Article Text
PDF
Abstract
Background There has been no prospective trial for treatment of immune-related pneumonitis (irP) occurred after immune checkpoint inhibitors (ICIs).
Methods In this single-arm phase II study, patients with cancer with grade ≥2 irP received oral prednisolone (1 mg/kg/day), tapered over 6 weeks. The primary endpoint was a pneumonitis control rate at 6 weeks from the start of the study treatment, defined as complete disappearance or partial improvement of irP in high-resolution CT of the chest.
Results Among 57 patients enrolled, 56 were included in the final analysis. The most frequent cause of irP was single ICI therapy (51.8%), followed by combination with chemotherapy plus ICI (39.3%). Thirty-five (62.5%) patients had grade 2 irP and 21 (37.5%) had grade ≥3. Fifty-one (91.1%) patients completed the study treatment while 5 discontinued the study treatment because of relapse of irP (n=1), death from cancer (n=1), occurrence of immune-related hepatitis (n=1), extension of the treatment duration more than 6 weeks (n=1), and attending physician’s decision (n=1). Six weeks after the start of the study treatment, 16 (28.5%) patients demonstrated complete recovery from irP, 35 (62.5%) had a partial improvement in irP, 1 (1.8%) had a relapse of irP, and 4 (7.1%) were not evaluable. The pneumonitis control rate at 6 weeks was 91.1% (95% CI, 80.7% to 96.1%). Twelve weeks after the start of the study treatment, 5 (8.9%), 27 (48.2%), and 15 (26.8%) patients demonstrated complete recovery, partial improvement, and relapse, respectively, and 9 (16.1%) were not evaluable. The pneumonitis control rate at 12 weeks was 57.1% (95% CI, 44.1% to 69.2%). During the observation period, 18 (32.1%) patients experienced a relapse of irP, and of those, 17 received re-treatment with corticosteroids. Grade ≥3 adverse events occurred in 10 (17.9%) patients, in which hyperglycemia was most frequent (n=6). There was no treatment-related death.
Conclusions In this first prospective study for irP, prednisolone at 1 mg/kg/day, tapered over 6 weeks, demonstrated a promising clinical benefit and manageable toxicity, suggesting a potential treatment option for irP.
Trial registration number jRCT: 1041190029.
- Autoimmunity
- Cytotoxicity, Immunologic
- Immune Checkpoint Inhibitors
- Immunotherapy
Data availability statement
Data are available upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Guidelines recommend high-dose corticosteroids for a certain period to treat immune-related pneumonitis (irP), which is based on clinical experience and consensus opinion. Treatment of irP has not been evaluated in any prospective trials, and there is no standardized, evidence-based treatment approach for irP.
WHAT THIS STUDY ADDS
This is the first prospective trial that evaluated the efficacy and safety of a 6-week tapering schedule of prednisolone at 1 mg/kg/day for patients with cancer with grade ≥2 irP. Pneumonitis control rate at 6 weeks was 91.1% (95% CI, 80.7% to 96.1%). During the observation period, 18 (32.1%) patients experienced a relapse of irP. Grade ≥3 adverse events occurred with 17.9% of patients; hyperglycemia was the most frequent adverse event (10.7%).
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE, OR POLICY
Prednisolone at 1 mg/kg/day, tapered over 6 weeks demonstrated a promising clinical benefit and manageable toxicity, and can be a potential treatment option for irP. The current study may provide a benchmark for future studies of irP treatment.
Background
Immune checkpoint inhibitors (ICIs) are increasingly used as new standard treatments for cancer therapy.1 Furthermore, several ICI therapeutic strategies have been developed, including single ICI therapy, combinations of ICI with chemotherapy and combinations of different ICIs.2 Unlike conventional chemotherapy that has direct cytotoxicity to tumor cells, ICIs exert their antitumoral effect by activating the antitumor responses of immune cells. Although the unique anticancer mechanism of ICI provides unprecedented clinical benefits for patients with cancer, there is also a serious issue of immune-related adverse events (irAEs). The irAEs are defined as a spectrum of specific side effects of ICIs that mimic autoimmune diseases.3 4 The irAEs occur in diverse organ systems and require organ-specific management and treatment.
Immune-related pneumonitis (irP) is a potentially serious irAE and influences the course of cancer treatment. In clinical trials, the incidence of irP is reported to range from 1% to 6%.5 6 However, irP is not uncommon in a real-world setting, and the incidence and mortality of irP is reported to range from 5% to 19% and from 2% to 27%, respectively.7–14 Thus, cancer therapy is often interrupted because of treatment for irP, even if patients avoid a fatal outcome.7
Treatment of irP has not been evaluated in any prospective trials, and there is no standardized, evidence-based treatment approach for irP. Guidelines recommend high-dose corticosteroids for a certain period to treat irP. However, these recommendations are based on clinical experience and consensus opinion, and the dose and duration of corticosteroid differs among guidelines.15–18 The American Society of Clinical Oncology and Society for Immunotherapy of Cancer recommend 1–2 mg/kg/day prednisone (or equivalent), tapered over 4–6 weeks for grade 2 irP and 1–2 mg/kg/day methylprednisolone intravenously (or equivalent), tapered over 4–6 weeks for grade ≥3 irP.15 16 The European Society for Medical Oncology recommends 2–4 mg/kg/day methylprednisolone (or equivalent), tapered over 6 weeks or more for grade≥3 irP.17
In clinical practice, the treatment approach of irP depends on each physician, and thus, the dose and duration of corticosteroid therapy can vary widely among patients. For example, a retrospective cohort study in the USA and Australia reported that patients with irP received a median prednisone dose of 50 mg (range, 20–80 mg) for a median duration of 68 days (range, 20–154 days).9 Our retrospective cohort study in Japan reported a median prednisolone dose of 30 mg (range, 20–60 mg) and a median duration of 91 days (IQR, 49.5–180.5 days) for treating irP.8 Insufficient corticosteroid therapy has potential risk of treatment failure.19 Conversely, an overlong duration of corticosteroid therapy is potentially harmful by increasing adverse events and interruption of subsequent cancer treatment. Elucidation of optimal corticosteroid therapy for irP in a prospective clinical trial is needed.
We hypothesized that high-dose corticosteroid therapy, tapered over fixed duration, would be efficacious of irP. The aim of this single-arm phase 2 study was to evaluate the efficacy and safety of a 6-week tapering schedule of prednisolone at 1 mg/kg/day for patients with cancer with irP.
Methods
Study design
This was a single-arm phase 2 study conducted in 15 hospitals in Japan between May 2019 and June 2022. Data were analyzed from January 2020 to September 2022. The study followed the ethical standards of the Declaration of Helsinki, and each patient provided written informed consent. The study was registered with the Japan Registry of Clinical Trials. This study followed the Transparent Reporting of Evaluations with Non-randomized Designs reporting guideline.
Patients
Patients with cancer who were diagnosed with grade 2 or higher irP based on the Common Toxicity Criteria for Adverse Events (CTCAE, V.5.0)20 and scheduled to receive oral corticosteroid therapy were included. The irP was defined as newly identified pulmonary parenchymal opacities at chest high-resolution CT (HRCT) after initiation of ICIs and the exclusion of other likely causes including bacterial pneumonia, pulmonary edema, radiation pneumonitis, or carcinomatous lymphangitis.21 Any types of cancer or ICIs and combination chemotherapy with ICIs were allowed in this study, excepting pneumonitis due to durvalumab following chemoradiotherapy for the chest. For patients with grade 3–4 irP, initial treatment with short-term intravenous corticosteroid therapy was allowed in advance of the study treatment at the investigators’ discretion. Exclusion criteria were grade 1 irP not requiring corticosteroid therapy, or patients who received immunosuppressants, antifibrotic agents, or corticosteroid therapy with a prednisolone-equivalent of ≥20 mg for comorbidities. The detailed inclusion and exclusion criteria are described in Trial Protocol in online supplemental file 1.
Supplemental material
Treatment
The study treatment was initiated with oral prednisolone at 1 mg/kg/day for 2 weeks. Then, the prednisolone dose was reduced to 0.7, 0.5, 0.2, 0.1, and 0 mg/kg/day in a stepwise manner for a week, for a total of 6 weeks of prednisolone treatment. The stepwise reduction of prednisolone was determined based on symptoms, chest imaging findings, respiratory function, and blood tests. Extension of the scheduled treatment duration was allowed for a maximum of 7 days if the investigator deemed this clinically necessary. The study treatment was stopped when patients demonstrated worsening progression in irP or more than 7 days extension of the treatment duration was required (Trial Protocol in online supplemental file 1).
Evaluation
The HRCT patterns of irP were classified into non-specific interstitial pneumonia (NSIP), organizing pneumonia (OP), hypersensitivity pneumonia (HP), eosinophilic pneumonia (EP), and diffuse alveolar damage, according to a position paper for drug-related pneumonitis by the Fleischner Society.21 After 6 and 12 weeks from the start of the treatment, the evaluation of the study treatment was assessed using chest HRCT as below; (1) recovery: complete disappearance of irP; (2) improvement: partial improvement compared with the onset of irP, but with residual abnormal finding; (3) unchanged: no change compared with the onset of irP; (4) worsening: worsening compared with the onset of irP; and (5) relapse: improvement compared with the onset of irP, but worsening compared with the best response during the treatment. The chest HRCT images were centrally reviewed by two independent radiologists who were blinded to all other patients’ data. Disagreements are resolved by consensus decision in collaboration with a third radiologist. Adverse events were evaluated using the CTCAE V.5.0.
Statistical analyses
The primary endpoint was pneumonitis control rate, defined as a percentage of patients who achieved recovery or improvement at 6 weeks from the start of the treatment. The secondary endpoints were pneumonitis control rate at 12 weeks from the start of the treatment, investigator-assessed pneumonitis control rate at 6 and 12 weeks, cumulative incidence of relapse of irP, time to the start of the next anticancer therapy from the start of the treatment, overall survival, and adverse events. In previous retrospective studies, patients with irP who received corticosteroid treatment demonstrated a pneumonitis control rate of approximately 80%.8 9 Based on the Southwest Oncology Group one-arm binominal design, the number of patients was set to assume a maximum irP control rate of 85% and a lower threshold of 70%, with a statistical power of 80% and an α error of 5% on one side, which would require the registration of 47 cases.22 The planned sample size was 55 patients for enrollment after taking dropouts into consideration.
Fisher’s exact test was used for the comparison of categorical variables. Kaplan-Meier analyses were used for overall survival. Gray’s test was used to analyze cumulative incidence of relapse of irP. Death prior to relapse was treated as a competing risk event. All values were analyzed using JMP V.13.2.0 (SAS Institute Japan, Tokyo, Japan), excluding Gray’s test using EZR V.1.55 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria).
Results
Patient characteristics
From May 2019 and June 2022, 57 patients with were enrolled, and 1 patient who was found not to have irP was excluded. Therefore, 56 patients with irP were included in the final analysis (figure 1). Patient characteristics are presented in table 1. The most frequent cancer type was non-small-cell lung cancer (NSCLC, 67.9%), followed by renal cell cancer (10.7%). Grade 2, 3, and 4 irP was present in 35 (62.5%), 19 (33.9%), and 2 (3.6%) patients, respectively. Additionally, 20 (35.7), 18 (32.1), 10 (17.9), and 8 (14.3) patients had the HRCT patterns of NSIP, OP, EP, and HP, respectively. The most frequent cause of irP was single therapy with anti-programmed death-1 (PD-1)/programmed death-ligand-1 (PD-L1) antibody (51.8%), followed by combination with chemotherapy plus anti-PD-1/PD-L1 antibody (39.3%). Among 21 patients having grade ≥3 irP, 10 (47.6%) patients received intravenous high-dose methylprednisolone therapy prior to the start of the protocol treatment (n=8, grade 3; and n=2, grade 4). These patients received 1 g of methylprednisolone for 3 days, except one who received 500 mg. All patients with grade 2 irP did not receive intravenous high-dose methylprednisolone therapy. A median observation time was 208 days (range, 45–784 days) at the data cut-off.
Patient characteristics
Chart of the study. irP, immune-related pneumonitis.
Treatment delivery
In a total of 56 evaluable patients, 51 (91.1%) completed the study treatment while 5 patients discontinued the study treatment because of the relapse of irP (n=1), death due to cancer progression (n=1), occurrence of immune-related hepatitis that required increasement of corticosteroid (n=1), extension of the scheduled treatment duration more than 6 weeks (n=1), and attending physician’s decision (n=1). The starting dose of prednisolone was a median of 60 mg (range, 40–60 mg), and the duration of the study treatment was a median of 42 days (range, 14–49 days).
Efficacy of 6-week prednisolone therapy
After 6 weeks from the start of the study treatment, 16 (28.5%) patients demonstrated complete recovery from irP, 35 (62.5%) had partial improvement in irP, 1 (1.8%) had relapse of irP, and 4 (7.1%) were not evaluable. The pneumonitis control rate at 6 weeks, the primary endpoint of the study, was 91.1% (95% CI, 80.7% to 96.1%) that exceeded both the preplanned maximum rate of 85% and a lower threshold of 70% (figure 2, online supplemental eTable 1). After 12 weeks from the start of the study treatment, 5 (8.9%), 27 (48.2%), and 15 (26.8%) patients demonstrated complete recovery, partial improvement, and relapse, respectively, and 9 (16.1%) were not evaluable. The pneumonitis control rate at 12 weeks was 57.1% (95% CI, 44.1% to 69.2%) (figure 3 and online supplemental eTable 1). The investigator-assessed pneumonitis control rate at 6 and 12 weeks was 91.1% (95% CI, 80.7% to 96.1%) and 57.1% (95% CI, 44.1% to 69.2%), respectively.
Supplemental material
Response and pneumonitis control rate at 6 weeks from the start of prednisolone therapy. Error bar indicates 95% CI.
Response and pneumonitis control rate at 12 weeks from the start of prednisolone therapy. Error bar indicates 95% CI.
Subgroup analyses
Patients with grade 2 irP had a pneumonitis control rate at 6 weeks of 97.1% (95% CI, 85.4% to 99.5%), which tended to be higher compared with grade ≥3 patients who had a pneumonitis control rate of 81.0% (95% CI, 60.0% to 92.3%, p=0.060). The pneumonitis control rate at 12 weeks did not differ between grade 2 (57.1%; 95% CI, 40.9% to 72.0%) and ≥3 (57.1%; 95% CI, 36.5% to 75.5%, p=1.000) (online supplemental eTable 2). Among 21 patients who had grade ≥3 irP, there was no significant difference in the pneumonitis control rate at 6 and 12 weeks either with or without intravenous high-dose methylprednisolone therapy prior to the study treatment (online supplemental eTable 3). There was no significant association of the severity of irP with the causative ICI therapies or cancer types. There was no significant difference in the pneumonitis control rate at 6 and 12 weeks between single ICI therapy and combination of ICI with other systemic anticancer therapies, between NSCLC and other types of cancers, and among HRCT patterns of irP (online supplemental eTable 4–7).
Adverse events
Adverse events of the study are shown in table 2. Any grade adverse events occurred in 37 (66.1%) patients. The most frequent any grade adverse event was hyperglycemia (42.9%), followed by insomnia (28.6%), infection (7.1%), adrenal dysfunction (5.4%), pneumothorax (5.4%), and constipation (5.4%). Grade 3 or 4 adverse events occurred in 10 (17.9%) patients, including hyperglycemia (n=6, 10.7%), infection (n=1, 1.8%), pneumothorax (n=1, 1.8%), esophageal ulcer (n=1, 1.8%), rash (n=1, 1.8%), and thrombocytopenia (n=1, 1.8%). One patient had both grade 3 hyperglycemia and urinary tract infection. There was no discontinuation of the study treatment due to adverse events and no treatment-related death.
Adverse events
Post protocol treatment
Twenty-nine (51.8%) patients resumed the systemic anticancer therapy as follows; cytotoxic agents (n=25), tyrosine kinase inhibitors (n=3), antibody-drug conjugates (n=2), concurrent chemoradiotherapy (n=1), and an anti-PD-L1 antibody (n=1). Three patients sequentially received different types of the anticancer therapies. A median time from the start of the study treatment to resumption of systemic anticancer therapy was 90 days (range, 32–580 days).
At the 12-week evaluation, 15 patients had a relapse of irP. Subsequently, three more patients had a relapse of irP. Overall, 18 (32.1%) patients experienced a relapse of irP during the observation period (figure 4). The severities of the relapsed irP were grade 1, n=6 (33.3%); grade 2, n=7 (38.9%); and grade 3, n=5 (27.8%). A median time from the start of the study treatment to relapse of irP was 68 days (range, 41–117 days). Among 18 patients who experienced a relapse of irP, 17 received re-treatment with corticosteroids, which improved the relapsed irP. Of these, 6 patients were still receiving corticosteroids at the data cut-off point. The median dose and duration of corticosteroids for re-treatment were 40 mg (range, 20–60 mg) and 146 days (range, 93–391 days), respectively. One patient who had grade 1 relapsed irP did not receive re-treatment with corticosteroid. The median overall survival from the start of the study treatment was 352 days (95% CI, 258 days to not estimable) (online supplemental eFigure 1).
Cumulative incidence of relapse of immune-related pneumonitis. Short bars indicate censored cases.
Discussion
This was the first prospective study to evaluate the efficacy and safety of corticosteroid therapy in patients with cancer with irP. Administration of oral prednisolone at 1 mg/kg/day, tapered over 6 weeks improved irP in 91.1% of the study patients. The efficacy of the 6-week prednisolone therapy was consistent across the severities of irP, the causative regimens of irP, and cancer types. Grade 3–4 adverse events, which were mostly corticosteroid-related and within allowable range, occurred in 17.9% of patients. Approximately 30% of the study patients experienced relapse of irP after the study treatment, which was controlled by re-treatment with corticosteroids. Furthermore, >50% of the patients resumed systemic anticancer therapy after the study treatment. The 6-week prednisolone therapy could be an effective and safe treatment option for irP.
The 6-week prednisolone therapy demonstrated sufficient efficacy with an irP improvement rate of >90%, which was comparable with or even better than those in previous observational studies. The improvement rate of irP after high-dose corticosteroid therapy was reported to be from 56% to 77% in retrospective studies.9–11 In those studies, the treatments for irP depended on physician discretion, and therefore, the dose and duration of corticosteroids widely varied between prednisolone-equivalent dose of 20 and 80 mg and between treatment duration of 11 and 154 days.9–11 Those studies included several patients who received relatively low-dose and/or short-duration corticosteroid therapy, which may provide a relatively low improvement rate, compared with the 6-week prednisolone therapy in the current study.
The 6-week prednisolone therapy demonstrated consistent efficacy across the severity of irP. There has been no clear evidence that demonstrates a difference in response to corticosteroid therapy among the severity of irP; therefore, the severity of irP is generally considered to be associated with resistance to corticosteroid therapy or poor outcome. In a retrospective cohort study, the efficacy of corticosteroid therapy was reported to be 93% in grade 2 irP and 64% in grade ≥3; however, the details of corticosteroid therapy, such as dose, duration, and oral/intravenous administration, for each severity were unknown.9 Here, the 6-week prednisolone therapy demonstrated a sufficient pneumonitis control rate at 6 weeks of 81% in grade ≥3 irP, which was better than that with grade ≥3 irP in the previous study.9 However, the pneumonitis control rate at 6 weeks in grade ≥3 patients tended to be lower compared with those with grade 2 who had a pneumonitis control rate of 97%. The treatment for grade ≥3 irP may have room for improvement with more intense therapy.
There is controversy regarding a need for intravenous high-dose methylprednisolone therapy prior to oral corticosteroid treatment. Despite the recommendation of guidelines, not all patients with grade ≥3 irP receive intravenous methylprednisolone in clinical practice.8 9 In the current study, approximately half of the patients with grade ≥3 irP did not receive high-dose methylprednisolone prior to the study treatment, which was not significantly associated with the pneumonitis control rate at 6 and 12 weeks. These data suggest that several patients with grade ≥3 irP would achieve irP control with sufficient dosing and duration of oral corticosteroid, even without intravenous high-dose methylprednisolone. However, the administration of intravenous methylprednisolone therapy was dependent on the investigators’ discretion in the current study, and therefore, there could be a selection bias whether to use it, based on clinical features other than severity of irP. Thus, whether or not to use this or for whom this is used should be further elucidated.
The relapse of irP after improvement by corticosteroid therapy requires careful attention in clinical practice. The cumulative incidence of the relapse of irP was 32.1% during the study period, which was numerically higher than the range reported in previous studies of 12.3% to 25.6%.8 9 19 However, all those studies included patients with grade 1 irP in the analysis for the relapse of irP, which may potentially have underestimated the relapse rate. In fact, when limited to grade ≥2 irP, the relapse rate was reported to be 42.3%.9 The current prospective study using the 6 weeks of prednisolone may provide a benchmark of the relapse rate of irP for future studies.
The current study had four main limitations. First, the evaluation method for the response to corticosteroid therapy in patients with irP was not validated. The efficacy was determined based only on chest HRCT findings, regardless of symptoms and laboratory findings. Furthermore, the assessment at 6 weeks after the start of the study treatment, that was intended to evaluate the early response to corticosteroid, was not necessarily suitable for the assessment of medium-term to long-term efficacy (ie, the assessment of relapse of irP). The optimal timing and method for the evaluation of irP needs to be discussed further. Second, the current study included irP that developed after treatment with the combination of ICI and other anticancer therapies. It is unknown whether irP that develops after single therapy with ICI (ie, pure irP) is different from irP that develops after a combination of ICI with other anticancer treatments (ie, complexed irP) regarding the clinical features and response to corticosteroid therapy. Additionally, when patients who receive a combination of ICI with cytotoxic chemotherapy develop pneumonitis, it is difficult to discriminate ‘true irP’ from pneumonitis caused by concomitant cytotoxic chemotherapy (ie, pseudo-irP). Third, only 3.6% of the study patients had grade 4 irP. Thus, the efficacy of 6-week prednisolone therapy for such severe irP needs to be elucidated further. Fourth, the current study evaluated only one fixed-dose and duration of corticosteroid therapy. It is possible that more intense treatment, such as increased dose and/or longer duration of corticosteroid, and addition of immunosuppressants, would improve the efficacy, especially the relapse after the corticosteroid treatment. Further studies are needed to elucidate the optimal treatment for irP.
Conclusions
In this first prospective study for irP, prednisolone at 1 mg/kg/day, tapered over 6 weeks, demonstrated a promising efficacy and manageable toxicity and thus could be a potential treatment option for patients with cancer with grade≥2 irP.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study was approved by the Institutional Review Board of Hamamatsu University School of Medicine (approved No. 19-037). Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We would like to acknowledge patients and their families. We thank Joe Barber Jr and Mike Herbert, PhD, from Edanz (https://jp.edanz.com/achttps://jp.edanz.com/ac) for editing a draft of this manuscript.
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors MK has full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. MK is the guarantor of the study. MK participated in concept and design, acquisition, analysis, and interpretation of data, statistical analyses, and drafting of the manuscript. NInu participated in concept and design, interpretation of data, drafting of the manuscript, and supervision. YI, HY, HH, YS, KF, TF, and NE participated in concept and design, interpretation of data, and drafting of the manuscript. KA, TU, MF, TM, SM, DH, MT, MI, HM, NIna, and YK participated in concept and design, data acquisition, and drafting of the manuscript. SF and SI: evaluation of chest imaging. SG: evaluation of chest imaging, conceptualization, and supervision. TS participated in concept and design, drafting of the manuscript, and supervision.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests No, there are no competing interests.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.