To evaluate the safety, tolerability and effectiveness of outpatient (office-based) laser ablation (OLA), with local anaesthetic, for non-muscle-invasive bladder cancer (NMIBC) in an elderly population with and without photodynamic diagnosis (PDD).
To compare the cost-effectiveness of OLA of NMIBC with that of inpatient cystodiathermy (IC).
Patients and Methods
We conducted a prospective cohort study of patients with NMIBC treated with OLA by one consultant surgeon between March 2008 and July 2011
A subgroup of patients had PDD before undergoing OLA.
Safety and effectiveness were determined by complications (In the immediate post operative period, at three days and at three months), patient tolerability (visual analogue score) and recurrence rates.
The long-term costs and cost-effectiveness of OLA and IC of NMIBC were evaluated using Markov modeling.
A total of 74 OLA procedures (44 white-light, 30 PDD) were carried out in 54 patients. The mean (range) patient age was 77 (52–95) years. More than half of the patients had more than three comorbidities. Previous tumour histology ranged from G1pTa to T3.
One patient had haematuria for 1 week which settled spontaneously and did not require hospital admission. There were no other complications.
The procedure was well tolerated with pain scores of 0–2/10.
Additional lesions were found in 21% of patients using PDD that were not found using white light.
At 3 months, the percentage of patients who had recurrence after OLA with white light and OLA with PDD were 10.6 and 4.3%, respectively. At 1 year, 65.1% and 46.9% of patients had recurrence.
The cost of OLA was found to be much lower than that of IC (£538 vs £1474), even with the addition of PDD (£912 vs £1844).
Over the course of a patient's lifetime, OLA was more clinically effective, measured in quality-adjusted life-years (QALY), than IC (0.147 [sd 0.059]) and less costly (£2576.42 [sd £7293.07]).
At a cost-effectiveness threshold of £30 000/QALY, as set by the National Institute for Health and Care Excellence, there was an 82% probability that OLA was cost-effective.
This is the first study to demonstrate the long-term cost-effectiveness of OLA of NMIBC.
The results support the use of OLA for the treatment of NMIBC, especially in the elderly.
Traditionally, non-muscle-invasive bladder cancer (NMIBC) has been treated by inpatient rigid cystoscopy and resection or cystodiathermy. The combination of a low risk of progression and a high risk of recurrence results in patients undergoing multiple procedures [1, 2]. Approximately 50% of these procedures are performed in patients aged >75 years  and the management of NMIBC in this age group is complicated by the significant risk of repeated admissions and interventions [4-6]. The UK National Confidential Enquiry into Peri-operative Deaths reported in 1997 that 102 patients a year died after cystoscopy (with or without biopsy/cystodiathermy) and 77% of all deaths from urological operations were in patients aged ≥70 years .
Frequent admissions for repeated procedures are also associated with significant costs to healthcare providers. In England and Wales, 39 712 inpatient endoscopic treatments of bladder lesions were performed, accounting for 80 923 inpatient bed-days between 2010 and 2011 . Consequently, the cost of managing patients with bladder cancer is the highest of all cancers: £55.39 m in 2005, of which NMIBC accounted for £35.25 m .
The use of cystodiathermy delivered under local anaesthetic to reduce the risks associated with general anaesthetic in this patient population has been previously explored [9-12] and is widely used, particularly in North America. Recent studies suggest its use is limited by tolerability as 12% of patients find the procedure painful . The use of laser technology, as an alternative to electrocautery, potentially alleviates this problem owing to the absence of electrical conduction and stimulation, and also causes less bleeding . The introduction of the neodymium:YAG laser for the management of recurrent NMIBC was limited by concerns about the risk of bowel injury because of deep laser penetration ; however, contemporary holmium:YAG laser technology is safer as it has a much shorter extinction length (0.4 mm in water) . Several studies have described its use in the day-surgery management of recurrent NMIBC [17-19]. All studies reported patient tolerability and satisfaction with local anaesthetic treatment, with a low rate of recurrence (10–25%) and follow-up of 3–33 months. We are the first group to report our experience of transferring local anaesthetic laser treatment from day surgery into an outpatient (office-based) setting.
In addition to using holmium:YAG laser to ablate tumours, we also used photodynamic diagnosis (PDD) or blue-light cystoscopy in a subset of patients. Rigid cystoscopy using PDD has been reported to detect additional tumours not seen with conventional white-light cystoscopy. The addition of PDD to outpatient laser ablation (OLA) could further increase the efficacy of recurrent NMIBC management by ensuring all abnormal tissue is ablated [20-23].
Although comparison of the costs associated with day-case local anaesthetic laser ablation and inpatient cystodiathermy (IC) has previously been reported [17, 19], further savings would seem possible when the procedure is performed in an outpatient setting rather than as day surgery. Furthermore, no study has yet considered the overall costs incurred over the course of a patient's lifetime, which may be significantly more than the costs associated with the initial management given the high recurrence rate in this population. The objectives of the present study were therefore to evaluate the safety, tolerability and effectiveness of OLA, with and without PDD, of NMIBC in an elderly population and to compare the cost-effectiveness of OLA of NMIBC with IC.
Patients and Methods
We conducted a prospective cohort study of patients with NMIBC treated with OLA by one consultant surgeon between March 2008 and July 2011. Elderly frail patients and patients with multiple comorbidities, for whom general anaesthesia would present a high risk, and small-volume recurrent tumours were offered OLA.
Exclusion criteria included first presentation of tumour, young age (<50 years), large tumours (>3 cm), tumours adjacent to the bladder neck, tumour invading the bladder muscle where patient is fit for treatment with curative intent and the presence of untreated urinary infection. In addition to patient demographics and tumour recurrence, data were recorded on comorbidities, initial histology, pain, tolerability and complications of the OLA procedures. In addition, length of hospital stay, complications and recurrence were recorded for any inpatient procedures that this cohort of patients underwent between March 2003 and October 2011.
Laser ablation was performed in the outpatient department by one consultant surgeon, assisted by a laser-trained band 6 nurse. Using an aseptic technique, 10 mL of Instillagel® (Farco-Pharma GmbH, Cologne, Germany) was administered before cystoscopy and a 16.5-F flexible video cystoscope was used to map the bladder with white light (Fig. 1A). A holmium:YAG laser (Olympus KeyMed Ltd., Southend-on-Sea, UK) with a 365- or 200-nm fibre at 0.6–0.8Js energy and rate of 10–15 Hz was used to ablate any tumours. Normal saline solution was used as irrigation fluid. Patients were asked to void before discharge.
From November 2009 to July 2011, a subgroup of patients underwent PDD before OLA. In these patients, 50 mL hexaminolevulinate acid (Hexvix GE; Amersham, Little Chalfont, UK; Photocure ASA, Oslo, Norway) was instilled into the patient's bladder 1 h before OLA. Patients were asked to void immediately before the procedure and a PDD-enabled 16.5-F flexible cystoscope was used to visualize the bladder initially with white light, then blue light (Fig. 1B). Any additional tumours seen under blue light were noted before laser ablation was performed.
All patients were given department contact details on discharge, had telephone follow-up 3 days after the procedure, and were reviewed at 3 months with check flexible cystoscopy.
Patient tolerability was determined by asking patients to rate their pain immediately after their procedure using a visual analogue pain scale (0 = no pain, 10 worst pain imaginable). Complications (e.g. pain, haematuria, UTIs and admissions) were assessed immediately after the procedure, during the telephone consultation at 3 days and at the time of their 3-month check cystoscopy. The recurrence rates at 3 months and 1 year using Kaplan–Meier curves were compared between OLA with white light and OLA with PDD [24, 25]. Statistical analysis was performed using Prism (GraphPad Software Inc., La Jolla, CA, USA).
Cost-effectiveness analysis was performed from a UK NHS perspective according to National Institute for Health and Care Excellence (NICE) guidelines on health technology appraisal . We used survival and health-related quality of life to determine the quality-adjusted life-years (QALY) pay-off for each intervention . This method allowed us to assess not only if OLA or IC was more clinically effective, but also if any additional effectiveness justified potentially larger costs associated with the more effective intervention.
The procedural costs of OLA and IC were calculated using manufacturer-supplied costs of equipment (Karl Storz Endoscopy Ltd., Slough, UK; Olympus Key Med Ltd) and data from our institution.
Markov modelling allowed us to simulate long-term survival, effectiveness and costs associated with OLA and IC. In our model, a patient can exist in one of three states (Recurrence and Procedure, Disease Free, and Death) (Fig. 2). A 1-year cycle was used and a 10-year time horizon was deemed sufficient to capture all treatment effects given the age and comorbidities of this patient population. We tested the appropriateness of this time horizon by performing a series of alternative analyses with different time horizons from 5 years to a lifetime time horizon in which modelling was continued until all the hypothetical patient cohort had died. The probability that a patient remained in the same state or moved to a different state, defined as the ‘transition probability’, was deduced from our recurrence data, applying previously described methods [28, 29]. Decision analytical modelling was performed using TreeAge 2009 (TreeAge Software Inc., Williamstown, MA, USA).
Annual age- and gender-specific mortality rates were obtained from the UK Government Actuarial Department life tables . Bladder cancer-related mortality for patients with NMIBC was taken from the largest published longitudinal UK study that explicitly reported cancer-specific hazard rates [31-35]. Owing to a lack of studies reporting peri-operative risk of patients undergoing IC, we assumed peri-operative mortality rates were similar to those of patients with similar demographic characteristics undergoing other minor operations under general anaesthetic [4-6]. Probabilistic sensitivity analysis was performed to quantify the combined uncertainty associated with all model variables on the results of our analysis [28, 36] in accordance with NICE guidance . A Monte Carlo simulation was performed in which analysis for each hypothetical cohort of 1000 patients was repeated 1000 times [28, 37].
Between March 2008 and July 2011, 74 OLA procedures (44 white-light, 30 PDD) were performed on 54 patients. The mean (range) patient age at presentation was 77 (52–95) years. The male : female ratio was 1.39:1. More than half of the patients had more than three comorbidities. Previous tumour histology ranged from G1pTa to T3 (Fig. 3) and all patients had low-volume recurrence at the time of OLA. Half of the eight patients on warfarin stopped taking this before the procedure, whilst the other half continued because of imperative indications.
All treatment was successfully completed on the day and no patients were admitted. OLA was well tolerated by all patients with a pain score of 0–2 and all patients would choose to have the procedure again. There were no adverse reactions to hexaminolevulinate acid. The four patients on warfarin therapy had successful treatment without complications. Early in the series, one patient with multiple tumours, not on warfarin, had haematuria after OLA which settled spontaneously and didn't necessitate hospital admission. Patients were telephoned on day 3 after the procedure and were followed up at 3-month check cystoscopy and there were no other reported cases of complications including pain, haematuria, retention, UTIs, or hospital admissions. Additional lesions were seen in 21% of patients with PDD that were not seen with white light.
At 3 months, 10.6% of patients who had OLA had recurrence compared with 4.3% of patients who had OLA with PDD. At 1 year, the figures were 65.1 and 46.9%, respectively.
The procedural cost of OLA was much lower than that of IC (£538 vs £1474). The addition of PDD clearly increased the costs of both OLA and IC, but the cost of OLA with PDD was still cost half the cost of IC with PDD (£912 vs £1844).
Markov analysis over a 10-year period showed that OLA was both more effective than IC (3.68 [sd 0.52] vs 3.56 [sd 0.50] QALYs) and less costly than IC (£3217.96 [sd £359.17] vs £5744.33 [sd £6760.76]). Further analysis at different time thresholds showed that OLA was still estimated to outperform IC both at 5 years (incremental effectiveness 0.067 [sd 0.026] QALY; cost reduction £2031.67 [sd £5357.85]) and on a lifetime time horizon (incremental effectiveness 0.147 [sd 0.059] QALY; cost reduction £2576.42, [sd £7293.07]).
At a cost-effectiveness threshold of £30 000/QALY set by NICE, there was an 81.9% probability that OLA was cost-effective (Figs 4, 5). Despite the increased cost with the addition of PDD to the OLA, it was still more cost-effective than IC, with a certainty of 79.2% (incremental effectiveness 0. 124 [sd 0.050] QALY; cost reduction £1961.56 [sd £6795.17]).
The present study has shown for the first time that laser ablation of NMIBC can be successfully delivered in an outpatient setting. The procedure appeared to be effective, well-tolerated, and associated with a low risk of complications. In addition, we have shown that PDD can be successfully used in combination with OLA, thereby potentially increasing its overall effectiveness. Not only does the procedure appear to be clinically effective but our analysis suggests the procedure is also cost-effective.
The results are consistent with previously published studies suggesting that local anaesthetic laser ablation in day-case settings can be clinically and cost-effective [14, 17-19]. Although laser is not the only treatment method that has been described for treating NMIBC under local anaesthetic, cystodiathermy appears less well tolerated with pain reported in ∼12% of patients . Randomized studies comparing laser ablation and diathermy have not to our knowledge been performed.
The patients in the present study were very frail with >50% of patients having three or more comorbidities, and 57% were aged >75 years. Despite this, the rate of complications was exceptionally low: no patients had a bladder perforation, developed urinary retention, or needed to be admitted to hospital. Pain scores were also very low despite the only analgesic used being intra-urethral Instillagel® (Farco-Pharma GmbH, Cologne, Germany), suggesting that complex local anaesthetic blocks do not appear necessary.
Furthermore, the finding that all the patients would undergo the procedure again suggested that it was extremely well tolerated. All the patients had previous experience of inpatient cystoscopy under general anaesthetic and they noted that particular benefits of the OLA technique were minimal pain and bleeding. Eight patients in this study were on warfarin and in four, this was stopped before OLA. Four patients did, however, undergo the procedure safely whilst still receiving the anticoagulation treatment warfarin.
The small size of our study population and the heterogeneous nature of the tumours treated make it difficult to draw strong conclusions about the relative effectiveness of OLA and IC in terms of tumour recurrence; however, recurrence rates after OLA in this study appeared to be similar to those after IC and within reported rates in the literature. Larger studies would be required to definitively answer questions relating to recurrence, however, one factor contributing to the high recurrence rates in patients with NMIBC is the possibility of a tumour being missed by the surgeon, despite diligent cystoscopy. The addition of PDD to laser ablation is an attractive solution to this problem in this population. We have demonstrated the feasibility of combining the two techniques and, interestingly, extra tumours were seen in 21% of patients under blue light. This is similar to our experience with the use of PDD under general anaesthetic with rigid cystoscopy .
The potential cost savings of this approach to the treatment of NMIBC in frail elderly patients are significant. The individual procedural cost of OLA was £538 compared with £1474 for IC. The tariff our trust receives for both IC and OLA is £852. Across the UK health service, in which >30 000 endoscopic treatments of bladder lesions are performed per year, the cost savings would be >£1.2 m from moving just 10% of these inpatient procedures to an outpatient setting. The addition of PDD increases the cost of both OLA (to £912) and IC (to £1844) but OLA is still significantly cheaper.
Our Markov model suggests with a high degree of certainty (81.9%) that OLA may be more cost-effective than IC over a 10-year period, and this result persists even when the time frame is shortened (5 years) or extended (lifetime). The reduced costs associated with OLA can largely be explained by the elimination of hospital stay and the significantly cheaper cost of clinic vs theatre time. The improved effectiveness assessed in terms of QALY pay-off is largely attributable to the reduction in peri-operative mortality and complications associated with IC.
This is the first study in this area where long-term costs and clinical effectiveness associated with a disease characterized by frequent recurrence and re-intervention have been considered. The ever-increasing financial constraints on health systems mean that this is important if a study is to facilitate the planning and provision of health services [28, 37].
The present study has some limitations. The group of patients included is highly selected to include elderly frail patients and patients with multiple comorbidities with tumours amenable to OLA. Clearly OLA is not appropriate for many patients with recurrent NMIBC, particularly those in whom accurate staging information is required from the surgical procedure. None of the patients in this study was a candidate for radical cystectomy and, in the majority, the previous tumour history was that of low grade recurrence. High-volume tumours and tumours at the bladder neck are difficult to treat by OLA, the latter because limitations to the angulation of the flexible cystoscopy prevent targeting of the laser onto the tissue. It is possible that single-shot intravesical chemotherapy after OLA could lead to further reduced recurrence rates but patients in this study were treated with laser alone. This policy was employed to minimize the risk of peri-operative complications.
There are also limitations inherent to all decision analytical models including the potential for oversimplification of the model and a tendency for modelling error to accumulate with time [28, 36]. We minimized this problem by limiting our base-case analysis to a 10-year horizon. Probabilistic sensitivity analysis was performed to assess the uncertainty associated with all model variables. Alternative analyses were performed to test the sensitivity of the results to model structure and key assumptions, including time horizon.
In conclusion, this study suggests that laser ablation of recurrent NMIBC can be safely performed in frail elderly patients or patients with multiple comorbidities at high anaesthetic risk in an outpatient setting. The technique appears to be well tolerated, oncologically robust and cost-effective. This study also demonstrates the feasibility of PDD as an adjunct in highlighting additional tumours that may otherwise be missed with white-light cystoscopy alone. The economic benefits to the NHS could be considerable if OLA was more widely used.