This post is an extraction of my Letter to the Editors of UHM (2019 Vol 46), and provided here for those who do not read this journal. I included some narrative on the nuances of clinical trial design to appreciate where this study sits within evidence-based medicine expectations.
Efficacy studies are those designed to determine maximum achievable treatment response in a tightly controlled research environment, and the capacity for any demonstrated effect in everyday practice. Clinical decision-making places increasingly emphasis on such high-level evidence, as do those who purchase health care. Efficacy data supporting hyperbaric medicine have long been in short supply. Practice decisions frequency rest on a mix of laboratory findings, the ‘matching’ of disease pathophysiology to a therapeutic mechanism, retrospective reports and uncontrolled prospective case series, alone or in any combination. When hyperbaric efficacy research does become available, then, it is deserving of particular attention, analysis and dissemination.
Oscarsson et al. have generated one such example that serves to elevate efficacy evidence for hyperbaric oxygen (HBO) treatment of less severe yet common forms of radiation cystitis. (1) The term radiation cystitis refers to a collection of signs and symptoms (Table). It is only the second randomized controlled trial to investigate HBO therapy for this condition (2) and the first to demonstrate a statistically significant healing advantage over standard care. Importantly, the trial was registered with International Committee of Medical Journal Editors approved trial registries. A clinical trial (defined as prospective assignment of participants to one or more health-related interventions to evaluate outcomes) registration is essential if results are to be considered for publication in ICMJE participating journals.
Of 223 patients assessed 87 met inclusion criteria and were subsequently enrolled. This 39% conversion ratio suggests that their reported improved outcomes are largely generalizable. Although not reported, the ‘Number Needed to Treat’ (an epidemiological measure used to communicate effectiveness of an intervention, and representing the average number of patients needed to be treated in order to produce one favorable outcome) was computed as an encouraging 3. This value was the same for the subjective Expanded Prostate Index Composite (EPIC) and objective Late Radiation Morbidity Grading Scheme (LRMGS) scores, rounded up by convention from 2.56 and 2.17, respectively. An identical NNT was reported in our HBO radiation proctitis randomized controlled trial. (3) One would hope that the authors’ decision to exclude from ‘Intention to Treat’ analysis those patients who withdrew consent immediately upon learning of their randomization is not be judged too harshly by evidence ‘purists’. Shorter intervals between diagnosis and initiation of HBO were associated with improved responses, as were reduced radiation therapy to HBO periods. This interval-related response has been reported elsewhere (4) and further supports HBO’s earlier application. Patients suffering concurrent radiation proctitis also experienced improvement in this condition, suggesting a unique benefit of systemically delivered HBO in the setting of multi-organ involvement. One would not expect a favorable response of more localized standard care to extend to other radiation-damaged organs and structures. Improved bladder findings per LRMGS scores add to the contention that HBO is uniquely disease modifying. (3, 5, 6) This effect serves to limit the frequency of, and in many cases eliminate altogether, the remitting-relapsing consequences of more common elements of standard management directed principally at relief of symptoms, such as arrest of bleeding while not overcoming its cause.
The absence of a sham component was unfortunate. Human behavior is influenced by what we know or believe, so blinding of patients to the treatment they will receive in a controlled trial is particularly important when response criteria are subjective, (7) which was the case with this study. Adoption of sham controls and blinding would have further elevated this work within the evidence-based medicine hierarchy. This decision eliminated the ability to blind patients, as it did LRMGS assessors. The authors’ arguments against sham were unconvincing, and neither example provided to suggest a study had been negatively impacted by inclusion of sham was correct. Sham control and double-blinding trial design was successfully incorporated into the study of HBOT for treatment of radiation proctitis. (3) Blinded sham controls may have also served to minimize the 16% drop-out rate post randomization in patients allocated to the control group, thereby permitting these patients to becoming eligible to receive HBO ‘off protocol’. Attempts to evaluate de facto indications for HBO at higher levels of scientific scrutiny are challenging. Providers may be ethically reluctant to commit patients to a sham exposure when clinical experience is associated with generally favorable responses. There is also the specter of medical-legal recourse. For example, a patient randomized to sham may suffer disease advancement during their study inclusion period, perhaps resulting in a fistula or bladder rupture, thereby necessitating a surgical procedure that arguably would not have been required if HBO had been delivered as ‘standard care’. These concerns can be lessened somewhat when treating chronic conditions by inclusion of a crossover arm. The benefit of crossover is that it assesses response in previously untreated control patients. A statistically significant improvement observed in crossover patients represents powerful confirmation of therapeutic effect. Its principal criticism is that it eliminates the potential to analyze an intervention’s enduring effect. When both groups have received active treatment, long-term comparisons are no longer possible. One might argue that for treatment of late radiation tissue injury, however, that any lessening or elimination of aforementioned remitting-relapsing characteristics represents an enduring effect surrogate, as noted in the proctitis RCT. (3) It is encouraging to note that eventual reporting of histologic data will involve blinded assessors.
The full extent of radiation-induced bladder injuries was not included. The authors believed that withholding HBO for more advanced cases would have been unethical. This position is somewhat difficult to reconcile as the authors note elsewhere that evidence supportive of HBO as treatment for radiation cystitis “is weak”.
References were well chosen; peer reviewed and reflective of the current era. This is refreshing for a hyperbaric publication, where inclusion of dated publications, non-peer reviewed textbook chapters and meeting Abstracts is too common. The former rather than the latter principally influence referring physicians and those who undertake literature reviews to generate clinical practice recommendations and guide reimbursement policy.
The hyperbaric dosing protocol was appropriate for this condition and slight inter-institutional variances of no consequence. Reported harms were those commonly anticipated in routine clinical practice. Each was minor in degree, largely self-limiting and supportive of the position that hyperbaric medicine is a well-tolerated, relatively safe and mastered medical technology.
Leading US commercial insurers and US Medicare approve the use HBO for radiation cystitis and reimburse accordingly. The basis for these policy decisions is the sum of modest yet consistent effectiveness data in the absence of prospective randomized efficacy studies. This new publication will certainly augment these payment positions. One also hopes that reimbursement policy will evolve to recognize HBO as essential standard care. A recent scoping report and meta-analysis, pre-dating the Oscarsson et al. publication, concludes that ‘using hyperbaric oxygen therapy early in the development of radiation cystitis may be associated with greater success’. (8) At present, health insurers commonly approve the use of HBO only after ‘standard care’ has proven unsuccessful. Of interest, none of the intra-vesical and other systemic agents used to treat radiation cystitis have been studied and proven efficacious to the level HBO now enjoys.
TABLE: RADIATION CYSTITIS MANIFESTATIONS
Frequency, urgency, decreased stream
Vascular telangiectasia, marked hyper-vascularity
Clot retention, obstruction
Reduced bladder capacity
Loss of mucosal integrity
Urethral stricture +/- fibrosis
Bladder neck contracture
Vesicovaginal fistula; colovesical fistula
Oscarsson N, Muller B, Rosen A, et al. Radiation-induced cystitis treated with hyperbaric oxygen therapy (rich-art): A randomised, controlled, phase 2-3 trial. Lancet Oncol 2019: http://dx.doi.org/10.1016/S1470-2045(19)30494-2
Shao Y, Lu GL, Shen ZJ, et al. Comparison of intravesical hyaluronic acid instillation and hyperbaric oxygen in the treatment of radiation-induced hemorrhagic cystitis. BJU Int. 2012;109(5):691-4
Clarke RE, Tenorio LMC, Hussey JR, et al. Hyperbaric oxygen treatment of chronic refractory radiation proctitis: A randomized and controlled double-blind crossover trial with long-term follow-up. Int J Radiation Oncology Biol Phys 2008:72(1):134-143
Dellis A, Papatsoris A, Kalentzos V, et al. Hyperbaric oxygen as sole treatment for severe radiation-induced haemorrhagic cystitis. Int Braz J urol 2017;43:489-495
Marx RE, Ehler WJ, Tayapongsak P, et al. Relationship of oxygen dose to angiogenesis induction in irradiated tissue. AM J SURG 1990;160(5):519-524
Svalestad J, Hellem S, Thorsen E, et al. Effect of hyperbaric oxygen treatment on irradiated oral mucosa: Microvessel density. Int J Oral Maxillofac Surg 2015;44(3):301-7
Day SJ. Blinding in clinical trials and other studies. BMJ 2000;321(7259):504
Cardinal J, Slade A, McFarland M, et al. Scoping review and meta-analysis of hyperbaric oxygen therapy for radiation-induced hemorrhagic cystitis. Curr Urol Rep 2018;19(6):38