Effect of pre-operative internal obturator muscle mass index in MRI on biochemical recurrence of prostate cancer patients after radical prostatectomy: a multi-center study

Background Recent reports show that the pre-operative or post-operative skeletal mass index (sarcopenia) affects survival rates for various cancers; however, the link between prostate cancer survival and sarcopenia is unclear. Therefore, this study examined the effect of the pre-operative internal obturator muscle (IOM) mass index on biochemical recurrence (BCR) of prostate cancer (PCa) patients who underwent radical prostatectomy. Methods In total, 222 patients, who underwent open, laparoscopic, or robot-assisted radical prostatectomy at seven centers in 2011 and were followed up for 5 years, were enrolled. BCR was examined in the context of pre-operative IOM mass index and BMI. Results The mean age of the patients was 67.82 ± 6.23 years, and the mean pre-operative prostate-specific antigen (PSA) level was 11.61 ± 13.22 ng/ml. There was no significant difference in baseline characteristics between the low and high IOM mass index groups (p > 0.05). Age, pre-op PSA level, ECE, and T-stage were associated with BCR (p = 0.049, p < 0.001, p = 0.001, p = 0.004, respectively). BMI, prostate volume, Gleason score, resection margin, N-stage, M-stage and IOM mass index was not associated with BCR (p > 0.05). Conclusions Pre-operative IOM mass index was not associated with BCR; however, long-term follow-up is necessary to evaluate cancer-specific and overall survival of PCa patients.


Background
Prostate cancer (PCa) is the most common cancer in men worldwide [1]. The incidence of PCa in Asia is among the lowest in the world [1]. However, the incidence in the Republic (Rep.) of Korea is increasing rapidly [2]. PCa mortality rates have fallen in high-resource countries, but have increased or remain stable in low-resource countries [1]. PCa-associated mortality in the Rep. of Korea is increasing along with incidence [2].
Sarcopenia is a condition characterized by progressive and generalized loss of muscle mass and strength [3]. Sarcopenia predicts drug toxicity and time-to-tumor progression in patients undergoing chemotherapy [4]. The impact of sarcopenia in cancer patients has been studied, and data suggest that sarcopenia is independently associated with post-operative outcome following resection of colorectal cancer, hepatocellular carcinoma, pancreatic cancer, and bladder cancer [5][6][7][8]. In addition, sarcopenia is a prognostic marker for disease recurrence and mortality in patients with urologic cancers [8][9][10][11].
Radical prostatectomy (RP) is the gold standard treatment for localized PCa. RP is often the treatment of choice for younger and less morbid patients. A previous report suggests that sarcopenia is associated with non-cancer-related death in PCa patients undergoing radiotherapy [12]. Conversely, Mason et al. report that sarcopenia is not independently associated with perioperative complications or oncologic outcomes after RP [13]. Therefore, there is a lack of evidence supporting an association between sarcopenia and oncologic outcomes in PCa. Generally, sarcopenia, skeletal mass index (SMI) was defined as the cross-sectional area of the rectus abdominis; internal, external, and transverse obliques; psoas; quadratus lumborum; and erector spinae muscles from L3 down. Most surgeon performed MRI imaging work-up for prostatectomy. However, preoperative CT imaging is the option and surgeon's preference. Unfortunately, cross sectional area of whole muscle from L3 down level cannot be taken in the MRI imaging. Internal obturator muscle (IOM) is the only measurable muscle in the pelvis MRI axial cut. Additionally, there was no study between pre-operative IOM mass index and oncologic outcomes in PCa. Therefore, we investigated the effect of pre-operative IOM mass index on biochemical recurrence (BCR) in Korean patients after RP.

Study population and data collection
In total, 222 PCa patients, who underwent open, laparoscopic, or robot-assisted RP at seven academic centers in 2011 and were followed up for 5 years, were enrolled. All patients had histologically confirmed primary adenocarcinoma of the prostate, and all had undergone pre-operative MRI within 3 months of surgery. Patients were excluded if they had previously received androgen deprivation therapy (ADT) or radiotherapy (RT). In addition, patients undergoing testosterone replacement Clinical and pathologic data included age, height, body weight, prostate-specific antigen (PSA) levels, prostate volume, pathologic T-, N-, and M-stage, receipt of adjuvant RT or ADT, and follow-up duration. BCR after RP was defined as a post-operative PSA level > 0.2 ng/mL. Patients undergoing adjuvant ADT or RT after RP were excluded in the BCR group. IOM mass index was measured by dividing the mean internal obturator muscle area (calculated from MRI) by height (squared). All internal obturator muscle area measurements were performed by a single radiologist using ImageJ (https:// imagej. nih. gov/ ij/) at the level of symphysis pubis (Cho, BS, Fig. 1). Median value of IOM mass index was used to determine the cutoff value between high and low IOM mass index. Using this median value (11.47708194 cm 2 /m 2 ), patients were classified into high or low IOM mass index group. BMI group were divided into 3 groups according to Korean BMI criteria (normal < 23 kg/m 2 ; 23 kg/ m 2 ≤ overweight < 25 kg/m 2 ; 25 kg/m 2 ≤ obese).

Statistical analysis
IOM mass index was divided into two groups according to a median cut-off value. BMI was divided to three groups according to Korean BMI criteria [14]. The baseline characteristics of the patients with a low IOM mass index and a high IOM mass index were compared using an independent t-test and the Chi-square test as appropriate. Comparisons of clinical and pathological parameters according to BMI were made using ANOVA and the Chi-square test as appropriate. The association between BCR and pre-operative IOM mass index or BMI values were examined using a Cox proportional hazards model and a Kaplan-Meier survival analysis. Statistical analyses were performed using the Statistical Package for Social Sciences, version 25 software (SPSS, Inc., Chicago, IL, USA). A p-value < 0.05 was considered statistically significant.

Baseline characteristics of the study population according to IOM mass index
The baseline characteristics of all the patients are shown in Table 1. The mean age was 67.82 ± 6.23 years, and the mean pre-operative PSA level was 11.61 ± 13.22 ng/mL. There was no significant difference in baseline characteristics between the low IOM mass index and high IOM mass index groups (p > 0.05).

Comparison of clinical and pathological parameters according to BMI
As shown in Table 2, there was no significant difference in clinical and pathological parameters according to BMI (p > 0.05).

Discussion
The results of the present study suggest that IOM mass index is not associated with BCR after RP in Korean men. The present study is the first to show no significant association between IOM mass index and BCR in men with RP. Thus, IOM mass index may not be a prognostic marker for BCR in Korean men with localized PCa undergoing RP. Previous studies suggest that high BMI is associated with increased risk of PCa [15,16]. A meta-analysis by Bergstrom et al. reports a 6% increase in the risk of PCa in overweight men and a 12% increase in obese men compared with men of normal weight [16]. Maclnnis et al. reported a weak positive association between BMI and risk of PCa; BMI was associated with moderate increase in the risk of advanced PCa [15].
The relationship between BMI and BCR after RP remains controversial. Freedland et al. reported that obesity among men treated with RP was associated with high-grade tumors, a trend toward increased risk of a positive surgical margin, and high BCR [17]. Magheli et al. reported that high BMI is associated with adverse pathological findings and is a strong independent predictor of BCR after RP [18]. Asmar et al. reported that both obesity and hypertension are associated with an increased risk of BCR after RP, independent of age at the  [20]. They suggested that their data provide evidence that obese men undergoing RP are not more likely to suffer PCa progression. In addition, Siddiqui et al. reported that obese patients appeared to have worse pathologic features at the time of prostatectomy; however, BMI did not appear to be an independent predictor of recurrence or survival after prostatectomy [21]. The present study found no association between BMI and BCR. Therefore, we believe that the relationship between BMI and BCR after RP remains unclear. Sarcopenia is a process associated with normal aging; however, it is exacerbated by the hypercatabolic state and inflammatory responses caused by malignancy [22]. A systematic review by Joglekar et al. investigated the impact of sarcopenia on outcome following surgical resection of cancer and reported that sarcopenia is an independent prognostic factor for both complications and survival following surgical resection [23]. Sarcopenia is a more objective and comprehensive pre-operative risk factor that predicts all-cause survival for bladder cancer after radical cystectomy (RC) [9,24]. With respect to PCa, several reports link sarcopenia and survival [12,25]. Thus, sarcopenia could be used to predict non-cancer-related death in men with PCa after RT [12] and may be a poor prognostic factor for CRPC treated with chemotherapy [25]. However, the association between sarcopenia and survival after RP is very unclear. Only one study was reported that sarcopenia was not significantly associated with the risks of BCR, systemic progression, or all-cause mortality after RP [13]. This study is the first study to evaluate the association between IOM and prognosis in all cancer patients. IOM mass index is the only measurable muscle mass index in pelvic MRI. So, there was no options to select the muscle mass index in pelvic MRI. However, long-term follow-up studies are necessary to identify (or not) any association between sarcopenia or IOM mass index and survival.
We recognize that this study has several limitations. First, we only measured internal obturator muscle mass, but not SMI. In general, SMI was defined as the crosssectional area of the rectus abdominis; internal, external, and transverse obliques; psoas; quadratus lumborum; and erector spinae muscles from L3 down. However, because this was a multi-institutional study, different authors have different protocols for pre-operative CT and MRI. Therefore, we selected only the internal obturator muscle area in the pelvis in the MRI, because of internal obturator muscle is the only measurable muscle in the pelvis MRI axial cut. So, we couldn't compared IOM mass with general SMI. Second, many patients who underwent RP at these institutions were excluded from the analysis due to differences in CT or MRI protocols. In some institutions, radiologists were not able to measure the internal obturator muscle at the same levels. Third, the follow-up period was too short to get an accurate picture of survival. In this case we did not examine cancer-specific or overall survival; studies may need follow-up data spanning more than 15 years to get an accurate picture of the association between IOM mass index and the survival of PCa patients.

Conclusions
Pre-operative IOM mass index was not associated with BCR; however, long-term (> 15 years) follow-up is necessary to better answer this still controversial question.