Skip to main content
The Oncologist logoLink to The Oncologist
. 2025 Jul 16;30(7):oyaf192. doi: 10.1093/oncolo/oyaf192

Real-world testing of serum lactate dehydrogenase among patients with metastatic colorectal cancer in the United States

Heinz-Josef Lenz 1,✉, Orsolya Lunacsek 2, Helene Ostojic 3, Xiaoyun Pan 4, Emmanuelle Dochy 5, Zdravko Vassilev 6, Nasreen Khan 7
PMCID: PMC12265455  PMID: 40580491

Abstract

Background

Serum lactate dehydrogenase (LDH) is a potential prognostic biomarker of outcomes in patients with metastatic colorectal cancer (mCRC). This retrospective, observational study assessed real-world LDH testing patterns and LDH as a prognostic factor for overall survival (OS) in US patients receiving chemotherapy for mCRC.

Methods

Patients with mCRC who initiated first-line chemotherapy between January 1, 2016, and November 30, 2022, were selected from a nationwide de-identified Electronic Health Record-derived database. LDH value was categorized based on laboratory reference ranges. The prognostic relationship between pretreatment LDH value and OS was assessed using Kaplan–Meier and multivariate Cox proportional-hazards models.

Results

Of 15 329 adult patients (median age 64 years), 3379 (22%) had LDH testing at or postindex; 21% had abnormal baseline values, while 40% had normal values. Patients with abnormal LDH levels were more likely to be female (47% abnormal LDH vs 41% normal LDH), ≥ 65 years of age (52% vs 49%), Black or African American (12% vs 7%), or reside in the Northeast (30% vs 21%). The median OS (95% CI) in patients with normal baseline LDH was 29.9 (28.3-31.7) vs 16.8 (14.8-18.2) months for those with abnormal LDH. In a multivariate Cox proportional-hazards model, patients with abnormal baseline LDH had higher risk of death (HR 1.91, P < .0001) after adjustment for demographic/clinical characteristics.

Conclusions

Abnormal baseline LDH levels were associated with shorter OS; however, only one-fifth of patients receiving chemotherapy underwent LDH testing. Efforts to increase LDH testing could be valuable in helping guide treatment decisions.

Keywords: metastatic colorectal cancer, lactate dehydrogenase, biomarker


Implications for Practice.

In this real-world study that included over 15 000 patients receiving chemotherapy for metastatic colorectal cancer (mCRC) in the United States, rates of lactate dehydrogenase (LDH) testing were low, with only one-fifth of patients having undergone testing. LDH above the upper limit of normal was associated with an increased risk of shorter overall survival; however, in the absence of specific guidance, LDH testing is currently suboptimal and not widely considered standard of care for patients with mCRC. The results of this study suggest that greater adoption of early LDH testing in clinical practice could help guide treatment decisions.

Introduction

Metastatic colorectal cancer (mCRC) carries a poor prognosis, with a relative 5-year survival rate in the United States of approximately 16% for patients with metastatic disease that has spread beyond the lymph nodes. This is in contrast to the survival rates for patients diagnosed at earlier stages, when the 5-year survival rate is approximately 74% for patients with regionally disseminated disease and 91% for localized disease.1 Standard first-line therapies for mCRC include immune checkpoint inhibitors (eg, pembrolizumab) or targeted therapies (eg, cetuximab, panitumumab, bevacizumab), with doublet (5-fluorouracil, leucovorin, and oxaliplatin [FOLFOX] or 5-fluorouracil, leucovorin, and irinotecan [FOLFIRI]) or triplet (5-fluorouracil, leucovorin, oxaliplatin, and irinotecan [FOLFOXIRI]) chemotherapy regimens.2 Contemporary approaches to treatment sequencing for patients with mCRC are primarily informed by the molecular features of the tumor, and can be used to direct an individualized treatment strategy.3-5 Key biomarkers for informing treatment selection for patients with mCRC include RAS (KRAS, NRAS) or BRAF mutational status, human epidermal growth factor receptor 2 status, and microsatellite instability.2,6

Although lactate dehydrogenase (LDH) is not currently considered as a routine biomarker for treatment selection in oncology clinical practice guidelines, several studies have evaluated its prognostic and predictive role as a nonspecific marker of tissue damage in several cancers, including mCRC.7 Increased serum LDH reflects metabolic changes in cancer cells and therefore acts as an indirect marker of tumor hypoxia, neoangiogenesis, and metastasis.7-9 A meta-analysis of 68 clinical studies involving 31 857 patients between 1978 and 2014 found that high levels of pretreatment LDH were associated with worse overall survival (OS). The hazard ratios (HRs) were 1.48 for all cancers and 1.52 among patients with CRC (the median cutoff for high LDH was 250 IU/L [range of 200-1000 IU/L]).10 Studies in patients with mCRC have suggested that LDH levels that are at or exceed the upper limit of normal (ULN) are both potential prognostic biomarkers and predictors of efficacy of antiangiogenic-based treatments.9,11,12 In a systematic review and meta-analysis of 1219 patients with mCRC, pretreatment LDH values exceeding the ULN were associated with a lack of therapeutic effect from bevacizumab-based first-line chemotherapy and from continued second-line chemotherapy that includes the antiangiogenic component.13 Therefore, LDH levels may also be a relevant factor for evaluating treatment efficacy and may help inform a more timely change in treatment approach.

Despite the established predictive and prognostic significance of LDH in several malignancies, little is understood about real-world testing patterns in mCRC. To the best of our knowledge, no US-based studies have quantified and characterized LDH testing or the potential relationship between LDH value and OS in a real-world setting. This study aimed to evaluate real-world LDH testing patterns in patients receiving chemotherapy-containing regimens for mCRC in the United States.

Methods

Study design and population

This retrospective, observational cohort study used the nationwide Flatiron Health, Inc. electronic mCRC database, which is a retrospective longitudinal Electronic Health Record (EHR)-derived database, comprising a probabilistic sample of de-identified patients diagnosed with mCRC. These clinical EHR records are derived from patient-level structured and unstructured data (physicians’ notes) for approximately 36 000 patients between 2013 and 2023, and are part of a larger real-world database originating from approximately 280 oncology community practices and academic research hospitals in the United States comprising 800 sites of care.

Eligible patients were those with mCRC, ≥18 years of age, who were initiated on a first-line chemotherapy-containing regimen between January 1, 2016, and November 30, 2022; the first-line start date was designated as the index date (Figure S1). Patients had to have data available for ≥ 3 months prior to the index date (defined as the baseline period) unless they were newly diagnosed with mCRC, and data available for ≥ 3 months following the index date unless deceased. The follow-up period was defined as the time period from the index date until loss to follow‑up (ie, last date of database activity), death, or the end of the study period, whichever occurred first. Patients with other non-CRC primary cancers (except nonmelanoma skin cancers) and/or involvement in clinical trials were excluded.

Objectives

This study assessed LDH testing patterns among patients with mCRC who were initiated on systemic chemotherapy. The main objectives were to: (1) describe patient demographics and clinical characteristics among those who were tested at least once after initiating first-line therapy and among those who did not get tested at all after initiation of therapy; (2) assess the relationship between baseline LDH levels (normal and abnormal) and OS; and (3) describe follow-up testing patterns after chemotherapy initiation, time interval between successive LDH tests, and LDH values over time by line of therapy.

Statistical analyses

Core exposure variables were patients with and without LDH testing after initiation of first-line therapy, and normal/abnormal baseline LDH level. Outcome variables were number of LDH tests, time to first LDH test postindex, time between successive LDH tests, and LDH values. Covariates used to describe patient demographics and clinical characteristics include age, sex, race, practice type, geographic region, and clinical characteristics, including Eastern Cooperative Oncology Group (ECOG) performance status, tumor stage, biomarkers, and line of treatment.

Abnormal and normal LDH values were defined based on individual patient testing and corresponding laboratory reference ranges; an LDH value was considered abnormal if it exceeded the upper limit of the reference range used by the individual laboratory, and normal otherwise. Lines of therapy was based on oncologist-guided algorithms (developed by Flatiron Health, Inc.) and applied to the EHR treatment data, and was generally defined as the first eligible drug episode plus other eligible drugs given within 28 days; an overlap of 3 days was permitted between oral agents and any other drug without being a single line.

Mean and standard deviation were reported for continuous variables, and frequencies and percentages were reported for categorical variables. The relationship between OS and normal vs abnormal baseline LDH levels was assessed using Kaplan–Meier and multivariate Cox proportional-hazards models controlling for gender, race, region, practice type, stage, index performance status, KRAS mutation status, and baseline LDH normality. All statistical analyses were conducted using SAS, Version 9.4 (SAS Institute Inc.) software.

Ethics

This analysis was a secondary data analysis based on patient healthcare information from the de-identified Flatiron Health, Inc. database, from which patients could not be deanonymized. As dictated by Title 45 Code of Federal Regulations (45 CFR 46.101(b)(4)) (available at https://www.govinfo.gov/content/pkg/CFR-2011-title45-vol1/pdf/CFR-2011-title45-vol1.pdf), this analysis was conducted under an exemption from Institutional Review Board oversight for US-based studies using de-identified healthcare records, and for which informed consent was not applicable.

Results

Patients

A total of 19 663 patients with mCRC initiated a first-line chemotherapy regimen between January 1, 2016, and November 30, 2022. Of these patients, 15 329 met the eligibility criteria and were included in the final study (Figure 1). FOLFOX ± bevacizumab (48%), capecitabine ± bevacizumab (18%), and FOLFIRI ± bevacizumab (15%) were the most common first-line regimens (Table S1).

Figure 1.

Flow chart showing patient numbers at each stage of the selection process, starting with the total number of patients with metastatic colorectal cancer in the Flatiron Health database, and ending with the total study population after various exclusion criteria were applied.

Attrition flow chart. aIncludes chemotherapeutic agents (capecitabine, floxuridine, 5-fluorouracil, irinotecan, leucovorin, and oxaliplatin) or biologic/targeted agents (bevacizumab, cetuximab, and panitumumab); bNew diagnosis with mCRC and the date of diagnosis were defined by Flatiron Health, Inc.; cDatabase rules allow first-line therapy to be initiated up to 14 days before the date of mCRC diagnosis; dNon-melanoma skin cancers were not considered. Abbreviations: CT, chemotherapy; mCRC, metastatic colorectal cancer.

In the total population (N = 15 329), the median age was 64 years (interquartile range 55-73 years), and patients were most frequently > 65 years of age (49%), White (61%), located in the South (43%), male (57%), commercially insured (50%), and received care at a community practice (89%) (Table 1). Overall, most patients were stage IV at initial diagnosis (59%); approximately two-thirds of patients had ECOG performance status 0 or 1 (Table 1). The most commonly reported site of CRC was the colon (72%). The proportion of patients who did not receive BRAF or KRAS biomarker testing at any time preindex to 30 days postindex was 45% and 35%, respectively. Baseline demographics and clinical characteristics were generally comparable among LDH-tested and non-LDH-tested patients, although LDH-tested patients were more frequently White and non-Hispanic, and less frequently resided in the Western region of the United States compared with non-LDH-tested patients (Table 1).

Table 1.

Baseline demographics and clinical characteristics.

Characteristics All patients
(N = 15 329)
LDH testing status after first-line therapy initiation
Tested
(n = 3379)
Not tested
(n = 11 950)
Age, yearsa
 Median (IQR) 64 (55-73) 64 (55-72) 64 (55-73)
Age group, n (%) years
 <50
 50-64
 ≥65
2120 (14)
5564 (36)
7507 (49)
481 (14)
1258 (37)
1640 (49)
1651 (14)
4349 (36)
5950 (50)
Sex, n (%)
 Male
 Female
8712 (57)
6615 (43)
1888 (56)
1490 (44)
6824 (57)
5125 (43)
Race, n (%)
 White
 Black/African American
 Other
 Unknown
9374 (61)
1767 (12)
2435 (16)
1753 (11)
2261 (67)
337 (10)
526 (16)
255 (8)
7113 (60)
1430 (12)
1909 (16)
1498 (13)
Ethnicity, n (%)
 Hispanic
 Other
 Unknown
1359 (9)
10 635 (69)
3335 (22)
203 (6)
2435 (72)
741 (22)
1156 (10)
8200 (69)
2594 (22)
Region, n (%)b
 South
 West
 Northeast
 Midwest
 Unknown
6586 (43)
2658 (17)
2266 (15)
1865 (12)
1954 (13)
1467 (43)
313 (9)
665 (20)
476 (14)
458 (14)
5119 (43)
2345 (20)
1601 (13)
1389 (12)
1496 (13)
Practice type, n (%)
 Academic
 Community
 Both
1540 (10)
13 592 (89)
197 (1)
367 (11)
2976 (88)
36 (1)
1173 (10)
10 616 (89)
161 (1)
Insurance type, n (%)
 Commercial
 Medicarec
 Otherd
 Unknown
7712 (50)
1823 (12)
4011 (26)
1783 (12)
1619 (48)
433 (13)
855 (25)
472 (14)
6093 (51)
1390 (12)
3156 (26)
1311 (11)
ECOG performance status, n (%)e
 0
 1
 2
 3
 Unknown
5424 (35)
4751 (31)
1326 (9)
298 (2)
3515 (23)
1250 (37)
1041 (31)
254 (8)
58 (2)
776 (23)
4174 (35)
3710 (31)
1072 (9)
240 (2)
2739 (23)
Stage IV disease at initial diagnosis, n (%) 9041 (59) 2059 (61) 6982 (58)
KRAS mutation status, n (%)f
 Positive
 Negative
 Unknown
 Not tested
4440 (29)
5008 (33)
455 (3)
5426 (35)
1001 (30)
1045 (31)
86 (3)
1247 (37)
3439 (29)
3963 (33)
369 (3)
4179 (35)
Time from diagnosis to index date, months
 Mean (SD)
 Median (IQR)
2.4 (5.3)
1.2 (0.7-1.9)
2.3 (4.8)
1.2 (0.6-1.8)
2.4 (5.4)
1.2 (0.7-2.0)

aPatients with a birth year of 1938 or earlier may have an adjusted birth year in Flatiron Health, Inc. datasets due to patient de-identification requirements; bRegions were defined by US Census Bureau Region; cMedicare is a federal and state government program that provides healthcare coverage to eligible individuals ≥ 65 years of age; d“Other” includes Medicaid (government program for those on low income) and is combined with multiple insurance types; eAt index (closest value in the 30-day preindex timeframe including index); fAny time preindex to 30 days postindex.

Abbreviations: ECOG, Eastern Cooperative Oncology Group; IQR, interquartile range; LDH, lactate dehydrogenase; SD, standard deviation.

LDH testing was performed in 3379 (22%) patients after their index date, while the remaining 78% of patients were not tested for LDH during therapy (83% of patients had also not been tested in the baseline period before starting chemotherapy).

Patient characteristics during follow-up

Median follow-up was 1.3 years from the index date among all patients (Table 2). Among all patients, 74% had received a total of 1 or 2 lines of therapy before the last follow-up day, and 60% of patients died (Table 2). Among LDH-tested and nontested patients, median follow-up time from index date was 1.6 years and 1.2 years, respectively; 66% and 76% received 1 or 2 lines of therapy, respectively; and 63% and 59% had died, respectively (Table 2).

Table 2.

Characteristics at follow-up.

Characteristics All patients
(N = 15 329)
LDH testing status after first-line therapy initiation
Tested
(n = 3379)
Not tested
(n = 11 950)
Median follow-up, years (IQR)
Mean follow-up, years (SD)
1.3 (0.6-2.4)
1.7 (1.4)
1.6 (0.9-2.7)
2 (1.5)
1.2 (0.6-2.2)
1.6 (1.4)
Duration of follow-up, years, n (%)
 <1
 1-<2
 2-<3
 3-<4
 4-<5
 ≥5
6066 (40)
4422 (29)
2376 (16)
1207 (8)
685 (4)
573 (4)
1038 (31)
988 (29)
646 (19)
338 (10)
194 (6)
175 (5)
5028 (42)
3434 (29)
1730 (14)
869 (7)
491 (4)
398 (3)
Deaths, n (%) 9232 (60) 2126 (63) 7106 (59)
Last line of therapy,an (%)
 1
 2
 3
 4
 ≥5
7048 (46)
4289 (28)
2177 (14)
1049 (7)
766 (5)
1206 (36)
1035 (31)
575 (17)
295 (9)
268 (8)
5842 (49)
3254 (27)
1602 (13)
754 (6)
498 (4)

aLines of therapy received before the last follow-up day.

Abbreviations: IQR, interquartile range; LDH, lactate dehydrogenase; SD, standard deviation.

LDH tests during follow-up

Among all LDH-tested patients (n = 3379), the median number of LDH tests per line of therapy ranged from 4 to 6 and the mean number of tests ranged from 6 to 10. Median duration of the line of therapy was 125-281 days, indicating that LDH testing was generally regularly carried out, on average every 3-4 weeks. The median time from line of therapy initiation to first LDH test was 7-10 days (Table 3).

Table 3.

LDH testing per person and over time by line of therapy for lines 1 to 5.

Therapy line
1 2 3 4 5
Total patients in the line,an 15 329 8281 3992 1815 766
Duration of the line, days, median (IQR) 281 (138-512) 202 (96-391) 133 (77-260) 118 (61-210) 125 (68-221)
Patients with ≥ 1 LDH test during the line, n (%) 2853 (19) 1527 (18) 712 (18) 303 (17) 152 (20)
Number of LDH tests per person during the line
 Mean (SD)
 Median (IQR)
10 (11)
6 (2-14)
9 (11)
5 (2-12)
7 (9)
5 (2-10)
6 (8)
4 (2-8)
6 (6)
4 (2-8)
Time to first LDH test, days, median (IQR) 9 (0-44) 7 (0-34) 8 (0-26) 9 (0-25) 10 (0-28)
Median LDH value (IQR) during the line, IU/L 206 (3-7891) 200 (55-9947) 214 (72-7355) 222 (91-8407) 242 (90-4160)

LDH tests performed on the start date of each line were included in calculations for that line.

aIncludes only patients who had an LDH test.

Abbreviations: IQR, interquartile range; LDH, lactate dehydrogenase; SD, standard deviation.

The mean and median number of LDH tests per person during the first line of therapy were 10 and 6 (interquartile range 2-14), respectively. Among those patients who received LDH testing, the number of tests performed decreased with progressive lines of therapy but remained regularly carried out (Table 3 and Figure S2).

Median LDH levels tended to increase across later lines of therapy (beyond second line) and ranged from 200-206 IU/L in the first 2 lines of therapy, and 214-242 IU/L in subsequent lines (Table 3).

Baseline LDH and overall survival

A total of 723/3379 (21%) LDH-tested patients had abnormal baseline LDH levels, 1336/3379 (40%) had normal baseline LDH, and 1320/3379 (39%) did not have any baseline LDH value (Table 4). Patients with abnormal baseline LDH values were more frequently female, ≥ 65 years of age, Black or African Americans, and resided in the Northeast region of the United States, when compared with those with normal baseline LDH (Table 4). The proportion of patients with ECOG performance status of 0 or 1 was lower in patients with abnormal baseline LDH compared with those who had normal baseline LDH (62% and 74%, respectively), and a greater proportion had stage IV disease at initial diagnosis (72% and 54%, respectively) (Table 4).

Table 4.

Baseline demographics and clinical characteristics by LDH value at baseline.

Characteristics LDH values at baseline (of all LDH-tested patients; n = 3379)
Abnormala LDH
(n = 723)
Normala LDH
(n = 1336)
Missing
(n = 1320)
Age, yearsb
 Median (IQR) 65 (55-72) 64 (55-73) 63 (54-71)
Age group, n (%) years
 <50
 50-64
 ≥65
95 (13)
253 (35)
375 (52)
178 (13)
500 (37)
658 (49)
208 (16)
505 (38)
607 (46)
Sex, n (%)
 Male
 Female
380 (53)
342 (47)
790 (59)
546 (41)
718 (54)
602 (46)
Race, n (%)
 White
 Black/African American
 Other
 Unknown
452 (63)
89 (12)
117 (16)
65 (9)
945 (71)
94 (7)
209 (16)
88 (7)
864 (65)
154 (12)
200 (15)
102 (8)
Ethnicity, n (%)
 Hispanic
 Other
 Unknown
39 (5)
507 (70)
177 (24)
72 (5)
935 (70)
329 (25)
92 (7)
993 (75)
235 (18)
Region, n (%)c
 South
 West
 Northeast
 Midwest
 Unknown
299 (41)
54 (7)
217 (30)
87 (12)
66 (9)
688 (51)
113 (8)
285 (21)
176 (13)
74 (6)
480 (36)
146 (11)
163 (12)
213 (16)
318 (24)
Practice type, n (%)
 Academic
 Community
 Both
39 (5)
677 (94)
7 (1)
43 (3)
1285 (96)
8 (1)
285 (22)
1014 (77)
21 (2)
Insurance type, n (%)
 Commercial
 Medicared
 Othere
 Unknown
334 (46)
87 (12)
200 (28)
102 (14)
619 (46)
178 (13)
340 (25)
199 (15)
666 (50)
168 (13)
315 (24)
171 (13)
ECOG performance status, n (%)f
 0
 1
 2
 3
 Unknown
232 (32)
217 (30)
72 (10)
25 (3)
177 (24)
577 (43)
412 (31)
98 (7)
19 (1)
230 (17)
441 (33)
412 (31)
84 (6)
14 (1)
369 (28)
Stage IV disease at initial diagnosis, n (%) 524 (72) 719 (54) 816 (62)
KRAS mutation status, n (%)g
 Positive
 Negative
 Unknown
 Not tested
203 (28)
241 (33)
27 (4)
252 (35)
400 (30)
414 (31)
33 (2)
489 (37)
398 (30)
390 (30)
26 (2)
506 (38)
Time from diagnosis to index date, months
 Mean (SD)
 Median (IQR)
1.9 (4.8)
0.9 (0.5-1.5)
2.3 (4.6)
1.2 (0.7-2.0)
2.4 (5.0)
1.2 (0.6-1.9)

aAbnormal and normal LDH values were defined through individual patient testing and corresponding laboratory reference ranges; an LDH value was considered normal if it fell within the reference range used by the individual laboratory; bPatients with a birth year of 1938 or earlier may have an adjusted birth year in Flatiron Health, Inc. datasets due to patient de-identification requirements; cRegions were defined by US Census Bureau Region; dMedicare is a federal and state government program that provides healthcare coverage to eligible individuals ≥ 65 years of age; e“Other” includes Medicaid (government program for those on low income) and is combined with multiple insurance types; fAt index (closest value in the 30-day preindex timeframe including index); gAny time preindex to 30 days postindex.

Abbreviations: ECOG, Eastern Cooperative Oncology Group; IQR, interquartile range; LDH, lactate dehydrogenase; SD, standard deviation.

Median OS was longer in patients with normal vs abnormal baseline LDH (29.9 months [95% confidence interval (CI), 28.3-31.7] vs 16.8 months [95% CI, 14.8-18.2]) (Figure 2). The relationship between abnormal baseline LDH values and worse OS was supported by an adjusted multivariate Cox proportional-hazards model (abnormal vs normal LDH HR 1.91; 95% CI, 1.70-2.15; P < .0001). A shorter OS was also noted in community practice (HR 4.22 [95% CI, 2.70-6.59; P < .0001] for community vs academic) and in patients with worse performance status or KRAS mutation (HR 1.38 [95% CI, 1.19-1.59; P < .0001] vs no KRAS mutation) (Table 5).

Figure 2.

Graph and data comparing survival in months of groups of patients with normal and abnormal LDH levels, showing survival curves and median overall survival times for both groups.

Kaplan–Meier graph of OS in patients with normal and abnormal LDH values at baseline. Abnormal and normal LDH values were defined through individual patient testing and corresponding laboratory reference ranges; an LDH value was considered normal if it did not exceed the upper limit of the reference range used by the individual laboratory. Abbreviations: CI, confidence interval; LDH, lactate dehydrogenase; OS, overall survival.

Table 5.

HRs for overall survival calculated using a multivariate Cox proportional-hazards model.

HR 95% CI P
Abnormal vs normal LDH 1.91 1.70-2.16 <.0001
Agea 1.02 1.01-1.02 <.0001
Sex
 Female vs male 0.89 0.79-0.99 .037
Race
 White vs Black/African American
 Other race vs Black/African American
 Unknown vs Black/African American
1.05
0.82
1.14
0.87-1.28
0.64-1.04
0.86-1.52
.610
.099
.358
Region
 Midwest vs South
 Northeast vs South
 West vs South
 Unknown vs South
0.90
0.90
0.87
3.55
0.75-1.07
0.78-1.04
0.70-1.08
2.47-5.08
.220
.149
.214
<.0001
Practice type
 Community vs academicb 4.22 2.70-6.59 <.0001
Payer provider
 Commercial vs Medicare
 Otherc vs Medicare
 Unknown vs Medicare
1.17
1.17
0.68
0.99-1.39
0.97-1.41
0.55-0.86
.073
.103
.001
CRC stage at diagnosis
 Stage III vs I/II
 Stage IV vs I/II
 Unknown vs I/II
1.14
1.14
1.33
0.94-1.39
0.96-1.36
0.94-1.89
.181
.143
.111
ECOG performance status
 1 vs 0
 2 vs 0
 3 vs 0
 Unknown vs 0
1.39
1.96
2.53
1.15
1.22-1.60
1.60-2.39
1.79-3.57
0.97-1.36
<.0001
<.0001
<.0001
.099
KRAS mutation status
 Positive vs negative
 Unknown vs negative
1.38
1.06
1.19-1.59
0.93-1.22
<.0001
.380
Months from diagnosis of mCRCa 1.00 0.99-1.01 .910

HRs for abnormal LDH, sex, race, region, practice type, stage, ECOG performance status, KRAS mutation status, and payer type were calculated using the multivariate Cox proportional-hazards model. HR calculations are vs “other” within the same covariate category, unless otherwise specified.

aContinuous variable; bAcademic category includes academic plus commercial for modeling; cIncludes Medicaid for modeling.

Abbreviations: CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; LDH, lactate dehydrogenase; mCRC, metastatic colorectal cancer.

Discussion

This real-world study of more than 15 000 patients receiving chemotherapy for mCRC in the United States demonstrated that overall rates of LDH testing are low in current US clinical practice, with only one-fifth of patients having undergone testing. These data suggest that LDH testing is not widely considered standard of care for patients with mCRC. Compared with non-LDH-tested patients, a greater proportion of LDH-tested patients were White and less frequently resided in the Western region of the United States. Compared with patients who had normal LDH values in the baseline period, a greater proportion of patients with abnormal LDH values were female, ≥ 65 years of age, Black or African American, and resided in the Northeast region of the United States. Interestingly, there were no disparities with respect to insurance type for LDH-tested vs nontested patients or with LDH test results.

Rates of LDH testing were generally consistent across lines of therapy, although the mean and median number of LDH tests per person decreased with progressive lines of therapy, presumably because of the decreasing duration of each line. The time intervals between successive LDH tests decreased over time, and although median LDH values increased with line of therapy, these were consistently < 250 IU/L.

The real-world cohort EHR data revealed an increase in serum LDH abnormality over time and with line of treatment, and this is of clinical relevance as it signifies a need to consider alternative treatment options earlier in the treatment paradigm. As seen for patients with poorer ECOG performance status and KRAS mutation, abnormal levels of LDH in those patients tested was associated with an increased risk of shorter OS. This finding is unsurprising given the known association of serum LDH with tumor hypoxia, neoangiogenesis, and metastasis, together with prior research associating high levels of pretreatment LDH with worse OS.7-10 Since prior retrospective studies have also indicated a potential prognostic and predictive role for serum LDH levels in treating patients with mCRC with chemotherapy and antiangiogenic agents, with abnormally high levels associated with a lack of efficacy,12-17 serum LDH level testing during treatment may help guide the subsequent use of alternative nonchemotherapeutic therapeutic agents.

In common with other observational studies, this study was limited by its retrospective nature, including the potential presence of unknown confounding factors, and the resulting lack of inherent ability to account or control for those. Test rates for LDH are shown to be low in current US clinical practice, and although this suggests a lack of standard practice, the study is limited to those LDH tests correctly entered and subsequently selected from the database. Test records were selected using a wide range of key words to minimize missed LDH test entries. However, some tests may still have been missed where they were incorrectly classified or hidden among other results comprising part of a panel of tests. Nonetheless, the real-world database used includes large numbers of patients across multiple geographic locations, thereby increasing the representativeness of the study population and enhancing the generalizability of the study’s findings.

In conclusion, the consistently low rates of LDH testing across lines of therapy observed in this real-world study have implications for clinical practice as they undermine the potential value of LDH as a prognostic and predictive marker in mCRC, particularly as abnormal baseline LDH levels are associated with shorter OS. Since LDH testing is a readily available, minimally invasive, and easily accessible test that can identify patients who potentially have worse outcomes, efforts to increase LDH testing could be valuable in helping guide treatment decisions.

Supplementary Material

oyaf192_suppl_Supplementary_Figures_S1-S2
oyaf192_suppl_Supplementary_Tables_S1

Acknowledgments

Medical writing and editorial support were provided by Robyn Fowler, PhD, and Matthew Naylor, PhD, of Luna, OPEN Health Communications, London, UK.

Contributor Information

Heinz-Josef Lenz, Division of Medical Oncology, Norris Comprehensive Cancer Center, Keck School of Medicine, University of Southern California, Los Angeles, CA 90033, United States.

Orsolya Lunacsek, Bayer HealthCare Pharmaceuticals, Whippany, NJ 07981, United States.

Helene Ostojic, Bayer Consumer Care AG, 4052 Basel, Switzerland.

Xiaoyun Pan, Bayer HealthCare Pharmaceuticals, Whippany, NJ 07981, United States.

Emmanuelle Dochy, Bayer Consumer Care AG, 1831 Machelen, Belgium.

Zdravko Vassilev, Bayer HealthCare Pharmaceuticals, Whippany, NJ 07981, United States.

Nasreen Khan, Bayer HealthCare Pharmaceuticals, Whippany, NJ 07981, United States.

Author contributions

Conceptualization, Heinz-Josef Lenz, Helene Ostojic, Emmanuelle Dochy, Zdravko Vassilev, Nasreen Khan; Data curation, Nasreen Khan; Formal Analysis, Orsolya Lunacsek, Nasreen Khan; Funding Acquisition, Nasreen Khan; Investigation, Orsolya Lunacsek, Nasreen Khan; Methodology, Heinz-Josef Lenz, Zdravko Vassilev, Nasreen Khan; Project Administration, Nasreen Khan; Resources, Nasreen Khan; Supervision, Nasreen Khan; Validation, Orsolya Lunacsek, Nasreen Khan; Visualization, Nasreen Khan; Writing—original draft, all authors; Writing—review & editing, all authors.

Funding

This study was funded by Bayer, who also funded medical writing and editorial support in accordance with Good Publication Practice (GPP) guidelines (www.ismpp.org/gpp-2022).

Conflicts of interest

O.L., H.O., X.P., E.D., Z.V., and N.K. are employees of Bayer. H.-J.L. has no disclosures to make.

Data availability

The data that support the findings of this study were originated by and are the property of Flatiron Health, Inc., which has restrictions prohibiting the authors from making the dataset publicly available. Requests for data sharing by license or by permission for the specific purpose of replicating results in this manuscript can be submitted to PublicationsDataAccess@flatiron.com.

References

  • 1. National Cancer Institute Surveillance, Epidemiology, and End Results Program. Cancer stat facts: colorectal cancer. Accessed December 10, 2024. https://seer.cancer.gov/statfacts/html/colorect.html. [Google Scholar]
  • 2. Morris VK, Kennedy EB, Baxter NN, et al. Treatment of metastatic colorectal cancer: ASCO guideline. J Clin Oncol. 2023;41:678-700. https://doi.org/ 10.1200/JCO.22.01690 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Goldberg RM, Rothenberg ML, Van Cutsem E, et al. The continuum of care: a paradigm for the management of metastatic colorectal cancer. Oncologist. 2007;12:38-50. https://doi.org/ 10.1634/theoncologist.12-1-38 [DOI] [PubMed] [Google Scholar]
  • 4. Grothey A, Sargent D.. Overall survival of patients with advanced colorectal cancer correlates with availability of fluorouracil, irinotecan, and oxaliplatin regardless of whether doublet or single-agent therapy is used first line. J Clin Oncol. 2005;23:9441-9442. https://doi.org/ 10.1200/JCO.2005.04.4792 [DOI] [PubMed] [Google Scholar]
  • 5. Peluso G, Incollingo P, Calogero A, et al. Current tissue molecular markers in colorectal cancer: a literature review. Biomed Res Int. 2017;2017:2605628. https://doi.org/ 10.1155/2017/2605628 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Ivanova M, Venetis K, Guerini-Rocco E, et al. HER2 in metastatic colorectal cancer: pathology, somatic alterations, and perspectives for novel therapeutic schemes. Life (Basel). 2022;12:1403. https://doi.org/ 10.3390/life12091403 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Forkasiewicz A, Dorociak M, Stach K, et al. The usefulness of lactate dehydrogenase measurements in current oncological practice. Cell Mol Biol Lett. 2020;25:35. https://doi.org/ 10.1186/s11658-020-00228-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Kong W, Zuo X, Liang H, et al. Prognostic value of lactate dehydrogenase in patients with hepatocellular carcinoma: a meta-analysis. Biomed Res Int. 2018;2018:1723184. https://doi.org/ 10.1155/2018/1723184 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Passardi A, Scarpi E, Tamberi S, et al. Impact of pre-treatment lactate dehydrogenase levels on prognosis and bevacizumab efficacy in patients with metastatic colorectal cancer. PLoS One. 2015;10:e0134732. https://doi.org/ 10.1371/journal.pone.0134732 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Zhang J, Yao YH, Li BG, et al. Prognostic value of pretreatment serum lactate dehydrogenase level in patients with solid tumors: a systematic review and meta-analysis. Sci Rep. 2015;5:9800. https://doi.org/ 10.1038/srep09800 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Marmorino F, Salvatore L, Barbara C, et al. Serum LDH predicts benefit from bevacizumab beyond progression in metastatic colorectal cancer. Br J Cancer. 2017;116:318-323. https://doi.org/ 10.1038/bjc.2016.413 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Scartozzi M, Giampieri R, Maccaroni E, et al. Pre-treatment lactate dehydrogenase levels as predictor of efficacy of first-line bevacizumab-based therapy in metastatic colorectal cancer patients. Br J Cancer. 2012;106:799-804. https://doi.org/ 10.1038/bjc.2012.17 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Feng W, Wang Y, Zhu X.. Baseline serum lactate dehydrogenase level predicts survival benefit in patients with metastatic colorectal cancer receiving bevacizumab as first-line chemotherapy: a systematic review and meta-analysis of 7 studies and 1,219 patients. Ann Transl Med. 2019;7:133. https://doi.org/ 10.21037/atm.2019.02.45 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Bar J, Spencer S, Morgan S, et al. Correlation of lactate dehydrogenase isoenzyme profile with outcome in patients with advanced colorectal cancer treated with chemotherapy and bevacizumab or cediranib: retrospective analysis of the HORIZON I study. Clin Colorectal Cancer. 2014;13:46-53. https://doi.org/ 10.1016/j.clcc.2013.11.002 [DOI] [PubMed] [Google Scholar]
  • 15. Cetin B, Kaplan MA, Berk V, et al. Prognostic factors for overall survival in patients with metastatic colorectal carcinoma treated with vascular endothelial growth factor-targeting agents. Asian Pac J Cancer Prev. 2012;13:1059-1063. https://doi.org/ 10.7314/apjcp.2012.13.3.1059 [DOI] [PubMed] [Google Scholar]
  • 16. Díaz R, Aparicio J, Gironés R, et al. Analysis of prognostic factors and applicability of Kohne’s prognostic groups in patients with metastatic colorectal cancer treated with first-line irinotecan or oxaliplatin-based chemotherapy. Clin Colorectal Cancer. 2005;5:197-202. https://doi.org/ 10.3816/ccc.2005.n.031 [DOI] [PubMed] [Google Scholar]
  • 17. Lin JT, Wang WS, Yen CC, et al. Outcome of colorectal carcinoma in patients under 40 years of age. J Gastroenterol Hepatol. 2005;20:900-905. https://doi.org/ 10.1111/j.1440-1746.2005.03893.x [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

oyaf192_suppl_Supplementary_Figures_S1-S2
oyaf192_suppl_Supplementary_Tables_S1

Data Availability Statement

The data that support the findings of this study were originated by and are the property of Flatiron Health, Inc., which has restrictions prohibiting the authors from making the dataset publicly available. Requests for data sharing by license or by permission for the specific purpose of replicating results in this manuscript can be submitted to PublicationsDataAccess@flatiron.com.


Articles from The Oncologist are provided here courtesy of Oxford University Press

RESOURCES