Prophylactic sodium thiosulfate reduces the risk and severity of acute kidney injury in oxaliplatin-based hyperthermic intraperitoneal chemotherapy

Abstract

BACKGROUND

Cytoreductive surgery (CRS) combined with hyperthermic intraperitoneal chemotherapy (HIPEC) has improved outcomes for patients with peritoneal carcinomatosis (PC). However, this treatment carries a significant risk of acute kidney injury (AKI), especially with cisplatin-based regimens. Sodium thiosulfate (ST) is established as a nephroprotective treatment in cisplatin-based HIPEC, but its efficacy in oxaliplatin-based HIPEC remains unexplored.

AIM

To evaluate the protective effects of ST in patients undergoing oxaliplatin-based HIPEC for colorectal PC.

METHODS

This retrospective single-center study included 119 patients who underwent CRS combined with oxaliplatin-based HIPEC for colorectal PC between 2018 and 2024. ST was introduced in 2022, and patients were grouped as ST-treated (n = 46) and non-ST-treated (n = 73) based on the treatment period. The primary endpoint was the incidence of AKI. Secondary outcomes were postoperative plasma creatinine levels, estimated glomerular filtration rate, severity of AKI, and rate of postoperative complications. Outcomes were analyzed using multivariate logistic regression, including sensitivity analysis.

RESULTS

The incidence of AKI was significantly lower in the ST-treated group (28%) compared with the non-ST-treated group (66%). Multivariate analysis confirmed ST as an independent factor associated with a reduced risk of AKI. Peak plasma creatinine and mean postoperative creatinine levels were also significantly lower in the ST-treated group. The rate of postoperative complications other than AKI was similar between the groups.

CONCLUSION

Prophylactic administration of ST significantly reduces the incidence and severity of AKI in patients undergoing CRS combined with oxaliplatin-based HIPEC. These findings suggest that ST may be considered as a nephroprotective strategy in oxaliplatin-based HIPEC protocols; however, further prospective studies are warranted to confirm these results.

Key Words: Hyperthermic intraperitoneal chemotherapy; Peritoneal carcinomatosis; Peritoneal surface malignancy; Oxaliplatin; Acute kidney injury; Sodium thiosulfate

Core Tip: This retrospective single-center study demonstrates that prophylactic sodium thiosulfate (ST) markedly reduces both the incidence and severity of acute kidney injury in patients undergoing cytoreductive surgery with oxaliplatin-based hyperthermic intraperitoneal chemotherapy for colorectal peritoneal carcinomatosis. Patients receiving ST showed significantly lower postoperative creatinine levels and fewer kidney injuries without an increased risk of other postoperative complications. These findings indicate a clear nephroprotective benefit in this setting and support consideration of ST in oxaliplatin-based hyperthermic intraperitoneal chemotherapy protocols, although future prospective studies are warranted to confirm these results.

INTRODUCTION

Peritoneal carcinomatosis (PC) is a severe manifestation of various malignancies and significantly affects patient survival. In colorectal cancer (CRC), the median survival ranges from 4-7 months with best supportive care or up to 12-23 months with palliative chemotherapy[1,2]. The introduction of cytoreductive surgery (CRS) combined with hyperthermic intraperitoneal chemotherapy (HIPEC) has significantly improved prognosis and led to a paradigm shift in treatment. Reported 5-year overall survival rates in CRC range from 25% to 45%[1-4]. However, this treatment carries a significant risk of postoperative complications[5-7] owing to the extensive surgery and the administration of cytotoxic agents. The combination of hyperthermia and chemotherapy results in a synergistic cytotoxic effect[8], although it may also increase the permeability of the peritoneum[9]. The extent of systemic drug uptake depends on multiple factors, including the concentration of the chemotherapeutic agent, its molecular size and properties, treatment duration, the size of the peritoneal defect, and other factors affecting the peritoneal diffusion barrier. Different chemotherapy protocols have been used in HIPEC for CRC, among which oxaliplatin-based HIPEC has demonstrated improved survival in selected patients[1,10,11]. Although its use has been discontinued in many centers after the PRODIGY-7 trial, oxaliplatin-based HIPEC is still frequently used in Sweden for PC of CRC origin.

A common postoperative complication in CRS-HIPEC is acute kidney injury (AKI), which has been reported in up to 48% of cases[12,13], with a mean incidence of 18% reported in a recent meta-analysis[14]. AKI may result from several factors, including surgical stress leading to systemic inflammatory response syndrome, hypotension, and hypovolemia, as well as the systemic effects of the chemotherapeutic agents used in the HIPEC procedure.

Cisplatin is the standard treatment used in HIPEC protocols for ovarian cancer, mesothelioma, and other soft tissue tumors but carries a high risk of postoperative kidney failure. Cisplatin is eliminated by the kidneys by both glomerular filtration and tubular secretion[15]. The accumulation of platinum in the kidneys starts a cascade of tubular toxicity, characterized by vascular vasoconstriction with secondary tissue damage, along with glomerular injury and interstitial inflammation. This cascade can eventually lead to AKI[16]. Oxaliplatin also has nephrotoxic effects[17-19], likely attributable to the platinum ion, although these effects are less pronounced than those observed with cisplatin.

Sodium thiosulfate (ST) exerts a nephroprotective effect by detoxifying platinum-induced reactive oxygen species and thereby reducing renal tubular cell injury[20]. As a result, ST has become the gold standard prophylactic treatment in patients receiving cisplatin-based HIPEC regimens[20,21], but it has not been routinely used in oxaliplatin-based protocols. While the nephroprotective effect of ST in cisplatin-based HIPEC is well recognized, its efficacy in preventing AKI in patients undergoing oxaliplatin-based HIPEC remains unexplored.

We hypothesized that the administration of ST in oxaliplatin-based HIPEC could reduce both the incidence and severity of AKI compared with standard treatment without ST. Therefore, this study aimed to evaluate the potential nephroprotective effect of prophylactic ST administration in patients undergoing oxaliplatin-based HIPEC for colorectal PC.

MATERIALS AND METHODS

Study design

Single-center retrospective study based on prospectively collected data.

Patient selection

All patients with colorectal PC who underwent CRS with oxaliplatin-based HIPEC at Skane University Hospital (Malmö, Sweden) between February 2018 and May 2024 were included. The study period was chosen based on stable perioperative routines, with no protocol changes implemented other than the introduction of ST treatment. Accordingly, goal-directed intraoperative fluid resuscitation, a standardized forced diuresis protocol, the anesthesia method, the intraperitoneal chemotherapy protocol, and the operation team remained consistent throughout the study period. Patient data were extracted from a local patient registry and supplemented by medical chart reviews when necessary. Patients were stratified into two groups based on the administration of prophylactic ST treatment: (1) Those who received prophylactic ST treatment (ST group, January 2022 to May 2024); and (2) Those who did not receive ST treatment [No ST (nST) group, February 2018 to December 2021].

Kidney function

Plasma creatinine (P-Cr) was analyzed using a standard enzymatic colorimetric assay with calibration traceable to primary reference material, with values assigned by isotope dilution mass spectrometry[22]. Estimated glomerular filtration (eGFR) rate based on creatinine was calculated using the Björk et al[23] revised creatinine-based eGFR equation.

Standard HIPEC treatment during the entire study period consisted of oxaliplatin 460 mg/m² intra-abdominally, combined with 5-fluorouracil 400 mg/m². All HIPEC procedures were performed using the open technique. Oxaliplatin was administered intraperitoneally for 30 minutes at a target temperature of 41.5-42.0 °C, diluted in 3 L of perfusate. These parameters were standardized and remained unchanged throughout the study period. The oxaliplatin dose was reduced in patients with a preoperative eGFR < 60 mL/minute/1.73 m² or those who had received prior neoadjuvant chemotherapy. Prophylactic intraoperative ST was introduced in January 2022. ST was administered intravenously using a two-phase intravenous protocol: A 9 g/m² bolus over 20 minutes at the start of HIPEC, followed by a 12 g/m² infusion over 6 hours. Urine output was routinely monitored postoperatively as part of standard intensive care unit and ward care. All patients were managed according to a standardized forced diuresis protocol targeting ≥ 2 mL/kg/hour for the first 48 hour, followed by diuretic treatment as needed to maintain a daily urine output > 1500 mL during the subsequent week. As urine output was protocol-driven, it was not considered a reliable indicator of intrinsic renal function. Thus, AKI classification was based solely on serum creatinine (S-Cr) criteria according to the Kidney Disease: Improving Global Outcomes (KDIGO) guidelines.

Data collection

Patient, tumor, and perioperative data were retrieved from the local patient registry, including surgical details, laboratory results, and clinical outcomes. Clinical and demographic variables included detailed medical history, smoking status, body mass index (BMI), age, sex, mean preoperative eGFR, and the use of neoadjuvant and adjuvant chemotherapy. BMI was calculated as weight in kilograms divided by height in meters squared and categorized as BMI ≥ 30 kg/m²[24]. Hypertension and diabetes mellitus were defined based on a documented medical diagnosis or ongoing pharmacological treatment. Smoking status was classified as current or former smoking at the time of surgery, as recorded in the medical charts. Surgical data, including the PC index, intraoperative blood loss, duration of surgery, intraoperative hypotension, lactate levels, and duration of postoperative inotropic support, were retrieved from medical records. Perioperative renal function was assessed using P-Cr and eGFR levels. Following CRS-HIPEC, P-Cr was measured daily for 10 days, whereas lactic acid levels were routinely measured for the first 48 hours postoperatively; these data were retrieved from medical records. Postoperative complications were recorded and graded according to the Clavien-Dindo (CD) classification. Furthermore, all episodes of sepsis were recorded separately.

The development of postoperative AKI was recorded and defined according to the KDIGO criteria, as any of the following: An increase in S-Cr by ≥ 0.3 mg/dL (≥ 26.5 μmol/L) within 48 hours, an increase in S-Cr to ≥ 1.5 times the baseline value within postoperative days (PODs) 0-7, or urine output ≤ 0.5 mL/kg/hour for at least 6 hours. AKI stages were classified as follows: Stage 1, increase in S-Cr ≥ 0.3 mg/dL (in 48 hours) or 1.5-1.9 times the baseline value within 7 days; stage 2, increase in S-Cr to 2.0-2.9 times the baseline value; and stage 3, increase in S-Cr to ≥ 3.0 times the baseline value, an absolute increase to ≥ 4.0 mg/dL, or initiation of renal replacement therapy regardless of previous KDIGO stages[25]. Preoperative CKD was classified according to the KDIGO eGFR categories as follows: G1, eGFR ≥ 90 mL/minute/1.73 m²; G2, eGFR 60-89 mL/minute/1.73 m²; G3a, eGFR 45-59 mL/minute/1.73 m²; G3b, eGFR 30-44 mL/minute/1.73 m²; G4, eGFR 15-29 mL/minute/1.73 m²; and G5, eGFR < 15 mL/minute/1.73 m²[25]. Classification of G1 and G2 was performed without albuminuria data owing to the unavailability of data.

Statistical analysis

Categorical variables were summarized as counts and proportions and compared using Pearson’s χ² test or binary logistic regression, as appropriate. Continuous variables were presented as medians with interquartile ranges and compared using the Mann-Whitney U test or Kruskal-Wallis test. Multivariable logistic regression was applied to estimate odds ratios for risk factors associated with postoperative AKI.

Daily postoperative P-Cr levels were analyzed using linear mixed-effects models. Patients were included as a random effect to account for repeated measurements. Treatment group and POD were included as fixed effects, and their interaction was assessed. Potential confounders and mediators for multivariable adjustment were identified using a directed acyclic graph (DAG), as shown in Supplementary Figure 1. Based on this analysis, adjustments were made for the following variables: Perioperative renal hypoperfusion (monitored as maximum lactate levels within 48 hours from the start of surgery, duration of intraoperative hypotension, and the need for postoperative inotropic support), surgical trauma (assessed by duration of surgery and intraoperative blood loss), standard or reduced dose of oxaliplatin, comorbidities associated with an increased risk of AKI (i.e., diabetes, history of myocardial ischemia or peripheral artery insufficiency, smoking, and hypertension), age, and postoperative complications. Duration of surgery, intraoperative blood loss, duration of postoperative inotropic support, and maximum lactate levels were dichotomized for analysis. Duration of intraoperative hypotension (i.e., systolic blood pressure < 80 mmHg) was also categorized and graded as 0, 1, or 2 according to the duration. Given the limited number of postoperative AKI events, multivariable analysis was performed with caution to minimize overfitting. Covariates were selected based on clinical relevance and DAG (Supplementary Figure 1). The number of variables included in the adjusted models was therefore restricted. Dichotomization of selected continuous variables was applied to facilitate model stability and clinical interpretation. Because of the multiple covariates considered, the multivariable analyses were supplemented with sensitivity analyses, presented in Supplementary Table 1. To further reduce the risk of overfitting, additional univariable and multivariable analyses including the six parameters with the greatest impact on the results are presented in Supplementary Table 2.

An ad hoc analysis of risk factors for postoperative complications was performed following the observation that complications were less frequent in the ST group. A two-sided P < 0.05 was considered statistically significant. Data were analyzed using SPSS version 28.0.0.0 (Statistical Package for the Social Sciences; IBM Corp., Armonk, NY, United States). The study was approved by the Ethical Committee in Uppsala, approval No. 2020/23504. Informed consent was waived by the Swedish Ethical Review Authority owing to the retrospective study design and the use of anonymized data. The study was conducted in accordance with the principles of the Declaration of Helsinki.

RESULTS

Demographic and operative characteristics

A total of 121 patients who underwent CRS and HIPEC with intraperitoneal oxaliplatin for PC of colorectal origin between February 2018 and May 2024 were identified. Of these, 2 were excluded owing to postoperative ureteral obstruction. The remaining 119 patients constituted the study population, of whom 46 patients received prophylactic ST treatment (ST group), and 73 did not receive ST-treatment (nST group). Patient characteristics were comparable between the groups, although there was a trend toward a higher proportion of patients in the nST group presenting with a PC index score > 15 compared with the ST group (P = 0.07). Importantly, baseline P-Cr and eGFR levels, as well as the distribution of baseline eGFR categories according to the KDIGO classification, were similar between the groups. Only 1 patient in each group had a history of renal failure, as illustrated in Table 1.

Table 1 Patient characteristics, including preoperative kidney function, stratified on perioperative treatment with prophylactic sodium thiosulfate or no sodium thiosulfate, n (%).

Variable	nST (n = 73)	ST (n = 46)	P value
Age			0.731
≤ 59	31 (42)	21 (46)
> 59	42 (58)	25 (54)
Sex			0.501
Male	38 (52)	21 (46)
Female	35 (48)	25 (54)
PCI score
0-8	36 (49)	31 (67)	Reference
9-15	21 (29)	10 (22)	0.192
> 15	16 (22)	5 (11)	0.07
Baseline creatinine: Median; IQR	68 (62-80)	69 (59-76)	0.603
Baseline eGFR: Median; IQR	77 (67-86)	76 (64-87)	0.943
Preoperative kidney function4,5
G1	14 (20.3)	11 (24.4)	Reference
G2	45 (65.2)	29 (64.4)	0.672
G3a	9 (13)	3 (6.7)	0.27
G3b	1 (1.4)	2 (4.4)	0.47
Preoperative comorbidities
BMI ≥ 30	20 (27)	10 (22)	0.491
Smoking	7 (10)	5 (11)	0.82
Diabetes mellitus	10 (14)	8 (17)	0.58
Diabetes mellitus - insulin	8 (11)	1 (2)	0.20
Cardiovascular disease	9 (12)	6 (13)	0.91
Hypertension	22 (30)	18 (39)	0.31
ASA
1	28 (38)	20 (44)	Reference
2	32 (44)	19 (41)	0.652
3	13 (18)	7 (15)	0.61

¹χ² test.

²Binary logistic regression analysis.

³Mann-Whittney U test.

⁴According to Kidney Disease: Improving Global Outcomes classification of estimated glomerular filtration rate.

⁵Missing in 5 patients.

nST: No sodium thiosulfate; ST: Sodium thiosulfate; PCI: Peritoneal carcinomatosis index; IQR: Interquartile range; eGFR: Estimated glomerular filtration rate; BMI: Body mass index; ASA: American Society of Anesthesiologists.

Perioperative data are presented in Table 2. A higher proportion of patients in the nST group received a reduced dose of intraperitoneal oxaliplatin compared to the ST group. The ST group also had a significantly shorter median operative time (413 minutes) compared with the nST group (601 minutes). Although no statistically significant differences were observed in the incidence of intraoperative hypotension or the need for inotropic support between the groups, the nST group showed significantly higher peak median postoperative lactic acid levels during the first 48 hours after surgery (2.7 mmol/L) compared with the ST group (1.9 mmol/L), as shown in Table 2. In addition, significantly higher mean lactate levels were observed in the nST group compared with the ST group on POD 0 (4.14 mmol/L vs 2.63 mmol/L) and POD 1 (2.72 mmol/L vs 2.10 mmol/L; P < 0.001), as shown in Supplementary Figure 2.

Table 2 Perioperative characteristics, stratified on perioperative treatment with prophylactic sodium thiosulfate or no sodium thiosulfate, n (%).

Variable	nST (n = 73)	ST (n = 46)	P value
Reduced dose of intraperitoneal chemotherapy
No	41 (56)	41 (89)	< 0.0011
Yes	32 (44)	5 (11)
Length of operation	601 (523-690)	413 (351-488)	< 0.0012
Intraoperative bleeding	600 (350-800)	700 (413-1000)	0.122
Intraoperative hypotension4: Minutes5
0	40 (55)	22 (49)	Reference
< 0 - < 7.12	11 (15)	4 (9)	0.523
≥ 7.12	22 (30)	19 (42)	0.27
Inotropic support6
≤ 32 hours	44 (64)	28 (61)	0.751
> 32 hours	25 (36)	18 (39)
Peak postoperative lactate median; (IQR)	2.7 (2.2-3.1)	1.9 (1.6-2.4)	< 0.0012
GI anastomosis
0	31 (42)	16 (35)	Reference
1	32 (44)	25 (54)	0.313
2	5 (7)	4 (9)	0.55
> 2	5 (7)	1 (2)	0.41
Stoma formation
No	33 (45)	28 (61)	Reference
Yes	38 (52)	15 (33)	0.063
Has prior to surgery	2 (3)	3 (6)	0.55

¹χ² test.

²Mann-Whittney U test.

³Binary logistic regression analysis.

⁴< 80 mmHg.

⁵One missing patient.

⁶Four missing patients.

nST: No sodium thiosulfate; ST: Sodium thiosulfate; IQR: Interquartile range.

Postoperative renal function

The nST group exhibited significantly higher median peak postoperative P-Cr levels (114 μmol/L) compared with the ST group (92 μmol/L; P < 0.001), as shown in Table 3. Moreover, mean daily P-Cr concentrations on PODs 1-10 were significantly higher in the nST-group than in the ST-group, as illustrated in Figure 1.

Figure 1 Preoperative and postoperative plasma creatinine levels stratified according to perioperative treatment with prophylactic sodium thiosulfate. nST: No sodium thiosulfate; ST: Sodium thiosulfate; POD: Postoperative day.

Table 3 Postoperative serum creatinine and acute kidney injury stratified by prophylactic sodium thiosulfate.

Variable	nST (n = 73)	ST (n = 46)	P value	OR (95%CI)
Peak postop creatinine median; (IQR)	114 (97-146)	92 (76-112)	< 0.0011	N/A
Max relative increase in creatinine percentage median; (IQR)	63 (45-111)	40 (26-53)	< 0.0011	N/A
Post op AKI	48 (66)	13 (28)	< 0.0012	0.21 (0.09-0.46)
Grade of AKI
No AKI	25 (34.2)	33 (71.7)	Reference	-
Grade 1	28 (38.4)	10 (21.7)	0.0042	0.27 (0.11-0.66)
Grade 2	17 (23.3)	3 (6.5)	0.0032	0.13 (0.04-0.51)
Grade 3	3 (4.1)	0 (0)	N/A3	N/A3

¹Mann-Whittney test.

²χ² test.

³Too few cases to be calculated.

nST: No sodium thiosulfate; ST: Sodium thiosulfate; AKI: Acute kidney injury; CI: Confidence interval; OR: Odds ratio; N/A: Not applicable; IQR: Interquartile range.

Overall, 61 patients (51%) in the cohort developed postoperative AKI. A significantly higher proportion of patients in the nST group developed postoperative AKI compared with the ST group (66% vs 28%; P < 0.001). This corresponds to an absolute risk reduction of 38% and a number needed to treat of 2.5. Although most AKI cases were grade 1, a significantly higher incidence of both grade 1 and grade 2 AKI was noted in the nST group compared with the ST group, as shown in Table 3 and Supplementary Figure 3. Irrespective of preoperative GFR, patients in the nST group showed a higher proportion of postoperative AKI compared with those in the ST group, as shown in Supplementary Figure 4. Two patients required postoperative dialysis, both belonging to the nST group. Both patients underwent reoperations: One for an anastomotic leak complicated by secondary sepsis on POD 7, and the other for intra-abdominal hemorrhage.

Adjusted multivariate logistic regression analyses demonstrated a significant reduction in the risk of AKI associated with ST treatment [odds ratio (OR) = 0.1, 95% confidence interval: 0.03-0.38], as demonstrated in Table 4. This association remained robust in sensitivity analyses (OR = 0.07), as shown in Supplementary Table 1. No other parameter was identified as a significant risk factor for AKI in the primary analysis. However, intraoperative hypotension exceeding 7.12 minutes emerged as a significant risk factor in the sensitivity analysis, as shown in Supplementary Table 1.

Table 4 Adjusted risk analysis of postoperative acute kidney injury, estimated by logistic regression.

Variable	No AKI (n = 58)	AKI (n = 61)	Univariate analysis		Multivariate analysis
			P value	OR (95%CI)	P value	OR (95%CI)
Sodium thiosulfate					< 0.001	0.10 (0.03-0.38)
No	25	48	Reference	-
Yes	33	13	< 0.001	0.21 (0.09-0.46)
Age					0.70	1.19 (0.46-3.24)
≤ 59	25	27	-	Reference
> 59	33	34	0.90	0.95 (0.46-1.97)
Sex					-	-
Male	29	30	Reference	-
Female	29	31	0.93	1.03 (0.50-2.12)
Cardiovascular comorbidities1
0	24	25	Reference	-	-	-
1	16	21	0.60	1.26 (0.53-2.97)	0.37	1.68 (0.54-5.23)
≥ 2	18	15	0.62	0.80 (0.33-1.94)	0.90	0.93 (0.30-2.90)
ASA classification
1	23	25	Reference	-	-	-
2	26	25	0.76	0.89 (0.40-1.95)
3	9	11	0.83	1.12 (0.39-3.20)
Reduced IP dose
No	40	42	Reference	-	-	-
Yes	18	19	0.99	1.01 (0.46-2.19)	0.13	0.41 (0.13-1.28)
Neoadjuvant chemo
No	43	49	Reference	-	-	-
Yes	15	12	0.42	0.70 (0.30-1.66)
Chemotherapy naive
No	33	36	Reference	-	-	-
Yes	25	25	0.82	0.92 (0.44-1.90)
Operation time
≤ 537 minutes	39	21	Reference	-	-	-
> 537 minutes	19	40	< 0.001	3.91 (1.83-8.37)	0.22	1.99 (0.67-5.96)
Intraoperative bleeding
≤ 600 mL	35	30	Reference	1.57 (0.76-3.25)	0.10	2.26 (0.86-5.94)
> 600 mL	23	31	0.22
Intraoperative hypotension (minute)3
None	34	28	Reference	-	-	-
0-7.12	8	7	0.92	1.06 (0.34-3.29)	0.51	0.60 (0.13-2.74)
> 7.12	15	26	0.07	2.11 (0.94-4.73)	0.06	2.82 (0.98-8.16)
Peak lactic acid level
< 2.4	34	28	Reference	-	-	-
≥ 2.4	24	33	0.17	1.67 (0.81-3.45)	0.92	0.95 (0.34-2.64)
Postoperative inotropy4
≤ 32 hours	35	37	Reference	-	-	-
> 32 hours	21	22	0.98	0.99 (0.47-2.11)	0.55	0.75 (0.28-1.97)
Complications
No complication	31	28	Reference	-	-	0.40 (0.13-1.19)
< 3b	24	22	0.97	1.02 (0.47-2.20)	0.10	-
≥ 3b	3	11	0.05	4.06 (1.03-16.1)	0.13	4.43 (0.64-30.5)
Sepsis with AKI	0	4	0.05	N/A2	-	-

¹Diabetes mellitus, hypertension, smoking, body mass index ≥ 30, history of myocardial infarction or peripheral arterial insufficiency.

²Too few cases to be calculated.

³One missing patient.

⁴Four missing patients.

AKI: Acute kidney injury; CI: Confidence interval; OR: Odds ratio; N/A: Not applicable; ASA: American Society of Anesthesiologists; IP: Intraperitoneal chemotherapy.

Postoperative complications

Postoperative complications occurred in 50.4% of the patients. The majority were classified as mild or moderate (CD < 3b), whereas 23% were serious complications (CD ≥ 3b). Patients in the ST group showed a significantly lower overall complication rate (33%) compared with the nST group (62%; P = 0.002). However, the spectrum of complications did not differ between the groups, as shown in Supplementary Table 3. A significantly shorter median length of hospital stay was noted in the ST group compared with the nST group (12 days vs 17 days; P < 0.001), whereas the duration of stay in the intensive care unit did not differ between the groups (Supplementary Table 2).

Ad hoc analysis

The ad hoc multivariable risk analysis of risk factors for postoperative complications identified an operative time exceeding 537 minutes as the only statistically significant predictor of complications (OR = 4.09, 95% confidence interval: 1.54-10.8), whereas ST treatment was not associated with a reduced risk of postoperative complications, as illustrated in Table 5.

Table 5 Adjusted risk analysis of postoperative complications (except acute kidney injury) estimated by logistic regression.

Variable	No complications (n = 59)	Complications (n = 60)	Univariate analysis		Multivariate analysis
			P value	OR (95%CI)	P value	OR (95%CI)
Sodium thiosulfate
No	28	45	Reference	-	-	-
Yes	31	15	0.002	0.30 (0.14-0.65)	0.29	0.56 (0.19-1.64)
Age
≤ 59	26	26	Reference	-	0.91	0.95 (0.39-2.30)
> 59	33	34	0.94	1.03 (0.50-2.13)	-	-
Sex
Male	26	33	Reference	-	-	-
Female	33	27	0.23	0.66 (0.31-1.33)
Cardiovascular comorbidities1
0	27	22	Reference	-	-	-
1	15	22	0.18	1.80 (0.76-4.27)	0.38	1.56 (0.57-4.26)
≥ 2	17	16	0.75	1.16 (0.48-2.80)	0.52	1.42 (0.49-4.16)
ASA classification
1	26	22	Reference	-	-	-
2	24	27	0.48	1.33 (0.60-2.93)
3	9	11	0.49	1.44 (0.51-4.12)
Reduced IP dose
No	42	40	Reference	-	-	-
Yes	17	20	0.6	1.24 (0.57-2.69)	0.6	0.77 (0.29-2.06)
Neoadjuvant chemotherapy
No	49	43	Reference	-	-	-
Yes	10	17	0.14	1.94 (0.80-4.68)
Chemotherapy naive
No	35	34	Reference	-	-	-
Yes	24	26	0.77	1.12 (0.54-2.31)
Operating time
≤ 537 minutes	41	19	Reference	-	-	-
> 537 minutes	18	41	< 0.001	4.92 (2.26-10.7)	0.005	4.09 (1.54-10.8)
Bleeding
≤ 600 mL	33	32	Reference	-	-	-
> 600 mL	26	28	0.78	1.11 (0.54-2.29)	0.82	0.90 (0.37-2.22)
Intraoperative hypotension (minute)
None	32	30	Reference	1.60 (0.51-5.04)	0.83	1.16 (0.31-4.36)
0-7.12	6	9	0.42	1.01 (0.46-2.24)	0.96	0.97 (0.38-2.45)
> 7.12	21	20	0.97	-	-	-
Peak lactic acid level
< 2.4	37	25	Reference	2.36 (1.13-4.92)	0.49	1.38 (0.55-3.43)
≥ 2.4	22	35	0.02
Postoperative inotropy
≤ 32 hours	36	36	Reference	-	-	-
> 32 hours	22	21	0.98	0.96 (0.45-2.03)	0.76	0.87 (0.36-2.10)

¹Diabetes mellitus, hypertension, smoking, body mass index ≥ 30, history of myocardial infarction or peripheral arterial insufficiency.

CI: Confidence interval; OR: Odds ratio; ASA: American Society of Anesthesiologists; IP: Intraperitoneal chemotherapy.

DISCUSSION

This study demonstrates a significantly reduced risk of AKI in patients undergoing oxaliplatin-based HIPEC with prophylactic ST treatment. AKI following CRS-HIPEC remains a major clinical concern, although reported incidence rates after CRS with oxaliplatin-based HIPEC vary widely, ranging from 2% to 62%[26-30]. A systematic review published in 2024 reported an incidence of 34.6% in patients treated with oxaliplatin-based HIPEC[31]. The rate of AKI in the present study was 51%, which is notably higher than previously reported rates. This discrepancy may, at least in part, be attributed to the relatively high standard oxaliplatin dose used at our center (460 mg/m²).

The nephrotoxic mechanism of cisplatin is largely attributed to the platinum ion, which inflicts a tubular injury. Exposure of renal tubular cells to platinum ions triggers a complex inflammatory cascade involving the upregulation of endothelial adhesion molecules, followed by leukocyte adhesion, activation, and migration, ultimately leading to ischemia, vascular injury, subsequent tubular damage, and cell death[32]. ST is a nonspecific pharmacological agent used in several clinical settings, including the management of systemic adverse effects associated with platinum-based chemotherapy. For example, its cation-chelating properties enable its use in the treatment of calciphylaxis in patients undergoing dialysis[33]. The chelating effect of ST reduces the exposure of renal tubular cells to free platinum ions and has been shown to exert a protective effect against cisplatin-induced nephrotoxicity[20,34,35].

Prophylactic administration of ST is well established in cisplatin-based HIPEC but has not been routinely adopted in oxaliplatin-based HIPEC protocols. However, several studies have demonstrated that oxaliplatin causes tubular injury through mechanisms similar to those observed in cisplatin, including renal tubular vacuolization, acute tubular necrosis, renal tubular acidosis, and AKI[17-19], effects that are most likely attributable to the platinum ion. Nevertheless, limited research has explored the efficacy of ST in preventing oxaliplatin-associated AKI.

The present study demonstrates a clear nephroprotective effect of ST in oxaliplatin-based HIPEC, reflected by lower postoperative creatinine levels, a reduced risk of AKI, and a reduced severity of AKI when it occurred. This protective effect was observed irrespective of preoperative kidney function. Additionally, no patients in the ST group required postoperative dialysis treatment, in contrast to 2 patients in the nST group. Although the number of dialysis cases was too small to permit definitive conclusions on the preventive effect of ST, this finding is in line with findings from previously published studies on cisplatin-based HIPEC, which reported that no ST-treated patients required dialysis compared with those who did not receive ST[34,36].

In addition to the nephrotoxic effects of intraperitoneal chemotherapy, several other factors contribute to AKI following CRS-HIPEC. Systemic inflammatory response and ischemia-reperfusion injury are well-known mechanisms of distant organ failure, including kidney failure, and are common features of CRS-HIPEC procedures. Other plausible risk factors include intraoperative hypotension, prolonged operation time, and preoperative comorbidities. A systematic review identified hypertension, diabetes mellitus, and preoperative CKD as risk factors for AKI, in addition to cisplatin use, following CRS-HIPEC[31]. A retrospective Portuguese study further identified the use of cisplatin, prolonged intensive care unit stay, and perioperative renal impairment as risk factors for AKI[37]. These and other relevant variables were adjusted for in the present study, and prophylactic ST administration remained a strong independent protective factor against AKI. This association also remained robust in sensitivity analyses.

Nevertheless, the multivariable analyses should be interpreted with caution, considering the limited number of AKI events, which limits the events-per-variable ratio and increases the risk of overfitting. To address this, covariates were selected based on clinical relevance and DAG. Dichotomization of selected continuous variables was used to improve model stability and acknowledge potential information loss. Importantly, the association between ST treatment and reduced AKI risk remained consistent across all sensitivity analyses, including a restricted model with fewer covariates.

The spectrum of postoperative complications was similar between the study groups. Multivariate analysis did not reveal a significant difference in the risk of postoperative complications. These findings suggest that ST treatment is not associated with an increased risk of postoperative complications. This observation is in line with most previous reports on ST use in cisplatin-based CRS-HIPEC[20,34,36]. However, Vachez et al[35] reported a higher incidence of septic complications in ST-treated patients compared with non-ST-treated patients in a retrospective study, although no differences were observed for any other complications.

This study has several limitations, mainly related to its retrospective design, although it was based primarily on prospectively collected data. Although the indications for surgery and intraoperative routines, including fluid resuscitation, remained unchanged throughout the study period, the retrospective nature of the study inherently limits the ability to control for biases and unmeasured confounding factors. However, all patients treated with CRS and oxaliplatin-based HIPEC during the study period were included, thereby minimizing selection bias related to patient inclusion. Consequently, the sample size reflects the total number of eligible patients rather than being based on a predefined power estimation. The study design may limit the precision of secondary analyses and introduces the possibility of time-period-related confounding. Data on perioperative intravenous contrast exposure and blood transfusions were not systematically available, although intravenous contrast-enhanced imaging was performed only during the workup of severe complications, and intraoperative blood loss was used as an indirect proxy for transfusion requirements. These factors may nevertheless influence postoperative renal function; therefore, residual confounding cannot be excluded. Furthermore, the impact of confounders was mitigated using multivariable analyses and backward sensitivity analyses. The classification of CKD stages was based solely on eGFR, as data on albuminuria were not available to further stratify CKD stages 1 and 2. However, baseline renal function did not differ between the groups.

A major strength of this study is the consistent use of the same P-Cr assay throughout the study period, as well as the application of a validated eGFR formula[23] applicable to the entire study population.

CONCLUSION

This study demonstrates a clear nephroprotective effect of prophylactic ST administration in oxaliplatin-based HIPEC, irrespective of preoperative kidney function. Moreover, no evidence of an increased or specific risk of postoperative complications associated with ST treatment was observed. These findings suggest that prophylactic ST may be beneficial in oxaliplatin-based HIPEC to reduce the incidence of postoperative AKI. However, prospective multicenter studies are warranted to confirm these findings before routine implementation.

ACKNOWLEDGEMENTS

The authors thank Professor Anders Christensson for his valuable feedback on kidney function measurements.