LAPAROSCOPIC AND ROBOTIC UROLOGY The Learning Curve Does Not Affect Positive Surgical Margin Status in Robot-Assisted Laparoscopic Prostatectomy Ekrem Islamoglu1*, Kaan Karamık1, Cagatay Ozsoy1, Husnu Tokgoz1, Mutlu Ates1, Murat Savas1 Purpose: To assess the oncologic results of our robot-assisted laparoscopic prostatectomy (RALP) cases and in- vestigate whether the learning curve (LC) affects the oncological outcomes. Materials and Methods: Between March 2015 and September 2017, 111 patients underwent RALP by a single surgeon in our clinic. The learning curve was analyzed using the moving average method. We compared the rate of positive surgical margins(PSM) and oncological outcomes, operation times, hematocrit changes and duration of hospitalization among the patients during and after the LC. Complications were also noted according to Clavien system. Result: LC analysis using the moving average method showed that the LC stabilized between cases 51–60. So, pa- tients were classified into two groups; 1-50 cases (Group 1) and 51-111 cases (Group 2). PSM rates were 36% for group 1 and 18% for group 2, and statistically different (p = 0,032). Extracapsular invasion (ECI) was significantly higher in group 1 (56,5%) than in group 2 (29,5%) (P = 0.005). Multiple logistic regression analysis revealed that presence of ECI was an independent factor for PSM associated with the groups (OR: 2.512; 95% CI: 1.055-5.979). Both operation time and duration of postoperative hospitalization were significantly reduced from group 1 to group 2. A total of 11 patients (10%) had complications and one of them (0.9%) required surgical intervention. Conclusion: We can conclude that at least 50 RALP cases are needed to gain proficiency even for an experienced surgeon in laparoscopic radical prostatectomy. Our study demonstrates that surgeons experience can affect the perioperative variables but the LC does not affect PSM status in RALP. Keywords: learning curve; positive surgical margin; robot-assisted prostatectomy INTRODUCTION Prostate cancer (PCa) is the most commonly diag-nosed cancer in men and it's also second-ranked cancer that results in death in the United States.(1) Ac- cording to the cancer statistics collected in Turkey, it is the second most common type of cancer in 11.8% of men, in all age groups.(2) Currently, radical prostatecto- my is the only surgical treatment of localized PCa. In 1980, Walsh et al. described retropubic radical prosta- tectomy (RRP) and this procedure remained the gold standard for a long time.(3) In 1992, Schuessler et al. performed the first laparoscopic radical prostatectomy (LRP) as an alternative to RRP.(4) Despite the develop- ment of laparoscopic technique in the following years, the LRP continued to be a long and complex surgery with a steep learning curve and even surgeons with high laparoscopy skills required a series of 40-100 cases to gain mastery.(5) The difficulties of the LRP brought along different quests and after the introduction of the da Vinci robotic surgery system, Binder et al. performed the first robot-assisted laparoscopic radical prostatecto- my (RALP) in May 2000.(6) With its three-dimensional magnified vision, enhanced ergonomics, computer fil- tration of tremors and scaled-down movement with the use of an endo-wrist instrument with seven degrees of freedom of range in motion, robotic surgery has initi- 1University of Health Sciences, Antalya Training and Research Hospital, Antalya, Turkey. *Correspondence: University of Health Sciences, Antalya Training and Research Hospital, Antalya, Turkey. Orcid id: 0000-0003-0693-0666. Phone: (+90)05327115759. Fax: +90 242 2494487. E-mail: meislamoglu@gmail.com. Received February 2018 & Accepted September 2018 ated a new era of radical prostatectomy. As a matter of fact, 85% of radical prostatectomies in the US have become robot-assisted, less than a decade after its intro- duction.(7) Of course, robotic surgery has some disad- vantages; high costs, inability to understand the tissue or suture tension due to lack of tactile sensation and col- lision of robotic arms with each other or assistant port are the major ones.(8) We assessed the results of our RALP cases performed by a single surgeon in our clinic and investigated wheth- er the learning curve affects the oncological outcomes. PATIENTS AND METHODS Our robotic team consists of a console surgeon, one as- sistant surgeon, one surgical nurse, one surgical tech- nician and one circulating nurse. Before the robotic prostate surgery, the console surgeon and the assistant surgeon had gained experience in open RRP and LRP cases. Between March 2015 and September 2017, 111 patients who underwent RALP by a single surgeon in our clinic, were enrolled in the study. Preoperative clinical data; including age, serum prostate-specific antigen (PSA), biopsy Gleason score and number of cores positive, were collected. The preoperative risk was determined by D'Amico risk stratification and patients were classi- fied as low, intermediate and high risk.(9) We performed Laparoscopic and Robotic Urology 333 Vol 15 No 06 November-December 2018 334 all RALP procedures via transperitoneal approach, us- ing 6 trocar ports of a conventional 4-arm da Vinci XI robotic system, beginning with initial dissection of the seminal vesicles and the prostate in a posterior fashion as described by Zorn et al.(10) Subsequently, returning to the anterior aspect of the prostate and separating the dorsal vein complex. The neurovascular bundle (NVB) was completely released and the prostate is dissected from the bladder neck. Urethrovesical anastomosis was done continuously, using two 15 cm 3-0 V-lock sutures and 18 French Foley catheter with 10 ml balloon was inserted. Bilateral pelvic lymphadenectomy (BPLND) was per- formed in all high risk and selected intermediate-risk patients according to Briganti nomogram.(11) None of the patients received neoadjuvant hormonal therapy. Perioperative parameters were recorded such as oper- ation time, intraoperative complications and whether BPLND or NVB preservation was done. Operation time was defined as skin to skin time in minutes and includes the docking and undocking time. Postopera- tive parameters including hematocrit change, duration of hospitalization and catheter removal date were noted. Pathological outcomes included pathological Gleason score; positive surgical margin (PSM) status; extraca- psular, lymphovascular, perineural and seminal vesicle invasion; as well as lymph node positivity. In order to classify the complications after surgery, the Clavien system, which provides standardization in the literature, was used.(12) Patient data were prospectively registered in a specific database that was accessible only to authorized people. The patients who provided a written informed consent document were assured regarding the confidentiality of their data. The data were analyzed in a retrospective way to evaluate the clinical and pathological outcomes. Our study was in accordance with the Helsinki Declara- tion and did not gain ethics committee permission as it included retrospective data. Statistical analysis The learning curve evaluation parameter was operation time. The learning curve was analyzed using the mov- ing average method.(13) We decided to use the 10-case moving average as the moving averages for less than 10 cases that exhibited excessive variation. Trends in the operation time can be unclear because of differenc- es between individual cases. With the moving average method, using the mean operation times, the individual changes are removed, and trends are clarified. The pa- tients were divided into two groups, one inside and one after the learning curve. Between two groups; statistical analysis was made us- ing IBM SPSS Statistics for Windows, Version 22.0 (IBM Corp., Armonk, NY). Fisher’s exact test and Pearson chi-square analysis performed for categorical variables. The normality assumptions were controlled by the Shapiro-Wilk test. The differences between two groups were evaluated with Student’s t-test for normal- ly distributed data or Mann-Whitney U test for non-nor- mally distributed data. Data are expressed as n(%), Table 1. Comparison of demographic and clinical characteristics of both groups Group 1(1-50) Group 2 (51-111) p-value Age, mean±SD 63.5 ± 6.4 64.7 ± 6.9 0.331 PSA, median(min-max) 7.8 (3.1-53.4) 6.5 (0.3-40) 0.406 Number of cores positive, median(min-max) 3 (1-10) 3 (1-11) 0.144 Biopsy Gleason score, n(%) 4-6 32 (64) 38 (62.3) 0.981 7 15 (30) 19(31.1) 8-10 3(6) 4 (6.6) Risk group, n(%) Low 18 (45) 28 (45.9) 0.932 Intermediate 15 (37.5) 24 (39.3) High 7 (17.5) 9 (14.8) Group 1(1-50) Group 2(51-111) p-value Surgical margin status, n(%) Negative 32 (64) 50 (82) 0.032 Positive 18 (36) 11(18) Extracapsular invasion, n(%) Yes 28 (56) 18 (29.5) 0.005 No 22 (44) 43 (70.5) Lymphovascular invasion, n(%) Yes 15 (30) 11(18) 0.139 No 35 (70) 50 (82) Perineural invasion, n(%) Yes 39 (78) 45 (73.8) 0.605 No 11(22) 16 (26.2) Seminal vesicle invasion, n(%) Yes 10 (20) 11(18) 0.792 No 40 (80) 50 (82) Pathological Gleason score, n(%) 4-6 19 (38) 22 (36.1) 0.922 7 27 (54) 35 (57.4) 8-10 4 (8) 4 (6.6) Lymph node positive, n(%) Yes 5 (25) 2 (13) 0.240 No 15 (75) 13 (87) Pathological stage, n(%) pT2 27 (54) 39 (64) 0.262 pT3 23 (46) 22 (36) pT3 13/23(56%) 6/22 (27%) Biochemical recurrence, n(%) Yes 6 (12) 4 (6.6) 0.999 No 44 (88) 57 (93.4) Duration of follow-up, months 15 (12-33) 6 (3-12) < 0.001 Additional treatment, n(%) 11(22) 6 (9.8) 0.153 Table 2.Comparison of clinical and pathological outcomes of both groups LC does not affect PSM status in RALP-Islamoglu et al. mean ± standard deviation(SD) or median (min-max), as appropriate. P values < 0.05 were considered statis- tically significant. Multivariate analysis was performed using logistic regression. RESULTS As shown in Figure 1, learning curve analysis using the moving average method showed that the learning curve stabilized between cases 51–60. So, we divided the patients into two groups, Group 1 (cases 1-50) and Group 2 (cases 51-111). Preoperative clinical characteristics of both groups are shown in Table 1 and two groups were similar to each other as seen. The entire cohort aged 64 years on me- dian (range 44-75), with a median PSA of 7.1 ng/ml. According to D'amico risk classification and biopsy Gleason scores, there was no difference between the two groups before surgery (P > 0.05). Pathological outcomes and clinical follow-up are shown in Table 2. PSM rates were 36% for group1 and 18% for group 2, and statistically different (P = 0.032). Additionally, extracapsular invasion (ECI) was sig- nificantly higher in group 1 (56.5%) than in group 2 (29.5%) (P = 0.005). There was no difference between the two groups in terms of pathologic Gleason score; lymphovascular, perineural, seminal vesicle invasion and pathologic T stage (P > .05). Multivariate analysis was performed to define whether ECI affected PSM, using logistic regression. As seen in Table 3, multiple logistic regression analysis revealed that presence of ECI was an independent factor for PSM associated with the groups (OR: 2.512; 95% CI: 1.055-5.979). A total of 35 patients underwent extended BPLND during RALP, 20 patients in group 1 (40%) and 15 pa- tients in group 2 (24%). Although lymph node positiv- ity was higher in group 1(25% and 13%, respectively), there was no statistically significant difference between two groups (P = 0.24). The median follow-up of en- tire patients was 11 months and biochemical recurrence (BCR) rates were 12% and 6,6% in group 1 and 2, re- spectively (p = 0.99). A total of 7 patients with positive lymph nodes received early hormone therapy and 10 patients with BCR directed to radiotherapy. Although patients in group 1 needed more additional treatment (22% and 9,8%, respectively), there was no statistically significant difference between two groups (P = 0.153). Patients with PSM are analyzed in subgroups and rela- tionship between PSM and pT stage, lymphovascular, perineural, extracapsular invasion are shown in Table 4. Operative parameters are shown in Table 5. Both oper- ation time (skin to skin time, as defined) and duration of postoperative hospitalization were significantly re- duced from group 1 to group 2 (P < 0.001). A factor that could affect the operation time is whether BPLND is performed or not. Since the BPLND performed pa- tients were equally distributed in two groups, the effects on the operation time were evaluated as being equal. Hematocrit decrease on the first postoperative day did not show a significant difference between two groups (P = 0.587). The complications were classified using the Clavien system (from 1 to 4) and shown in Table 6.(14) A total of 11 patients (10%) (each one suffering single event) had complications and one of them (0.9%) required surgical intervention. This patient had adhesions due to previous peritonitis surgery, and ileal perforation was detected when he was explored due to the acute abdomen on the 2nd postoperative day. The ileum was repaired as pri- mary and patient was discharged on the postoperative 9th day. Two patients with urethra-vesical anastomo- sis stenosis and one patient with urethral stricture were treated with endoscopic intervention. DISCUSSION Radical prostatectomy is a complex surgery combined of extraction and reconstruction of tissues. Robotic sur- gery has the advantage to simplify this complex pro- cedure with its excellent three-dimensional vision and high motion range endo-wrist instruments. After the introduction of RALP, there has been a rapid increase in daily practice and interest in robotic surgery. But it should be kept in mind that, as with any new surgical technique RALP also have a learning curve. There is no consensus regarding the optimal way of detecting the learning curve of a surgical procedure but traditionally, the operative time has been widely used to assess this. Zorn et al. suggested that 120 RALP cases are needed to achieve a skin-to-skin operation time under 4 hours. (15) Ou et al. reported that the console time becomes gradually shorter with every 50 cases experience in their study.(16) In our study; operation time, described as Table 3. Multiple logistic regression analysis for the effect of ECI on surgical margin status Wald p-value Odds Ratio (95%CI) Surgical margin status 0.972 0.324 1.635(0.615-4.344 Extracapsular invasion 4.331 0.037 2.512(1.055-5.979) PSM (n:29) Group 1(1-50) Group 2 (51-111) p-value Lymphovascular invasion, n(%) No 10(55.6) 9(81.8) 0.234 Yes 8(44.4) 2(18.2) Perineural invasion, n(%) No 2(11.1) 3(27.3) 0.339 Yes 16(88.9) 8(72.7) Extracapsular invasion, n(%) Yes 0(0) 6(54.5) 0.001 No 18(100) 5(45.5) Pathological Stage n(%) T2 5(27.8) 5(45.5) * T3A 6(33.3) 4(36.4) T3B 7(38.9) 2(18.2) Table 4. Comparison of pathological parameters in patients with PSM between two groups LC does not affect PSM status in RALP-Islamoglu et al. Laparoscopic and Robotic Urology 335 Vol 15 No 06 November-December 2018 336 skin-to-skin time, was used to detect the learning curve. We used the moving average method to find the cut-off point for a learning curve in RALP, as used in the liter- ature before.(17,18) We found that at least 50 RALP cases are needed to gain proficiency even for an experienced surgeon in laparoscopic radical prostatectomy. To our understanding, reducing docking time with the more rapid determination of trocar positions and placement affected the operation time as well as improvements in surgical technique. The independent risk factors for disease recurrence and progression after radical prostatectomy are the presence of PSM, preoperative PSA, pathologic Gleason score and seminal vesicle involvement.(19) Among these, PSM is the only factor dependant on surgical experience. The main purpose of any urologist performing radical pros- tatectomy should be to reduce PSM rate and prevent disease recurrence. In the most extensive literature re- view, Novara et al. reported a 15% mean rate of PSMs in RALRP series published between 2008 and 2011 (each including >100 cases), with a range of 6.5–32% and concluded that PSM rate is higher in men with a more advanced pathologic stage.(20) Our study included a single surgeon with open and laparoscopic surgical background in radical prostatectomy and yielded that PSM rate has decreased dramatically from 36% to 18% after first 50 patients. But the higher rate of ECI in group 1 directed us to make a multivariate analysis to find out the effect of ECI on PSM. In this analysis, we found that the presence of extracapsular invasion was an independent factor for PSM and affected the PSM rates in both groups. From this point of view, we feel that the learning curve does not play a significant role in pathologic outcomes of RALP. We performed video documentation in all RALPs and reviewed our records in correlation with pathological reports to improve our technique. According to discus- sions with pathologists we made minor modifications in our technique after 40 patients. After dissecting sem- inal vesicles from a posterior approach we continued dissection until neurovascular bundle and apex appear, then turned in anterior approach and completed NVB dissection, dorsal venous complex ligation, and urethral incision in a traditional way. In Tewari's series, this ret- roapical technique decreased the authors' rate of PSMs from 4.4% to 1.4% and we believe that this modifica- tion had an effect in our lower PSM rate in the second group.(21) Although positive margins in prostate cancer are con- sidered an adverse oncologic outcome, their long-term impact on survival is highly variable and largely influ- enced by other risk modifiers.(22) BCR rates for RALP differs with follow-up time in different series. Propiglia et al. and Asimakopoulos et al. reported their BCR rates as 2,0% and 4,4% respectively, with a 12 months fol- low-up.(23,24) Park et al. and Ploussard et al. found BCR rates as 13,1% within 19 months and 10,3% within 15 months follow-up, respectively.(25,26) In our study, we had a BCR rate as 12% in Group 1 with 15 months of follow-up and 6,6% in Group 2 with a 6 months fol- low-up. The rates of need for additional therapy were 22% and 6,6% (p = 0,153) in Group 1-2, respective- ly. Although follow-up period was short these findings were consistent with the literature. We classified our complications using Clavien system and our 10% complication rate is within average when compared with newer series ranging from 5,08% to 19,6%.(16,27) In Patel's series with 2500 cases, a single surgeon had low complication rates of 5,08% in a large volume center and demonstrated a tendency to decrease with increasing experience of the surgeon.(28) This study has some limitations. Although it was based on a prospective database the study was retrospective. Also, the follow-up period is relatively short and onco- logical outcomes such as BCR require further observa- tion. Lastly, the cohort was small with 111 cases and a study with larger sample size could demonstrate the effect of learning curve on PSM better. Table 5. Comparison of operative parameters of both groups Group 1(1-50) Group 2 (51-111) p-value Operation time, (mean ± SD) min 257.1 ± 32.7 174.4 ± 41.3 < 0.001 Duration of hospitization (min-max) 4 (2-9) 3 (2-8) < 0.001 Hematocrit decrease median (min-max) 4.5 (0.7-10.9) 3.3 (0.5-14.5) 0.587 BPLND, n(%) 20 (40) 15 (24) 0.376 NVB preservation, n(%) 18 (36) 29 (47) 0.403 Number (by event) Detail Clavien 1 5 Lymphocele(2), urine leakage(1), intraoperative tachycardia(1), umblical wound infection(1), Clavien 2 2 Blood transfusion(2) Clavien 3 3 Urethral stricture(1), urethra-vesical anastomosis stenosis (2) Clavien 4 1 Ileum perforation(1) Table 6. Classification of complications occurred in entire patients using Clavien system Figure 1.Time taken to perform RALP in each case. Moving av- erage curve of RALP. LC does not affect PSM status in RALP-Islamoglu et al. CONCLUSIONS We can conclude that at least 50 RALP cases are needed to gain proficiency even for an experienced surgeon in laparoscopic radical prostatectomy. Our study demon- strates that surgeons experience can affect the periop- erative variables but the learning curve does not affect PSM status in RALP. ACKNOWLEDGEMENT The authors would like to thank Dr.Basak Oguz and ap- preciate her support for the statistical analysis of this study. CONFLICT OF INTEREST No competing financial interests exist. REFERENCES 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2018. CA Cancer J Clin. 68, 7-30. 2. Gültekin M, Boztaş G. Türkiye kanser istatistikleri. In: Saglık Bakanlıgı, Türkiye Halk Saglıgı Kurumu, 2014; 43. 3. Walsh PC, Partin AW, Epstein JI. Cancer control and quality of life following anatomical radical retropubic prostatectomy: results at 10 years. J Urol. 1994;152(5 Pt 2), 1831-6. 4. Schuessler WW, Schulam PG, Clayman RV, Kavoussi LR. Laparoscopic radical prostatectomy: initial short-term experience. Urology. 1997;50, 854-7. 5. Ahlering TE, Skarecky D, Lee D, Clayman RV. Successful transfer of open surgical skills to a laparoscopic environment using a robotic interface: initial experience with laparoscopic radical prostatectomy. J Urology. 2003;170(5), 1738-41. 6. Binder J, Kramer W. Robotically‐assisted laparoscopic radical prostatectomy. BJU Int. 2001;87, 408-10. 7. Lepor H. Status of radical prostatectomy in 2009: Is there medical evidence to justify the robotic approach? Rev Urol. 2009;11, 61. 8. Kural AR, Atug F. The applications of robotic surgery in urology. Turk J Urol. 2010; 36, 248-58. 9. D'amico AV, Whittington R, Malkowicz SB, et al. Biochemical outcome after radical prostatectomy, external beam radiation therapy, or interstitial radiation therapy for clinically localized prostate cancer. Jama. 1998; 280, 969-74. 10. Zorn KC, Gofrit ON, Orvieto MA, et al. Robotic-assisted laparoscopic prostatectomy: functional and pathologic outcomes with interfascial nerve preservation. Eur Urol. 2007; 51(3), 755-63. 11. Briganti A, Larcher A, Abdollah F, et al. Updated nomogram predicting lymph node invasion in patients with prostate cancer undergoing extended pelvic lymph node dissection: the essential importance of percentage of positive cores. Eur Urol. 2012; 61, 480-7. 12. Gandaglia G, Bravi CA, Dell’Oglio P, et al. The impact of implementation of the European Association of Urology Guidelines Panel recommendations on reporting and grading complications on perioperative outcomes after robot-assisted radical prostatectomy. Eur Urol. 2018; https://doi.org/10.1016/j. eururo.2018.02.025 13. Diggle PJ. Time series: A biostatistical introduction. In: Clarendon Press, Oxford. 1990: p 257 14. Clavien PA, Barkun J, De Oliveira ML, et al. The Clavien-Dindo classification of surgical complications: five-year experience. Ann Surg. 2009; 250, 187-96. 15. Zorn KC, Orvieto MA, Gong EM, et al. Robotic radical prostatectomy learning curve of a fellowship-trained laparoscopic surgeon. J Endourol. 2007; 21, 441-7. 16. Ou YC, Yang CR, Wang J, et al. The learning curve for reducing complications of robotic‐ assisted laparoscopic radical prostatectomy by a single surgeon. BJU Int. 2011; 108, 420-5. 17. Fukumoto K, Miyajima A, Hattori S, et al. The learning curve of laparoendoscopic single-site adrenalectomy: an analysis of over 100 cases. Surg Endosc. 2017; 31, 170-7. 18. Kayano H, Okuda J, Tanaka K, Kondo K, Tanigawa N. Evaluation of the learning curve in laparoscopic low anterior resection for rectal cancer. Surg Endosc. 2011; 25, 2972-9. 19. Wieder JA, Soloway MS. Incidence, etiology, location, prevention and treatment of positive surgical margins after radical prostatectomy for prostate cancer. J Urology. 1998; 160, 299- 315. 20. Novara G, Ficarra V, Mocellin S, et al. Systematic review and meta-analysis of studies reporting oncologic outcome after robot-assisted radical prostatectomy. Eur Urol. 2012; 62:382–404. 21. Tewari AK, Srivastava A, Mudaliar K, et al. Anatomical retroapical technique of synchronous (posterior and anterior) urethral transection: a novel approach for ameliorating apical margin positivity during robotic radical prostatectomy. BJU Int. 2010;106: 1364–73. 22. Yossepowitch O, Briganti A, Eastham JA, et al. Positive surgical margins after radical prostatectomy: A systematic review and contemporary update. Eur Urol. 2014; 65, 303-13. 23. Porpiglia F, Morra I, Chiarissi ML, et al. Randomised controlled trial comparing laparoscopic and robot-assisted radical prostatectomy. Eur Urol. 2013; 63, 606-14. 24. Asimakopoulos AD, Miano R, Di Lorenzo N, Spera E, Vespasiani G, Mugnier C. Laparoscopic versus robot-assisted bilateral LC does not affect PSM status in RALP-Islamoglu et al. Laparoscopic and Robotic Urology 337 Vol 15 No 06 November-December 2018 338 nerve-sparing radical prostatectomy: comparison of pentafecta rates for a single surgeon. Surg Endosc. 2013; 27, 4297-304. 25. Park B, Kim W, Jeong BC, et al. Comparison of oncological and functional outcomes of pure versus robotic-assisted laparoscopic radical prostatectomy performed by a single surgeon. Scand J Urol. 2013; 47, 10-18. 26. Ploussard G, de la Taille A, Moulin M, et al. Comparisons of the perioperative, functional, and oncologic outcomes after robot-assisted versus pure extraperitoneal laparoscopic radical prostatectomy. Eur Urol. 2014; 65, 610-19. 27. Lasser MS, Renzulli J, Turini GA, Haleblian G, Sax HC, Pareek G. An unbiased prospective report of perioperative complications of robot- assisted laparoscopic radical prostatectomy. Urology. 2010; 75, 1083-9. 28. Coelho RF, Palmer KJ, Rocco B, et al. Early complication rates in a single-surgeon series of 2500 robotic-assisted radical prostatectomies: report applying a standardized grading system. Eur Urol. 2010; 57, 945-52. LC does not affect PSM status in RALP-Islamoglu et al.