Transoral robotic surgery-based therapy for HPV-related oropharyngeal squamous cell carcinoma

Introduction

The prevalence of oropharyngeal squamous cell carcinoma (OPSCC) in Australia continues to rise, distinguishing it from other head and neck cancers (1,2). Human papillomavirus (HPV), predominantly subtypes HPV-16 and HPV-18, is the primary causative factor contributing to 12–85% of cases (1,2). Patients typically present at a younger age, exhibit milder symptoms, and develop locoregionally advanced disease earlier in their disease course compared to traditional smoking, or alcohol related cases. Treatment with any modality consistently demonstrates excellent survival outcomes (3). Given these patients have longer life expectancies, there is a shift in focus, where functional outcomes and quality of life is of greater importance.

Historically, the preference for radiotherapy-based approaches was driven by the high morbidity and mortality associated with open surgical approaches. However, these organ-preserving techniques (radiotherapy, with or without chemotherapy) are not without their own short and long-term side effects. Radiotherapy short and long-term toxicities can include mucositis, xerostomia, dysgeusia, swallowing dysfunction, osteoradionecrosis, stricture, neuropathy, and other systemic side effects which collectively, can cause long-term swallowing dysfunction and functional sequelae that impact on quality of life (4).

In 2006, Weinstein et al. first described utilising transoral robotic surgery (TORS) for the resection of head and neck pathology, and with its subsequent FDA approval in 2009 for its application in the treatment of head and neck cancers, this minimally invasive surgical technique has marked a transformative shift in this landscape (5,6). TORS enhances visualisation and dissection in a challenging anatomical area permitting en-bloc tumour resection and encourages adjuvant therapy rationalisation (4,5). However, while TORS asserts a minimally invasive technique, it requires postgraduate training and an experienced head and neck surgeon to be safe and oncologically effective (6,7).

Current treatment strategies for early-stage HPV-related OPSCC (HPVOPSCC) aim for single-modality treatment, such as primary surgical treatment with TORS and neck dissection or definitive radiotherapy. Locoregionally advanced disease necessitates dual-modality treatment, incorporating chemotherapy with definitive radiotherapy, or adjuvant radiotherapy post-surgical resection in the presence of poor prognostic features on histopathology (8).

While there is Australian literature exploring the use of TORS in oropharyngeal cancer, most of these are not specific for HPV-associated disease and include patients with both HPV positive and HPV-negative disease (9-13). The oncological and treatment outcomes of a fellowship trained robotic surgeon in Australia is described. We present this article in accordance with the STROBE reporting checklist (available at https://www.theajo.com/article/view/10.21037/ajo-24-15/rc).

Methods

Patients and selection

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Research Ethics Governance Information System (2021/ETH00217). Due to the retrospective nature of the research, the requirement for informed consent was waived. A retrospective review was conducted of data collected from patient records held within the senior author’s patient database for patients with HPVOPSCC undergoing primary surgical treatment between 2017 and 2022. Patients were classified with HPV-related disease if their biopsy, taken from the primary tumour or suspicious neck node, tested positive for p16, a surrogate marker for detecting HPV oncoprotein. All patients were clinically assessed by a qualified head and neck surgeon and underwent a fluorine-18-fluorodeoxyglucose positron emission tomography (PET) with diagnostic computer tomography (CT) neck and chest with intravenous contrast prior to presentation at a tertiary centre Head and Neck Cancer Multidisciplinary Team Meeting (MDT).

All patients with HPVOPSCC were presented at a Head and Neck Cancer Multidisciplinary Team Meeting (MDT) meeting and, given the comparable oncologic outcomes for both primary non-surgical therapy and surgical therapy, the pros and cons of each treatment method was discussed with the patient. Patients were given the option of primary surgical therapy if, on examination by the senior author, the primary tumour was considered resectable to clear margins and the resection would not lead to functional compromise in speech or swallowing. This encompassed T1–T2 and select T3 tumours. In cases considered T3 disease, this upstaging was due to the tumour being of an exophytic nature without invasion into local structures.

Additional factors influencing surgical resectability included; ability to obtain adequate intra-oral exposure to allow for robotic access, feasibility and safety for a patient to stop anti-platelet or anti-coagulant medication for a period of four weeks, and if the patient was deemed medically fit to undergo a general anaesthetic. Patients were not offered surgery as a treatment option if there was radiological evidence of >N1 disease [as per American Joint Committee on Cancer 8th edition (AJCC8) staging system], tumour extension into the soft palate, extra-nodal extension, or metastatic disease. Additional exclusion criteria comprised of patients who presented with carcinoma of unknown primary and patients who had a history of previous head and neck cancer treatment.

Clinical and pathologic features

Tumour staging adhered to the revised 8th edition of the AJCC staging system for malignant head and neck tumours (14). Collected data included age at time of diagnosis, sex, HPV status (by both p16 staining and HPV ISH testing), smoking (defined as never, ≤10 pack years, or >10 pack years) and alcohol consumption (grams per day), as well as clinical and pathologic tumor-node-metastasis (TNM) classification from MDT discussions. Data from final histopathology reports collected included closest mucosal resection margins, number and locations of involved lymph nodes, and presence of high-risk pathological features such as perineural invasion (PNI), lymphovascular invasion (LVI), and extranodal extension (ENE).

Surgical treatment

Patients meeting inclusion criteria specified were treated surgically by the senior author utilising the Da Vinci^® Surgical Platform Xi system, with the oropharynx exposed using the Feyh-Kastenbauer-WeinsteinO’Malley (FK-WO) retractor. Tumour resection involved en-bloc removal with additional margins in concerning areas (guided by macroscopic assessment by the surgeon +/– confirmation through intraoperative frozen section) removed at time of surgery and sent as additional specimens to assess clearance. Neck dissection was performed either concurrently or within two weeks of the primary tumour resection depending on access to the surgical robot and theatre time availability. Patients with a lateralised primary tonsil tumour underwent a selective neck dissection (SND) involving removal of levels II–IV while patients who had a base of tongue primary underwent bilateral SND. Prior to a patient developing an oropharyngeal fistula as a complication, the N+ neck was treated with a modified radical neck dissection (MRND) including levels I–V. Each nodal level underwent separate histopathological examination. The lingual and facial artery branches of the external carotid were ligated at the time of the neck dissection in all patients.

Adjuvant therapy

Surgical histopathology reports were reported by a dedicated head and neck pathologist and reviewed at a MDT to determine the need for adjuvant therapy. Indications for adjuvant therapy were aligned with NCCN guidelines and based on high-risk features on histopathology. Indications for adjuvant radiotherapy included patients with tumours demonstrating PNI, LVI, greater than two involved lymph nodes, or one node greater than 3 cm in size. A relative indication for adjuvant radiotherapy was close surgical margins (defined as tumour presence <2 mm from mucosal edge). Concurrent chemotherapy (chemoradiotherapy) was indicated in cases of positive margins or ENE (15). The radiotherapy dose and fields were individualised for each patient. Involved nodal levels received 60 Gy in 30 fractions delivered over 6 weeks, whereas uninvolved levels received 54 Gy. If the neck was pN0, radiotherapy was delivered to the primary site alone. If the primary site required treatment, it received 60 Gy in 30 fractions. Similarly, if the primary tumour had clear margins and no adverse features, adjuvant radiotherapy targeted areas in the neck in the presence of pathological upstaging of the radiologically assessed neck disease.

Follow-up

Long-term follow-up was scheduled at three-monthly intervals for the first two years, fourth monthly in the third year, and six monthly thereafter to five years. A PET with diagnostic CT neck and chest scan was performed at 3 months post completion of treatment, again at 12 months post treatment, and then yearly until 5 years as routine onco-surveillance.

Outcomes

Primary outcomes

The primary outcome was to assess complete resection of the tumour through the absence of positive margins. A positive margin was defined as presence of tumour at specimen mucosal edge on final histopathology report. In cases where the main specimen indicated tumour at mucosal edge, but there was a further separate margin sent that correlated with the area of concern and was negative, this was considered an overall negative margin. Additional primary outcomes included biopsy proven disease recurrence, and survival analysis.

Secondary outcomes

The main secondary outcome was to determine TORS-related complications. Post-op haemorrhage was recorded based on the Mayo Clinic staging system (16), and defined as a primary haemorrhage if bleeding occurred within 24 hours of surgery, or secondary haemorrhage if bleeding occurred greater than 24 hours following surgery. Additional TORS-related complications recorded were salivary fistula necessitating surgical intervention, and dysphagia requiring PEG insertion.

Data relating to complications arising post neck dissection was recorded and included; neck haematoma requiring surgical evacuation, chyle leak detected by positive lipids and chylomicrons result on neck fluid sample, seroma requiring bed-side or surgical drainage, and neck collection requiring intravenous antibiotics and drainage.

Statistical analysis

Statistical analyses were performed using SPSS for Windows version 25 (IBM Corporation, New York, USA). Descriptive analysis was used to summarise patient and tumour characteristics (age at operation, gender, tumour site, staging, smoking and alcohol consumption, margin status, tumour features, complications, use of adjuvant therapy). Time-to-event data (disease specific survival, overall survival, time to recurrence) were calculated from the date of TORS surgery and described using the Kaplan-Meier method.

Results

Patient characteristics

Patient demographics, summarised in Table 1, include 41 patients in the final analysis with the majority being male (n=38, 92.7%). The average age at diagnosis was 58.1 (SD 9.1) years. The median follow-up duration was 51 [interquartile range (IQR) 24] months. Further patient demographic details are provided in Table 1.

Tumour characteristics and treatment data

Tumour characteristic and treatment details are presented in Table 2. Majority of patients had a tonsillar primary (n=34, 82.9%) and were cT1N1 at time of diagnosis. Final histopathological analysis demonstrated a slight upstaging of disease to pT2N1. Most tumours (n=35, 85.4%) were both p16 and HPV ISH positive, with only 6 patients lacking ISH testing. Most patients (n=21, 51.2%) underwent a selective neck dissection (SND). The average lymph node yield for this cohort was 36.6 (95% CI: 34.1–39.1) for the ipsilateral SND, with an average of 0.8 involved lymph node (95% CI: 0.4–1.2).

High risk pathological features were detected in 10 (24.4%) patients with PNI, LVI, and ENE identified in 1 (2.4%), 5 (12.2%), and 4 (9.8%) patients, respectively. Adjuvant therapy was indicated in 15 patients (36.6%). Among them, 4 (9.8%) received adjuvant radiotherapy (RT) exclusively for a close mucosal margin defined <2 mm from mucosal edge. After MDT discussion, there were an additional 3 patients (7.3%) who received adjuvant RT despite no poor prognostic features on histopathology but because risk of recurrence was deemed higher due to immunosuppression (transplant or HIV positive patients). RT was provided to 17 (41.4%) patients in total, and chemoradiotherapy (CRT) was provided to one (2.4%) patient for ENE. There were three patients that did not receive adjuvant chemotherapy as indicated (for ENE), either for medical reasons or due to patient preference. Additional tumour clinical and pathological characteristics are summarised in Table 2.

Primary outcomes: oncological outcomes

The positive margin rate was 4.9% (n=2), leading to further surgical margin excision in both patients that ultimately returned negative margins without subsequent need for adjuvant therapy. The disease specific-survival and overall survival at 3- and 5-year was 100% and 95.1%, respectively. In total, 3 patients (7.3%) experienced disease recurrence. The locoregional recurrence rate (LRR) was 4.9% (n=2), of which one subsequently died from both locoregionally advanced disease and further metastatic disease recurrence. This patient had no poor prognostic features on their initial histopathology, surgical margins were >4 mm and they had one positive node <3 cm but developed an early locoregional recurrence 9 months post-surgery. They underwent combined chemoradiotherapy but again developed locoregional recurrence and distant metastatic disease soon after completion of their second definitive treatment. The second locoregional recurrence was identified in the parapharyngeal space on routine oncologic surveillance four years post-operatively. The initial histopathology showed close surgical margins and the MDT recommendation had been for adjuvant radiotherapy which the patient declined. They underwent combined chemoradiotherapy and remain free of disease at last follow-up. One patient (2.4%) had isolated distant metastatic disease recurrence with subsequent mortality Table 3. summarises the location of disease recurrence, treatment, and outcome. Figure 1 demonstrates survival analysis with regards to disease-specific survival (A) and recurrence free survival (B).

Figure 1 Kaplan-Meier graphs demonstrating disease-specific survival (A) and recurrence free survival (B) of patients with HPVOPSCC undergoing TORS. HPVOPSCC, human papillomavirus-related oropharyngeal squamous cell carcinoma; TORS, transoral robotic surgery.

Secondary outcome: complications

TORS related complications occurred in 5 patients (12.1%). There were 2 patients (4.9%) that had a secondary post-TORS haemorrhage defined as bleeding occurring greater than 24 hours following surgery. One patient (2.4%) had a minor bleed managed conservatively. Another patient (2.4%) had an intermediate haemorrhage requiring operative intervention. TORS-related swallowing dysfunction occurred in two patients (4.9%) who required a short-term PEG for less than 12 months. Salivary fistula occurred in one patient (3.1%) that required return to the operating room for local flap repair. Neck dissection related complications occurred in 5 patients (12.1%). Two patients experienced a chyle leak requiring operative intervention, and one patient had a neck haematoma that required surgical evacuation. One patient required on ward management for serial drainage of a seroma. A further patient required delayed readmission for needle aspiration of an infected collection with intravenous antibiotics therapy.

Discussion

This study adds to the available Australian literature demonstrating oncological outcomes post robotic resection specifically for HPVOPSCC. We demonstrate excellent disease specific survival rates of 100% at 3 years, and 95.1% at 5 years with low rates of LRR (4.9%). Our recurrence free survival remained high at 92.7%. Our results align with larger cohort studies utilising primary TORS for HPVOPSCC. Brody et al. established international benchmarks by analysing 634 patients over a 10-year period and demonstrated a 5-year OS rate of 91.2%, recurrence free survival of 86.1%, and a LRR of 7.5% (17). Similarly, Moore et al. conducted an important study of over 300 patients undergoing TORS and adjuvant therapy for OPSCC and demonstrated disease specific survival rates of 96% at 3 years, and 86% at 5 years, however this was not HPV specific (18). Other studies, including a systematic review on TORS for early-stage OPSCC, reported comparable OS and disease-free survival rates of 95% and 90%, respectively (8,19). We present figures paralleled to other institutions, with maintenance of high disease specific survival rates and low rates of recurrence.

The aim of surgical management of HPVOPSCC through a transoral approach is to excise the primary tumour en-bloc with clear margins. Presence of positive margins has been shown to be a risk factor for distant metastatic disease recurrence and subsequent disease related mortality (17). Our positive margin rate on histopathology analysis was 4.9% (n=2), with both patients having further surgical margin excision, and ultimately returning negative margins. This permitted both patients to avoid adjuvant therapy, and neither patient has experienced disease recurrence. One patient, during surgery was noted to have early invasion of the medial pterygoid upstaging them to T4 disease. The en-bloc resection margin was positive but the further cuff of medical pterygoid resected at the time of surgery was negative for disease. This patient was classified as a close rather than involved margin and underwent adjuvant radiotherapy. This is comparable to the lower range of the positive margin rate ranging from 0–26% in the existing literature (8,17,19-21).

An additional benefit of TORS-based therapy is the opportunity to de-escalate adjuvant therapy and reduce radiotherapy dosage to areas of lower risk. Particularly, as the population facing HPVOPSCC are younger with longer life expectancies, long-term toxic side effects can be debilitating and severely impact quality of life (4,17,19). TORS is often criticised when a large proportion of patients require adjuvant treatment, especially with CRT (tri-modality treatment). In our study, 17 patients (41.4.%) underwent adjuvant RT with only one patient (2.4%) undergoing CRT. Therefore, most patients were able to avoid adjuvant treatment and associated morbidity which is significantly lower than larger cohort studies that indicate varying rates adjuvant therapy application (RT: 40–70%; CRT: 24–57%) (8,12,17). Results from clinical trials such as the E3311 and PATHOS study will facilitate more consistent adjuvant treatment strategies with the aim of maintaining high survival rates and improved quality of life (22,23).

TORS-related complications are a significant concern, particularly life-threatening haemorrhage. Our study shows a complication rate of 12.1% (n=5), in the lower range of 5–30% reported in previous studies (6,10,24). We had two patients (4.9%) with a secondary post-operative haemorrhage with one requiring return to the operating theatre for arrest of haemorrhage. This compares favourably to a range of 0–10.4% reported in the literature (6,24). Structural and functional swallowing dysfunction is another concern with TORS-related resection and can require short-term or long-term gastrostomy tube insertion for maintenance of nutrition. There were 2 patients (4.9%) that required this intervention in our study for a short period of time (<12 months). This is in the lower range when compared to similar cohort studies (6,10,13). We anticipate with meticulous patient selection and surgeon experience, TORS-related complications will continue to be a rare occurrence.

This study is mainly limited by its retrospective design and relatively small sample size. Furthermore, as there was a single surgeon performing all TORS procedures, there is both an element of reproducibility and reliability, but also a possible selection bias. Disease-related mortality was rare (n=2), making meaningful associations with regards to smoking status, and recurrence on survival analysis difficult and therefore excluded from this study. Additionally, functional outcomes such as evaluation of swallowing function and quality of life scores were not routinely recorded in our database and could not be included within this study.

Conclusions

TORS-based treatment in Australia for HPVOPSCC can be an effective and safe treatment modality. This study has demonstrated excellent oncological outcomes that are comparable to previously established international benchmarks.

Acknowledgments

Funding: None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://www.theajo.com/article/view/10.21037/10.21037/ajo-24-15/rc

Data Sharing Statement: Available at https://www.theajo.com/article/view/10.21037/ajo-24-15/dss

Peer Review File: Available at https://www.theajo.com/article/view/10.21037/ajo-24-15/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://www.theajo.com/article/view/10.21037/ajo-24-15/coif). J.C. serves as an unpaid editorial board member of the Australian Journal of Otolaryngology from January 2019 to December 2024. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Research Ethics Governance Information System (2021/ETH00217). Due to the retrospective nature of the research, the requirement for informed consent was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Morris G, Shukla S, National Centre for Monitoring Cancer (Australia), Australian Institute of Health and Welfare. Head and neck cancers in Australia. Pg 49. Available online: https://www.aihw.gov.au/reports/cancer/head-neck-cancers-australia
2. Floros P, Rao A, McCloy RA, et al. Crawford JA, Gallagher RM. Altered presentation of oropharyngeal cancer, a 6-year review. ANZ J Surg 2021;91:1240-5. [Crossref] [PubMed]
3. Vidal L, Gillison ML. Human papillomavirus in HNSCC: recognition of a distinct disease type. Hematol Oncol Clin North Am 2008;22:1125-42. vii. [Crossref] [PubMed]
4. Sansoni ER, Gross ND. The role of transoral robotic surgery in the management of oropharyngeal squamous cell carcinoma: a current review. Curr Oncol Rep 2015;17:432. [Crossref] [PubMed]
5. Lörincz BB, Jowett N, Knecht R. Decision management in transoral robotic surgery: Indications, individual patient selection, and role in the multidisciplinary treatment for head and neck cancer from a European perspective. Head Neck 2016;38:E2190-6. [Crossref] [PubMed]
6. Weinstein GS, O'Malley BW Jr, Magnuson JS, et al. Transoral robotic surgery: a multicenter study to assess feasibility, safety, and surgical margins. Laryngoscope 2012;122:1701-7. [Crossref] [PubMed]
7. Gross ND, Holsinger FC, Magnuson JS, et al. Robotics in otolaryngology and head and neck surgery: Recommendations for training and credentialing: A report of the 2015 AHNS education committee, AAO-HNS robotic task force and AAO-HNS sleep disorders committee. Head Neck 2016;38:E151-8. [Crossref] [PubMed]
8. Mahmoud O, Sung K, Civantos FJ, et al. Transoral robotic surgery for oropharyngeal squamous cell carcinoma in the era of human papillomavirus. Head Neck 2018;40:710-21. [Crossref] [PubMed]
9. Huang J, Phillips N, Nightingale J, et al. Trans-oral robotic surgery: a safe and effective tool in head and neck surgery in an Australian rural setting. ANZ J Surg 2021;91:2345-51. [Crossref] [PubMed]
10. Hirshoren N, Ruskin O, Fua T, et al. Transoral robotic surgery: implementation as a tool in head and neck surgery - a single-centre Australian experience. ANZ J Surg 2018;88:1129-34. [Crossref] [PubMed]
11. Krishnan G, Mintz J, Foreman A, et al. The acceptance and adoption of transoral robotic surgery in Australia and New Zealand. J Robot Surg 2019;13:301-7. [Crossref] [PubMed]
12. Nichols AC, Theurer J, Prisman E, et al. Radiotherapy versus transoral robotic surgery and neck dissection for oropharyngeal squamous cell carcinoma (ORATOR): an open-label, phase 2, randomised trial. Lancet Oncol 2019;20:1349-59. [Crossref] [PubMed]
13. Dowthwaite S, Panizza B, O’Neill J, et al. Outcomes of primary trans-oral surgical management of early tonsillar squamous cell carcinoma with risk-adapted adjuvant radiotherapy. Aust J Otolaryngol 2019;2:7. [Crossref]
14. Amin M, Edge S, Green F, et al. AJCC Cancer Staging Manual 8th ed. New York: Springer. 2017.
15. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines®) Head and Neck Cancers. Version 2.2022. 2022. Available online: https://www.nccn.org/professionals/physician_gls/pdf/head-and-neck.pdf
16. Pollei TR, Hinni ML, Moore EJ, et al. Analysis of postoperative bleeding and risk factors in transoral surgery of the oropharynx. JAMA Otolaryngol Head Neck Surg 2013;139:1212-8. [Crossref] [PubMed]
17. Brody RM, Shimunov D, Cohen RB, et al. A benchmark for oncologic outcomes and model for lethal recurrence risk after transoral robotic resection of HPV-related oropharyngeal cancers. Oral Oncol 2022;127:105798. [Crossref] [PubMed]
18. Moore EJ, Van Abel KM, Price DL, et al. Transoral robotic surgery for oropharyngeal carcinoma: Surgical margins and oncologic outcomes. Head Neck 2018;40:747-55. [Crossref] [PubMed]
19. Kelly K, Johnson-Obaseki S, Lumingu J, et al. Oncologic, functional and surgical outcomes of primary Transoral Robotic Surgery for early squamous cell cancer of the oropharynx: a systematic review. Oral Oncol 2014;50:696-703. [Crossref] [PubMed]
20. Funk RK, Moore EJ, García JJ, et al. Risk factors for locoregional relapse after transoral robotic surgery for human papillomavirus-related oropharyngeal squamous cell carcinoma. Head Neck 2016;38:E1674-9. [Crossref] [PubMed]
21. Yver CM, Shimunov D, Weinstein GS, et al. Oncologic and survival outcomes for resectable locally-advanced HPV-related oropharyngeal cancer treated with transoral robotic surgery. Oral Oncol 2021;118:105307. [Crossref] [PubMed]
22. Owadally W, Hurt C, Timmins H, et al. PATHOS: a phase II/III trial of risk-stratified, reduced intensity adjuvant treatment in patients undergoing transoral surgery for Human papillomavirus (HPV) positive oropharyngeal cancer. BMC Cancer 2015;15:602. [Crossref] [PubMed]
23. Ferris R, Flamand Y, Weinstein G, et al. Phase II Randomized Trial of Transoral Surgery and Low-Dose Intensity Modulated Radiation Therapy in Resectable p161 Locally Advanced Oropharynx Cancer: An ECOG-ACRIN Cancer Research Group Trial (E3311). J Clin Oncol 2021;40:138-49. [Crossref] [PubMed]
24. Chia SH, Gross ND, Richmon JD. Surgeon experience and complications with Transoral Robotic Surgery (TORS). Otolaryngol Head Neck Surg 2013;149:885-92. [Crossref] [PubMed]