial therapy (especially antibiotics and antifungal drugs), intravenous hydration, blood product transfusions, total parenteral nutrition, and/or venous sampling [1–4]. The choice of insertion site—whether the internal jugular vein (IJV) or subclavian vein (SCV), or the basilic vein and brachial vein (BBV) of the right or left upper arm, may contribute to the emergence of catheter related (CR)-complications [5–7]. This represents an important practical concern for treating physicians, who generally select the site based on personal preference or institutional practice. In patients with acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL), data on hemorrhagic adverse events associated with the upfront use of peripherally inserted central catheters (PICCs) for long-term access remain limited, particularly in severely thrombocytopenic patients [8].
We read with great interest the report by Zhang K et al. published in Annals of Hematology in 2025 9, which examined the safety of PICC insertion in hematological-oncology patients with a platelet count ≤ 50 × 10⁹/L and/or coagulation dysfunction aiming to identify risk factors for bleeding within 24 h after PICC placement. The retrospective study by Zhang et al. included a total of 1511 hematological patients (median age at diagnosis: 37 years, range 17–85 years) with AML (n = 909)/ALL (n = 308)/other hematological malignancies (remaining cases) all presenting with a platelet count ≤ 50 × 109/L and undergoing PICC insertion between February 2017 and January 2020. Thus, all patients had a PICC inserted: 1443 were polyurethane-PICCs (95,5%), and 68 were silicone-PICCs (4,5%). A platelet count < 25 × 109/L was observed in 52% (n = 786) of patients, whereas 48% (n = 725) had a platelet count between 25 × 10⁹/L and 50 × 10⁹/L. PICC-related bleeding events (defined as either minor bleeding, i.e., petechiae, purpura, and/or exudation, or clinically significant bleeding, i.e., deep-tissue bleeding, limb swelling, severe hemodynamic instability, or fatal bleeding) were recorded. Among the 1217 patients with acute leukemia (AML + ALL), hemorrhagic adverse events occurred in 49 patients corresponding to a bleeding rate of 4% within 24 h of PICC implantation. All hemorrhagic side effects were classified as minor bleeding. Patients in the bleeding-group had a significantly lower platelet count before PICC insertion compared with those in the non-bleeding group (14 × 109/L [range, 1–50 × 109/L] versus 26 × 109/L [range, 1–50 × 109/L] respectively, P < 0.01). In addition, pre-PICC platelet transfusions were significantly more frequent in the bleeding group than in the non-bleeding group (20 patients [36%] versus 324 patients [22%] respectively, P = 0.016). Multivariate analysis identified three independent risk factors for bleeding events within 24 h after PICC insertion: prothrombin time ≥ 13 s, platelet count ≤ 25 × 109/L, and ADL (Activities of Daily Living) score ≤ 95 points. According to the authors, clinical decisions regarding CVC implantation should therefore be made cautiously, taking into account the patient’s clinical status, symptom severity, and treatment needs, particularly in those at higher risk. Studies with longer follow-up are warranted to comprehensively evaluate both early and late-onset major CR-complications including mechanical events (such as bleedings), symptomatic deep-vein thrombosis (s-DVT) and bloodstream infections (BSI) to better support the decision-making process, particularly in patients with acute leukemia.
In our tertiary hospital in southern Italy (the Hematology Unit of the Federico II University Medical School of Naples), PICC implantation has been adopted as a routine minimally invasive procedure for patients with hematological diseases [5–7]. Recently, we performed a real-life analysis based on the registry database of the Unit focusing on the impact of different CVC sites, i.e., BBV versus SCV and IJV, on major CR-complications in acute leukemia patients 10. The primary endpoint of the study was the cumulative incidence of grade 3 or above of hemorrhagic and vascular access complications (VAC), i.e.hematoma and/or arterial injury, s-DVT and BSI according to the CTCAE (National Cancer Institute Common Terminology Criteria for Adverse Events, Version 4) from catheter placement through day 30. We reviewed the clinical charts of patients with acute leukemia who underwent CVC placement intended to remain in-situ for at least 30 days after catheterization, as part of intent-to-cure anticancer therapy, i.e., hematological remission induction chemotherapy. The observation period spanned from 1 January 2012 to 31 December 2021. A total of 336 PICCs were inserted at hematologic diagnosis into 336 patients through the BBV in 115 cases (34%; BBV-cohort), IJV in 111 cases (33%; IJV-cohort), and SCV in 110 cases (32,7%; SCV-cohort). Overall, there were 170 males and 166 females with a median age of 53 years (range, 18–72 years). There were 208 patients with AML and 128 patients with ALL; at the baseline, median platelets count was 41 × 109/L with a range of 3,0–275 × 109/L. Age, sex, prothrombotic risk factors, underlying hematological malignant disease, eastern cooperative oncology group (ECOG) performance status and blood cell counts were well balanced according to catheter-site groupings (Table 1). In the BBV-cohort, the most frequently implanted device was a 5-French (Fr) double lumen catheter (open-ended, non-valved power-injectable polyurethane PICC with a flexible tip; ProPic MEDCOMP, Harleysville, USA) inserted through the basilic vein in 89% of cases and through the brachial vein in the remaining 11% of cases. The median number of venipuncture attempts was 1 (range, 1–3). Tip placement was located at the cavo-atrial junction in 93 (81%) cases and at the lower third of the superior vena cava in 22 (19%) cases. In the centrally inserted central catheters (CICC) setting (IJV + SCV: 221/336, 66%), the most commonly positioned device—both at the IJV and SCV sites—was a 7 Fr triple-lumen catheter (Arrow®, Teleflex, 3015 Carrington Mill Boulevard Morrisville, NC, USA). All patients received parenteral infusions of anticancer therapy through the CVC (chemotherapy regimens based on cytarabine [55%], methotrexate-asparaginase [39%], and/or fludarabine [6%]) and/or supportive care; all the infusions were evenly distributed among the three cohorts (Table 2). Regarding chemotherapy-induced hematological toxicity, severe thrombocytopenia (platelet count < 10 × 109/L) occurred in 336 (100%) patients and severe neutropenia (neuthrophil count < 5 × 109/L) in 336 (100%) patients. The median duration of severe thrombocytopenia was 10 days in each cohort with ranges of 7–15 days in the BBV-cohort, 7–14 days in the SCV-cohort, and 8–16 days in the IJV-cohort. The median duration of severe neutropenia was 20 days in each cohort with ranges of 8–22 days in the BBV-cohort, 8–24 days in the SCV-cohort, and 8–23 days in the IJV-cohort. The median duration of in-situ CVC placement was 30 days (Table 2). The frequency of bleeding complications (all ≥ grade 3 according to CTCAE) associated with catheter placement differed among the three insertion-site (P = 0.015; Fig. 1 A), with 3 events (hematoma) in the BBV-cohort, 10 events (hematoma, 7; VAC of arterial puncture, 3) in the SCV-cohort, and 12 events (hematoma, 9; VAC of arterial puncture, 3) in the IJV-cohort. CR-sDVT occurred in 6 patients in the BBV-cohort, 12 patients in the SCV-cohort and 15 patients in the IJV-cohort (5,2% versus 11% versus 13,5%, respectively). The incidence rate of sDVT was 1,9, 2,7 and 5,1 per 1000 catheter-days in the BBV-cohort, SCV-cohort, and IJV-cohort, respectively (P = 0.052; Fig. 1 B). When thrombotic events occurred, the median interval between catheter insertion and thrombotic episodes was 16 days (range, 8–30 days) for the IJV-cohort and 20 days (range, 15–29 days) for the SCV-cohort versus 9,5 days (range, 7–10 days) for the BBV-cohort. CR-BSI occurred in 4 patients in the BBV-cohort, 8 patients in the SCV-cohort and 12 patients in the IJV-cohort (3,5% versus 7,2% versus 10,8%;). The incidence rate of BSI was 1,3, 4 and 4 per 1000 catheter-days in the BBV-cohort, SCV-cohort and IJV-cohort, respectively (P = 0.074; Fig. 1 C). When BSI occurred, the median time from catheter implantation to BSI event, was 7 days (range, 6–8 days) for the BBV cohort, compared to 16 days (range, 8–24 days) in the SCV cohort and 15 days (range, 7–22 days) in the IJV cohort. The pathogens most frequently isolated were coagulase-negative staphylococci. In our experience, a detailed protocol was followed during PICC implantation 1. First, bilateral ultrasonography (US) of all the arm and neck veins was performed before insertion. The selection of the appropriate vein in the upper-mid arm (the “green zone” of the zone insertion method) was based on the vein’s diameter in millimeters measured on US, which had to be at least equal to the catheter in Fr. Second, a clear identification on the US of the median nerve and brachial artery was required before venipuncture was performed with US guidance. Third, hand washing, aseptic technique, and maximal barrier protection were strictly observed. Fourth, US of the IJV was also performed during catheter advancement. Finally, we used an intracavitary electrocardiographic method to assess the tip position and secured the PICC with a sutureless device. In addition, a standardized compression protocol was applied in all patients, consisting of sterile gauze covering the puncture site with moderate pressure using an elastic bandage, manual compression of the puncture site for a full 20 min, and cyanoacrylate glue application at the puncture site when needed. A dedicated central venous access team managed all PICCs following standardized procedures from insertion to weekly follow-up.
Table 1.
Baseline patient characteristics according to catheter insertion site (modified from Picardi et al., Ref#10)
| Variable | Total (%) | BBV-cohort (%) | IJV-cohort (%) | SCV-cohort (%) | P value |
|---|---|---|---|---|---|
| No. of patients | 336 | 115 | 111 | 110 | |
|
Median age, years Range |
53 18–72 |
53 18–72 |
54 18–72 |
54.5 22–70 |
0.71 |
| Male | 170 (51) | 58 (50) | 56 (51) | 56 (51) | 0.67 |
| Prothrombotic risk factors | |||||
| BMI > 25 | 36 (11) | 13 (11) | 12 (11) | 11 (10) | 0.96 |
| Smoke | 60 (18) | 22 (19) | 20 (18) | 18 (16) | 0.85 |
| Hypertension | 69 (19) | 23 (20) | 24 (21) | 22 (20) | 0.97 |
| Diabetes | 26 (7) | 9 (8) | 10 (9) | 7 (6) | 0.71 |
| Hematological disease | |||||
| AML | 208 (62) | 71 (62) | 69 (62) | 68 (61) | |
| ALL | 128 (38) | 44 (38) | 42 (38) | 42 (38) | 0.98 |
| ECOG PS status | |||||
| 0–1 | 286 (85) | 97 (84) | 95 (85) | 94 (85) | |
| 2–3 | 50 (15) | 18 (16) | 16 (15) | 16 (15) | 0.97 |
| Blood cell count | |||||
| Platelets, x 103/mm3 | |||||
| Median, range | 41, 3.0–275 | 36, 3.0–275 | 42.5, 9.0–232 | 40, 5.0–200 | 0.92 |
| WBC, x 103/mm3 | |||||
| Median, range | 3.4, 0.2–147 | 3.2, 0.97–96 | 3.9, 1.0–147 | 3.5, 0.2–100 | 0.75 |
| Neutrophils, x 103/mm3 | |||||
| Median, range | 0.5, 0.4–12.34 | 0.45, 0.4–7.5 | 0.56, 0.8–12.34 | 0.48, 0.75–5.5 | 0.69 |
| Hemoglobin, g/dl | |||||
| Median, range | 9.5, 5.9–12.6 | 9.7, 5.9–12.1 | 9.5, 6.6–12.6 | 9.3, 6–12 | 0.81 |
Unless otherwise indicated, data refer to the number of patients
BBV: basilic and brachial veins; IJV: internal jugular vein; SCV: subclavian vein; BMI: body max index; AML: acute myeloid leukemia; ALL: acute lymphoblastic leukemia; ECOG PS: eastern cooperative oncology group performance status; WBC: white blood cell
Table 2.
Characteristics of implanted catheters and their use (modified from Picardi et al. Ref# 10)
| Variable | Total (%) N = 336 |
BBV-cohort (%) N = 115 |
IJV-cohort (%) N = 111 |
SCV-cohort (%) N=110 |
|
|---|---|---|---|---|---|
| Devices insertion place | |||||
| Hematology ward bedside | 120 (35) | 92 (80) | 17 (15) | 11 (10) | |
| ICU | 216 (65) | 23 (20) | 94 (95) | 99 (90) | |
| Device type | |||||
| Single lumen | 20 (6) | 20 (17) | NA | NA | |
| Double lumen | 111 (33) | 90 (78) | 10 (9) | 11 (10) | |
| Triple lumen | 205 (61) | 5 (5) | 101 (91) | 99 (90) | |
| 4 French | 40 (12) | 40 (35) | NA | NA | |
| 5 French | 70 (21) | 70 (61) | NA | NA | |
| 6 French | 5 (1) | 5 (4) | NA | NA | |
| 7 French | 190 (57) | NA | 96 (86) | 94 (85) | |
| 8 French | 31 (9) | NA | 15 (14) | 16 (15) | |
| Venous access | |||||
| Basilic | 73 (22) | 73 (63) | NA | NA | |
| Brachial | 42 (12) | 42 (37) | NA | NA | |
| Internal jugular | 111 (33) | NA | 111 (100) | NA | |
| Subclavian | 110 (33) | NA | NA | 110 (100) | |
| Right side | 227 (67) | 88 (76) | 74 (67) | 65 (59) | |
| Left side | 109 (33) | 27 (24) | 37 (33) | 45 (41) | |
| Attempts at venipuncture, n | |||||
| Median, range | 1, 1–4 | 1, 1–4 | 1, 1–3 | 1, 1–3 | |
| Tip location | |||||
| Cavo-atrial junction | 252 (75) | 93 (81) | 80 (72) | 79 (72) | |
| Lower third of superior vena cava | 84 (25) | 22 (19) | 31 (28) | 31 (28) | |
| Interval from CVC implantation to chemotherapy start, day | |||||
| Median | 1 | 1 | 1 | 1 | |
| Range | 0.5–1.5 | 0.5–1 | 0.5–1.5 | 0.5–1 | |
| Chemotherapy regimen | |||||
| Cytarabine-based | 188 (56) | 63 (55) | 62 (56) | 63 (57) | |
| MTX-asparaginase-based | 128 (38) | 45 (39) | 42 (38) | 41 (37) | |
| Fludarabine-based | 20 (6) | 7 (6) | 7 (6) | 6 (5.5) | |
| Chemotherapy-induced hematological toxicity | |||||
| Severe thrombocytopenia | 336 (100) | 115 (100) | 111 (100) | 110 (100) | |
| Severe neutropenia | 336 (100) | 115 (100) | 111 (100) | 110 (100) | |
| Duration of catheterization—days | |||||
| Median | 30 | 30 | 30 | 30 | |
| Range | 1–30 | 6–30 | 1–30 | 2–30 | |
Unless otherwise indicated, data refer to the number of patients. BBV: basilic and brachial veins; IJV: internal jugular vein; SCV: subclavian vein; ICU: intensive care unit
Cytarabine-based regimen included 3 + 7 protocol (cytarabine 200 mg/m2 in continuous i.v. infusion for 7 days and daunorubicin 60 mg/m2 i.v. on days 1, 3 and 5) (ref# 10)
Fludarabine-based regimen included FLAG-Ida protocol (cytarabine 2000 mg/m2 daily as a 3-h i.v. infusion for 4 days plus fludarabine 30 mg/m2 daily as a 30-min i.v. infusion for 4 days plus idarubicin 12 mg/m2 daily as a 1-h i.v. infusion on days 2–4) (ref# 10)
MTX-asparaginase-based regimen included the followings: GIMEMA LAL-1913 (idarubicin, vincristine, dexamethasone, asparaginase, cyclophosphamide, cytarabine, 6-mercaptoturine, and methotrexate; for the specific schedules see ref# 23), and NILG ALL 10/07 (idarubicin, vincristine, dexamethasone, L-asparaginase, cyclophosphamide, cytarabine, 6-mercaptoturine, and methotrexate; for the specific schedules see ref# 10)
Fig. 1.
A. Cumulative numbers of peripherally inserted central catheter (PICC)-related versus centrally inserted central catheters (CICC)-related bleeding events in patients with acute leukemia and long-term central venous catheter in-situ (for at least 30 days after catheterization). P-value was calculated using Pearson’s χ2 test comparing BBV-cohort versus the combined cohorts SCV + IJV. BBV, basilic/brachial vein; SCV, subclavian vein; IJV, internal jugular vein. B. Cumulative numbers of peripherally inserted central catheter (PICC)-related versus centrally inserted central catheters (CICC)-related symptomatic deep venous thrombosis in patients with acute leukemia and long-term central venous catheter in-situ (for at least 30 days after catheterization). P-value was calculated using Pearson’s χ2 test comparing BBV-cohort versus the combined cohorts SCV + IJV. BBV, basilic/brachial vein; SCV, subclavian vein; IJV, internal jugular vein. C. Cumulative numbers of peripherally inserted central catheter (PICC)-related versus centrally inserted central catheters (CICC)-related bloodstream infection in patients with acute leukemia and long-term central venous catheter in-situ (for at least 30 days after catheterization). P-value was calculated using Pearson’s χ2 test comparing BBV-cohort versus the combined cohorts SCV + IJV. BBV, basilic/brachial vein; SCV, subclavian vein; IJV, internal jugular vein
In conclusion, PICCs represent a feasible and minimally invasive option for central venous access in patients with acute leukemia undergoing chemotherapy induction, with a favorable safety profile in terms of CR-bleeding, sDVT, and BSI in our real-world experience 9, 10. However, these findings should be interpreted with caution due to the retrospective, single-center design of the study and, importantly, the presence of relevant confounding factors, including differences in catheter characteristics (caliber and number of lumens) and clinical conditions at the time of insertion. These imbalances may have influenced the observed complication rates and limit the ability to draw definitive comparative conclusions across insertion sites. Therefore, prospective, multicenter studies with appropriate adjustment for these variables are warranted to better define the optimal vascular access strategy in this high-risk population.
Author Contribution
M. Picardi designed the research; M.Picardi and C. Giordano, performed the research and wrote the paper; A. Vincenzi and N. Pugliese analysed data and prepared the figures; R. Della Pepa, E. Cimmino, A. Scarpa and M. Persico collected the data; F. Pane and M. Picardi performed the final revision of the manuscript.
Funding
No funds were received for this research.
Data availability
No datasets were generated or analysed during the current study.
Declarations
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Scoppettuolo G, Pittiruti M (2014) Ultrasound-guided placement of peripherally inserted central venous catheters. In: Lumb P, Karakitsos D (eds) Critical Care Ultrasound. Elsevier-Saunders, Philadelphia, pp 89–94 [Google Scholar]
- 2.Caira M, Morselli M, Candoni A et al (2008) Invasive fungal infections in patients with acute myeloid leukemia and in those submitted to allogeneic hemopoietic stem cell transplant: who is at highest risk? Eur J Haematol 81(3):242–243. 10.1111/j.1600-0609.2008.01096.x [DOI] [PubMed] [Google Scholar]
- 3.Candoni A, Simeone E, Facchinelli D et al (2014) Multicentre surveillance study on antifungal combination therapy for invasive fungal diseases in hematologic patients: the SEIFEM real-life combo study. Mycoses 57(6):342–350. 10.1111/myc.12161 [DOI] [PubMed] [Google Scholar]
- 4.Bucaneve G, Micozzi A, Picardi M et al (2014) Results of a multicenter, controlled, randomized clinical trial evaluating the combination of piperacillin/tazobactam and tigecycline in high-risk hematologic patients with cancer with febrile neutropenia. J Clin Oncol 32(14):1463–1471. 10.1200/JCO.2013.51.6963 [DOI] [PubMed] [Google Scholar]
- 5.Picardi M, Della Pepa R, Cerchione C et al (2019) A frontline approach with peripherally inserted versus centrally inserted central venous catheters for induction chemotherapy of acute myeloid leukemia: a randomized comparison. Clin Lymphoma Myeloma Leuk 19(4):e184–e194 [DOI] [PubMed] [Google Scholar]
- 6.Giordano C, Picardi M, Vincenzi A et al (2025) Intravascular complications of central venous catheterization in chronic hematological diseases: low risk with peripherally inserted catheters in a large retrospective study. Clin Lymphoma Myeloma Leuk S2152–2650(25):04222–04223. 10.1016/j.clml.2025.09.011 [DOI] [PubMed] [Google Scholar]
- 7.Picardi M, Giordano C, Pane F et al (2025) Thrombotic complications of central venous catheterization with peripherally inserted catheters in classical Hodgkin lymphoma. Hemasphere 9(11):e70259. 10.1002/hem3.70259 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Morano SG, Coppola L, Latagliata R et al (2014) Early and late complications related to central venous catheters in hematologic malignancies: a retrospective study of 1102 patients. Mediterr J Hematol Infect Dis 6:e2014014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhang K, Liu Y, Wang H et al (2025) Safety and risk factors of PICC insertion in hematologic malignancy patients with thrombocytopenia: a retrospective study. Ann Hematol 104:6005–6011. 10.1007/s00277-025-06717-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Picardi M, Giordano C, Della Pepa R et al (2023) Intravascular complications of central venous catheterization by insertion site in acute leukemia during induction chemotherapy: lower risk with peripherally inserted catheters. Cancers (Basel) 15(7):2147. 10.3390/cancers15072147 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Picardi M, Pagliuca S, Chiurazzi F et al (2012) Early ultrasonographic detection of septic thrombophlebitis as an indicator for central venous catheter removal in neutropenic patients with bloodstream infection. Ann Oncol 23(8):2122–2128. 10.1093/annonc/mdr588 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
No datasets were generated or analysed during the current study.