Patience Required
A few weeks ago, Delcath Systems (Nasdaq: DCTH) reported another strong quarter, but the stock continued to fight the headwinds caused by its May announcement that Hepzato would be entering the Medicaid National Drug Rebate Agreement (NDRA or sometimes called 340B) starting July 1st. Hospitals participating in 340B, known as Disproportionate Share Hospitals (DSH), receive a 23% discount off Hepzato’s $187,500 list price. During its Q2 call, Delcath’s CEO, Gerard Michel, mentioned that early evidence suggests about 50% of Hepzato kit purchases were made by DSH, which should result in a blended average price of roughly $167,500 per kit for 3Q2025.
The company lowered its FY2025 guidance from $94-$98 million to $93-$96 million, representing a small decrease. After reporting nearly $44 million in revenue in 1H2025, the company now expects around $50 million in 2H2025 revenue based on its guidance. If Delcath investors are looking for one reason behind the sharp decline in the stock, it’s likely the lack of significant growth for 2H2025, more specifically 3Q2025, as short-term impacts of its participation in 340B are being absorbed.
Delcath ended Q2 with 20 active sites. The company revised its site activation guidance from 30 sites by year-end to 25-28, while maintaining its goal of 40 active sites by the end of 2026. During his prepared remarks, CEO Michel stated that the average number of Hepzato procedures performed per month per site in Q2 was two (later during the Q&A, he mentioned the average was slightly over two). With this new information, we can update our “Hepzato Math” formula, which we first shared in May, to see what 2H2025 sales could look like.
Using 25 active sites by year-end (bottom end of the company’s guidance), spreading the five new sites across Q3 and Q4, and bumping utilization up slightly in Q4, we arrive at Hepzato revenue of $22 million for Q3 and $26 million for Q4. Once Chemosat sales are included, approximately $1.75 million a quarter, Q3 revenue should be around $24 million, and Q4 around $28 million.
Delcath investors should be reassured that the temporary blunting of growth from 340B will be absorbed in Q3, and meaningful growth should return as early as Q4. Based on Delcath’s guidance of 40 active sites by the end of 2026, and assuming modest monthly utilization growth, Delcath is expected to reach an annualized run rate of nearly $185 million as it exits FY2026.
Driving Utilization
Delcath will take price increases; it may expand beyond 40 active sites, but the number that matters most for growth in the Hepzato Math formula is utilization. How does Delcath grow the utilization of Hepzato? There are multiple ways, and they are all somewhat intertwined. The first is familiarity. Hepzato is a drug-device combination. The kit is priced like an orphan drug, but using the kit is a complicated procedure that requires a team of specialists: a radiologist, a perfusionist, and an anesthetist. Think of these specialists as members of a band; like any band, the more they practice a song together, the better they sound. The same applies to Hepzato. The more procedures a team performs, the more skilled they become, which leads to better patient outcomes, increasing the likelihood they’ll use it more often. It’s no surprise that the “power users” of Hepzato (Moffitt, Jefferson, etc.) are sites that participated in the Phase 3 FOCUS study, with teams that have the most experience. As new sites start using Hepzato and their teams gain experience, we expect the average monthly utilization to steadily rise.
The second way Delcath can increase Hepzato’s utilization is through sequencing. Although we’ve discussed how important sequencing could be for Hepzato’s medium-term growth, it wasn’t until the company’s second-quarter call that they explicitly addressed sequencing in their prepared remarks. CEO Michel stated, “Another potential area of development includes the combination sequence with immunotherapy agents such as immune checkpoint inhibitors… the upcoming readout from the randomized Phase 2 CHOPIN trial is expected to inform the feasibility of these combination approaches.” With top-line data from CHOPIN expected in October, and assuming these results are positive, Delcath will have prospective sequencing data that its medical science liaisons can share with oncologists and radiologists. Sequencing could shift Hepzato from a second-line therapy (which we assume it is for most active sites) to a more prominent 1B position, with systemic therapy occupying the 1A spot. Moving up in the treatment hierarchy for sites that adopt the sequencing approach should help increase Hepzato’s monthly utilization numbers.
The last way Delcath can increase the use of Hepzato is to expand beyond metastatic uveal melanoma (mUM). The company is conducting studies with Hepzato in metastatic colorectal and breast cancer, but in his prepared remarks, CEO Michel also mentioned, “There is strong interest in intrahepatic cholangiocarcinoma and metastatic cutaneous melanoma, among others.” IPI/NIVO is commonly used for metastatic cutaneous melanoma. Although the CHOPIN study is only enrolling mUM patients, data showing the benefit of sequencing IPI/NIVO and Hepzato could lead to additional off-label use for Hepzato.
Patience Rewarded
Delcath’s stock has fallen from $18 to $10 since the company announced its 340B news. Even a strong Q2 could only generate a day or two of positive momentum before the stock dropped again to $10. While Delcath’s Hepzato business remains robust, many investors find the opportunity cost of holding through a stagnant Q3 too high. Still, for those willing to patiently hold Delcath through this transient quarter, the potential rewards should be significant.
One Scenario We Hadn’t Anticipated
Last week, RenovoRx (Nasdaq: RNXT) announced its Q2 results, which included just over $400,000 in RenovoCath sales. This marks only the second quarter of commercial sales for RenovoCath. Sales may be modest, but the number of new cancer centers approved to purchase the device increased from five in Q1 to thirteen in Q2, suggesting more meaningful growth could be on the horizon. Still, investors shouldn’t focus too much on the income statement at this stage. The company’s CEO, Shaun Bagai, said during the conference call that 2025 will be a “learning year” for the company commercially, and the significant sales increase is likely to occur in 2026.
While the income statement matures, investors were eager to receive data from the second planned interim analysis of the company’s ongoing Phase 3 study with RenovoGem (RenovoCath + gemcitabine) in patients with locally advanced pancreatic cancer (LAPC). We previously discussed the potential scenarios investors might expect from this second interim data: “A statistically significant survival benefit from the second interim could send the stock soaring. The stock should also react favorably if the second interim demonstrates an improving statistical trend in survival, providing optimism that statistical significance can be achieved in the final analysis next year. Finally, there’s a possible “soft landing” scenario if the statistics trend the wrong way for survival in the second interim, but safety remains significantly better than standard-of-care (SoC).” However, there was one scenario we hadn’t anticipated, and that was receiving no data at all.
In the company’s Q2 press release, it announced that “The DMC (Data Monitoring Committee) has concluded its review and has recommended that the Company continue with the trial. To avoid compromising the integrity of the trial with the FDA, and after discussions with the DMC and consultation with its regulatory advisors, RenovoRx elected to defer publishing the interim data.” CEO Bagai added during the quarterly call, “…we’re nearing the end of the trial, and there’s enthusiasm for the potential positivity, we didn’t want to introduce any bias at this stage in the game.” There’s no question that releasing interim data in the middle of an ongoing pivotal study can bias the study, which is why it’s uncommon. Yet, Renovo released data from its first interim, setting the precedent that it would do the same with the second interim. But maybe the company had a good reason for releasing data from the first and also for not releasing data from the second.
First Good Reason
When Renovo started its registrational Phase 3 trial, it only had limited uncontrolled data from fifteen LAPC patients in a Phase 1 study. It had not conducted a proper Phase 2 study with RenovoGem before launching Phase 3. As a result, the first interim analysis from Phase 3, triggered after 26 events (deaths), can be seen almost as Phase 2 data, offering preliminary evidence of RenovoGem’s potential benefit versus SoC in LAPC. During the Q2 call CEO Bagai described the rationale for sharing data from the first interim, “We went to the phase three, not having done the phase two…taking a very early look at 30% of the data (first interim) would give us confidence that we are on the right track almost in lieu of a phase two trial.”
In that initial interim, RenovoGem demonstrated an impressive six-month median overall survival (mOS) improvement compared to SoC, IV gemcitabine, and nab-paclitaxel (21.5 months versus 15.5 months). The p-value was an encouraging p=0.051, but still below the strict significance threshold of p<0.0001. Nonetheless, like any positive Phase 2 study, it provided the company and stakeholders with confidence in the potential for Phase 3 success, and importantly, enabled the company to raise the necessary funds to continue the trial.
Second Good Reason
The second interim analysis was triggered after 52 events. Investors expected this update to include similar detailed data as the first, such as the primary endpoint mOS results, the associated p-value, and safety information. However, the DMC recommended that, to maintain the study’s integrity, the company should withhold releasing any data from this interim. As mentioned earlier, CEO Bagai indicated on the call that concerns about potential bias influenced the DMC’s advice. As a result, without data, investors are left to interpret what the DMC’s recommendation might mean for the chances of success in the final analysis.
It is reasonable to assume that RenovoGem did not reach the statistical threshold of an improvement in mOS with p<0.008 compared to SoC; otherwise, the study would have been stopped for efficacy. This leaves two other possible scenarios: either the p-value for mOS for RenovoGem versus SoC improved to less than p=0.051, or it worsened. Remember, the p-value for significance in the final analysis (86 events) is p<0.048.
In the scenario where the p-value from the second interim improves over the first interim (for example, p=0.045), indicating a strong trend that RenovoGem is likely to show a statistically significant improvement in mOS in the final analysis, the most notable bias from releasing data would be study participants refusing to stay on the SoC arm. LAPC is a death sentence; extending survival by months is the goal. If study participants learn that RenovoGem likely offers meaningful survival benefits compared to SoC, those assigned to the SoC arm would have a strong reason to opt out and seek treatment at a hospital offering RenovoCath commercially. The commercial availability of RenovoCath, launched this year, marks a significant difference between the first and second interim analyses. If RenovoGem demonstrated an improving statistical trend in mOS versus SoC in the second interim, and the company shared those data, it could unquestionably impact the study’s integrity.
In a scenario where the p-value from the second interim is worse than the first interim (for example, p=0.06), indicating a strong trend that RenovoGem is unlikely to show a statistically significant improvement in mOS in the final analysis, the most noticeable risk of releasing data would be patients choosing to withdraw from the study entirely, believing RenovoGem offers little benefit compared to SoC. A worsening p-value indicates that the six-month survival benefit observed with RenovoGem in the first interim has narrowed, but it would be surprising if the numerical survival benefit has completely evaporated. As long as survival is still numerically in RenovoGem’s favor, and there remains an outside chance of success in the final analysis (otherwise DMC would have stopped the study for futility), it shouldn’t have a meaningful impact on the ability to recruit patients into the study.
Body Language
In our view, it seems most plausible that the DMC recommended against sharing data from the second interim analysis because there is a positive statistical signal indicating that RenovoGem is likely to meet the mOS in the final analysis. However, this assumption requires investors to make a considerable leap of faith. Additionally, it demands a significant amount of patience—something that, as we discussed earlier with Delcath, is currently in very short supply.
During their recent call, the company stated that as of August 12th, 95 patients had been randomized and 61 events had occurred. The study aims to enroll a total of 114 patients, with the final analysis expected to take place after the 86th event. Renovo has indicated that they anticipate the last patient will be enrolled and randomized by the end of this year or early 2026. We estimate that top-line data will not be available any earlier than late 2026, and more likely in 1H2027.
Even if investors accept the notion that the DMC’s body language suggests the study is trending toward success, as we pointed out, it still requires a great deal of faith and patience until the final data are released. However, the potential rewards from a successful Phase 3 study in locally advanced pancreatic cancer (LAPC) would be substantial. We remind investors that in 2024, NovoCure (Nasdaq: NVCR) gained >$1b in market cap after showing a two-month survival improvement with their Tumor Treatment Fields (TTF) technology in a Phase 3 LAPC study.
Tug-O-War: Science Vs Stock
In public biopharma companies, there is often a tension—whether obvious or subtle—between the scientific and financial aspects of the business. The scientific side is generally more cautious about sharing positive data, prioritizing peer approval through podium presentations at scientific symposia and peer-reviewed publications. This side often underappreciates the cost of capital. In contrast, the financial side tends to be more cavalier about sharing positive data, focusing on public approval as reflected in the stock price while being acutely aware of the cost of capital.
Following the meeting with the DMC, it appears that the scientific side has taken precedence, as they decided not to release any data from the second interim analysis. However, if the company’s stock continues to languish and its balance sheet starts to show signs of vulnerability, its resolve not to share data, assuming positive, will be tested.