Flexibility-tilt agreement best predicts shoulder balance following posterior spine fusion for adolescent idiopathic scoliosis.

PURPOSE: We propose a novel concept, called flexibility-tilt agreement (FTA), which could be applied intra-operatively to improve shoulder balance following posterior spine fusion for Adolescent Idiopathic Scoliosis (AIS). We retrospectively applied this concept to a cohort with thoracic-only curves, seeking to: (1) evaluate the impact of FTA and other peri-operative variables on post-operative shoulder balance, and (2) evaluate deformity characteristics associated with achieving FTA. METHODS: A single-institution registry was queried for patients undergoing PSF from 2000 to 2017 with main thoracic and double thoracic curves with at least 2-year follow-up. Flexibility was defined as the Cobb angle of the unfused proximal thoracic curve (cephalad to chosen UIV) on pre-operative bender. Tilt refers to post-operative UIV tilt angle. FTA means these values cancel (Flexibility + Tilt = 0 ± 5°). Logistic regression was performed to determine the association between peri-operative variables and shoulder balance. RESULTS: One hundred and sixty-one patients were included, mean age 13.6 years old, and 47-month mean follow-up. FTA was achieved in 74 (46%) patients and was associated with post-operative (OR = 4.59) and final (OR = 6.98) medial shoulder balance with a threshold of 6° (AUC = 0.77, p = 0.038). FTA was the best predictor of shoulder balance of all tested variables. CONCLUSION: Patients that showed flexibility-tilt agreement, or FTA, had vastly increased odds of medial and lateral shoulder balance at a minimum of 2-year follow-up for all thoracic curves. Future studies can evaluate whether applying FTA to determine intra-operative corrective maneuvers prospectively leads to improvements in shoulder balance. LEVEL OF EVIDENCE: Level III-retrospective comparative study.

L1-pelvic angle: a convenient measurement to attain optimal deformity correction.

PURPOSE: (1) Evaluate the associations between L1-pelvic angle (L1PA) and both sagittal vertical axis (SVA) and T1-pelvic angle (T1PA), and (2) assess the clinical impact of L1PA. METHODS: A single-institution retrospective cohort study was undertaken for patients undergoing adult spinal deformity (ASD) surgery from 2013 to 2017. Ideal L1PA was defined as (0.5xPelvic Incidence)-21. Pearson correlation was performed to compare L1PA, SVA, and T1PA. Univariate/multivariate regression was performed to assess the effect of L1PA on mechanical complications, controlling for age, BMI, and postoperative pelvic incidence-lumbar lordosis mismatch (PI/LL). Due to the overlapping nature of patients with pseudarthrosis and rod fracture, these patients were analyzed together. RESULTS: A total of 145 patients were included. Mean preoperative L1PA, SVA, and T1PA were 15.5 ± 8.9°, 90.7 ± 66.8 mm, and 27.1 ± 13.0°, respectively. Mean postoperative L1PA, SVA, and T1PA were 15.0 ± 8.9°, 66.7 ± 52.8 mm, and 22.3 ± 11.1°, respectively. Thirty-six (24.8%) patients achieved ideal L1PA. Though the correlation was modest, preoperative L1PA was linearly correlated with preoperative SVA (r2 = 0.16, r = 0.40, 95%CI = 0.22-0.60, p < 0.001) and T1PA (r2 = 0.41, r = 0.62, 95%CI = 0.46-0.76, p < 0.001). Postoperative L1PA was linearly correlated with postoperative SVA (r2 = 0.12, r = 0.37, 95%CI = 0.18-0.56, p < 0.001) and T1PA (r2 = 0.40, r = 0.62, 95%CI = 0.45-0.74, p < 0.001). Achieving ideal L1PA ± 5° was associated with a decreased risk of rod fracture/pseudarthrosis on univariate and multivariate regression (OR = 0.33, 95%CI = 0.12-0.86, p = 0.024). No association between achieving ideal L1PA and patient-reported outcomes was observed. CONCLUSION: L1PA was modestly correlated with SVA and T1PA, and achieving ideal L1PA was associated with lower rates of rod fracture/pseudarthrosis. Future studies are warranted to better define the clinical implications of achieving a normal L1PA. LEVEL OF EVIDENCE: III.

Clinical and radiographic outcomes following correction of idiopathic scoliosis in adolescence vs young adulthood.

PURPOSE: The natural history of adolescent idiopathic scoliosis (AIS) has been well documented, but the impact of age at the time of surgical correction is relatively understudied. In this study, we matched patients undergoing surgical correction of adult idiopathic scoliosis (AdIS) with a cohort of AIS patients to compare: (1) coronal and sagittal radiographic correction, (2) operative variables, and (3) postoperative complications. METHODS: A single-institution scoliosis registry was queried for patients undergoing idiopathic scoliosis surgery from 2000-2017. INCLUSION CRITERIA: patients with idiopathic scoliosis, no previous spine surgery, and 2-year follow-up. AdIS patients were matched 1:2 with AIS patients based on Lenke classification and curve characteristics. Independent sample t-test and Chi-square test was used to analyze the data. RESULTS: 31 adults underwent surgical correction of idiopathic scoliosis and were matched with 62 adolescents. Mean age of adults was 26.2 ± 11.05, mean BMI was 25.6 ± 6.0, and 22 (71.0%) were female. Mean age of adolescents was 14.2 ± 1.8, mean BMI was 22.7 ± 5.7, and 41(66.7%) were female. AdIS had significantly less postoperative major Cobb correction (63.9% vs 71.3%, p = 0.006) and final major Cobb correction (60.6% vs 67.9%, p = 0.025). AdIS also had significantly greater postoperative T1PA (11.8 vs 5.8, p = 0.002). AdIS had longer operative times (p = 0.003), higher amounts of pRBCs transfused (p = 0.005), longer LOS (p = 0.016), more ICU requirement (p = 0.013), higher overall complications (p < 0.001), higher rate of pseudarthrosis (p = 0.026), and more neurologic complications (p = 0.013). CONCLUSION: Adult patients undergoing surgical correction of idiopathic scoliosis had significantly worse postoperative coronal and sagittal alignment when compared with adolescent patients. Adult patients also had higher rates of complications, longer operative times, and longer hospital stays. LEVEL OF EVIDENCE: III.

The importance of Hounsfield units in adult spinal deformity surgery: finding an optimal threshold to minimize the risk of mechanical complications.

Background: Low bone mineral density (BMD) is a well-established risk factor for mechanical complications following adult spinal deformity (ASD) surgery. Hounsfield units (HU) measured on computed tomography (CT) scans are a proxy of BMD. In ASD surgery, we sought to: (I) evaluate the association of HU with mechanical complications and reoperation, and (II) identify optimal HU threshold to predict the occurrence of mechanical complications. Methods: A single-institution retrospective cohort study was undertaken for patients undergoing ASD surgery from 2013-2017. Inclusion criteria were: ≥5-level fusion, sagittal/coronal deformity, and 2-year follow-up. HU were measured on 3 axial slices of one vertebra, either at the upper instrumented vertebra (UIV) itself or UIV ±4 from CT scans. Multivariable regression controlled for age, body mass index (BMI), postoperative sagittal vertical axis (SVA), and postoperative pelvic-incidence lumbar-lordosis mismatch. Results: Of 145 patients undergoing ASD surgery, 121 (83.4%) had a preoperative CT from which HU were measured. Mean age was 64.4±10.7 years, mean total instrumented levels was 9.8±2.6, and mean HU was 153.5±52.8. Mean preoperative SVA and T1PA were 95.5±71.1 mm and 28.8°±12.8°, respectively. Postoperative SVA and T1PA significantly improved to 61.2±61.6 mm (P<0.001) and 23.0°±11.0° (P<0.001). Mechanical complications occurred in 74 (61.2%) patients, including 42 (34.7%) proximal junctional kyphosis (PJK), 3 (2.5%) distal junctional kyphosis (DJK), 9 (7.4%) implant failure, 48 (39.7%) rod fracture/pseudarthrosis, and 61 (52.2%) reoperations within 2 years. Univariate logistic regression showed a significant association between low HU and PJK [odds ratio (OR) =0.99; 95% confidence interval (CI): 0.98-0.99; P=0.023], but not on multivariable analysis. No association was found regarding other mechanical complications, overall reoperations, and reoperations due to PJK. HU below 163 were associated with increased PJK on receiver operating characteristic (ROC) curve analysis [area under the curve (AUC) =0.63; 95% CI: 0.53-0.73; P<0.001]. Conclusions: Though several factors contribute to PJK, it appears that 163 HU may serve as a preliminary threshold when planning ASD surgery to mitigate the risk of PJK.

Does Low-Grade Versus High-Grade Bilsky Score Influence Local Recurrence and Overall Survival in Metastatic Spine Tumor Surgery?

BACKGROUND AND OBJECTIVES: Spinal cord compression caused by spinal tumors is measured using the epidural spinal cord compression scale, also known as the Bilsky score. Whether Bilsky score predicts short-/long-term outcomes remains unknown. The objectives were to determine the correlation of Bilsky score 0-1 vs 2-3 with regards to (1) preoperative presentation, (2) perioperative variables, and (3) long-term outcomes. METHODS: A single-center, retrospective evaluation of a cohort of patients undergoing metastatic spine surgery was performed between 01/2010 and 01/2021. Multivariable logistic/linear/Cox regression were performed controlling for age, body mass index, race, total decompressed levels, tumor size, other organ metastases, and postoperative radiotherapy/chemotherapy. RESULTS: Of 343 patients with extradural spinal metastasis, 92 (26.8%) were Bilsky 0-1 and 251 (73.2%) were Bilsky 2-3. Preoperatively, patients with Bilsky 2-3 lesions were older ( P = .008), presented more with sensory deficits ( P = .029), and had worse preoperative Karnofsky Performance Scale (KPS) ( P = .002). Perioperatively, Bilsky 2-3 patients had more decompressed levels ( P = .005) and transpedicular decompression ( P < .001), with similar operative time ( P = .071) and blood loss ( P = .502). Although not statistically significant, patients with Bilsky 2-3 had more intraoperative neuromonitoring changes ( P = .412). Although rates of complications ( P = .442) and neurological deficit ( P = .852) were similar between groups, patients with Bilsky 2-3 lesions had a longer length of stay ( P = .007) and were discharged home less frequently ( P < .001). No difference was found in 90-day readmissions ( P = .607) and reoperation ( P = .510) Long-term: LR ( P =.100) and time to LR (log-rank; P =0.532) were not significantly different between Bilsky 0-1 and Bilsky 2-3 lesions. However, patients with Bilsky 2-3 lesions had worse postoperative KPS ( P < .001), worse modified McCormick scale score ( P = .003), shorter overall survival (OS) (log-rank; P < .001), and worse survival at 1 year ( P = .012). Bilsky 2-3 lesions were associated with shorter OS on multivariable Cox regression (hazard ratio = 1.78, 95% CI = 1.27-2.49, P < .001), with no significant impact on time to LR (hazard ratio = 0.73, 95% CI = 0.37-1.44, P = .359). CONCLUSION: Bilsky 2-3 lesions were associated with longer length of stay, more nonhome discharge, worse postoperative KPS/modified McCormick scale score, shorter OS, and reduced survival at 1 year. Higher-grade Bilsky score lesions appear to be at a higher risk for worse outcomes. Efforts should be made to identify metastatic spine patients before they reach the point of severe spinal cord compression..

Is lower screw density associated with mechanical complications in adult spinal deformity surgery?

PURPOSE: To assess the impact of screw density on: (1) rod fracture/pseudarthrosis, (2) proximal/distal junctional kyphosis/failure (PJK/DJK/PJF), and (3) deformity correction judged by sagittal vertical axis (SVA) and T1-pelvic angle (T1PA). METHODS: A single-center, retrospective cohort study of patients undergoing adult spinal deformity (ASD) surgery from 2013 to 2017 was undertaken. Screw density was calculated by dividing the number of screws placed by the total instrumented levels. Screw density was then dichotomized at our calculated mean density to ≥ 1.65 and < 1.65. Outcomes consisted of mechanical complications and the amount of correction obtained. RESULTS: 145 patients underwent ASD surgery with 2-year follow-up. Mean screw density (range) was 1.6 ± 0.3 (1.00-2.00). The most common levels with missing screws were L2 (n = 59, 40.7%), L3 (n = 57, 39.3%) and L1 (n = 51, 35.2%), located mainly along the concavity in 113(80.0%) patients and apices in 98 (67.6%) patients. Rod fracture/pseudarthrosis: 23/32 (71.8%) patients with rod fracture and 35/46 (76.0%) with pseudarthrosis had missing screws within two levels of the rod fracture/pseudarthrosis. Logistic regression showed no significant association between screw density and rod fracture/pseudarthrosis. PJK/F: 15/47 (31.9%) patients with PJK and 9/30 (30.0%) with PJF had missing screws within three levels of the upper instrumented vertebra (UIV). Logistic regression showed no significant association between screws density and PJK/F. Correction obtained: linear regression failed to show any significant association between screw density and SVA or T1PA correction. CONCLUSION: These findings showed that no significant association was found between screw density and mechanical complications or the amount of correction obtained, though approximately 3 out of 4 patients with rod fracture/pseudarthrosis had missing screws at or within two levels of the pathology. The prevention of mechanical complications is likely multifactorial and subject to both patient's characteristics and surgical techniques. LEVEL OF EVIDENCE: III.

Dysplasia Epiphysealis Hemimelica in the Lower Extremity.

BACKGROUND: Because of the rarity of dysplasia epiphysealis hemimelica (DEH), little is known about the relationship between disease classification and clinical symptoms or patient outcomes. This studies therefore aims to characterize DEH of the lower extremity and correlate radiographic classification to presenting symptomatology and need for surgical intervention. METHODS: A multi-center, retrospective review of all patients with DEH of the lower extremity over a 47-year period was conducted. Demographic data, presenting complaints, treatments, and symptoms at final follow-up were recorded. Radiographs were reviewed to classify lesions using the Universal Classification System for Osteochondromas (UCSO) and document the presence of solitary or multiple lesions within the involved joint. Correlative statistics were used to determine whether presenting complaints, lesion location or radiographic classification predicted the need for surgery or a pain-free outcome. RESULTS: Twenty-eight patients met inclusion criteria with an average age at presentation of 7.8 years. The ankle was the most commonly affected joint with 20/28 patients (71%) having lesions of the talus, distal tibia, or distal fibula. Patients with chief complaints of pain were more likely to undergo surgery than those with complaints of a mass or deformity ( P =0.03). Ankle lesions were more likely to be managed operatively than those of the hip or knee ( P =0.018) and all 12 patients with talar lesions underwent surgery. Neither the number of lesions nor lesion classification was predictive of surgical intervention or a pain-free outcome after surgery. Patients presenting with pain were more likely to have a pain-free outcome (11/14 patients) after surgery ( P =0.023) whereas all patients presenting with deformity who underwent surgery had pain at final follow-up. CONCLUSIONS: Although no single radiographic characteristic of DEH was predictive of surgical intervention or outcome, painful lesions of the ankle, and lesions of the talus were more likely to be managed operatively. Although surgery does not always result in a pain-free outcome, the operative management of painful lesions was more likely to provide a pain-free outcome than surgery for deformity or a mass.

Unplanned Readmission Is Associated With Decreased Overall Survival and Performance After Metastatic Spine Surgery.

STUDY DESIGN: Retrospective case-control study. OBJECTIVE: In a cohort of patients undergoing metastatic spine surgery, we sought to: (1) identify risk factors associated with unplanned readmission, and (2) determine the impact of an unplanned readmission on long-term outcomes. SUMMARY OF BACKGROUND DATA: Factors affecting readmission after metastatic spine surgery remain relatively unexplored. MATERIALS AND METHODS: A single-center, retrospective, case-control study was undertaken of patients undergoing spine surgery for extradural metastatic disease between 02/2010 and 01/2021. The primary outcome was 3-month unplanned readmission. Preoperative, perioperative, and tumor-specific variables were collected. Multivariable Cox regression was performed, controlling for tumor size, other organ metastasis, and preoperative/postoperative radiotherapy/chemotherapy. RESULTS: A total of 357 patients underwent surgery for spinal metastases with a mean follow-up of 538.7±648.6 days. Unplanned readmission within 3 months of surgery occurred in 64/357 (21.9%) patients, 37 (57.8%) were medical, 27 (42.2%) surgical, and 21 (77.7%) were related to their spine surgery. No significant differences were found regarding demographics and preoperative variables, except for insurance, where most readmitted patients had private insurance compared with nonreadmitted patients ( P =0.021). No significant difference was found in preoperative radiotherapy/chemotherapy. Regarding perioperative exposure variables, readmitted patients had a higher rate of postoperative complications (68.8% vs. 24.2%, P <0.001) and worse postoperative Karnofsky Performance Score ( P =0.021) and Modified McCormick Scale ( P =0.015) at the time of first follow-up. On multivariate logistic regression, postoperative complications were associated with increased readmissions (odds ratio=1.38, 95% CI=1.25-1.52, P <0.001). Regarding the impact of unplanned readmission on long-term tumor control, unplanned readmission was associated with shorter time to local recurrence (log-rank; P =0.029) and reduced overall survival (OS) (log-rank; P <0.001). On multivariate Cox regression, other organ metastasis [hazard ratio (HR)=1.48, 95% CI=1.13-1.93, P =0.004] and 3-month readmission (HR=1.75, 95% CI=1.28-2.39, P <0.001) were associated with worsened OS, with no impact on LR. Postoperative chemotherapy was significantly associated with longer OS (HR=0.59, 95% CI=0.45-0.77, P <0.001). CONCLUSIONS: Postoperative complications were associated with unplanned readmission following metastatic spine surgery. Furthermore, 3-month unplanned readmission was associated with a shorter time to local recurrence and decreased OS. These results help surgeons understand the drivers of readmissions and the impact of readmissions on patient outcomes. LEVEL OF EVIDENCE: 3.

Cranially Directed Upper Instrumented Vertebrae Screw Angles Are Associated With Proximal Junctional Kyphosis in Adult Spinal Deformity Surgery.

STUDY DESIGN: A retrospective cohort study. OBJECTIVE: To evaluate the impact of the upper instrumented vertebral (UIV) screw angle in adult spinal deformity (ASD) surgery on: (1) proximal junctional kyphosis/failure (PJK/F), (2) mechanical complications and radiographic measurements, and (3) patient-reported outcome measures (PROMs). SUMMARY OF BACKGROUND DATA: The effect of UIV screw angle in ASD surgery on patient outcomes remains understudied. MATERIALS AND METHODS: A single-institution, retrospective study was undertaken from 2011 to 2017. UIV screw angle was trichotomized into positive: cranially directed screws relative to the superior endplate (2°≤θ), neutral: parallel to the superior endplate (-2°<θ<2°), and negative: caudally directed screws relative to the superior endplate (-2°≥θ). The primary outcome was PJK/F. Secondary outcomes included remaining mechanical complications, reoperation, and PROMs: Oswestry Disability Index, Numeric Rating Scale (NRS) back/leg, and EuroQol. Regression controlled for age, body mass index, postoperative sagittal vertical axis (SVA), and pelvic incidence lumbar-lordosis mismatch. RESULTS: Among 145 patients undergoing ASD surgery, UIV screw angles were 35 (24.1%) cranially directed, 24 (16.6%) neutral, and 86 (59.3%) caudally directed. PJK occurred in 47(32.4%) patients. Positive screws were independently associated with increased PJK [odds ratio (OR)=4.88; 95% CI, 1.85-13.5, P =0.002] and PJF (OR=3.06; 95% CI, 1.32-12.30, P =0.015). Among 108 (74.5%) patients with lower thoracic UIV, PJK occurred in 38 (35.1%). Cranially directed screws were independently associated with an increased odds of PJK (OR=5.56; 95% CI, 1.86-17.90, P =0.003) with a threshold of 0.2° (area under the curve =0.65; 95% CI, 0.54-0.76, P <0.001), above which the risk of PJK significantly increased. No association was found between positive screw angle and PJF (OR=3.13; 95% CI, 0.91-11.40, P =0.073). Because of the low number of patients with an upper thoracic UIV (N=37, 25.5%), no meaningful conclusions could be drawn from this subgroup. There was no association between UIV screw angle and remaining mechanical complications, reoperations, postoperative SVA and T1-pelvic angle, or PROMs. CONCLUSIONS: Cranially directed UIV screw angles increased the odds of PJK in patients with lower thoracic UIV. Meticulous attention should be paid to the lower thoracic UIV screw angle to mitigate the risk of PJK in ASD.

The postoperative course of mechanical complications in adult spinal deformity surgery.

PURPOSE: (a) Describe the time course of each mechanical complication, and (b) compare radiographic measurements and preoperative patient-reported outcome measures (PROMs) among each mechanical complication type. METHODS: A single-institution case-control study was undertaken of patients undergoing adult spinal deformity (ASD) surgery from 2009-2017. Exposure variables included patient demographics, operative variables, radiographic measurements, and preoperative PROMs, including Oswestry Disability Index (ODI), Numeric Rating Scale Back/Leg-pain scores (NRS-Back/Leg), and EuroQol-5D (EQ-5D). The primary outcomes were occurrence of a mechanical complication and time to complication. Due to overlapping occurrence, rod fracture and pseudarthrosis were grouped into one composite category. RESULTS: 145 patients underwent ASD surgery and were followed for at least 2 years. 30/47 (63.8%) patients with proximal junctional kyphosis (PJK) required reoperation, whereas 27/31 (87.1%) patients with pseudarthrosis/rod fracture required reoperation (63.8% vs. 87.1%, Χ2 = -0.23, 95% CI -0.41, -0.05, p = 0.023). Cox regression showed no significant difference in time to reoperation between PJK and rod fracture/pseudarthrosis (HR = 0.97, 95% CI 0.85-1.11, p = 0.686). Distal junctional kyphosis (DJK) (N = 3; 2 reoperation) and implant failures (N = 4; 0 reoperations) were rare. Patients with PJK had significantly lower Hounsfield Units preoperatively compared to those with pseudarthrosis/rod fracture (138.2 ± 43.8 vs. 160.3 ± 41.0, mean difference (MD) = -22.1, 95% CI -41.8, -2.4, p = 0.028), more prior fusions (51.1% vs. 25.8%, Χ2 = 0.253, 95% CI 0.41, 0.46, p = 0.026), fewer instrumented vertebrae (9.2 ± 2.6 vs. 10.7 ± 2.5, MD = -1.5, 95% CI -2.7, -0.31, p = 0.013), and higher postoperative thoracic kyphosis (TK) (46.3 ± 12.7 vs. 34.9 ± 10.6, MD = 11.4, 95% CI 5.9, 16.9, p < 0.001). Higher postoperative C7 sagittal vertical axis (SVA) did not achieve a significant difference (80.7 ± 72.1 vs. 51.9 ± 57.3, MD = 28.8, 95% CI -1.9, 59.5, p = 0.066). No differences were seen in preoperative PROMs. CONCLUSION: Patients with pseudarthrosis/rod fracture had a higher reoperation rate compared to those with PJK with similar time to reoperation. Moreover, patients with PJK had higher postoperative TK, lower Hounsfield Units, more prior fusions, and fewer instrumented levels compared to those with pseudarthrosis/rod fracture. The results of this single-institution study suggest that even though mechanical complications are often analyzed as a single group, important differences may exist between them. LEVEL OF EVIDENCE: III.

Establishing quasi-steady state operations of microphysiological systems (MPS) using tissue-specific metabolic dependencies.

Microphysiological systems (MPS), consisting of tissue constructs, biomaterials, and culture media, aim to recapitulate relevant organ functions in vitro. MPS components are housed in fluidic hardware with operational protocols, such as periodic complete media replacement. Such batch-like operations provide relevant nutrients and remove waste products but also reset cell-secreted mediators (e.g. cytokines, hormones) and potentially limit exposure to drugs (and metabolites). While each component plays an essential role for tissue functionality, MPS-specific nutrient needs are not yet well-characterized nor utilized to operate MPSs at more physiologically-relevant conditions. MPS-specific nutrient needs for gut (immortalized cancer cells), liver (human primary hepatocytes) and cardiac (iPSC-derived cardiomyocytes) MPSs were experimentally quantified. In a long-term study of the gut MPS (10 days), this knowledge was used to design operational protocols to maintain glucose and lactate at desired levels. This quasi-steady state operation was experimentally validated by monitoring glucose and lactate as well as MPS functionality. In a theoretical study, nutrient needs of an integrated multi-MPS platform (gut, liver, cardiac MPSs) were computationally simulated to identify long-term quasi-steady state operations. This integrative experimental and computational approach demonstrates the utilization of quantitative multi-scale characterization of MPSs and incorporating MPS-specific information to establish more physiologically-relevant experimental operations.

Interconnected Microphysiological Systems for Quantitative Biology and Pharmacology Studies.

Microphysiological systems (MPSs) are in vitro models that capture facets of in vivo organ function through use of specialized culture microenvironments, including 3D matrices and microperfusion. Here, we report an approach to co-culture multiple different MPSs linked together physiologically on re-useable, open-system microfluidic platforms that are compatible with the quantitative study of a range of compounds, including lipophilic drugs. We describe three different platform designs - "4-way", "7-way", and "10-way" - each accommodating a mixing chamber and up to 4, 7, or 10 MPSs. Platforms accommodate multiple different MPS flow configurations, each with internal re-circulation to enhance molecular exchange, and feature on-board pneumatically-driven pumps with independently programmable flow rates to provide precise control over both intra- and inter-MPS flow partitioning and drug distribution. We first developed a 4-MPS system, showing accurate prediction of secreted liver protein distribution and 2-week maintenance of phenotypic markers. We then developed 7-MPS and 10-MPS platforms, demonstrating reliable, robust operation and maintenance of MPS phenotypic function for 3 weeks (7-way) and 4 weeks (10-way) of continuous interaction, as well as PK analysis of diclofenac metabolism. This study illustrates several generalizable design and operational principles for implementing multi-MPS "physiome-on-a-chip" approaches in drug discovery.