Abstract
Anaplastic lymphoma kinase (alk) is an oncogenic driver in non-small-cell lung cancer (nsclc). Chromosomal rearrangements involving the ALK gene occur in up to 4% of nonsquamous nsclc patients and lead to constitutive activation of the alk signalling pathway. ALK-positive nsclc is found in relatively young patients, with a median age of 50 years. Patients frequently have brain metastasis.
Targeted inhibition of the alk pathway prolongs progression-free survival in patients with ALK-positive advanced nsclc. The results of several recent clinical trials confirm the efficacy and safety benefit of crizotinib and ceritinib in this population.
Canadian oncologists support the following consensus statement: All patients with advanced nonsquamous nsclc (excluding pure neuroendocrine carcinoma) should be tested for the presence of an ALK rearrangement. If an ALK rearrangement is present, treatment with a targeted alk inhibitor in the first-line setting is recommended. As patients become resistant to first-generation alk inhibitors, other treatments, including second-generation alk inhibitors can be considered.
Keywords: ALK, anaplastic lymphoma kinase, non-small-cell lung cancer, molecular testing, targeted inhibition
INTRODUCTION
Standard first-line systemic treatment for advanced non-small-cell lung cancer (nsclc) is chemotherapy with a 2-drug combination including a platinum compound and a non-platinum drug. Response rates range from 20% to 40%, and median time to progression is 4–5 months1.
Decades of research have revealed a pivotal role for tyrosine kinases as key regulators of the signalling pathways that control cell growth and differentiation. Deregulation of tyrosine kinase–mediated signalling occurs frequently in cancer and is believed to drive disease initiation and progression.
Chromosomal rearrangements involving the tyro-sine kinase anaplastic lymphoma kinase (ALK) gene in nsclc were first described in 20072. The most common is a rearrangement resulting in a small inversion within chromosome 2p, leading to the expression of a chimeric tyrosine kinase in which the N-terminal half of the echinoderm microtubule-associated protein-like 4 (EML4) is fused to the intracellular kinase domain of ALK. EML4-ALK has active oncogenic properties.
ALK is a hotspot for other translocation events, and many variations of the ALK rearrangements exist. ALK can also be altered through point mutations in the kinase domain and by overexpression of the full-length protein3,4. ALK rearrangements occur in up to 4% of nsclc patients4,5. ALK rearrangements and point mutations are present in many other cancer types as well3,4,6,7. The physiologic role played by ALK has not yet been well defined.
Patients with nsclc whose tumours harbour ALK rearrangements (ALK-positive) are typically younger than those without this rearrangement and tend to be light or never-smokers4. Central nervous system metastases are the most common site of progression in ALK-positive nsclc patients; such metastases occur in approximately 50% of patients8.
Several oral small-molecule tyrosine kinase inhibitors of alk have been developed. Crizotinib is a first-generation inhibitor of surface membrane receptor tyrosine kinases including alk, hepatocyte growth factor receptor, and ros (encoded by the ROS1 gene)9. ALK-positive patients eventually acquire resistance to crizotinib through a variety of molecular mechanisms, including secondary mutations in the ALK tyrosine kinase domain, ALK gene amplification, and activation of other signalling pathways10,11. Ceritinib is a second-generation atp-competitive tyrosine kinase inhibitor of alk9.
ALK TESTING
Detection of tumours harbouring ALK rearrangements is necessary to select patients for treatment with alk inhibitors. Testing for ALK gene rearrangements should be performed in all patients eligible for targeted therapy at time of diagnosis of advanced incurable nonsquamous nsclc in which a component of adenocarcinoma is noted or suspected12. Eligible pathologic diagnoses include adenocarcinoma, large-cell carcinoma, sarcomatoid carcinoma, adenosquamous carcinoma, and nsclc not otherwise specified. Not recommended for routine testing are “pure” squamous cell, small-cell, and other neuroendocrine carcinomasb (for example, large-cell neuroendocrine carcinoma). ALK testing can be considered for atypical cases such as non-smoking patients with squamous cell carcinoma. If possible, testing should be conducted before chemotherapy is initiated. Switching therapies can be considered if an ALK rearrangement is identified after chemotherapy is initiated. A tissue sample from the primary tumour or a metastasis is equally suitable for the analysis.
Until recently, the standard methodology for detection of ALK gene rearrangements in the United States was fluorescence in situ hybridization (fish) using the U.S. Food and Drug Administration–approved Vysis ALK Break Apart FISH Probe Kit (Abbott Molecular, Abbott Park, IL, U.S.A.). That test is both costly and challenging to implement across molecular pathology laboratories. The Food and Drug Administration subsequently approved an ALK immunohistochemistry (ihc) assay [Ventana ALK (D5F3) CDx Assay: Ventana Medical Systems (Roche Group), Tucson, AZ, U.S.A.]. A network of pulmonary and molecular pathologists working in cancer centres across Canada addressed the challenge of standardization and optimization of detection tests for ALK-positive nsclc and supports the use of an appropriately validated ihc assay to screen for ALK- positive nsclc17. In many Canadian centres, ALK-positive and -equivocal cases by ihc are subsequently tested by ALK fish. Positive ALK ihc is sufficient for obtaining access to alk inhibitors in some Canadian provinces; however, confirmation of weak ihc staining by fish should remain the standard practice. ALK ihc therefore requires high levels of reliability and ongoing external quality assurance; its implementation in pathology laboratories should follow strict validation standards.
EVIDENCE DEMONSTRATING THE BENEFITS OF TARGETED INHIBITION OF ALK-POSITIVE TUMOURS
PROFILE 1001
The phase i profile 1001 trial evaluated single-agent crizotinib in the treatment of locally advanced or metastatic ALK-positive nsclc18, showing that crizotinib is well-tolerated, with rapid, durable responses in patients with ALK-positive nsclc.
Of 143 ALK-positive patients, 87 experienced an objective response [60.8%; 95% confidence interval (ci): 52.3% to 68.9%]. Median time to objective response was 7.9 weeks (range: 2.1–39.6 weeks), and median duration of response was 49.1 weeks (95% ci: 39.3 to 75.4 weeks). Overall, median progression-free survival (pfs) was 9.7 months (95% ci: 7.7 to 12.8 months). It was 18.3 months (95% ci: 8.3 months to not reached) in the 24 patients receiving first-line crizotinib, and 9.2 months (95% ci: 7.3 to 12.7 months) in the 125 patients receiving crizotinib as second-line or later treatment18.
Treatment-related adverse events, mostly grade 1 or 2, occurred in 97% of patients overall. The most common treatment-related grades 1 and 2 adverse events were visual effects, nausea, diarrhea, constipation, vomiting, and peripheral edema. The most common treatment-related grades 3 and 4 adverse events were neutropenia (n = 9, 6%), elevated alanine aminotransferase (n = 6, 4%), hypophosphatemia (n = 6, 4%), and lymphopenia (n = 6, 4%)18.
PROFILE 1007
The multicentre, randomized phase iii efficacy and safety study profile 1007 compared crizotinib with standard-of-care chemotherapy in patients with platinum pre-treated ALK-positive advanced nsclc19. This second-line trial confirmed the efficacy and safety benefits of crizotinib for the treatment of patients with ALK-positive disease in that setting.
The 347 participating patients were randomized 1:1 to crizotinib (n = 173) or to pemetrexed or docetaxel (n = 174). Patients continued on treatment until progressive disease (independent radiology review), unacceptable toxicity, or withdrawal of consent. The primary endpoint was pfs. The independent radiology review used the Response Evaluation Criteria in Solid Tumors to determine pfs. At the discretion of the investigator, patients could continue treatment beyond progression if they were experiencing clinical benefit.
Median pfs was 7.7 months for patients randomized to crizotinib and 3.0 months for patients randomized to chemotherapy [hazard ratio (hr): 0.49; p < 0.0001]. The overall response rate for patients on crizotinib was 65% (95% ci: 58% to 72%) compared with 20% (95% ci: 14% to 26%) for chemotherapy (odds ratio: 3.39; p < 0.0001). The median overall survival (os) was 20.3 months for crizotinib and 22.8 months for chemotherapy (hr: 1.02; p = 0.5394), but was confounded by crossover, because 108 patients on the chemotherapy arm (62%) crossed over to receive crizotinib in a separate trial (A8081005)19.
No new or unexpected adverse events were recorded. Patient-reported outcomes and quality of life favoured crizotinib19.
PROFILE 1014
The phase iii profile 1014 trial compared first-line crizotinib with chemotherapy in 343 patients with advanced ALK-positive nonsquamous nsclc who had received no prior systemic treatment for advanced disease20, confirming that crizotinib is superior to chemotherapy in patients with previously untreated ALK-positive advanced nsclc.
Patients were randomly assigned to receive crizotinib or intravenous chemotherapy with cisplatin or carboplatin and pemetrexed. The primary endpoint was pfs as assessed by independent radiologic review. Median pfs with crizotinib, at 10.9 months, was significantly longer than it was with chemotherapy (7.0 months; hr: 0.45; 95% ci: 0.35 to 0.60; p < 0.001). Objective response rates were 74% for crizotinib and 45% for chemotherapy (p < 0.001). Median os was not reached in either group because of crossover (hr: 0.82; 95% ci: 0.54 to 1.26; p = 0.36)20.
No new or unexpected adverse events were recorded. Crizotinib was associated with a greater reduction in lung cancer symptoms and a greater improvement in quality of life.
TREATMENT BEYOND PROGRESSION
When patients develop progressive disease on crizotinib, either systematically or to brain, consideration should be given to treating oligometastases with local surgery or radiation, and continuing patients on an alk inhibitor if they are experiencing continued clinical benefit—particularly if most of the disease remains under good control.
A retrospective analysis of the effect of continued alk inhibition with crizotinib beyond progression in patients with advanced ALK-positive nsclc observed a survival benefit21. Of the 194 crizotinib-treated patients, 120 continued crizotinib beyond progression. Those patients experienced a significantly longer median os (29.6 months from starting crizotinib) than the 10.8 months experienced by those who did not continue treatment after progression (hr: 0.30; 95% ci: 0.19 to 0.46; p < 0.0001). Patients who were treated beyond progression had a significantly longer median os from the time of disease progression: 16.4 months versus 3.9 months for those who stopped crizotinib at progression (hr: 0.27; 95% ci: 0.17 to 0.42; p < 0.0001)21. This finding might reflect disease biology, but is still an important concept applicable in this patient setting.
ASCEND-1, -2, AND -3
The pivotal phase i ascend-1 trial, evaluating the efficacy and safety of ceritinib, was conducted in 20 centres in 11 countries22. Of the 246 participating patients with advanced ALK-positive nsclc, 163 had already received an alk inhibitor, and 83 were alk-inhibitor naïve. The response rate for all patients was 61.8%; it was 56.4% for pre-treated patients, and 72.3% for inhibitor-naïve patients. The median pfs for all patients in the trial was 6.9 months; for the alk-inhibitor-naïve patients, median pfs was not reached at 18 months23.
Ceritinib also demonstrated central nervous system activity. Of the 28 ascend-1 patients with measurable brain metastases at baseline, 10 (36%) experienced a partial response24. As a result of the trial, the Food and Drug Administration approved ceritinib in April 2014 for patients with advanced ALK-positive nsclc after treatment with crizotinib.
The single-arm phase ii ascend-2 trial evaluated the efficacy of ceritinib in patients with advanced ALK-positive nsclc who had progressed on both standard chemotherapy and crizotinib. It confirmed the efficacy of ceritinib, showing an overall response rate of 38.6% and a pfs of 5.7 months25. The intracranial response rate was 35%, similar to that in ascend-1.
The single-arm phase ii ascend-3 trial evaluated the efficacy of ceritinib in treatment-naïve patients with advanced ALK-positive nsclc26. The published pfs was an impressive 11.1 months, and the response rate was 36.3%. Of the 124 patients participating in ascend-3, 50 (40.3%) had brain metastases, and of those 50, 27 (54%) had received radiation26. The intracranial response was 58.8% as determined by blinded independent central review.
The ascend-1, -2, and -3 trials confirmed that ceritinib is an effective alk inhibitor and has activity against brain metastases. Ceritinib was approved by Health Canada as a second-line option for patients who have progressed after crizotinib.
OPTIMAL SEQUENCING OF THERAPY
A retrospective study evaluated a cohort of patients with advanced ALK-positive nsclc treated sequentially with both crizotinib and ceritinib. Ceritinib was associated with significant antitumour activity in a crizotinib-resistant population27. For the 73 ALK-positive patients, median pfs on crizotinib was 8.2 months (95% ci: 7.4 to 10.6 months), and median combined pfs for sequential treatment was 17.4 months (95% ci: 15.5 to 19.4 months). Although the study was retrospective and did not include a comparator arm, it sets a benchmark for future clinical trials of sequenced tyrosine kinase inhibitors.
A PROMISING FUTURE WITH AN EVOLVING TREATMENT PARADIGM
New second- and third-generation alk inhibitors will be entering the ALK-positive advanced nonsquamous nsclc treatment paradigm. As a result, patients who become resistant to first-generation alk inhibitors will have other treatment options available.
Alectinib, a second-generation alk inhibitor, has significant efficacy for treating brain metastasis. A small study demonstrated the efficacy of alectinib in 4 patients with ALK-positive nsclc and leptomeningeal metastases who had already received both crizotinib and ceritinib; median pfs was not reached at 18 months28.
Sequential treatment with multiple alk inhibitors is a possibility. We look forward to the results of randomized controlled trials that will evaluate and compare the various alk inhibitors with each other and with chemotherapy agents to help determine the most effective treatment sequences for ALK-positive patients.
RECOMMENDATION
As Canadian physicians involved in the treatment of patients with lung cancer in Canada, we recommend that all patients with advanced nsclc having the appropriate nonsquamous histologic features who are eligible for targeted therapy be tested for the presence of an ALK rearrangement. Positive alk ihc is sufficient for obtaining access to alk inhibitors in some Canadian provinces. Confirmation by fish of weak or equivocal ihc staining remains the standard practice. Treatment with crizotinib, a targeted alk inhibitor, in the first-line setting is recommended for ALK-positive nsclc. Data support the use of ceritinib for second-line treatment after progression on crizotinib.
ACKNOWLEDGMENTS
The authors thank Chrystal Palaty from Metaphase Health Research Consulting for her project management and administrative assistance in the preparation of this article. Chrystal Palaty was compensated by an unrestricted educational grant from Pfizer. During the entire process, Pfizer did not influence the article in any way and neither read nor reviewed the article. None of the authors were compensated for their work on this article.
Footnotes
CONFLICT OF INTEREST DISCLOSURES
We have read and understood Current Oncology’s policy on disclosing conflicts of interest, and we declare the following interests: BM, RA, SB, NB, DNI, RJ, and WM serve on advisory boards for Novartis, Pfizer, and Roche; JA has given talks for, or served on advisory boards for, AstraZeneca, Boehringer Ingelheim, Bristol–Meyers Squibb, Merck, Novartis, and Pfizer; DB serves on an advisory board for Pfizer; CB serves on advisory boards for AstraZeneca, Boehringer Ingelheim, Bristol–Meyers Squibb, Merck, and Pfizer; JD has served or serves on advisory boards for AstraZeneca, Merck, Pfizer, and Roche; SAL has received honoraria from both Novartis and Pfizer; GL serves on the advisory boards of AstraZeneca, Novartis, Pfizer, Roche, and Takeda; MST received honoraria and research grants from Pfizer Canada; VH serves on advisory boards for Amgen, AstraZeneca, Boehringer Ingelheim, Bristol–Meyers Squibb, Eli Lilly, Merck, Novartis, Roche, and Pfizer. DGB, PCheema, PCheung, VC, and SKR declare that they have no conflicts of interest.
REFERENCES
- 1.Schiller JH, Harrington D, Belani CP, et al. Comparison of four chemotherapy regimens for advanced non-small-cell lung cancer. N Engl J Med. 2002;346:92–8. doi: 10.1056/NEJMoa011954. [DOI] [PubMed] [Google Scholar]
- 2.Soda M, Choi YL, Enomoto M, et al. Identification of the transforming EML4-ALK fusion gene in non-small-cell lung cancer. Nature. 2007;448:561–6. doi: 10.1038/nature05945. [DOI] [PubMed] [Google Scholar]
- 3.Shaw AT, Solomon B. Targeting anaplastic lymphoma kinase in lung cancer. Clin Cancer Res. 2011;17:2081–6. doi: 10.1158/1078-0432.CCR-10-1591. [DOI] [PubMed] [Google Scholar]
- 4.Hallberg B, Palmer RH. Mechanistic insight into alk receptor tyrosine kinase in human cancer biology. Nat Rev Cancer. 2013;13:685–700. doi: 10.1038/nrc3580. [Erratum in: Nat Rev Cancer 2013;13:820] [DOI] [PubMed] [Google Scholar]
- 5.Koivunen JP, Mermel C, Zejnullahu K, et al. EML4-ALK fusion gene and efficacy of an alk kinase inhibitor in lung cancer. Clin Cancer Res. 2008;14:4275–83. doi: 10.1158/1078-0432.CCR-08-0168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Murugan AK, Xing M. Anaplastic thyroid cancers harbor novel oncogenic mutations of the ALK gene. Cancer Res. 2011;71:4403–11. doi: 10.1158/0008-5472.CAN-10-4041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ogawa S, Takita J, Sanada M, Hayashi Y. Oncogenic mutations of ALK in neuroblastoma. Cancer Sci. 2011;102:302–8. doi: 10.1111/j.1349-7006.2010.01825.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Rangachari D, Yamaguchi N, VanderLaan PA, et al. Brain metastases in patients with EGFR-mutated or ALK-rearranged non-small-cell lung cancers. Lung Cancer. 2015;88:108–11. doi: 10.1016/j.lungcan.2015.01.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Shaw AT, Kim DW, Mehra R, et al. Ceritinib in ALK-rearranged non-small-cell lung cancer. N Engl J Med. 2014;370:1189–97. doi: 10.1056/NEJMoa1311107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Katayama R, Shaw AT, Khan TM, et al. Mechanisms of acquired crizotinib resistance in ALK-rearranged lung cancers. Sci Transl Med. 2012;4:120ra17. doi: 10.1126/scitranslmed.3003316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Doebele RC, Pilling AB, Aisner DL, et al. Mechanisms of resistance to crizotinib in patients with ALK gene rearranged non-small cell lung cancer. Clin Cancer Res. 2012;18:1472–82. doi: 10.1158/1078-0432.CCR-11-2906. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lindeman NI, Cagle PT, Beasley MB, et al. Molecular testing guideline for selection of lung cancer patients for egfr and alk tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology. Arch Pathol Lab Med. 2013;137:828–60. doi: 10.5858/arpa.2012-0720-OA. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Aisner DL, Marshall CB. Molecular pathology of non-small cell lung cancer: a practical guide. Am J Clin Pathol. 2012;138:332–46. doi: 10.1309/AJCPFR12WJKCEEZZ. [DOI] [PubMed] [Google Scholar]
- 14.Nakamura H, Tsuta K, Yoshida A, et al. Aberrant anaplastic lymphoma kinase expression in high-grade pulmonary neuroendocrine carcinoma. J Clin Pathol. 2013;66:705–7. doi: 10.1136/jclinpath-2012-201329. [DOI] [PubMed] [Google Scholar]
- 15.Ionescu DN. Impact of the College of American Pathologists, the International Association for the Study of Lung Cancer, and the Association for Molecular Pathology clinical practice guidelines for EGFR and ALK testing in lung cancer in Canada. Curr Oncol. 2013;20:220–6. doi: 10.3747/co.20.1568. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Montagut C, Arriola E, Galvan AB, et al. ALK chromosomal alterations in neuroendocrine tumors [abstract 10585] J Clin Oncol. 2011;29 [Available online at: http://meetinglibrary.asco.org/content/81898-102; cited 12 April 2016] [Google Scholar]
- 17.Cutz JC, Craddock KJ, Torlakovic E, et al. Canadian anaplastic lymphoma kinase study: a model for multicenter standardization and optimization of ALK testing in lung cancer. J Thorac Oncol. 2014;9:1255–63. doi: 10.1097/JTO.0000000000000239. [DOI] [PubMed] [Google Scholar]
- 18.Camidge DR, Bang YJ, Kwak EL, et al. Activity and safety of crizotinib in patients with ALK-positive non-small-cell lung cancer: updated results from a phase 1 study. Lancet Oncol. 2012;13:1011–19. doi: 10.1016/S1470-2045(12)70344-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Shaw AT, Kim DW, Nakagawa K, et al. Crizotinib versus chemotherapy in advanced ALK-positive lung cancer. N Engl J Med. 2013;368:2385–94. doi: 10.1056/NEJMoa1214886. [DOI] [PubMed] [Google Scholar]
- 20.Solomon BJ, Mok T, Kim DW, et al. First-line crizotinib versus chemotherapy in ALK-positive lung cancer. N Engl J Med. 2014;371:2167–77. doi: 10.1056/NEJMoa1408440. [Erratum in: N Engl J Med 2015;373:1582] [DOI] [PubMed] [Google Scholar]
- 21.Ou SHI, Jänne PA, Bartlett CH, et al. Clinical benefit of continuing alk inhibition with crizotinib beyond initial disease progression in patients with advanced ALK-positive nsclc. Ann Oncol. 2014;25:415–22. doi: 10.1093/annonc/mdt572. [DOI] [PubMed] [Google Scholar]
- 22.Kim DW, Mehra R, Tan DSW, et al. Ceritinib in advanced anaplastic lymphoma kinase (ALK)-rearranged (ALK+) non-small cell lung cancer (nsclc): results of the ascend-1 trial [abstract 8003^] J Clin Oncol. 2014;32:5s. doi: 10.1200/JCO.2013.49.4757. [Available online at: http://meetinglibrary.asco.org/content/128749-144; cited 1 April 2016] [DOI] [Google Scholar]
- 23.Felip E, Kim D, Mehra R, et al. Efficacy and safety of ceritinib in patients (pts) with advanced anaplastic lymphoma kinase (ALK)-rearranged (ALK+) non-small cell lung cancer (nsclc) Ann Oncol. 2014;25(suppl 4):iv426–70. [Google Scholar]
- 24.Shaw A, Mehra R, Tan DSW, et al. Ceritinib (LDK378) for the treatment of patients with ALK-rearranged (ALK+) non-small cell lung cancer (nsclc) and brain metastasis (bm) in the ascend-1 trial. Neuro Oncol. 2014;16(suppl 5):39. doi: 10.1093/neuonc/nou240.32. [DOI] [Google Scholar]
- 25.Mok T, Spigel D, Felip E, et al. ascend-2: a single-arm, open-label, multicenter phase 2 study of ceritinib in adult patients (pts) with ALK- rearranged (ALK+) non-small cell lung cancer (nsclc) previously treated with chemotherapy and crizotinib (crz) [abstract 8059] J Clin Oncol. 2015;33 [Available online at: http://meetinglibrary.asco.org/content/144861-156; cited 1 April 2016] [Google Scholar]
- 26.Felip E, Orlov S, Park K, et al. ascend-3: a single-arm, open-label, multicenter phase ii study of ceritinib in alki-naive adult patients (pts) with ALK- rearranged (ALK+) non-small cell lung cancer (nsclc) [abstract 8060] J Clin Oncol. 2015;33 [Available online at: http://meetinglibrary.asco.org/content/145907-156; cited 1 April 2016] [Google Scholar]
- 27.Gainor JF, Tan DS, De Pas T, et al. Progression-free and overall survival in ALK-positive nsclc patients treated with sequential crizotinib and ceritinib. Clin Cancer Res. 2015;21:2745–52. doi: 10.1158/1078-0432.CCR-14-3009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Gainor JF, Sherman CA, Willoughby K, et al. Alectinib salvages cns relapses in ALK-positive lung cancer patients previously treated with crizotinib and ceritinib. J Thorac Oncol. 2015;10:232–6. doi: 10.1097/JTO.0000000000000455. [DOI] [PMC free article] [PubMed] [Google Scholar]