Foregut malignancy after discordant fecal immunochemical test: A cohort study at a United States safety-net hospital

Table 1 Baseline characteristics of discordant and concordant groups, n (%) or median (interquartile range)

Characteristic	Discordant (n = 172)	Concordant (n = 207)	P value
Age	61 (55-67)	61 (56-66)	0.78
Male sex	63 (36.6)	100 (48.3)	0.03
Race/ethnicity
Hispanic	129 (75.0)	147 (71.0)	0.42
White non-Hispanic	4 (2.3)	15 (7.2)	0.03
Asian	9 (5.2)	11 (5.3)	1.00
Black	12 (7.0)	9 (4.3)	0.37
Other/unknown	18 (10.5)	25 (12.1)	0.75
Foregut symptoms	76 (44.2)	79 (38.2)	0.25
Anemia at FIT	60 (34.9)	77 (37.2)	0.67
Hemoglobin (g/L)	128 (112-140)	131 (114-145)	0.18
MCV	88.6 (83.3-91.0)	89.3 (83.8-93.1)	0.10
Cirrhosis	18 (10.5)	58 (28.0)	< 0.001
NSAID/aspirin use	61 (35.5)	57 (27.5)	0.12
Smoking status
Never	145 (84.3)	164 (79.2)
Former	21 (12.2)	28 (13.5)
Current	6 (3.5)	15 (7.2)
Ever-smoker	27 (15.7)	43 (20.8)	0.23
Alcohol use
None	121 (70.3)	118 (57.0)
Low-risk drinking	22 (12.8)	18 (8.7)
Heavy drinking/AUD	9 (5.2)	35 (16.9)	< 0.001
Any alcohol use	31 (18.0)	53 (25.6)	0.08
Not documented	20 (11.6)	36 (17.4)
Insurance type			0.045
Medi-Cal	112 (65.1)	111 (53.6)
Medicare	18 (10.5)	20 (9.7)
Uninsured	10 (5.8)	12 (5.8)
Private	2 (1.2)	0 (0.0)
Other	10 (5.8)	28 (13.5)
Not documented	20 (11.6)	36 (17.4)
Gastric biopsy obtained	162 (94.2)	193 (93.2)	0.83
Same-day bidirectional	79 (45.9)	101 (48.8)	0.61

P values by Fisher’s exact test (categorical) or Mann-Whitney U test (continuous). Insurance compared by χ² test. Low-risk drinking defined as social or moderate alcohol consumption within recommended limits. Heavy drinking/alcohol use disorder defined as alcohol intake exceeding recommended limits or diagnosed alcohol use disorder. FIT: Fecal immunochemical test; MCV: Mean corpuscular volume; NSAID: Nonsteroidal anti-inflammatory drug; AUD: Alcohol use disorder.

Table 2 Esophagogastroduodenoscopy findings by colonoscopy concordance group (n = 379), n (%)

EGD finding	Discordant (n = 172)	Concordant (n = 207)	P value
Foregut malignancy1	11 (6.4)	0 (0.0)	< 0.001
Premalignant lesion2	4 (2.3)	16 (7.7)	0.02
Esophagitis	18 (10.5)	19 (9.2)	0.73
Gastritis	40 (23.3)	32 (15.5)	0.07
Peptic ulcer	10 (5.8)	9 (4.3)	0.64
GAVE	0 (0.0)	4 (1.9)	0.13
Gastric polyp	14 (8.1)	19 (9.2)	0.86
H. pylori positive	18 (10.5)	23 (11.1)	0.87
No abnormal findings	94 (54.7)	119 (57.5)	0.60

¹All 11 foregut malignancies occurred in the discordant group (complete separation). See Table 3 for adjusted analysis.

²Premalignant lesions were defined as gastric intestinal metaplasia, Barrett’s esophagus, or esophageal/gastric dysplasia. All cases in this cohort were gastric intestinal metaplasia (n = 19) or Barrett’s (n = 1); no dysplasia was identified.

P values by Fisher’s exact test. GAVE: Gastric antral vascular ectasia; H. pylori: Helicobacter pylori; EGD: Esophagogastroduodenoscopy.

Table 3 Firth penalized logistic regression: Predictors of foregut malignancy (exploratory analysis, n = 379)

Variable	aOR	95%CI	P value
Discordant status	26.14	2.16-315.82	< 0.001a
Age (per year)	1.05	1.01-1.10	0.036a
Male sex	1.45	0.46-4.57	0.554
Anemia	1.71	0.49-6.01	0.439
Foregut symptoms	3.52	0.96-12.90	0.067
NSAID/aspirin use	1.00	0.30-3.34	0.995
Cirrhosis	1.36	0.23-8.17	0.758

^aP < 0.05.

Model includes all listed variables simultaneously. Covariates selected a priori based on clinical relevance. Firth penalized maximum likelihood was used because all 11 foregut malignancies occurred in the discordant group (complete separation). P values are from penalized likelihood-ratio tests and confidence intervals are Wald-type. This analysis is exploratory given the small number of events (n = 11) relative to the number of covariates. The wide confidence interval for discordant status reflects the limited precision inherent to rare-event analyses with complete separation. aOR: Adjusted odds ratio; NSAID: Nonsteroidal anti-inflammatory drug.

Table 4 Patient-level characteristics of the 11 foregut malignancy cases (demographics, timing, and clinical presentation)

Patient	Age	Sex	Interval (days)	Same-day	Foregut Sx	Anemia	EGD indication	Suspected before EGD	Cancer site	Smoking	Alcohol
1	66	F	0	Yes	Yes	No	GERD	No	Gastric	Never	None
2	54	M	0	Yes	Yes	Yes	GIM/H. pylori surveillance	No	Gastric	Never	Social
3	56	M	36	No	Yes	No	GERD	No	Gastric	Never	None
4	69	F	54	No	No	Yes	GERD	No	Gastric (NET)	Never	None
5	62	F	75	No	Yes	N/A	IDA workup, esophageal thickening on CT	Yes	Gastric (NET)	Never	None
6	70	M	96	No	No	Yes	Anemia workup	No	Gastric	Never	None
7	71	F	102	No	No	No	Abnormal CT-gastric wall thickening	Yes	GEJ	Never	None
8	69	F	121	No	Yes	No	Abnormal CT-gastric outlet obstruction	Yes	Gastric	Never	None
9	55	M	237	No	Yes	No	GERD	No	Gastric	Never	None
10	70	M	280	No	Yes	Yes	Dysphagia	No	Esophageal	Never	None
11	69	F	350	No	No	No	GIM surveillance	No	Gastric	Never	None

EGD: Esophagogastroduodenoscopy; GERD: Gastroesophageal reflux disease; F: Female; M: Male; NET: Neuroendocrine tumor; GIM: Gastric intestinal metaplasia; H. pylori: Helicobacter pylori; N/A: Not applicable; IDA: Iron deficiency anemia; CT: Computed tomography; GEJ: Gastroesophageal junction; Sx: Symptoms.

Table 5 Patient-level characteristics of the 11 foregut malignancy cases (imaging, staging, and outcomes)

Patient	Prior imaging	Imaging finding	Stage (TNM)	Metastatic	Treatment	Vital status	Prior UGI surg
1	None	N/A	Not staged	No	None	Alive	No
2	None	N/A	Not staged	No	None	Alive	No
3	None	N/A	Not staged	No	None	Alive	No
4	None	N/A	T1N0M0	No	Endoscopic resection	Alive	No
5	CT	Esophageal thickening	T2N0M0	No	Surgery	Alive	No
6	None	N/A	T1N0M0	No	None	Alive	No
7	CT	Gastric wall thickening, lymphadenopathy	T2N0M0	No	Surgery + chemotherapy	Alive	No
8	CT	Gastric outlet obstruction	T3N0M0	No	Surgery + chemotherapy	Deceased	No
9	None	N/A	T3N1M0	No	Chemotherapy + radiation	Deceased	No
10	None	N/A	T1N0M0	No	None	Alive	No
11	None	N/A	Not staged	No	None	Alive	No

Patients are listed in order of colonoscopy-to-esophagogastroduodenoscopy interval. All 11 patients had discordant colonoscopy (fecal immunochemical test-positive with non-explanatory colonoscopy). None of the 11 patients were ever-smokers, and none had heavy alcohol use or alcohol use disorder. None had prior upper gastrointestinal surgery. Fecal immunochemical test quantitative values were not available for any patient. N/A: Not applicable; CT: Computed tomography; UGI: Upper gastrointestinal.