John S. McGrath, Department of Urology, Surgical Health Services Research Unit, Royal Devon and Exeter NHS Trust Exeter, Exeter EX8 5AB, UK. e-mail: email@example.com
Study Type – Therapy (case series)
Level of Evidence 4
What's known on the subject? and What does the study add?
Guidance from the UK National Institute for Health and Clinical Excellence (NICE) on the use of intraoperative cell savage (ICS) has been in place for over 3 years and recommends its routine usage in all patients undergoing radical pelvic urological surgery. The current series describes the contribution of ICS to contemporary blood conservation strategies and the goal of ‘bloodless’ cystectomy.
• To describe a 10-year experience of intra-operative cell salvage (ICS) during radical cystectomy at a regional cancer centre.
PATIENTS AND METHODS
• Between 1st January 2001 and 31st December 2010, 213 consecutive patients underwent radical cystectomy and pelvic lymphadenectomy for bladder cancer, with an ICS suction device used in theatre.
• Surgery was performed by one of three consultant surgeons using an open technique with lymph node clearance to the iliac bifurcation. Orthotopic bladder substitution was performed in 25% of patients overall.
• ICS data were collected prospectively on an electronic database and the institutional database was then cross-referenced with a complete review of patients' medical records, laboratory results and radiological investigations retrospectively.
• Data collected included patient demographics, haemoglobin levels before and after surgery, the volume of ICS blood collected and re-infused, complications related to ICS usage, the volume of allogeneic red blood cells (RBCs) transfused, length of stay and overall patient survival at 3 and 5 years after surgery.
• In all 213 cases described, ICS was used without complication, with no recorded episodes of device failure and no complications related to the use of cell salvage.
• Overall, 91% of patients received ICS blood and 28% of patients avoided any further transfusion products.
• The median (range) follow-up for the cohort was 24 (9–119) months.
• Seventy percent of the transfusion requirement for patients who underwent surgery in 2001 was met using allogeneic RBC transfusion but by 2010, as blood loss markedly reduced, ICS blood was able to provide ∼70% of overall transfusion requirements. As a consequence, the percentage of patients avoiding an allogeneic RBC transfusion significantly increased during the 10-year period, such that 70% of patients avoided allogeneic RBC transfusion in 2010 compared with only 10–20% in the period 2001–2003
• The overall survival rate at 3 and 5 years was 58% and 49%, respectively.
• In conclusion, the use of ICS during radical cystectomy is safe; it is capable of meeting the majority of or, in some cases, the total blood product requirement for individual patients. As a result, it decreases the need for allogeneic RBC transfusion and hence the associated risks. Current follow-up shows no apparent risk of decreased long-term survival from an oncological perspective.
• The authors advocate routine availability of ICS for all major urological oncology cases.
National Institute for Health and Clinical Excellence
Open radical cystectomy and extended pelvic lymphadenectomy is the ‘gold standard’ treatment for non-metastatic carcinoma invading bladder muscle and high-risk non-muscle-invasive bladder cancer [1,2]. Despite improvements in surgical technique, radical cystectomy is associated with the potential for significant blood loss. The median blood loss during radical cystectomy in centres submitting data to the BAUS Cancer Registry cystectomy dataset for years 2008 (43 centres contributed data) and 2009 (57 centres contributed data) was 1000 mL and 500 mL, respectively [3,4]. Patients undergoing cystectomy are commonly elderly, with multiple pre-existing co-morbidities and they may be anaemic before surgery . With the advent of neo-adjuvant chemotherapy, a significant proportion of patients may also be immuno-compromised and have lower preoperative haemoglobin levels. Traditionally, donated allogeneic red blood cell (RBC) transfusion has been required to replace losses; however, donor blood transfusion can be associated with transfusion reactions and has the potential to transmit blood-borne infections. An increasing concern is the fact that it may result in immunosuppression with poorly understood, but possibly adverse, effects on urological cancer recurrence [6–8]. Furthermore, cost and availability are issues with donor blood transfusion as well as the emerging concern regarding the ‘unknown unknowns’ such as prions.
Intra-operative cell salvage (ICS) offers an efficacious, alternative technique for blood replacement. Guidance from the UK National Institute for Health and Clinical Excellence (NICE) on the use of ICS has been in place for over 3 years and recommends its routine usage in all patients undergoing radical pelvic urological surgery . The technique enables intra-operative collection and re-infusion of the patient's own washed red blood cells at the time of surgery. Barriers to adoption appear to be related to concerns regarding the theoretical, but to date unsubstantiated, risk of tumour dissemination resulting from reinfusion of malignant cells. Additionally, clinical dogma has conventionally associated the requirement for ICS with procedures of catastrophic blood loss or when a routine procedure has unexpectedly resulted in major bleeding, rather than as a normal adjunct in major open pelvic surgery. As an integral part of our blood conservation strategy, our unit has made it standard practice since 1996 to use ICS with auto-transfusion in all patients undergoing radical cystectomy. The present paper describes our long-term experience with ICS during radical cystectomy and considers the contribution of ICS to contemporary blood conservation strategies and the goal of ‘bloodless’ cystectomy.
Between 1st January 2001 and 31st December 2010, 213 consecutive patients underwent radical cystectomy and pelvic lymphadenectomy for bladder cancer, with an ICS suction device used in theatre. Patient demographics and details of pathological staging are summarized in Table 1. Patients undergoing pelvic exenteration for non-bladder primary tumours were excluded. Surgery was performed by one of three consultant surgeons (J.M., R.M. and R.P.) using an open technique with lymph node clearance to the iliac bifurcation. Orthotopic bladder substitution was performed in 25% of patients overall. Neo-adjuvant chemotherapy was introduced as a standard of care in 2006; exclusions included renal impairment, superficial disease, intractable symptoms and non-transitional primary tumour. The percentage of patients per year receiving chemotherapy varied from 70 to 90%, dependent on the rate of exclusions during the analysis period. ICS was discussed with all patients during the consent process before surgery. This included written information at the preparation-for-surgery visit and then signed consent before the surgical procedure. Three cell salvage machines were used over the 10-year study period: the BRAT (COBE later SORIN Group, Gloucester, UK), the ELECTA (SORIN Group, Gloucester, UK) and the Cell Saver 5+ (Haemonetics Ltd, Glasgow, UK). All devices were run in ‘automatic’ mode. Briefly, to obtain cell salvaged blood: an ICS suction device collects the blood in the operative field, if a decision is made to re-infuse the blood the collected blood is washed with normal saline, rinsed and spun through a centrifuge thus isolating red blood cells ready for re-infusion. The processing time is ∼2 min. Leukocyte depletion filters were not used during re-infusion.
Table 1. Patient characteristics
Total number of patients
Median age (Range), years
Male : female patients (%)
166 (80) : 47 (20)
Type of urinary diversion
Ileal conduit, n (%)
Neobladder, n (%)
Median length of stay (range), days
30-day mortality, n (%)
Final pathological staging, n (%)
T stage unknown
Node-positive, n (%)
The ICS data were collected prospectively on an electronic database (Microsoft Excel). The institutional database was then cross-referenced with a complete review of patient's medical records, laboratory results and radiological investigations retrospectively. Data collected included patient demographics, pre- and postoperative haemoglobin levels, the volume of ICS blood collected and re-infused, complications related to ICS usage, the volume of allogeneic RBC transfused, length of stay and overall patient survival at 3 and 5 years after surgery. (Total estimated blood loss is not routinely recorded on the database by intention, because it is generally an unreliable measure of actual blood loss. Factors such as irrigation, urine, washout and loss in swabs cause significant variation in estimated losses.)
In all 213 cases described ICS was successfully used, with no recorded episodes of device failure and no complications related to the use of cell salvage. Overall, 91% of patients received ICS blood and 28% of patients avoided any further transfusion products. The median (range) follow-up for the cohort was 24 (9–119) months.
Figure 1 shows the trend in total blood administration during the series, with a mean replacement of 2250 mL per patient in the first 5 years of the series, decreasing to ∼600 mL per patient in the final 2 years. It can also be seen that 70% of the transfusion requirement for patients who underwent surgery in 2001 was met using allogeneic RBC transfusion but, by 2010 as blood loss markedly reduced, ICS blood was able to provide ∼70% of overall transfusion requirements.
As a consequence, the percentage of patients avoiding an allogeneic RBC transfusion significantly increased during the 10-year period such that 70% of patients avoided allogeneic RBC transfusion in 2010 compared with only 10–20% in 2001–2003 (Fig. 2).
The mean preoperative, postoperative and discharge haemoglobin levels, as well as the respective haematocrits, are summarized in Table 2. The median length of stay decreased from 18.4 days in 2001 to 10.7 days in 2010. Two patients suffered cardiac events in the peri-operative period (one fatal and one non-fatal myocardial infarction). The overall survival rate at 3 and 5 years was 58% and 49%, respectively.
Table 2. Haemoglobin levels
1 day after surgery
Median haemoglobin (range) g/dL
Median haematocrit (range)
The present series, analysing patients undergoing radical cystectomy, shows an increasing volume of surgical cases treated at our institution over the 10-year period described. During this time, the total blood loss and total transfusion requirements have steadily declined, in keeping with national statistics which report that the total number of transfused blood products in the UK has decreased consistently over the last 10 years . The reasons are likely to be multifactorial. Centralization of pelvic surgery occurred mid-series with an overall increase in the number of operated cases, as well as a major reconfiguration of the operating personnel. The use of technology also changed, with adoption of operating headlights, haemostatic energy devices and goal-directed fluid therapy as just a few examples. More latterly, enhanced recovery protocols were introduced, beginning ‘piece-meal’ in 2007 and then fully implemented in 2010.
It can be seen that a majority of, but not all, patients undergoing radical cystectomy now avoid allogeneic blood transfusion and the associated risks. This is attributable to the progressive reduction in total blood loss coupled with the corresponding ability of ICS to meet the majority of the transfusion requirements.
Considerable progress towards ‘bloodless’ cystectomy has therefore been made in the 10-year period but two key challenges remain. The first is to understand the barriers to widespread adoption of the technique in all centres offering radical pelvic surgery, given the obvious benefits in reducing the reliance on allogeneic RBC transfusion and the associated risks and cost. The second is to identify the techniques that will allow us to ‘close the gap’ for the remaining 30% of patients in order to eliminate the need for allogeneic RBC and make the concept of ‘bloodless’ cystectomy a reality.
BARRIERS TO ADOPTION
In April 2008, NICE issued guidelines supporting the use of ICS during radical prostatectomy or radical cystectomy . Although there is considerable evidence to support the use of ICS in vascular, cardiac and orthopaedic surgery , the evidence to support its use during urological oncological surgery and, specifically cystectomy, is limited. Studies evaluating the technique have focused on the use of cell salvage in radical prostatectomy probably because there is an established test (serum PSA) to guide follow-up and indicate the presence of tumour recurrence. By contrast, there is a lack of evidence-based guidelines governing the appropriate follow-up of patients who undergo cystectomy, thereby making comparison of recurrence rates difficult . The authors are aware of three small, uncontrolled, retrospective series describing the outcomes of patients undergoing radical cystectomy who received ICS blood [13–15]. The largest and most contemporary series included 65 patients with a median 2-year follow-up period, where there was no reported adverse effect on overall patient survival .
In this context, the present series of 213 patients under going radical cystectomy with cell salvage available (of which 192 patients received ICS blood) is the largest contemporary UK-based evaluation of the technique to date. There were no complications related to ICS recorded, which is consistent with previously published series [13–15]. Potential complications associated with ICS include non-immune haemolysis, air embolus, febrile non-haemaolytic transfusion reactions, mis-transfusion, coagulopathy and contamination. The present series did not investigate the complications related to allogeneic transfusion within the cystectomy cohort. However, the Cleveland clinic performed a 5-year retrospective review of adverse reactions related to ICS vs allogeneic transfusion and found the adverse event rate to be low in both groups but with a fivefold difference favouring ICS (0.027% vs 0.14%, respectively) . This supports the conclusion that ICS re-infusion is at least as safe and potentially safer than allogeneic blood transfusion. A major safety advantage for ICS is the fact that the salvaged blood does not leave the patient's side, thus eliminating the potential for mis-transfusion.
Oncological safety has been cited as a potential barrier to adoption. The evidence to date has not substantiated the theoretical concern of potential tumour dissemination by the use of ICS in oncological surgery. Circulating tumour cells are detectable in patients with advanced, non-metastatic bladder cancer but their role in tumour recurrence and metastasis post-cystectomy is currently poorly understood . As far as we are aware, there are no published data regarding the presence and related risk of tumour cells in ICS blood at radical cystectomy. Stoffel et al. showed that PSA-expressing cells were detectable in the serum preoperatively in 13% of 112 patients undergoing radical prostatectomy. Peri-operatively, 88% of the patients had these cells present in the cell salvage reservoir. However, at 3–5weeks after surgery, the cells were no longer detectable and ICS was not associated with a greater long-term risk of biochemical failure in their cohort. By contrast, allogeneic blood transfusion has been associated with immunomodulation and increased tumour recurrence in other malignancies [19,20]. As mentioned above, ICS has been extensively investigated in radical prostate cancer surgery; in the largest study, which included over 1000 patients, cell salvage did not increase recurrence rates in the 25.5% of patients who received ICS (median follow-up of 40.2 months) . Similar results were found in other retrospective studies of urological cancers [13–15,18,22–24]. The present findings add to the body of evidence that long-term follow-up of patients undergoing cystectomy with cell salvage (without a leukocyte depletion filter) showed no apparent risk of reduced overall survival when compared with previously published series . The rationale for the use of leukocyte depletion filters in obstetric practice is to reduce amniotic fluid embolus risk by reducing the amount of amniotic fluid in the ICS blood . Although the use of a filter may seem intuitive in oncological surgery, there is currently no evidence of a conferred benefit in terms of patient outcome compared with not using a filter. In a gynaecological oncological study of 50 patients undergoing radical surgery, viable tumour cells were identified in the cell salvaged blood. After passing the blood through leukocyte depletion filters no tumour cells were seen; however, tumour cell fragments were present which were unable to cause metastases . Irradiation of ICS blood is the only other theoretical method of eliminating viable circulating tumour cells. Again, the need for this is unproven, the technique is expensive and it necessitates the blood leaving the patient's side, potentially increasing the risk of mis-match though incorrect identification. NICE guidelines, however, recommend the use of a leukocyte depletion filter and we suggest that the issue is an area requiring further scientific or clinical research to address the current uncertainty.
CLOSING THE GAP
Further reductions in the reliance on allogeneic RBC transfusion are likely to come from an array of techniques and devices positioned within the context of an overall blood conservation strategy. The usage of these various measures is typically determined by availability, cost and clinician preference.
Acute normovolaemic haemodilution and preoperative autologous donation have previously been reviewed by the International Study of Peri-operative Transfusion group . The group concluded that the quality of supporting data was of a variable standard, but that both techniques may offer a modest reduction in the demand for RBC transfusion.
Anti-fibrinolytic drugs, such as aprotinin and tranexamic acid, have previously been shown to reduce the degree of surgical blood loss but concerns about increased cardiac events led to the withdrawal of the former, and neither would be appear to be in regular use.
Preoperative administration of erythropoietin or iron enables optimization of haemoglobin levels, reducing the need for RBC transfusion. The timing of treatment can be difficult in planned early oncological surgery but the routine use of neo-adjuvant chemotherapy in patients undergoing radical cystectomy for transitional cell carcinoma may afford a larger ‘therapeutic window’ to allow scheduling of administration.
Improvements in anaesthetic and surgical techniques will continue to yield results in reducing the overall blood loss from surgical interventions. Permissive hypotension, maintenance of normothermia and goal-directed fluid therapy are now commonplace. The increasing use of minimal access surgery, electrocautery energy devices and topically applied therapeutic haemostatic agents are also contributing to reduced intra-operative blood loss.
Of prime importance are blood transfusion triggers and transfusion protocols. A more tolerant approach to postoperative anaemia in patients without significant cardiac co-morbidity has undoubtedly reduced the use of blood products.
LIMITATIONS OF THE PRESENT SERIES
The present series is retrospective, non-randomized and has no internal control group against which to compare clinical outcome data. However, given the acknowledged risks associated with allogeneic RBC transfusion and the recognized benefits of cell salvage, we feel it would be difficult ethically to sanction a randomized controlled trial. All patients were consented for the use of ICS during their procedure and we agree with other authors, who have also reviewed the literature, that the only contraindication to cell salvage based on current evidence should be patient refusal . We believe that this observational study of a large consecutive and contemporary cohort of patients undergoing radical cystectomy with similar transfusion thresholds represents current practice and identifies clearly the niche for ICS in patients undergoing radical cystectomy.
In conclusion, the use of ICS during radical cystectomy is safe; it is capable of meeting the majority of or, in some cases, the total blood requirement for individual patients. As a result, it decreases the need for allogeneic RBC transfusion and hence the associated risks. Current follow-up shows no apparent risk of decreased long-term survival from an oncological perspective. The authors advocate routine availability of ICS for all major urological oncology cases.