Outcomes of excessive alcohol drinkers without baseline evidence of chronic liver disease after 15 years follow-up: Heavy burden of cancer and liver disease mortality.

BACKGROUND: Most long-term heavy drinkers do not have clinically evident chronic liver disease (CLD). However, at any time-point, their risk of developing CLD remains unknown. We aimed to evaluate the long-term outcomes of a group of heavy drinkers, without evidence of CLD at baseline.
METHODS: A cohort of 123 long-term heavy drinkers without CLD were prospectively recruited in 2002 and retrospectively followed until 2018.
RESULTS: At baseline (2002), median alcohol consumption was 271±203g/day during 21.5±20 years, 65% being abstinent during the previous 1.75±5 months. Patients were followed for 14±3 years. During follow-up, 53% reported any alcohol intake. Alcohol consumption during follow-up associated weakly with either 1- or 6-months previous abstinence at baseline. Until 2018, progression to CLD occurred in 6%, associating with years of alcohol intake during follow-up (OR 1.15 [1.01-1.31]) and baseline alkaline-phosphatase (OR 1.05 [1.01-1.10]). During follow-up, being abstinent for at least 1 year positively associated with CLD-free survival. 27% died (55% of cancer-mostly oropharyngeal cancer, 27% of cardiovascular disease, and 9% of liver disease), with a mean age of 71 years [69-74] (10 years less than the expected in the Portuguese population). Achieving abstinence for at least 1 year positively associated with overall survival, while smoking, and hepatic steatosis at baseline associated negatively.
CONCLUSION: Long-term heavy drinkers seemed to have a decreased life expectancy compared with the overall Portuguese population. Cancer was the main cause of death. Our results suggest that progression to CLD depends mostly on continued alcohol intake. Alcohol abstinence, even if temporary, seems to decrease the risks of CLD and mortality.

Introduction

Alcohol consumption is the third preventable cause of mortality, and the 7th cause of mortality, accounting for 5% of deaths worldwide [1]. Any amount of alcohol intake is associated with increased mortality [1]. Alcohol consumption is associated with more than 200 diseases [2], being the liver the single organ most commonly affected by excessive drinking [3]. Alcohol intake is responsible for up to 50% of cases of liver cirrhosis worldwide [4], with geographical differences: alcohol contributes to 15% of cirrhosis-related admissions in Africa, 48% in the US, and 72% in Europe [5]. However, only 15%-20% of heavy-drinkers will develop liver cirrhosis [2, 3]. The amount of alcohol intake seems critical in inducing liver disease. Indeed, there seems to be a dose-dependent effect between the amount of alcohol intake and the risk of developing chronic liver disease (CLD) [6], as well as, a threshold effect, with alcohol intake higher than 30g/day in men and 20g/day in women being considered potentially hepatotoxic [7]. The duration of alcohol consumption necessary to develop liver cirrhosis is unknown [8, 9] but there seems to be a linear increase with time. Development of liver cirrhosis has been described after just 4 years of alcohol consumption, however, the prevalence of cirrhosis increases exponentially after 8–12 years of consumption [10]. Several known co-factors increase the risk for alcohol-associated liver cirrhosis, such as hepatitis C virus infection, obesity, diabetes-mellitus, genetic factors (for example, polymorphisms in the PNPLA3 and TM6SF2 genes), smoking, and low coffee intake [3]. On the other hand, abstinence is known to decrease the risk for progression to cirrhosis, and to improve the prognosis of patients with cirrhosis [11, 12].

Little is known regarding the long-term outcomes of alcohol harmful consumers without known liver disease. Are those patients still at risk for developing CLD? We followed, for 15 years, a well-characterized cohort of harmful alcohol consumers, without evidence of CLD at baseline, in order to evaluate their long-term outcomes and to identify risk factors for mortality and further development of CLD.

Material and methods

Study design and patients

Patients with harmful alcohol consumption and evidence of alcohol-dependence, followed in an Alcohol Rehabilitation Clinic, were prospectively and consecutively enrolled in a clinical protocol performed in 2002, with data partially reported in 2005 [13], and retrospectively re-evaluated in 2018. At enrollment, all patients attending outpatient visits at Centro Regional de Alcoologia do Sul (CRAS) were asked to participate in the study and were evaluated at a Hepatology visit.

At baseline, inclusion criteria were: 1) alcohol consumption of at least 50 grams per day during at least 5 years, 2) age greater than 18 years, and 3) agreement to participate in the study. Abstinence at recruitment was accepted. Exclusion criteria were: 1) presence of liver disease not related to alcohol (i.e. infection by hepatitis B/C virus, primary biliary cholangitis, autoimmune hepatitis, primary sclerosing cholangitis, Wilson’s disease, hemochromatosis, or α1-antitrypsin deficiency), 2) treatment with potentially hepatotoxic drugs within 6 months of enrollment, and 3) active cancer. After thorough clinical evaluation by a dedicated Hepatologist, patients with clinical evidence of CLD or with altered liver blood tests were also excluded. CLD was defined by the presence of a past medical history of hepatic decompensation (jaundice, ascites, hepatic encephalopathy, and/or variceal bleeding), or histological confirmation of liver disease of greater severity than steatosis (liver biopsy was performed in 16 patients). Requisites for the absence of CLD included: the absence of a history of hepatic decompensation, absence of physical stigmata of chronic liver disease (i.e. cutaneous signs, hepatosplenomegaly, gynecomastia, and testicular atrophy), and at least two normal determinations of blood liver tests (aminotransferases, alkaline phosphatase, total bilirubin, prothrombin time and albumin) with an exception of an isolated rise in γ-glutamyl transpeptidase (GGT). Isolated increase of GGT was not an exclusion criteria, since it is a marker of alcohol consumption and not of liver injury, translating hepatic enzyme induction rather than liver cell injury [14]. Of the 262 patients screened, 130 did not have evidence of CLD, 103 had alcohol-associated liver disease, and the remaining 29 had either other liver diseases or mild increase in liver enzymes (that is, an increase lower than 3 times the reference value). Isolated increase of GGT was found in 39 patients (31.7%) without evidence of liver disease. Patients to whom there were clinical doubts regarding the presence of liver disease motivating the need for a liver biopsy were excluded. Those patients could not be classified as absence of liver disease (according to protocol) even if the biopsy was normal, because sampling error could not be excluded. Lastly, in the group of selected patients without evidence of CLD, we retrospectively applied the non-invasive score FIB-4 [15] to assess advanced fibrosis, and the ALBI-FIB4 score [17] to assess the risk of hepatic decompensation, at baseline.

Data were collected through an interviewer-administered semi-structured questionnaire including alcohol intake assessment, clinical examination, and blood tests.

All participants gave written informed consent, and the study was approved by the Ethics committee of the institutions involved: Comissão de Ética do Centro Académico de Lisboa—CAML in 2002 and again in 2018 (ID number 13/18).

Baseline evaluation

Demographics and alcohol consumption data

In 2002, the questionnaire included demographic data (age, gender, race) and details regarding alcohol intake. Patients were questioned about total duration of alcohol consumption, abstinence periods, beverage type (wine, beer, or spirits), and average daily amount of alcohol intake (expressed in grams of pure ethanol). The baseline questionnaire also included smoking habits, medication, and family history of alcohol abuse or liver disease.

Clinical and biochemical assessment

A physical examination was performed including height (kg), weight (m), body mass index (BMI) (kg/m2), and the presence of stigmata of alcohol consumption or CLD.

The following blood tests were performed: blood count and liver tests, namely alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), GGT, bilirubin, albumin, and prothrombin time. To exclude other causes of CLD, viral serologies (hepatitis B and C), autoantibodies screen, iron metabolism, ceruloplasmin, and α1-antitrypsin serum levels were performed. Renal function, serum sodium and potassium, lipid profile, and serum fasting glucose were also collected. All were analyzed at the Hospital’s Central Pathology Department, using standardized methods.

All patients had a previous abdominal ultrasound performed at the rehabilitation clinic, within the previous 6 months. The presence of hepatomegaly, steatosis, splenomegaly, and ascites were recorded.

Follow-up

Patients were re-evaluated in 2018. The end of follow-up was the date of the last clinical interview or the date of death. All data were obtained from the computerized national health database (including hospital admissions or outpatient visits, primary care visits, blood tests, and imaging results), from consultation of the records in the Alcohol Rehabilitation Clinic at 2011, and through a survey performed by telephone interview to the patient or patient´s relatives. The parameters evaluated were: 1) co-morbidities such as diabetes-mellitus, dyslipidaemia, cardiovascular disease, or cancer; 2) assessment of alcohol intake to estimate lifetime alcohol exposure, abstinence periods, duration and quantification of alcohol intake and type of beverage, 3) medication; 4) diagnosis of liver cirrhosis or CLD, and the number of liver decompensations/hospitalizations; 5) mortality and cause of death.

Patients were considered to have developed CLD when they presented abnormal liver tests for more than 6 months, imaging findings compatible with CLD, or presented with hepatic decompensation (jaundice, ascites, hepatic encephalopathy, and/or variceal bleeding).

Statistical analysis

The primary endpoint was the estimation of the risk for developing alcohol-related CLD and associated risk factors, in long-term harmful alcohol consumers without baseline CLD. The secondary endpoint was the evaluation of overall mortality, causes of death, and associated risk factors.

Continuous variables were expressed as median±IQR or mean±standard deviation (SD) depending on the absence or presence of a normal distribution (according to Kolmogorov-Smirnov test), respectively. Continuous variables with normal distribution were compared using the independent t-Student test. Categorical variables were described using frequencies and compared using the chi-square test. The independence of the associations of variables with the development of CLD and survival were assessed by multivariable logistic regression analysis. Kaplan-Meier survival curves and Cox-regression were used to assess the outcomes CLD free- and overall-survival. Analyses were performed using IBM Statistical Package for the Social Sciences (SPSS) software version 21. Two-tailed p values less than 0.05 were considered statistically significant.

Results

General characteristics and demographics of the study cohort

At baseline, 130 patients with harmful alcohol consumption without evidence of CLD were recruited from an alcohol rehabilitation clinic. Of those, 7 were lost for follow-up, and 123 were included in the analysis. One hundred and one were male (82%) and the mean age was 45±11years. The mean duration of follow-up was 14±3years (maximum 16 years). Although, at baseline, no patient had evidence of CLD, 11 (9%) had previous biochemical changes compatible with asymptomatic ambulatory alcoholic hepatitis. At the time of recruitment, 2002, transient elastography was still not available for clinical practice, and it was deemed unethical to perform a liver biopsy to confirm the absence of liver disease. To better characterize this population, we retrospectively applied FIB-4 [15], a well accepted noninvasive score for liver fibrosis assessment. FIB-4 >3.25 (cutoff suggesting significant fibrosis) was present in one patient, and FIB-4 >1.45 (cutoff used to exclude fibrosis) was present in 9 patients. Interestingly, no patient with FIB-4 >1.45 progressed to CLD on follow-up. We also applied the ALBI-FIB-4 [16] score and all patients presented a low risk for liver disease decompensation (the mean score was -3.91±0.51, minimum -5.308 and maximum -2.359). Concerning family history, 26 patients (21%) reported a family history of CLD and 75 (61%) of alcohol use disorder. Eighty-one (66%) were smokers, 48 (46%) overweight/obese (28 obese, 23%) and 6 patients had type 2 diabetes mellitus (5%). The demographic and clinical characteristics are presented in Table 1.

Table 1

Patient characteristics at inclusion.

	All Patients	No liver events on follow-up	Liver events on follow-up	p
Patients (N)	123	116	7
Age (years)*	45.5±11	45±11	52±9	0.124
Male sex	101 (82%)	95 (83%)	5 (71%)	0.607
Body mass index (kg/m2)*	25±4	25±4	25±4	0.982
Obese/overweight	56 (46%)	53 (46%)	3 (43%)	0.849
Presence of alcohol stigmata	93 (76%)	100 (86%)	6 (86%)	1.000
Smoking habits	81 (66%)	75 (65%)	6 (86%)	0.420
Family history of alcoholism	75 (61%)	74 (64%)	7 (100%)	0.096
Family history of liver disease	28 (23%)	25 (22%)	2 (33%)	0.618
Alcohol intake per day (g)	271±203	286±207	237±192	0.449
Years of alcohol intake	21.5±20	20±18	38±36	0.423
Total alcohol intake (tons)	2.5±2.4	2.4±2.4	2.6±2.7	0.982
Type of beverage consumed:
Wine	95 (77%)	88 (76%)	7 (100%)	0.349
Beer	90 (73%)	86 (74%)	4 (57%)	0.384
Spirituous drinks	89 (72%)	85 (73%)	4 (57%)	0.394
Multiple types	96 (78%)	90 (78%)	5 (71%)	0.666
Months of alcohol abstinence	1.75±5	1.5±5	2±3	0.703
Hepatic steatosis on ultrasound	74 (60%)	70 (60%)	(5) 71%	1.000
Previous pattern of ambulatory alcoholic hepatitis	11 (9%)	10 (9%)	1 (14%)	0.490
Laboratorial data:
Mean corpuscular volume (fL)*	94±6	94±6	97±6	0.274
Platelet count (cells/uL)	239500±76750	237000±78000	271000±82000	0.423
Alanine aminotransferase (IU/L)	15±8	15±7	18±18	0.842
Aspartate aminotransferase (IU/L)	17±13	17±13	17±12	0.284
γ-glutamyl transpeptidase (IU/L)	32±63	32±58	41±81	0.912
Alkaline phosphatase (IU/L)	73±27	72±24	113±60	0.078
Total bilirubin (mg/dL)	0.4±0.2	0.4±0.2	0.3±0.3	0.704
Albumin (g/L)*	43±4	43±4	41±2	0.123
INR*	0.9±0.1	0.9±0.1	0.9±0.1	0.174
Total cholesterol (mg/dL)*	211±44	212±44	189±45	0.182
Triglycerides (mg/dL)	125±105	126±104	125±123	0.982
Urea (mg/dL)	29±14	29±12	22±11	0.031
Creatinine (mg/dL)	0.7±0.2	0.7±0.2	0.6±0.1	0.103
Glucose (mg/dL)	80±20	81±22	72±17	0.910
Non-invasive scores of fibrosis
FIB-4 >1.45	9 (7%)	9 (8%)	0 (0%)	0.444

Data is expressed as mean±standard deviation (*), when variables presented a normal distribution or median±IQR when variables did not present a normal distribution (according to Kolmogorof-Smirnov test).

Evaluation at baseline

At inclusion, the median duration and amount of daily alcohol intake were 21.5±20years (5–70) and 271±203g/day (50–1470), respectively. 80 patients (65%) reported abstinence, for 1.75±5months. Considering the pattern of consumption, most patients (n = 96, 78.1%) consumed more than one type of beverage, 77% consumed wine (n = 95), 73% beer (n = 90), and 72% spirits (n = 89).

Evaluation at the end of follow-up

At the end of follow-up, 65 (53%) patients reported some alcohol consumption during the period of follow up. Relapse of alcohol intake did not associate with gender, duration, or amount of alcohol intake previous to baseline, type of beverage, family history of alcoholism, or CLD. Relapse of alcohol consumption presented a very weak association with duration of abstinence at baseline (Table 2). A previous abstinence period of 6 months did not outperform 3 or 1-month previous abstinence.

Table 2

Duration of alcohol abstinence as a predictor of alcohol relapse.

	Se	Sp	PPV	NPV	OR with 95%CI, p
> 1 month	46%	26%	70%	58%	0.306 [0.142–0.658], p = 0.004
> 3 months	31%	44%	64%	61%	0.347 [0.165–0.729], p = 0.008
> 6 months	14%	63%	61%	41%	0.276 [0.114–0.668], p = 0.007

Se, sensitivity; Sp, specificity; PPV, positive predictive value; NPV, negative predictive value; OR, odds ratio.

During follow-up, the median period of abstinence was 10±15years (0–42), whereas the median duration and amount of alcohol intake were 1.5±12years (0–16) and 32±176g/day (0–470), respectively.

The most relevant comorbidities were: cardiovascular disease in 68 patients (55%), cancer in 30 (24%), and type 2 diabetes-mellitus in 19 (15%).

Development of type 2 diabetes-mellitus positively associated with age (52±11 vs. 44±11years, p = 0.004), and previous years of alcohol consumption (31±15 vs. 24±12years, p = 0.020), whereas inversely associated with baseline smoking (32% vs. 72%, p = 0.001). Of note, we did not have information regarding the smoking status on follow up. Type 2 diabetes-mellitus did not associate with alcohol amount, type of beverage consumption, being overweight/obese, nor with the presence of liver steatosis on ultrasound. On multivariable analysis, the only identified independent risk factors for the development of diabetes-mellitus were baseline smoking (OR 0.181, 95%IC [0.521–0.063]) and age (OR 1.054, 95%IC [1.002–1.109]).

Development of cardiovascular disease positively associated with baseline BMI (26±4 vs. 24±3kg/m2, p = 0.002), average daily alcohol intake during follow-up (158±276 vs. 85±11g, p = 0.050), and total alcohol intake during follow-up 643±137 vs. 261±411kg, p = 0.034). Cardiovascular disease negatively associated with liver tests at inclusion (GGT: 54±72 vs. 106±176IU/L, p = 0.042). On multivariable analysis, the only identified independent risk factor for the development of cardiovascular disease was BMI (OR 1.181, 95%IC [1.054–1.323]).

During follow-up, 30 patients developed cancer: 11 oropharyngeal, 6 gastrointestinal, 3 urothelial, 3 lung, 2 breast, 2 skin, 1 prostate, 1 leiomyosarcoma, and 1 leukemia. Cancer development positively associated with the following variables at inclusion: age (49±10 vs. 44±11years, p = 0.015), smoking (83% vs. 60%, p = 0.026), presence of alcoholism stigmata (90% vs. 72%, p = 0.049) namely facial angiectasias (48% vs. 25%, p = 0.036), and presence of liver steatosis on ultrasound (87% vs. 51%, p = 0.010). On multivariable analysis, the identified independent risk factors for cancer development were facial angiectasias (OR 6.99, 95%IC [1.71–28.57]), smoking (OR 6.6.7, 95%IC [1.34–33.33]), and liver steatosis (OR 6.41, 95%IC [1.14–35.71]).

Risk factors for the development of CLD

CLD development

During follow-up, 7 patients (6%) progressed to CLD, of whom 5 presented with liver decompensation. Clinical events were ascites in 4 (57%), hepatic encephalopathy in 1 (14%), and esophageal varices-related upper gastrointestinal bleeding in 1 (14%). Other complications of CLD such as spontaneous bacterial peritonitis, refractory ascites, hepatic hydrothorax, hepatorenal syndrome, or hepatocellular carcinoma, were not reported.

Progression to CLD associated with alcohol consumption during follow-up, but not with alcohol consumption previous to inclusion or total burden of alcohol intake. Indeed, progression to CLD was positively associated with years of consumption during follow up (14±12 vs. 5±7years, p = 0.002), and inversely with abstinence >1 year during follow up (43% vs. 79%, p = 0.047). The total lifelong duration of heavy alcohol intake was 46±21 with a minimum of 16 years and a maximum of 78 years. In this population, heavy alcohol intake for longer than 50 years had a good accuracy for the development of CLD, as 9 in 10 patients with this duration of consumption developed CLD (57% sensitivity and 89% specificity) (Fig 1).

Fig 1

Duration of alcohol intake and development of chronic liver disease.

Very early markers of liver disease might be higher ALP (113±60 vs. 72±24IU/L, p = 0.078) and lower urea (22±11 vs. 29±12mg/dL, p = 0.031), since both associated with progression to CLD. In multivariate analysis, the independent risk factors for progression to CLD were years of alcohol consumption during follow-up (OR 1.15, 95%IC [1.01–1.31]) and ALP (OR 1.05, 95%IC [1.01–1.10]).

CLD-free survival

In order to allow an analysis of time to progression, Kaplan Meier curves were computed. CLD-free survival was not affected by the amount or duration of alcohol consumption before baseline. However, achieving at least one year of abstinence during follow up associated with higher CLD-free survival (15.6±0.2 vs. 14.7±0.6 years, p = 0.019) (Fig 2). The hazard ratio for CLD-free survival, after Cox regression analysis, was 3.01 [0.65–13.97] (p = 0.159) for achieving at least one year of abstinence during follow up, 1.04 [1.01–1.08] (p = 0.004) for ALP at baseline, and 0.87 [0.77–0.99] (p = 0.039) for urea at baseline.

Fig 2

Alcohol intake pattern on follow-up modulates chronic liver disease-free survival.

Kaplan-Meier curves showing chronic liver disease-free survival according to the abstinence of at least one year and binge-drinking on follow-up. The log-rank test was used to compare the two curves.

Risk factors for mortality

During follow-up, 33 patients (27%) died. The causes of death were cancer (n = 18, 55%), cardiovascular diseases (n = 9, 27%), liver disease (n = 3, 9%), car accident (n = 2, 6%), and pneumonia (n = 1, 3%) (Fig 3).

Fig 3

Causes of death.

The mean age at death was 71±1years (41–77). Factors associated with longer survival were abstinence for at least one year during follow-up (73±1 vs. 65±2years, p = 0.013) and beer consumption at baseline (74±1 vs. 66±2years, p = 0.001). On the contrary, factors associated with decreased survival were smoking (69±2 vs. 74±1, p = 0.017) and hepatic steatosis on baseline ultrasound (68±2 vs. 75±2, p = 0.015) (Fig 4).

Fig 4

Factors associated with overall survival.

Kaplan-Meier curves showing the overall survival according to alcohol intake pattern, smoking habits, and presence of liver steatosis on ultrasound. The log-rank test was used to compare the two curves.

Discussion

We present a 15-year longitudinal study on long-term heavy-drinkers without CLD. These patients represent the largest subgroup of heavy-drinkers on clinical practice (particularly in Primary Care Clinics), however, little is known regarding their real risk of progression to CLD and their co-morbidities [3, 17]. Many patients who died from alcohol-related CLD had previous non-liver related recurrent admissions to the hospital, which might have been missed opportunities for early intervention [18]. As such, it is pressing to better understand the prognosis of these patients and to identify early predictors of progression to CLD.

The first striking finding, from this cohort, is that patients who survived up to 25 years of heavy-drinking without developing evidence of CLD, and hence who were more likely resistant to the dismal effects of alcohol on the liver, are still at increased risk for CLD. 6% of those patients did progress to CLD during a follow-up of 15 years. Interestingly, lifelong heavy alcohol intake for 50 years seems to almost invariably result in CLD, since 9 in 10 patients with this long-term intake did progress to CLD. Furthermore, in this cohort, liver-related mortality was the third cause of death, representing 9% of all deaths, which is 10-times higher than the burden of liver death in Portugal (0.9%) [19].

When we evaluate a heavy drinker without CLD, the previous amount, duration, and type of alcohol intake do not seem to predict progression to CLD. The risk of developing CLD seems to be determined by the forthcoming consumption of alcohol. Indeed, for each year of heavy alcohol intake during follow-up, the estimated risk of CLD increased by 15%, compared to patients who remained abstinent. A large Danish study also corroborated that recent drinking (i.e. in the previous decade) seems to better predict progression to CLD, as compared to earlier drinking [20]. Similarly, achieving abstinence, even if temporarily, for as short as one year, strongly reduced the risk for CLD and overall mortality. As such, this study highlights that interventions to promote abstinence is highly recommended, and that even small achievements can impact the long-term prognosis of harmful alcohol consumers.

Regarding the controversial topic of the type of alcoholic beverages consumption as a risk factor for CLD [20-25], no association was found in these patients. Furthermore, in the heavy drinking range, the amount of alcohol consumed did not seem to impact the risk for CLD. We did not have information regarding binge-drinking.

This cohort suggests that blood urea and ALP might be very early markers of liver disease. Since urea is produced in the liver from the metabolism of amino-acids, it is possible that even in the very early stages of liver disease, levels and activity of urea cycle enzymes are decreased [26] leading to decreased synthesis of urea [27]. Furthermore, mild ALP elevation might represent subclinical intrahepatic cholestasis, a known histological feature associated with a worse prognosis in alcohol-related liver disease [11, 28].

In recent literature, a strong association has been suggested between metabolic factors such as obesity and type 2 diabetes-mellitus and increased risk of CLD [29-33], which did not occur in this cohort.

Also of note was the heavy burden of cancer in these patients. Cancer was the number one cause of mortality, responsible for more than half of the cases of death, which is about two times higher the expected in the Portuguese population [19]. The main cancer was oropharyngeal cancer, which, as expected [21, 25, 30], showed a strong association with smoking. This is of concern since drinkers are more likely to smoke than nondrinkers [34, 35]. Further studies should address if this population should be submitted to a personalized screening program.

Finally, the mean age of death was 71 years, which is 10 years younger than the reported life expectancy in Portugal [36]. This result goes in agreement with a cohort study with 19002 alcohol-dependent individuals from Denmark, which showed a potential loss of 10 years of life when compared with 186767 controls from the Danish Civil Registration System [37]. Interestingly, our cohort had a lower rate of progression (6% vs. 25%) than the Danish cohort, probably related to the fact that these were patients without evidence of CLD baseline, suggesting a significant degree of resistance to liver disease development.

Similarly, a recent Sweden population-based cohort study on 3453 patients who performed liver biopsy for suspicion of alcohol-related liver disease, showed that patients with normal liver histology or simple steatosis, presented a 2–3 fold increased mortality compared to the matched general population, particularly if continuing heavy drinking [38].

The strongest factors associated with mortality were smoking and hepatic steatosis. Of note, baseline hepatic steatosis was associated with overall mortality, even though it did not associate with the liver outcome, cardiovascular diseases, or diabetes-mellitus. However, hepatic steatosis did associate with cancer, which might suggest that alcohol-related liver steatosis might induce a systemic procarcinogenic environment [39]. Once again, alcohol abstinence, for as short as one year, showed a positive impact on survival.

The study strengths include its prospective design, the long duration of follow-up, the extensive data collection, and the evaluation of lifetime alcohol exposure. Furthermore, all patients were initially assessed by a Hepatologist, and hence we are confident regarding the exclusion of liver disease at baseline, and in our ability to control for several potential confounders. To strengthen the confidence of exclusion of CLD, we retrospectively applied noninvasive scores of fibrosis, and only one patient presented FIB-4 suggestive of advanced liver fibrosis. Furthermore, we applied the ALBI-FIB4 score and no patient presented a high-risk score for liver decompensation. Still, the current study has some limitations. First, the risks of bias due to the nature of alcohol history questionnaires with recall issues. Indeed, we tried to be thorough in the evaluation of alcohol consumption, applying a semi-structured questionnaire to patients and their relatives, while confronting them with the information collected from medical records from the national database, as well as information collected from the medical records from the alcohol rehabilitation clinical at the middle of the follow up. Second, liver biopsies confirming the absence of CLD at baseline and on follow-up were not performed, as it would be deemed unethical to submit healthy subjects to an invasive procedure. Furthermore, transient hepatic elastography was also not performed because it was not accessible at the time of inclusion of the patients. However, liver biopsy is not routinely performed in these patients in the real-life clinical practice, and many centers (mainly in the primary care set) still do not have easy access to transient hepatic elastography. Thus this study may help clinicians to guide evidence-based clinical decisions in their daily practice, in patients without evidence of CLD by conventional diagnostic approaches. Third, the development of CLD during follow-up was a rare event, thus making the comparisons weaker. We also included a small number of women. Lastly, the analysis of the effect of variables and confounders on outcomes has the risk of bias by the lack of control to time to progression, which was taken into account in the evaluation of the survival curves.

In conclusion, this study showed that even long-term heavy-drinkers with no evidence of CLD, and periods of abstinence, can still progress to end-stage liver disease, which depends more on forthcoming alcohol intake rather than previous alcohol intake. Furthermore, achieving alcohol abstinence, even for a short period of time, has a positive impact on the risk of progression for CLD and mortality.

In conclusion, for clinical practice, our results suggest that special attention should be made regarding the risk for cancer, with personalized screening programs and intensive anti-smoking interventions, and also those periods of alcohol abstinence are paramount, even with interspersed periods of relapse.

23 Mar 2021

PONE-D-21-05110

Outcomes of excessive alcohol drinkers after 15 years follow-up: heavy burden of cancer and liver disease mortality

PLOS ONE

Dear Dr. Machado,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Special attention should be given to the issue of retrospective evaluation of alcohol drinking history during long-term follow-up: As pointed out by Reviewers 1 and 3, more details on drinking history should be provided or - if those are lacking - this limitation should be stated more clearly in the Discussion.

I concur with Reviewer 2 that ROC analysis does not seem appropriate for dichotomous variables (Figure 1).

Please submit your revised manuscript by May 07 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Rudolf E. Stauber, MD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for including your ethics statement:  "This study was approved by the CHULN and CAM ethic committee (ID number 13/18). All patients gave their informed, written consent to participate in the study.".

Please amend your current ethics statement to include the full name of the ethics committee/institutional review board(s) that approved your specific study.

Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

For additional information about PLOS ONE ethical requirements for human subjects research, please refer to http://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research.

3. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed the survey or questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information. If the questionnaire is published, please provide a citation to the (1) questionnaire and/or (2) original publication associated with the questionnaire.

4. Please provide the name of the alcohol rehabilitation clinic.

5. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please delete it from any other section.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

Reviewer #3: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

Reviewer #3: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I consider this an important study since alcohol-related health issues are significant and we lack enough data on long-term follow up in heavy drinkers. In my opinion, since it is quite challenging to enroll drinkers in such studies, I am fine with the partly retrospective aspect of the study and partly missing data. It is also natural that transient elastography was not available. However, the real and important data should be shown, and the retrospective study design should be visible already in the abstract. An important limitation of this retrospective study design is the diagnosis of “chronic liver disease” (CLD), since it cannot actually be ruled out based on the used conventional diagnostic approaches. Both lab tests and routine ultrasound can overlook manifest liver cirrhosis in up to 50%. Pending a critical revision, the paper could be considered again for publication since it provides important clinical read outs in a long-term followed up cohort of heavy drinkers.

Major comments/questions:

1. It is correct that we do not know enough about heavy drinkers outcome with no liver disease. But the absence of CLD cannot be done efficiently based on the diagnostic criteria as mentioned above. In addition, APRI score, although quite usefuly in the setting of viral hepatitis, performs very poorly in ALD. IN contrast, FIB4 can be accepted, and I recommend to solely stratify the cohort based on FIB4 since no other data such as ELF are available. Fib4 could also be combined with the available albumin since the recently introduced Albi Fib4 score seems to perform even better.

2. The strength of the study is the rather careful medical history and the descriptions of the various causes of death. I miss a clear description of these findings either in supplemental table or even in the major text body.

3. The risk to develop ALD is described to be as high as 15% per year. This seems to by far to highy. E.g. in other cohorts of heavy drinkers, 20% developed cirrhosis after 15 years of heavy drinking.

4. Baseline data: Again it is not very clear but important for this study how patients were enrolled: It reads that “Patients with harmful alcohol consumption and evidence of alcohol-dependence, followed in an Alcohol Rehabilitation Clinic, were prospectively and consecutively enrolled in a clinical protocol performed in 2002 ref 13 and retrospectively re-evaluated in 2018. What means retrospectively in this context? How was liver disease excluded? I am also confused when looking at Ref 13: Herewith some data:

5. Ref 13 (Martins A et al Eur J Gastroenterol Hepatol 2005; 17 (10):1099-1104.). In this study, completely different patient numbers are provided

6. It is also interesting that obviously an an isolated rise in GGT was not considered an exclusion criterium. Why? How many patients were enrolled based on these criterium.

7. It is further pointed out that follow up data were retrospectively obtained form a national database. IT would be interesting to learn how data on alcohol consumption are regularly entered in this data base. Who is doing this? How often did patients undergo alcohol detoxification?

8. Since alcohol drinking history is essential for this study, more details should be provided, given the fact that it is generally challenging for patients to provide a detailed alcohol history with stops and goes. How was duration of alcohol consumption assessed. How do patients remember an alcohol drinking history of the last 10 years or so? In line with this, a 53% of relapse appears to be rather low for heavy drinkers

9. On page 12 it is stated that 9% had previous AH while in the inclusion criteria, previous decompensation was ruled out?

10. What about the numbers of diabetes and obesity?

11. Smoking is shown as risk factor for DM II with an OR 0.181 Does smoking protect from diabetes? There are studies continueosly showing a tight association between smoking and alcohol consumption? What are the numbers for this cohort?

Reviewer #2: This is an interesting prospective study that studies the natural history of heavy drinkers without evidence of liver disease and evaluates their clinical outcomes, risk of death and of developing liver disease during 15 years of follow up. Risk factors were also identified.

From this standpoint, the study is important because little is know about the risk of developing CLD in heavy drinkers. However, there are some issues that need further clarification:

- the data in table 1 are presented as mean±SD, which - based on the figures, seem not quite appropriate (at least for some of the variables. Could you please look back to the distribution of the variables and use median±IQR/range when needed. Also, in some cases, non parametric tests should have been used.

- regarding the data presented in figure 1, I think that the AUROC analysis is not the right choice for what was intended by the authors. AUROC is usually used with continuous variables, not with dichotomial ones. I believe a regression analysis (or a basic chi-square test) would be better choices.

- the manuscript also need a further English polish ti increase its readability.

Reviewer #3: Mariana Machado and colleagues present a single-center study from Lisbon on the natural history of 130 patients with a history of excessive drinking who present at an alcohol rehabilitation unit, but without evidence of chronic liver disease or cancer. The authors are commended for managing to recall almost all (7 losses to follow up) participants after 14±3 years for repeated investigations and a detailed description of alcohol use during follow up.

While there is a need for longitudinal studies on the natural history of patients with harmful drinking, taking fluctuations in alcohol use over the years into account, the study in its current form has substantial methodological and reporting issues which limits interpretation of results. Please find below my questions, comments and suggestions for improvement.

MAJOR:

• Please use the STROBE guidelines for reporting: https://www.equator-network.org/reporting-guidelines/strobe/

• Title: the short titles accurately states that development of CLD during follow up is the outcome of interest, while the long title does not.

• Lay summary, first point. It is not surprising that ongoing rather than past alcohol intake dictates the risk of liver cirrhosis development when you specifically exclude anyone with evidence of liver disease at baseline. Therefore I suggest to delete this sentence.

• The authors stress the point that alcohol consumers on average has a 10 yr reduced life expectancy. Their way of comparing cohort data with a country average is not really methodologically correct – the correct comparator would be 5:1 or 10:1 gender, age, time and geography matched random sample from the population. Therefore, I would propose not to emphasize the 10 yr reduction in life expectancy as much as the authors do – e.g. it appears in both abstract and bullet points.

• Introduction: The statement “Liver disease is the main alcohol-related chronic illness” gives the wrong impression, since liver disease – as the authors show – does not account for the majority of alcohol-related harm. I would propose instead to stress that the the liver is the single organ most commonly affected by excess drinking.

• Methods: Reference 13 is referred to as the original study. However, it describes a cohort of only 76 patients without evidence of liver disease at baseline, but this manuscript includes n=130 ?

• Methods: Please describe in more detail the in- and exclusion criteria. Which liver blood tests were considered? If fibrosis stage F1 on liver biopsy, the patient were considered to have ALD and excluded? From the text is seem like a very heterogeneous cohort, since many with compensated cirrhosis at baseline may have normal liver blood tests (see e.g. Table 1 in Mueller, World J Gastroenterol 2014 October 28; 20(40): 14626-14641). Those patients would be included unless they had a liver biopsy? In contrast, many with no or minimal fibrosis may have elevated transaminases and alkaline phosphatase – they would be excluded?

• Methods: Similarly, the definition of the primary outcome, development of CLD, is poorly described. The current description can encompass patients with anything from steatohepatitis without fibrosis, all the way to decompensated cirrhosis.

• Methods: The authors state that inclusions were in 2002 and reassessment in 2018, which would result in 16 years between baseline and reevaluation. Yet, in methods they state a follow up of 14 ±3 yrs?

• Clearly, the alcohol habits of those who die in between baseline and follow up can not be evaluated, or did the authors have information from other sources on alcohol behaviour during follow up?

• Methods: When the primary endpoint is risk of CLD, the authors should account for how they handle non-CLD death as a competing risk?

• Methods: Baseline CLD and cancer are excluded, but cardiovascular or metabolic disease are not, so the comparison between deaths from CLD, cancer and CVD is unbalanced.

• Results: seven patients (6%) progressed to CLD, which is lower than in studies of similar populations (for example Holst, Addiction 2017, vol 112, where approximately 25% die from liver related death). Could the authors discuss this? Would more accurate liver tests – like FibroScan, which was widely available in 2018 – have resulted in higher proportion being detected with CLD?

• Results: It makes less sense to discuss biomarkers of liver disease, when abnormal liver blood tests were an exclusion criterium.

• Discussion: It is problematic that there is such a long time period between baseline and follow up with regards to obtaining alcohol history. I fear that recall bias may play a strong role, since those who have developed CLD will be more prone to remember alcohol drinking periods than those who have not. Can the authors discuss this limitation?

• I recommend a language revision by a native english speaker. Examples from the abstract: “Most long-term heavy drinkers do not present chronic liver disease” (present with CLD?); “being abstinent for at least 1 year positively modeled CLD-free survival” (correlated or associated with CLD-free survival?); “showed an impressive decreased life expectancy” (impressive has positive connotations, which is in stark contrast to decreased life expenctancy).

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

30 Mar 2021

We thank the editor and the reviewers for the constructive critiques and for the opportunity to improve our manuscript. We hope the reviewers now find it worthy for publication. It follows the point-by-point response to the editors and reviewers comments.

EDITORS’ COMMENTS

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

We performed the necessary changes.

2. Thank you for including your ethics statement: "This study was approved by the CHULN and CAM ethic committee (ID number 13/18). All patients gave their informed, written consent to participate in the study.".

Please amend your current ethics statement to include the full name of the ethics committee/institutional review board(s) that approved your specific study.

Once you have amended this/these statement(s) in the Methods section of the manuscript, please add the same text to the “Ethics Statement” field of the submission form (via “Edit Submission”).

For additional information about PLOS ONE ethical requirements for human subjects research, please refer to http://journals.plos.org/plosone/s/submission-guidelines#loc-human-subjects-research.

We added to the methods the following: “Comissão de Ética do Centro Académico de Lisboa _ CAML in 2002 and again in 2018.”

3. Please include additional information regarding the survey or questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed the survey or questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information. If the questionnaire is published, please provide a citation to the (1) questionnaire and/or (2) original publication associated with the questionnaire.

We did not use a specific questionnaire, rather we performed a detailed clinical history in an outpatient Hepatology visit.

4. Please provide the name of the alcohol rehabilitation clinic.

We added to the methods the following sentence: “At enrollment all patients attending outpatient visits at Centro Regional de Alcoologia do Sul (CRAS) were asked to participate in the study and evaluated at a Hepatology visit with laboratorial evaluation.”

5. Your ethics statement should only appear in the Methods section of your manuscript. If your ethics statement is written in any section besides the Methods, please delete it from any other section.

We removed that information from other sections besides the Methods.

REVIEWER 1

I consider this an important study since alcohol-related health issues are significant and we lack enough data on long-term follow up in heavy drinkers. In my opinion, since it is quite challenging to enroll drinkers in such studies, I am fine with the partly retrospective aspect of the study and partly missing data. It is also natural that transient elastography was not available. However, the real and important data should be shown, and the retrospective study design should be visible already in the abstract. An important limitation of this retrospective study design is the diagnosis of “chronic liver disease” (CLD), since it cannot actually be ruled out based on the used conventional diagnostic approaches. Both lab tests and routine ultrasound can overlook manifest liver cirrhosis in up to 50%. Pending a critical revision, the paper could be considered again for publication since it provides important clinical read outs in a long-term followed up cohort of heavy drinkers.

We thank the reviewer for these encouraging words. We agree that the ruling out of “chronic liver disease” has flaws, but we consider that we have done as best as possible at that time to achieve it.

Major comments/questions:

1. It is correct that we do not know enough about heavy drinkers outcome with no liver disease. But the absence of CLD cannot be done efficiently based on the diagnostic criteria as mentioned above. In addition, APRI score, although quite usefully in the setting of viral hepatitis, performs very poorly in ALD. IN contrast, FIB4 can be accepted, and I recommend to solely stratify the cohort based on FIB4 since no other data such as ELF are available. Fib4 could also be combined with the available albumin since the recently introduced Albi Fib4 score seems to perform even better.

We understand the reviewer considerations regarding the exclusion of chronic liver disease. We do agree with the limitations of conventional diagnostic approaches to rule it out. We go even further, since even the invasive gold standard evaluation with liver biopsy can misclassify liver fibrosis in up to on third of the patients (due to sample error and inter pathologist variability). However, the conventional diagnostic approaches mirrors real-life medical practice, and it may be the most useful information to help guiding evidence-based clinical decisions that clinicians have to make in their daily practice. Having said that, we were very thorough in our evaluation. An expert in Hepatology with special interest in alcohol-associated liver disease evaluated all patients. We needed two normal blood tests and ultrasound (except for isolated steatosis). Finally we retrospectively applied fibrosis scores, which confirmed low probability of advanced fibrosis in all patients. As such, we believe that we are the most confidants we could be regarding exclusion of liver disease, with the available tools at the time of enrollment (2002).

We followed the reviewer suggestion and removed the data on APRI and applied the ALBI-FIB4 score, and all patients presented a low risk score for liver decompensation. We added the following sentence in the methods: “We also applied the ALBI-FIB-4 score and all patients presented a low risk of decompensation of liver disease (the mean score was -3.91±0.51, minimum -5.309 and maximum -2.359).

2. The strength of the study is the rather careful medical history and the descriptions of the various causes of death. I miss a clear description of these findings either in supplemental table or even in the major text body.

We thank the reviewer for this suggestion. Accordingly, in the results, we changed the sentence: “The 3 main causes of death were cancer (n=18.55%), cardiovascular diseases (n=9.27%), and liver disease (n=3. 9%) (Figure 4).” to “The causes of death were cancer (n=18, 55%), cardiovascular diseases (n=9, 27%), liver disease (n=3, 9%), car accident (n=2, 6%), and pneumonia (n=1, 3%) (Figure 4).”

3. The risk to develop ALD is described to be as high as 15% per year. This seems to by far to highy. E.g. in other cohorts of heavy drinkers, 20% developed cirrhosis after 15 years of heavy drinking.

We thank the reviewer for this comment, as we were not clear conveying the message. Only 6% progressed to clinically evident chronic liver disease after 15 years of follow up. Compared to patients who remained abstinent, the risk of developing chronic liver disease increased 15% per year of alcohol consumption on follow up.

In the lay summary, we added to the sentence “For each year of heavy alcohol intake during follow-up, the risk of developing chronic liver disease increased 15%, compared to patients who remained abstinent.”

4. Baseline data: Again it is not very clear but important for this study how patients were enrolled: It reads that “Patients with harmful alcohol consumption and evidence of alcohol-dependence, followed in an Alcohol Rehabilitation Clinic, were prospectively and consecutively enrolled in a clinical protocol performed in 2002 ref 13 and retrospectively re-evaluated in 2018. What means retrospectively in this context? How was liver disease excluded?

We added the following statement, to the methods, regarding baseline data: “At enrollment all patients attending outpatient visits at Centro Regional de Alcoologia do Sul (CRAS) were asked to participate in the study and were evaluated at an Hepatology ambulatory clinic with laboratorial evaluation.”

The retrospective evaluation is explained in the methods: “All data were obtained from the computerized national health database (including hospital admissions or consultations, primary care appointments, blood tests, and imaging results), from consultation of the records in the Alcohol Rehabilitation Clinic at 2011, and through a survey performed by telephone interview with the patient or patient´s relatives.” At the end of follow up, liver disease was excluded as explained: “Patients were considered to have developed CLD when they presented abnormal liver tests for more than 6 months, imaging findings compatible with CLD, or presented with hepatic decompensation (jaundice, ascites, hepatic encephalopathy, and/or variceal bleeding).”

5. I am also confused when looking at Ref 13 (Martins A et al Eur J Gastroenterol Hepatol 2005; 17 (10):1099-1104.). In this study, completely different patient numbers are provided.

This cohort of patients was enrolled at 2002, as part of a study regarding genetic susceptibility for alcohol-associated liver disease. For the manuscript Martins A et al Eur J Gastroenterol Hepatol 2005; 17 (10):1099-1104, only 76 patients were included in the group of heavy drinkers without chronic liver disease, because for inclusion we needed two laboratorial evaluations with normal liver disease and many patients did not have available other laboratorial evaluation apart from the one performed at recruitment. In the following year, many patients evaluated did perform a second laboratorial evaluation 6 months apart with normal liver tests. As such, the pool of heavy drinkers without evidence of chronic liver disease increased. We referenced that manuscript because it is the same cohort, with the same methodology. However, for clarity, we have changed to “with data partially reported in 2005 [13]”

6. It is also interesting that obviously an isolated rise in GGT was not considered an exclusion criterium. Why? How many patients were enrolled based on these criterium.

We did not exclude patients with isolated rise in GGT, because we consider that GGT in this setting may be only a marker of alcohol consumption, rather than alcohol-associated liver disease. Striking elevations of serum GGT activities can be observed in patients with a high alcohol intake over a prolonged period, in patients without liver disease other than fatty liver. The increase in GGT seems to be primarily due to hepatic enzyme induction in the endoplasmic reticulum, rather than to liver cell injury. Isolated increase in GGT was present in 39 patients (31.7%). Interestingly, higher GGT levels did not associate with an increased risk of progression to chronic liver disease.

To clarify this, we added the following sentence to the methods: “Isolated increase of GGT was not an exclusion criteria, since it is a marker of alcohol consumption and not of liver injury, translating hepatic enzyme induction rather than liver cell injury {Nishimura, 1983 #55}. Isolated increase of GGT was found in 39 patients (31.7%).”

7. It is further pointed out that follow up data were retrospectively obtained form a national database. IT would be interesting to learn how data on alcohol consumption are regularly entered in this database. Who is doing this? How often did patients undergo alcohol detoxification?

8. Since alcohol drinking history is essential for this study, more details should be provided, given the fact that it is generally challenging for patients to provide a detailed alcohol history with stops and goes. How was duration of alcohol consumption assessed? How do patients remember an alcohol drinking history of the last 10 years or so? In line with this, a 53% of relapse appears to be rather low for heavy drinkers.

The national database consists of the outpatients visits in primary care, as well as outpatient visits and hospitalizations in public hospitals. Working in a public hospital, a physician has access to the medical records from other public institutions, regarding the patient they are following or did follow. Because all patients were evaluated in a Hepatology visit in our institution, we did have access to their medical records. We do realize that information regarding alcohol consumption in medical records most of the times is not thorough. We also, did consult the medical records of the alcohol rehabilitation clinic at 2011. As such, the history of alcohol intake during follow up was collected by a thorough questionnaire to the patient and their relatives, confronted with the information collected from medical records. We do recognize that there may be recall issues in the assessment of alcohol consumption, and that was now emphasized in the discussion. Accordingly, we added to the discussion: “Still, the current study also has some limitations. First, the risk of bias due to the nature of alcohol history questionnaires with recall issues. Indeed, we tried to be thorough in the evaluation of alcohol consumption, applying a structured questionnaire to patients and their relatives, while confronting them with the information collected from medical records from the national database, as well as information collected from the medical records from the alcohol rehabilitation clinical at the middle of the follow up.”

We did not register how often patients did undergo alcohol detoxification.

All patients were followed in an alcohol rehabilitation clinic, which may explain the rather low percentage of relapse in those heavy drinkers.

9. On page 12 it is stated that 9% had previous AH while in the inclusion criteria, previous decompensation was ruled out?

We did not account for asymptomatic ambulatory alcohol hepatitis pattern as hepatic decompensation. We amended the phrase to: “9% had previous biochemical changes compatible with asymptomatic ambulatory alcoholic hepatitis”

10. What about the numbers of diabetes and obesity?

We added the following to the results: “Eighty-one (66%) were smokers, 48 (46%) overweight/obese (28 obese, 23%) and 6 patients had type 2 diabetes mellitus (5%)”.

11. Smoking is shown as risk factor for DM II with an OR 0.181 Does smoking protect from diabetes? There are studies continuously showing a tight association between smoking and alcohol consumption? What are the numbers for this cohort?

We did found an inverse association between baseline smoking and the development of type 2 diabetes mellitus, which is not in agreement with literature. Of note, we do not have information regarding smoking status on follow up, nor the number of cigarettes smoked per day, and that may explain these results. This is now stressed in the results: “Of note, we did not have information regarding the smoking status on follow up.” Baseline smoking did not associate with alcohol consumption relapse.

REVIEWER 2

This is an interesting prospective study that studies the natural history of heavy drinkers without evidence of liver disease and evaluates their clinical outcomes, risk of death and of developing liver disease during 15 years of follow up. Risk factors were also identified.

From this standpoint, the study is important because little is know about the risk of developing CLD in heavy drinkers. However, there are some issues that need further clarification.

We thank the reviewer for these encouraging words.

1. The data in table 1 are presented as mean±SD, which - based on the figures, seem not quite appropriate (at least for some of the variables. Could you please look back to the distribution of the variables and use median±IQR/range when needed. Also, in some cases, non parametric tests should have been used.

We thank the reviewer for this pertinent comment. We changed it accordingly.

2. Regarding the data presented in figure 1, I think that the AUROC analysis is not the right choice for what was intended by the authors. AUROC is usually used with continuous variables, not with dichotomial ones. I believe a regression analysis (or a basic chi-square test) would be better choices.

We agree with the reviewer, and we changed it accordingly.

3. The manuscript also need a further English polish it increase its readability.

We carefully edited the text to improve readability.

REVIEWER 3

Mariana Machado and colleagues present a single-center study from Lisbon on the natural history of 130 patients with a history of excessive drinking who present at an alcohol rehabilitation unit, but without evidence of chronic liver disease or cancer. The authors are commended for managing to recall almost all (7 losses to follow up) participants after 14±3 years for repeated investigations and a detailed description of alcohol use during follow up.

While there is a need for longitudinal studies on the natural history of patients with harmful drinking, taking fluctuations in alcohol use over the years into account, the study in its current form has substantial methodological and reporting issues which limits interpretation of results. Please find below my questions, comments and suggestions for improvement.

We thank the reviewer for these encouraging words.

1. Please use the STROBE guidelines for reporting: https://www.equator-network.org/reporting-guidelines/strobe/

We thank the reviewer for this comment. We made the necessary changes to follow the STROBE guidelines for reporting:

Item No Recommendation Page No

Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title or the abstract 3

(b) Provide in the abstract an informative and balanced summary of what was done and what was found

Introduction

Background/rationale 2 Explain the scientific background and rationale for the investigation being reported 6

Objectives 3 State specific objectives, including any prespecified hypotheses 7

Methods

Study design 4 Present key elements of study design early in the paper 8

Setting 5 Describe the setting, locations, and relevant dates, including periods of recruitment, exposure, follow-up, and data collection 8

Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection of participants. Describe methods of follow-up 8

(b) For matched studies, give matching criteria and number of exposed and unexposed n.a.

Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, and effect modifiers. Give diagnostic criteria, if applicable 9,10

Data sources/ measurement 8* For each variable of interest, give sources of data and details of methods of assessment (measurement). Describe comparability of assessment methods if there is more than one group 9,10

Bias 9 Describe any efforts to address potential sources of bias 10

Study size 10 Explain how the study size was arrived at n.a.

Quantitative variables 11 Explain how quantitative variables were handled in the analyses. If applicable, describe which groupings were chosen and why 11

Statistical methods 12 (a) Describe all statistical methods, including those used to control for confounding 11

(b) Describe any methods used to examine subgroups and interactions n.a.

(c) Explain how missing data were addressed n.a.

(d) If applicable, explain how loss to follow-up was addressed 8

(e) Describe any sensitivity analyses 12

Results

Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers potentially eligible, examined for eligibility, confirmed eligible, included in the study, completing follow-up, and analysed 12

(b) Give reasons for non-participation at each stage 12

(c) Consider use of a flow diagram n.a.

Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, social) and information on exposures and potential confounders Table 1

(b) Indicate number of participants with missing data for each variable of interest n.a.

(c) Summarise follow-up time (eg, average and total amount) 12

Outcome data 15* Report numbers of outcome events or summary measures over time 13

Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted estimates and their precision (eg, 95% confidence interval). Make clear which confounders were adjusted for and why they were included Table 1

(b) Report category boundaries when continuous variables were categorized n.a.

(c) If relevant, consider translating estimates of relative risk into absolute risk for a meaningful time period n.a.

Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, and sensitivity analyses n.a.

Discussion

Key results 18 Summarise key results with reference to study objectives 17

Limitations 19 Discuss limitations of the study, taking into account sources of potential bias or imprecision. Discuss both direction and magnitude of any potential bias 20

Interpretation 20 Give a cautious overall interpretation of results considering objectives, limitations, multiplicity of analyses, results from similar studies, and other relevant evidence 20

Generalisability 21 Discuss the generalisability (external validity) of the study results 20,21

Other information

Funding 22 Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on which the present article is based 2

2. Title: the short titles accurately states that development of CLD during follow up is the outcome of interest, while the long title does not.

We changed the title to: “Outcomes of excessive alcohol drinkers without baseline evidence of chronic liver disease after 15 years follow-up: heavy burden of cancer and liver disease mortality”

3. Lay summary, first point. It is not surprising that ongoing rather than past alcohol intake dictates the risk of liver cirrhosis development when you specifically exclude anyone with evidence of liver disease at baseline. Therefore I suggest to delete this sentence.

Thank you for the suggestion. Accordingly, we changed the sentence to: ongoing alcohol, and deleted: “rather than past”

4. The authors stress the point that alcohol consumers on average has a 10 yr reduced life expectancy. Their way of comparing cohort data with a country average is not really methodologically correct – the correct comparator would be 5:1 or 10:1 gender, age, time and geography matched random sample from the population. Therefore, I would propose not to emphasize the 10 yr reduction in life expectancy as much as the authors do – e.g. it appears in both abstract and bullet points.

We agree with the reviewer, as such we removed that sentence from the lay summary. In the abstract we reduced the emphasis of the statement, from “showed an impressive decrease” to “seemed to have a decreased”.

5. Introduction: The statement “Liver disease is the main alcohol-related chronic illness” gives the wrong impression, since liver disease – as the authors show – does not account for the majority of alcohol-related harm. I would propose instead to stress that the liver is the single organ most commonly affected by excess drinking.

We thank the reviewer for this comment, and we changed it accordingly.

6. Methods: Reference 13 is referred to as the original study. However, it describes a cohort of only 76 patients without evidence of liver disease at baseline, but this manuscript includes n=130?

This cohort of patients was enrolled at 2002, as part of a study regarding genetic susceptibility for alcohol-associated liver disease. For the manuscript Martins A et al Eur J Gastroenterol Hepatol 2005; 17 (10):1099-1104, only 76 patients were included in the group of heavy drinkers without chronic liver disease, because for inclusion we needed two laboratorial evaluations with normal liver disease and many patients did not have available other laboratorial evaluation apart from the one performed at recruitment. In the following year, many patients evaluated did perform a second laboratorial evaluation 6 months apart with normal liver tests. As such, the pool of heavy drinkers without evidence of chronic liver disease increased. We referenced that manuscript because it is the same cohort, with the same methodology. For clarity, we have changed to “with data partially reported in 2005 [13]”.

7. Methods: Please describe in more detail the in- and exclusion criteria. Which liver blood tests were considered? If fibrosis stage F1 on liver biopsy, the patient were considered to have ALD and excluded? From the text is seem like a very heterogeneous cohort, since many with compensated cirrhosis at baseline may have normal liver blood tests (see e.g. Table 1 in Mueller, World J Gastroenterol 2014 October 28; 20(40): 14626-14641). Those patients would be included unless they had a liver biopsy? In contrast, many with no or minimal fibrosis may have elevated transaminases and alkaline phosphatase – they would be excluded?

We added the following to the methods, which we hope clarifies the inclusion and exclusion criteria: “Requisites for the absence of CLD included: the absence of a history of hepatic decompensation, absence of physical stigmata of chronic liver disease (i.e. cutaneous signs, hepatosplenomegaly, gynecomastia, and testicular atrophy), and at least two normal determinations of blood liver tests (aminotransferases, alkaline phosphatase, total bilirubin, protrhombin time and albumin) with an exception of an isolated rise in γ-glutamyl transpeptidase (GGT). Of the 262 screened patients, 130 did not have evidence of CLD, 103 had alcohol-associated liver disease, and the remaining 29 had either other liver diseases or mild increase in liver enzymes (that is, an increase lower than 3 times the reference value). Isolated increase of GGT was not an exclusion criteria, since it is a marker of alcohol consumption and not of liver injury, translating hepatic enzyme induction rather than liver cell injury{Nishimura, 1983 #55}. Isolated increase of GGT was found in 39 patients (31.7%) without evidence of liver disease. Patients who underwent liver biopsy raised clinical doubts regarding the presence of liver disease, and hence were not classified as not having liver disease even if the biopsy was normal, because of concerns of sampling error.

8. Methods: Similarly, the definition of the primary outcome, development of CLD, is poorly described. The current description can encompass patients with anything from steatohepatitis without fibrosis, all the way to decompensated cirrhosis.

We agree with the reviewer, one limitation of the study is the retrospective design of the follow up. As such, we considered clinically evident chronic liver disease, either by persistent abnormal liver tests, imaging evidence of chronic liver disease or clinical decompensation of liver disease, as stated in the methods.

9. Methods: The authors state that inclusions were in 2002 and reassessment in 2018, which would result in 16 years between baseline and reevaluation. Yet, in methods they state a follow up of 14 ±3 yrs?

Follow up was smaller than 16 years because many patients died, thus reducing the mean follow up.

10. Clearly, the alcohol habits of those who die in between baseline and follow up can not be evaluated, or did the authors have information from other sources on alcohol behaviour during follow up?

Thank you for the observation. We agree, and added the following sentence: “Indeed, we tried to be thorough in the evaluation of alcohol consumption, applying a structured questionnaire to patients and their relatives, while confronting them with the information collected from medical records from the national database, as well as information collected from the medical records from the alcohol rehabilitation clinical at the middle of the follow up.”

11. Methods: When the primary endpoint is risk of CLD, the authors should account for how they handle non-CLD death as a competing risk?

We did that sensitive analysis by performing Kaplan-Meier curves of CLD-free survival, with Cox-regression for confounders.

12. Methods: Baseline CLD and cancer are excluded, but cardiovascular or metabolic disease are not, so the comparison between deaths from CLD, cancer and CVD is unbalanced.

We agree with the reviewer, the comparison may be unbalanced, but that would favor CVD. However, we did find unexpectedly high CLD and cancer mortality, and not CVD, which further strengthen the burden of CLD and cancer mortality.

13. Results: seven patients (6%) progressed to CLD, which is lower than in studies of similar populations (for example Holst, Addiction 2017, vol 112, where approximately 25% die from liver related death). Could the authors discuss this? Would more accurate liver tests – like FibroScan, which was widely available in 2018 – have resulted in higher proportion being detected with CLD?

Thank you for your observation. We believe the low risk of progression to CLD can be explained by the fact that this is a highly selected population, of patients without evidence of CLD baseline, even though with a median alcohol consumption of 271±203g/day during 21.5±20 years. This cohort represents a population particularly resistant to CLD, and even so, 6% still progress to CLD. The study mentioned by the reviewer does not exclude patients already with CLD when referred to an alcohol rehabilitation clinic, which might explain the higher liver-related death rate. In accordance we have now added the phrase: Interestingly, our cohort had a lower rate of progression (6% vs 25%) than the Danish cohort, probably related to the fact that these are patients without evidence of CLD baseline, suggesting a significant degree of resistance to liver disease development.

14. Results: It makes less sense to discuss biomarkers of liver disease, when abnormal liver blood tests were an exclusion criterion.

We understand the reviewer’s point, however we believe this analysis is still relevant. First of all, we did accept isolated increase in GGT, and the fact that it did not associate with an increased risk for progression to CLD validates our choice. Second, even though a parameter is in the normal range, differences in the normal range may still have subtle impact in the risk of disease. One such example is the impact of body mass index in the development of hepatic steatosis. We know that even in the normal range, the higher the BMI, the higher the risk of developing hepatic steatosis. Indeed, we did find that the higher the alkaline phosphatase, the higher the risk of developing chronic liver disease, which may be explained by subclinical cholestasis.

15. Discussion: It is problematic that there is such a long time period between baseline and follow up with regards to obtaining alcohol history. I fear that recall bias may play a strong role, since those who have developed CLD will be more prone to remember alcohol drinking periods than those who have not. Can the authors discuss this limitation?

We agree, we discussed it: First, the risk of bias due to the nature of alcohol history questionnaires with recall issues. Indeed, we tried to be thorough in the evaluation of alcohol consumption, applying a structured questionnaire to patients and their relatives, while confronting them with the information collected from medical records from the national database, as well as information collected from the medical records from the alcohol rehabilitation clinical at the middle of the follow up.”

16. I recommend a language revision by a native English speaker. Examples from the abstract: “Most long-term heavy drinkers do not present chronic liver disease” (present with CLD?); “being abstinent for at least 1 year positively modeled CLD-free survival” (correlated or associated with CLD-free survival?); “showed an impressive decreased life expectancy” (impressive has positive connotations, which is in stark contrast to decreased life expectancy).

We carefully edited the text to improve readab

Submitted filename: Response to reviewers - 2021.03.30.docx

Click here for additional data file.

30 Apr 2021

PONE-D-21-05110R1

Outcomes of excessive alcohol drinkers without baseline evidence of chronic liver disease after 15 years follow-up: heavy burden of cancer and liver disease mortality

PLOS ONE

Dear Dr. Machado,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

Please follow the recommendations of Reviewer #3 in order to further improve the presentation of your interesting findings.

The abbreviation 'LC' should be replaced by 'cirrhosis' (not 'liver cirrhosis' as there is no other cirrhosis).

==============================

Please submit your revised manuscript by Jun 14 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Rudolf E. Stauber, MD

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I am satisfied with the author's response and modifications of the manuscript.

Acceptance recommended.

Reviewer #2: (No Response)

Reviewer #3: Thank you for the revised version of this manuscript. I have just a few additional suggestions and comments:

General:

• Please throughout the manuscript state both absolute numbers and proportions. The cohort is relatively small, so the sole reporting of percentages gives a wrong impression.

• Language: English editing is still needed. There remain grammar errors and sentence structures which gives a poor impression, and at certain places makes the text difficult to even understand.

• LC is not a common abbreviation; I suggest to write liver cirrhosis in full throughout.

Results:

- There is a lack of time to progression for all outcomes, except for the Kaplan Meier plots. This hinders accurate interpretation, and should be stated in the text.

- Figure 1: The ROC analysis is not the correct method to assess prognostic accuracy as it is not time-dependent. Should be replaced with Harrell’s C or time-dependent ROC curves.

Discussion:

- Based on the low event rate and relatively high survival for this type of patient I still suspect a high degree of survivors bias. The authors correctly state: “The first striking finding, from this cohort, is that patients who have no evidence of CLD despite being heavy-drinkers for 25 years are still at increased risk for CLD”. I suggest to more clearly acknowledge and discuss the limitation of survivors bias and that this is not an inception cohort. Rather, included participants have already survived >20 years of heavy drinking without developing evidence of chronic liver disease and this population is therefore selected to be less likely to develop liver disease during the next 14 years, than the average patients with heavy drinking.

- Regarding excess cancer risk due to combined alcohol and smoking. The sentence “This population should probably be submitted to a personalized screening program.” is unsubstantiated as this cohort is not suited to investigate such topic.

- Similarly the discussion about beer drinking having a possible protective effect is purely speculative and this study is not powered or designed to make such assumptions. The effect of beer may also be that it is the type of beverage containing the lowest percentage of alcohol; or just a common type I error.

- The discussion of lack of liver biopsy may be replaced with lack of accurate confirmatory tests – LB or LSM – to ensure the absence of liver disease.

- I am not sure I understand the sentence “These patients represent the largest subgroup of heavy-drinkers on clinical practice”. I guess it depends on which healthcare setting one operates in?

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

9 May 2021

RESPONSE TO THE REVIEWERS

We thank the editor and the reviewers for the constructive critiques and for the opportunity to improve our manuscript. We hope the reviewers now find it worthy for publication. It follows the point-by-point response to the editors and reviewers comments.

RESPONSE TO REVIEWER 3

Thank you for the revised version of this manuscript.

I have just a few additional suggestions and comments:

General:

• Please throughout the manuscript state both absolute numbers and proportions. The cohort is relatively small, so the sole reporting of percentages gives a wrong impression.

R: It was changed accordingly.

• Language: English editing is still needed. There remain grammar errors and sentence structures, which gives a poor impression, and at certain places makes the text difficult to even understand.

R: The English was carefully edited.

• LC is not a common abbreviation; I suggest to write liver cirrhosis in full throughout.

R: Thank you for the suggestion. It was changed accordingly.

Results:

• There is a lack of time to progression for all outcomes, except for the Kaplan Meier plots. This hinders accurate interpretation, and should be stated in the text.

R: It was stated in the text: “In order to allow for an analysis of time to progression, Kaplan Meyer curves were computed.” We also added in the discussion: “Lastly, the analysis of the effect of variables and confounders on outcomes has the risk of bias by the lack of control to time to progression, which was taken into account in the evaluation of the survival curves.”

• Figure 1: The ROC analysis is not the correct method to assess prognostic accuracy as it is not time-dependent. Should be replaced with Harrell’s C or time-dependent ROC curves.

R: We agree that ROC analysis was not adequate to evaluate the abstinence time, as such we removed it in the previous revision of the manuscript.

Discussion:

• Based on the low event rate and relatively high survival for this type of patient I still suspect a high degree of survivors bias. The authors correctly state: “The first striking finding, from this cohort, is that patients who have no evidence of CLD despite being heavy-drinkers for 25 years are still at increased risk for CLD”. I suggest to more clearly acknowledge and discuss the limitation of survivors bias and that this is not an inception cohort. Rather, included participants have already survived >20 years of heavy drinking without developing evidence of chronic liver disease and this population is therefore selected to be less likely to develop liver disease during the next 14 years, than the average patients with heavy drinking.

This is the main interest of this cohort. Albeit it is a highly selected cohort of patients less likely to develop CLD, since they already survived more than 20 years of heavy drinking without developing evidence of CLD, 6% still progresses. The risk of progression is not related with past alcohol consumption, rather than on ongoing alcohol intake.

R: We changed the sentence to: “The first striking finding, from this cohort, is that patients who survived up to 25 years of heavy-drinking without developing evidence of CLD, and hence who were more likely resistant to the dismal effects of alcohol on the liver, are still at increased risk for CLD.”

• Regarding excess cancer risk due to combined alcohol and smoking. The sentence “This population should probably be submitted to a personalized screening program.” is unsubstantiated as this cohort is not suited to investigate such topic.

R: We agree with the reviewer, and as such, we changed the sentence to: “Further studies should address if this population should be submitted to a personalized screening program.”

• Similarly the discussion about beer drinking having a possible protective effect is purely speculative and this study is not powered or designed to make such assumptions. The effect of beer may also be that it is the type of beverage containing the lowest percentage of alcohol; or just a common type I error.

R: We removed the paragraph regarding the discussion about the possible protective effect of beer drinking.

• The discussion of lack of liver biopsy may be replaced with lack of accurate confirmatory tests – LB or LSM – to ensure the absence of liver disease.

R: Accordingly, we added the following sentences: “Furthermore, transient hepatic elastography was also not performed because it was not accessible at the time of inclusion of the patients. However, liver biopsy is not routinely performed in these patients in the real-life clinical practice, and many centers (mainly in the primary care set) still do not have easy access to transient hepatic elastography.”

• I am not sure I understand the sentence “These patients represent the largest subgroup of heavy-drinkers on clinical practice”. I guess it depends on which healthcare setting one operates in?

R: We clarified the sentence as following: “These patients represent the largest subgroup of heavy-drinkers on clinical practice (particularly in Primary Care Clinics)”

Submitted filename: RESPONSE TO THE REVIEWERS - 2021.05.09.docx

Click here for additional data file.

12 May 2021

Outcomes of excessive alcohol drinkers without baseline evidence of chronic liver disease after 15 years follow-up: heavy burden of cancer and liver disease mortality

PONE-D-21-05110R2

Dear Dr. Machado,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Rudolf E. Stauber, MD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

The issues raised by Reviewer 3 have been answered satisfactorily.

Page 18: please correct Kaplan Meyer --> Kaplan-Meier.

Reviewers' comments:

17 May 2021

PONE-D-21-05110R2

Outcomes of excessive alcohol drinkers without baseline evidence of chronic liver disease after 15 years follow-up: heavy burden of cancer and liver disease mortality

Dear Dr. Machado:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Professor Rudolf E. Stauber

Academic Editor

PLOS ONE