key: cord-0868489-blsohibp
authors: Amaddeo, Giuliana; Brustia, Raffaele; Allaire, Manon; Lequoy, Marie; Hollande, Clémence; Regnault, Hélène; BLAISE, Lorraine; Ganne-carrié, Nathalie; Séror, Olivier; Larrey, Edouard; Lim, Chetana; Scatton, Olivier; El mouhadi, Sanaa; Ozenne, Violaine; Paye, François; Balladur, Pierre; Dohan, Anthony; Massault, Pierre-Philippe; Pol, Stanislas; Dioguardi burgio, Marco; Vilgrain, Valérie; Sepulveda, Ailton; Cauchy, Francois; Luciani, Alain; Sommacale, Daniele; Leroy, Vincent; Roudot-thoraval, Francoise; Bouattour, Mohamed; Nault, Jean-Charles
title: Impact of COVID-19 on the management of hepatocellular carcinoma in a high-prevalence area
date: 2020-10-31
journal: JHEP Rep
DOI: 10.1016/j.jhepr.2020.100199
sha: 483b9657e88da09d633fbe4d2a6a60e31bc9604b
doc_id: 868489
cord_uid: blsohibp

BACKGROUND: Patients affected by HCC represent a vulnerable population during the COVID-19 pandemic and may suffer from the unusual allocation of healthcare resources. The aim of this study was to determine the impact of the COVID-19 pandemic on the management of HCC patients within six French referral centers of the metropolitan area of Paris. MATERIALS AND METHODS: We performed a multicenter, retrospective, cross-sectional study on the management of patients affected by HCC during the first six weeks of COVID-19 pandemic (exposed), compared to the same period in 2019 (unexposed). Were included all patients discussed in multidisciplinary tumor meeting (MTB) and/or undergoing radiological or surgical programmed procedure during the study period, in a curative or palliative intent. Endpoints were the number of patients with a modification in the treatment strategy, or a delay in decision-to-treatment. RESULTS: After screening, n=670 patients were included (n=293 Exposed to COVID, n=377 Unexposed to COVID). A decrease of the numbers of patients with HCC presented in MTB in 2020 (p=0.034) and with a first diagnosis of HCC (n=104 Exposed to COVID, n=143 Unexposed to COVID, p=0.083) was find. Modification in the treatment strategy was observed in 13.1% of patients, with no differences between the two periods. Nevertheless 21.5% versus 9.5% of patients experienced a treatment delay longer than 1 month in 2020 compared to 2019 (p<0.001). In 2020, 7.1% (21/293) of patients had a diagnosis of an active COVID-19 infection: 11 (52.4%) were hospitalized, and 4 (19.1%) died. CONCLUSIONS: In a metropolitan area highly impacted by COVID-19 pandemic, we observed a decreased number of cases of HCC, and similar rates of modification in treatment strategy, but with a treatment delay significantly longer in 2020 versus 2019.

, has no true precedent in modern times and is a rapidly evolving crisis worldwide. Cancer remains a heavy burden with more than 18 million cases diagnosed in 2018 according to the Globocan reports while the prevalence of cancer is largely beyond 43 million people(2).

Cancer patients represent a vulnerable population due to the acquired immunodeficiency, and are at increased risk of COVID-19-related serious events (intensive care admission, requirement for mechanical ventilation, or death)(3).

In a report from China on 72,314 COVID-19-positive patients, the crude-fatality rate was 5.3% among cancer patients with higher mortality rates among those aged over 70 years (4, 5) . In a Italian cohort of 335 infected patients, 20.3% of patients COVID-related who died had an active cancer (6) .

The context is complexified by the unusual allocation of healthcare resources to the pandemic, which might be responsible of collateral damages to the health care system on which patients depend: screening interruption, treatments cancelled or downgraded, followup delay, and patient's fear(7). According to the EASL guidelines, management and care of patients affected by hepatocellular carcinoma (HCC), the fourth cause of cancer-related death worldwide, should be maintained, even with minimal exposure to medical staff, including systemic treatments and liver transplantation workup(8). Similarly, the French Association for the Study of the Liver (AFEF) board suggests to maintain of the curative treatments of HCC in dedicated hospital units, preserved from COVID-19 patients(9). In France, the metropolitan area of Paris(10) is heavily impacted by the outbreak and nonurgent procedures might be re-scheduled, with an approximate delay of 1-2 months(9).

The aim of this study was to determine the impact of the COVID-19 pandemic on the management (compliance with MTB decisions on the time and type of treatment) of patients affected by HCC within six French academic referral centers of the metropolitan area of Paris. 

-Adult patients (>18 y old) affected by HCC (histological and/or radiological diagnosis according to the EASL criteria(12)), who received during the inclusion period:

-A proposal of treatment in multi-disciplinary tumour board (MTB) meetings.

OR -A programmed surgical or radiological procedure, such as liver resection (LR) or any interventional radiology (IR) procedure: percutaneous ablation (radiofrequency, microwaves or irreversible electroporation), Trans-arterial-chemo-embolization (TACE);

Selective internal radiation therapy (SIRT).

Patients fulfilling the inclusion criteria and exposed to the pandemic between March 6 th to April 17 th 2020 were considered as cases (Exposed_COVID), and those fulfilling the same inclusion criteria between March 6 th to April 17 th 2019 (Unexposed_COVID) were considered as controls.

The impact of the pandemic on the management of the target population affected by HCC The treatment strategy was considered "downgraded" in case of treatment switch from "curative intent" to "palliative intent" or "best supportive care".

In the cohort of patients exposed to COVID-19 in 2020, were identified patients with a 

No a priori sample size calculation was realized, and the whole cohort of patients according to the inclusion criteria was included.

Descriptive results were expressed as median with interquartile range [IQR] for continuous variables and as numbers (percentages) for categorical data. Baseline characteristics of patients and HCC were compared between the two-period cohorts using Mann Whitney test for continuous data, and chi square test or Fisher exact test for qualitative variables.

Variables significantly different between the two cohorts were used as variables for adjustment.

The main criteria, proportion of modified treatment was compared between the two period cohorts and expressed as crude Odd Ratio (OR) with its 95% confidence interval (95% CI).

Baseline characteristics found different between the two period cohorts (with p value <0.10)

were tested in a logistic regression model giving adjusted OR (aOR) with its 95% CI. The main criteria was analyzed in the whole study population proposed for surgical, radiological and medical treatment and separately in patients with a first diagnosis of HCC and patients in follow-up for a known HCC. A logistic regression model was used to test variables independently associated with main criteria aORs are presented with their 95% CI.

No multiple imputations were used. A p value ≤0.05 was considered significant.

All analyses were performed using IBM SPSS V26 and Stata V13.0 (Statacorp, College Station, Texas).

J o u r n a l P r e -p r o o f

We screened 1661 patients in six centers for eligibility: either presented in a MTB (n=543 Exposed_COVID, n=804 Unexposed_COVID) or with a programmed surgical or radiological procedure in intention to treat between March 6 th to April 17 th in 2019 and 2020. A detailed flowchart is represented in Figure 1 . Among them 723 were excluded because not affected by HCC, and further 268 given the absence of any treatment proposal (follow-up or requiring further explorations). Six hundred seventy patients were included, and represented the study population (n=293 Exposed_COVID, n=377 Unexposed_COVID). The population's gender was male in 82.6% of cases, and the median age of 67 [60-74] years: no differences were observed in terms of patients' characteristics (Child-Pugh score, MELD) or tumor patterns (maximum diameter, portal vein invasion, AFP) between the two periods ( Table 1 ).

Among the group of patients exposed to the pandemic, 7.1% (n=21) had a diagnosis of COVID-19 infection.

The absolute number of patients affected by HCC -including those with a first diagnosis - We observed no significant differences in the decision-to-treatment interval between 2019 and 2020 (Table 1) for each treatment class. Nevertheless, a higher rate of patients experienced a treatment delay longer than one month in 2020 compared to 2019 (21.5%, n=63 versus 9.5%, n=36, respectively; p<0.001). The reasons for this delay were different according to the study period (Table 1) . When focusing on the COVID period, a higher rate of patients requiring an interventional procedure experienced a delay longer than one month Overall, thirty-six patients (5.4%) accepted to be included within a study protocol, with no differences in the inclusion rates between the two periods (n=12, 4.1% versus n=24, 6.4% in 2020 versus 2019 respectively, p=0.228).

Finally, apart from a higher rate of canceled consultations, the outpatient models have changed with a significantly greater use of teleconsultation during the pandemic (7.8% n=21 versus 1.4% n=5, respectively, p<0.001) (table 1) .

A modification in the treatment strategy (between the treatments proposed during MTB and those finally received) was reported in 13.1% (n=88) of patients, with no differences between the two periods (13.3%, n=39 in 2020 versus 13%, n=49 in 2019; p=0.91) ( Figure 3 ).

No differences were observed in the treatment distribution: neither of the treatment intent (Table 1) .

• Subgroup with active HCC, first diagnosis (Table 2 (Table 2) .

• Subgroup with active, recurrent HCC (Table 3) This subgroup of 423 patients (n=189 in 2020 versus n=234 in 2019) presented similar characteristics between the two periods. Nevertheless, the rate of patients with a delay of treatment longer than one month was significantly higher in 2020 versus 2019 (n=44, 23.3% versus n=11, 4.7%, respectively; p<0.001) ( 

In univariate analysis, the diagnosis status (first diagnosis or follow-up), the type of treatment proposed and the period (2019 vs 2020) were associated with a significant delay or change in strategy of treatment at the univariate analysis. On multivariate analysis, only the period (2020) was independently associated with delay or change in strategy (aOR = 9.661 (CI95% : 2.85-32.72), p<0.001) ( Table 4) .

When focusing on variables associated with a treatment delay longer than one month, the period and the type of treatment proposed were found to be significant at the univariate analysis. After adjusting, only the period (2020) was independently associated (aOR = 9.323 ((CI95%:2.74-31.69) p<0.001). (Table 4 )

The COVID-19 pandemic has suddenly shattered the processes of health care in general, and for patients affected by cancer in particular. In France, similarly to other Western countries, J o u r n a l P r e -p r o o f planned clinical activities were reduced and postponed to minimize the risk of viral transmission but also given the reattribution of health professional to support COVID units.

Scientific Societies recommendations (14) try to deal with the current situation, but the real impact of COVID pandemic in HCC management is still unknown. A recent multicentric Italian study reported the outcomes of COVID19 infection in cirrhotic patients (15) . In contrast, our study focused on HCC and the treatment received in this period.

In this study, involving six referral academic centers within the metropolitan area of Paris, highly affected by the pandemic, we observed a significant decrease over the weeks in the The rate of patients with a treatment delay longer than one month was significantly higher in 2020, compared to 2019, and this observation is supported by the multivariate analysis, in which the period 2020 was found to be a strong independent predictor of treatment delay or cancel. This finding could be explained by a decrease access to operating room, interventional radiology facilities, post-operative intensive care units and ventilators, forcing physicians to delay the patient's treatment. However, we found a shorter interval between presentation in multidisciplinary tumor board and treatment in 2020 compared to 2019. This difference could be explained by the high numbers of treatments delayed in 2020 after the period considered. Consequently, the few patients treated during the COVID19 period in 2020 has a shorter delay to treatment compared to 2019. Most of the patients with a treatment delayed more than one month have a recurrent HCC. These data suggested than physician tend to treat more quickly patients with a new HCC during the COVID19 period.

We could hypothesize than patients with new HCC tends to be treated quickly due to the absence of knowledge of tumor biology and pressure of patients receiving a recent diagnostic of cancer. In patients with HCC recurrence, the physician tends to know the tumor biology and could adapt the delay of treatment accordingly.

Not surprisingly, the significant shift towards the rates of teleconsultation in 2020 versus 2019 is related to social distancing measures as well as patient's fear, due to the pandemic.

The impact of teleconsultation on the patient-physician relation, as well as on patient's understanding of HCC diagnosis and treatment should be further explored.

Finally, 7.1% of patients exposed to the pandemic were affected by COVID, and almost half of them required hospitalization. With four deaths (19%,) the crude-fatality rate in this study population is significantly higher than what reported in China(4,5), Italy (6) or UK (17) We have to underline that the COVID19 diagnosis was performed by clinicians and was not a systematic prospective assessment of patient's symptoms with a predefined PCR or CT scanbased diagnostic algorithm.

The main limit of this study is represented by the uncertain delay of treatment report in 2020, (at the time of the study, some 7% of patients had not received the treatment planned) preventing to calculate a reliable delay to be compared to 2019. With this study, we could not assess the impact on patient survival of treatment delay and of the reduced rates of MTB presentations. Moreover, the observations reported in this study could not be generalized to areas with a low incidence of COVID19 infection.

To better understand the impact of COVID19 on HCC management, we will need to gather the reasons for the reduction in MTB presentations and the delay in treatments. These data This study is a very early snapshot (six weeks) of the French lockdown (ten weeks): it offers the first report of a homogeneous population affected by HCC within a network of highvolume academic French centers in COVID19 pandemic area.

Based on the available data, the pandemic seems to impact the management of patients affected by HCC in one fourth of the population, due to a delay in treatment realization but not with a modification in treatment strategy.

The mid-term follow-up of this cohort will inform about the impact of the pandemic on the long-term HCC management, waitlist dropout and mortality. 

Comments to the author(s):

Reviewer #1: The authors studied the impact of covid-19 pandemic on management of HCC in Paris, by measuring the compliance with multidisciplinary tumor meeting's decision on the time and type of treatment in HCC patients in the same period of 2019 and 2020. The study demonstrated that a higher number of HCC patients had a treatment delay, and lower number of patients were diagnosed in the first 6 weeks of covid-19 pandemic compared with the same period in 2019. The finding is supporting the negative impact of covid-19 pandemic on HCC management. concerns 1. Table 2 are suggesting the curative treatment for patients with a first diagnosed HCC was given more quickly in 2020 than that in 2019, which is NOT supporting your conclusion.

We agree with reviewer, however we think that the shorter interval between presentation in multidisciplinary tumor board and treatment in 2020 compared to 2019 is due to the high numbers of treatments delayed after the lockdown period in 2020. Consequently, the lower numbers of patients treated during the lockdown period has a shorter delay to treatment in 2020 compared to 2019. We added several sentences in the discussion to highlight this point: "However, we found a shorter interval between presentation in multidisciplinary tumor board and treatment in 2020 compared to 2019. This difference could be explained by the high numbers of treatments delayed in 2020 after the period considered. Consequently, the few patients treated during the COVID19 period in 2020 has a shorter delay to treatment compared to 2019." 2. The number of patients with MTB-to-treatment interval > 1 month in those with first diagnosis of HCC was 19 (18.3%) in 2020, which is not different from that in 2019 (17.5%). While, there were more patients with MTB-to-treatment interval > 1month and with a recurrent active HCC in 2020 than those in 2019, suggesting the delayed treatment was more likely happened in patients with a recurrence active HCC. Could you discuss it.

We thank the reviewer for his comment. Obviously, most of the delayed treatments were in patients presenting with a recurrent HCC. We added several sentences in the discussion to discuss this point: "Most of the patients with a treatment delayed more than one month have a recurrent HCC. These data suggested than physician tend to treat more quickly patients with a new HCC during the COVID19 period. We could hypothesize than patients with new HCC tends to be treated quickly due to the absence of knowledge of tumor biology and pressure of patients receiving a recent diagnostic of cancer. In patients with HCC recurrence, the physician tends to know the tumor biology and could adapt the delay of treatment accordingly" 3. The type of the planned treatment which have been delayed or not delayed for one month in patients with a recurrence active HCC should be analyzed to demonstrated which type of treatments is most likely to be delayed.

We thank the reviewer for his precision. From the group of patients during the COVID period (2020) we selected the subgroup with active recurrent HCCC. As suggested, we analyzed the treatment differences between those with MTB-to-treatment interval >1 and <1month.

We added this detail within the revised version of the manuscript.

When focusing only on the COVID period, we observed differences between the classes of treatment with a MTB-to-treatment delay > 1month or <1month (Suplementary table 2) .

Overall 

We agree with the reviewer and add a sentence in the discussion: "The lower number of new HCC diagnosed during the COVID19 period could be also explained by the travel limitation which impaired patients from other area to reach the medical centers in Paris"

Reviewer #2 Amaddeo et al. reported a cross-sectional retrospective study on pattern of management of HCC during COVID-19 pandemic (6 Mar to 17 Apr 2020) in multiple centres of Paris and compared to the same period in 2019.

Key findings were: 1) low proportion of treatments in 2020 (43.7%) as compared to 56.7% in 2019; 2): longer treatment delay was observed in 2020 as compared to 2019; 3) Higher rate of telemedicine in 2020 (7.8%) as compared to 1.4% in 2019; 4) No difference in the treatment given compared to prosed in tumour board between the two periods. 5) in subgroup analyses, HCC diagnosed in 2020 tended to be larger in tumour size.

Major comments 1. The comparison between the two periods may be subject to bias due to lack of matching though there is no statistically significant difference in patients' characteristics.

We agree with the reviewer, however the two groups were very closed in term of main features. Moreover, we preferred not to perform matching between the two groups in order to provide a detailed description of HCC clinical care and assess the consequence of COVID19 pandemic in HCC management in real life.

2. Most of the findings have been reported by another Italian group which somehow has limited the novelty of the paper.

We agree with the reviewer, an Italian group reported the outcomes of COVID19 infection in patients with cirrhosis. The originality of our study is the focus on HCC and the deep detail about treatment received in this period.

We add sentences in the discussion to highlight this point: "A recent multicentric Italian study reported the outcomes of COVID19 infection in cirrhotic patients (doi: 10.1016/j.jhep.2020.06.001). In contrast, our study focused on HCC and the treatment received in this period."

3. More breakdown of delayed treatment modality should be given; Are those delayed treatment procedurerelated or medication-related?

When focusing on the COVID period, a higher rate of patients requiring an interventional procedure experienced a delay longer than one month (interventional procedure: <1M n=100 (54.3%) vs >1M n=75 (68.8%)), compared to those requiring medical treatment ( J o u r n a l P r e -p r o o f BSC= best supportive care, LS= liver surgery, IR= interventional radiology, MTB = multidisciplinary tumor board *LT and the related workup were excluded J o u r n a l P r e -p r o o f 

The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies

Clinical Practice Guidelines: Management of hepatocellular carcinoma

Clinical impact of COVID-19 on patients with cancer (CCC19): a cohort study

COVID-19 epidemic: Proposed alternatives in the management of digestive cancers: A French intergroup clinical point of view

High rates of 30-day mortality in patients with cirrhosis and COVID-19

Management of primary hepatic malignancies during the COVID-19 pandemic: recommendations for risk mitigation from a multidisciplinary perspective

The Lancet Gastroenterology and Hepatology

Assumptions for Disparities in Case-Fatality Rates of Coronavirus Disease (COVID-19) Across the Globe

Could you please discuss the reason why the treatment modification was not different between 2019 and 2020 in your cohorts, We thank the reviewer for his comments. The period of the current study was the beginning of the lockdown in France

The table 2 needs to be double-checked or filled with number of patients with missing data (for example, Alfa-feto-protein)

However these data are still speculative and it is difficult to have the true infection rate of the population in the same period. So, we prefer not to speculate about a higher infection rate in HCC patients than in the general population It is also difficult to compare outcomes of patients with HCC and COVID19 with patients with HCC and without COVID19. Among all the HCC patients infected by COVID19 (n=21)

2%) was lower than observed (34% at 30 days) in cirrhotic patients (mostly without HCC) in an Italian multicentric study

To note, less patients required hospitalization in our series (52.4%) compared to the Italian cohort (96%) suggesting a different severity profile at baseline