陳民章博士：藥明康德，讓“不可能三角”成為可能｜逐夢(mèng)者說(shuō)｜Bilingual

但鮮有人知的是，早在多年以前，藥明康德就洞察到了這一重要的產(chǎn)業(yè)趨勢(shì)，并開(kāi)始提前布局。

成立于2018年、專(zhuān)注多肽和寡核苷酸藥物的CRDMO平臺(tái)——WuXi TIDES，就是藥明康德化學(xué)業(yè)務(wù)平臺(tái)悄然邁出的一步“大棋”。

陳民章博士認(rèn)為，這里的CRDMO，有著兩個(gè)維度。

第一個(gè)維度是明線，自然是順著多肽和寡核苷酸藥物早期研究、中期開(kāi)發(fā)、后期商業(yè)化生產(chǎn)的箭頭往前走：

在前兩個(gè)階段，藥明康德的客戶(hù)數(shù)均實(shí)現(xiàn)了兩位數(shù)的增長(zhǎng)；

在第三個(gè)階段，藥明康德在過(guò)去的12個(gè)月里完成了2000多批次的生產(chǎn)，2019年以來(lái)，已完成10個(gè)新化學(xué)實(shí)體的工藝性能確認(rèn)（PPQ），成功率高達(dá)100%。

另一個(gè)維度則是暗線——依托豐富的化學(xué)合成經(jīng)驗(yàn)，WuXi TIDES建立了一條穩(wěn)健的供應(yīng)鏈。

用陳民章博士的話(huà)說(shuō)：

“我們可以把供應(yīng)鏈一直推到合成這些藥物所需的最小單體，而且還可以生產(chǎn)出特殊的定制分子?！?/strong>

這樣一條自給自足的供應(yīng)鏈，可以用更快的速度，更低的成本來(lái)合成所需的藥物分子。

具體是怎樣“定制”的呢？

說(shuō)到這里，陳民章博士想起了一次讓他印象深刻的合作經(jīng)歷。

那是一家生物技術(shù)公司，當(dāng)時(shí)正在開(kāi)發(fā)一款治療心血管疾病的寡核苷酸藥物。

其中一個(gè)重要步驟是，需要給藥物連上一種叫做GalNAc的“向?qū)Х肿印?，帶著藥物前往人體內(nèi)正確的部位。

問(wèn)題在于，他們需要的分子是一個(gè)“限定款”，能夠提供原料的供應(yīng)商寥寥無(wú)幾，因此成本極為高昂。

就在心急如焚之時(shí)，他們找到了藥明康德。

為了讓客戶(hù)趕上重要的項(xiàng)目時(shí)間節(jié)點(diǎn)，藥明康德迅速召集了6個(gè)團(tuán)隊(duì)投入到這個(gè)項(xiàng)目中。

陳民章博士坦言，那次任務(wù)極具挑戰(zhàn)：

“這種藥物分子非常復(fù)雜。能調(diào)動(dòng)這么多的團(tuán)隊(duì)去平行推進(jìn)，這在其他地方基本上是不存在的?！?/strong>

盡管如此，團(tuán)隊(duì)的科學(xué)家們?nèi)匀挥y而上，順利完成了任務(wù)。

他們不僅替客戶(hù)單獨(dú)合成了所需的GalNAc分子，還平行推進(jìn)著其他步驟：一邊優(yōu)化合成路徑，減少副產(chǎn)物；一邊開(kāi)發(fā)制劑工藝，將API損耗嚴(yán)格控制在1%以?xún)?nèi)。

最終，在短短14個(gè)月內(nèi)，兩款候選藥物均被推進(jìn)到IND（臨床試驗(yàn)申請(qǐng)）階段。

藥明人用行動(dòng)，再次詮釋了什么是“藥明速度”。

他們用盡全力守護(hù)的，不止是客戶(hù)的希望，更是千千萬(wàn)萬(wàn)患者的希望。

類(lèi)似的案例還有很多。由于藥明康德在多肽和寡核苷酸藥物領(lǐng)域的一體化能力，從原料藥到制劑開(kāi)發(fā)，不同團(tuán)隊(duì)之間能形成極佳的協(xié)同效應(yīng)。

截至2025年上半年，使用TIDES平臺(tái)原料藥及制劑一體化服務(wù)的分子數(shù)已達(dá)69個(gè)，與兩年前相比，這個(gè)數(shù)字已經(jīng)翻了一倍多。

極致運(yùn)營(yíng)：從俄羅斯方塊說(shuō)起

隨著合作伙伴和項(xiàng)目數(shù)量的快速增加，也給公司的日常運(yùn)營(yíng)帶來(lái)了新的挑戰(zhàn)。

怎樣才能合理調(diào)配產(chǎn)能，更高效地完成客戶(hù)的項(xiàng)目呢？

陳民章博士給出了一個(gè)意想不到的比喻：俄羅斯方塊。

在這款有著40年歷史的經(jīng)典游戲里，玩家需要將由四個(gè)小方塊組成的不同形狀填至下方的空隙。一旦某一行被填滿(mǎn)，就可以消除得分。

“我們安排反應(yīng)釜的過(guò)程，有點(diǎn)像俄羅斯方塊，只是更加復(fù)雜?！?/p>

在實(shí)際的合成中，每一個(gè)批次都有特殊的要求，需要反應(yīng)釜、過(guò)濾區(qū)、烘干區(qū)等進(jìn)行不同的排列組合，每個(gè)組合就像是不同形狀的俄羅斯方塊。

與此同時(shí)，整個(gè)生產(chǎn)基地的總產(chǎn)能是固定的。

就像游戲畫(huà)面里，能輾轉(zhuǎn)騰挪的空間就這么大。

換句話(huà)說(shuō)，如果在固定的產(chǎn)能空間里，盡可能填入更多的反應(yīng)組合，留出最少的空窗，就能將生產(chǎn)基地的使用率提升到極致。

據(jù)陳民章博士回憶，2004年藥明康德只有一個(gè)生產(chǎn)車(chē)間，36臺(tái)反應(yīng)釜。

那時(shí)只需要一位員工，用最簡(jiǎn)單的Excel?軟件，就可以排好生產(chǎn)計(jì)劃。

而如今，面對(duì)數(shù)千位客戶(hù)截然不同的項(xiàng)目需求，藥明康德的規(guī)模已經(jīng)擴(kuò)大到了當(dāng)初難以想象的量級(jí)：

五個(gè)生產(chǎn)基地、四十個(gè)生產(chǎn)車(chē)間、800多臺(tái)反應(yīng)釜；每年要生產(chǎn)超過(guò)三萬(wàn)批次的產(chǎn)品、平均每天約100批次……

如果還繼續(xù)安排專(zhuān)人用Excel?來(lái)排產(chǎn)，既不高效，也不合理。

更棘手的是，由于藥明康德的生產(chǎn)規(guī)模過(guò)于龐大以及原料藥生產(chǎn)流程的復(fù)雜性，市場(chǎng)上也沒(méi)有一個(gè)現(xiàn)成的軟件可以購(gòu)買(mǎi)。

面對(duì)這一難題，陳民章博士與同事們快速達(dá)成共識(shí)：

“既然市場(chǎng)上沒(méi)有，我們?yōu)槭裁床桓纱嘧约洪_(kāi)發(fā)一個(gè)系統(tǒng)呢？”

說(shuō)干就干。不久后，這個(gè)以“俄羅斯方塊”為靈感的系統(tǒng)誕生了——

只要輸入需求，它就會(huì)實(shí)時(shí)排好生產(chǎn)計(jì)劃。而且，即便有些項(xiàng)目發(fā)生了變動(dòng)，或是需要加入新的項(xiàng)目，這個(gè)系統(tǒng)也會(huì)實(shí)時(shí)生成新的計(jì)劃。

在智能系統(tǒng)的輔助下，藥明康德的反應(yīng)釜使用率從2020年的60%，提升到了今天的72%。考慮到清洗、維修所必需的時(shí)間，這個(gè)比例已接近實(shí)際所能觸達(dá)的極限。

更重要的是，從60%到72%，使用率的提升幅度達(dá)到了五分之一，也就是20%——

如果以目前40個(gè)生產(chǎn)車(chē)間計(jì)，使用率提升20%就相當(dāng)于憑空多出來(lái)了8個(gè)車(chē)間的產(chǎn)能，這是相當(dāng)難得的。

此外，藥明康德的生產(chǎn)已經(jīng)實(shí)現(xiàn)高度的自動(dòng)化——多肽藥物的自動(dòng)化生產(chǎn)比例達(dá)到80%，小分子藥物的自動(dòng)化生產(chǎn)比例更是達(dá)到90%。

對(duì)于這個(gè)系統(tǒng)解決的種種痛點(diǎn)，陳民章博士十分滿(mǎn)意——

它不但能幫工人“減負(fù)”、降低勞動(dòng)強(qiáng)度，人為操作的錯(cuò)誤風(fēng)險(xiǎn)也會(huì)明顯減少。

“而且每一次操作都有記錄，有很完整的數(shù)據(jù)?！?/p>

正是這些從小處著手的細(xì)密?chē)?yán)謹(jǐn)，筑就了藥明康德數(shù)十年來(lái)長(zhǎng)期穩(wěn)健發(fā)展的基石，守住了維系客戶(hù)信賴(lài)的源泉。

“我經(jīng)常和團(tuán)隊(duì)講，只要客戶(hù)親眼看到藥明康德每一天是怎么工作的，就會(huì)對(duì)合作充滿(mǎn)信心。”陳民章博士笑著補(bǔ)充道。

展望未來(lái)：要把“藥明經(jīng)驗(yàn)”種進(jìn)全球基地

說(shuō)到藥明康德的未來(lái)發(fā)展，陳民章博士給出了一個(gè)關(guān)鍵詞——全球化。

目前，藥明康德在新加坡、瑞士庫(kù)威、美國(guó)特拉華州米德?tīng)栴D的基地都在如火如荼的建設(shè)中，以滿(mǎn)足各地產(chǎn)業(yè)客戶(hù)的迫切需求、造福全球患者。其中，新加坡基地的產(chǎn)能還會(huì)以每年一個(gè)新車(chē)間的速度持續(xù)提升。

同時(shí)，陳民章博士深知，光是完成地理上的“布局”還遠(yuǎn)遠(yuǎn)不夠。

更為重要的，是怎么把“藥明經(jīng)驗(yàn)”復(fù)制到全球各個(gè)基地。

回望之前走過(guò)的25年，藥明康德已經(jīng)積累了大量工藝開(kāi)發(fā)和生產(chǎn)制造的經(jīng)驗(yàn)。

在規(guī)模效應(yīng)下，如果一個(gè)新車(chē)間需要100多名技術(shù)員和工人，藥明康德很容易在各個(gè)基地間調(diào)配資源，將有豐富經(jīng)驗(yàn)的團(tuán)隊(duì)迅速投入到新車(chē)間的建設(shè)中，很快將車(chē)間運(yùn)營(yíng)起來(lái)。

而在全球化的布局下，這些經(jīng)驗(yàn)將在全球基地之間無(wú)縫傳遞，更高效地為全球客戶(hù)提供服務(wù)，助力將更多新藥、好藥早日帶到患者身邊。

人們常說(shuō)，“質(zhì)量、速度、成本”是不可能三角，三者最多只能得其二。

“但我看未必?！?/strong>

陳民章博士認(rèn)為，藥明康德就是一個(gè)特例——

“我們的CRDMO商業(yè)模式，結(jié)合TIDES業(yè)務(wù)的高速發(fā)展，精益運(yùn)營(yíng)，以及全球的產(chǎn)能網(wǎng)絡(luò)，就可以同時(shí)實(shí)現(xiàn)這三點(diǎn)，也必將在未來(lái)更好地滿(mǎn)足客戶(hù)的需求?！?/p>

《道德經(jīng)》中有句話(huà)：“天下難事，必作于易；天下大事，必作于細(xì)?！?/p>

25年來(lái)，藥明康德之所以能做成“不可能”的難事與大事，之所以敢向“讓天下沒(méi)有難做的藥，難治的病”發(fā)起挑戰(zhàn)，靠的從來(lái)不是一鳴驚人、一步登天，而是從一件件正確的小事做起，先做到精細(xì)，再做到極致。

正如公司那句樸實(shí)無(wú)華的核心價(jià)值觀——“做對(duì)的事，把事做好”。

Making the “Impossible Triangle” Possible at WuXi AppTec

Editor’s Note: “Every drug can be made and every disease can be treated”. Over the past twenty-five years, from a single laboratory to a global network spanning Asia, Europe, and North America; from early chemical synthesis services to an integrated platform that connects Research (R), Development (D), and Manufacturing (M); from its very first customer to thousands of partners across more than thirty countries, WuXi AppTec has never stopped moving forward to realize its enduring vision.

In tribute to the journey and in celebration of new beginnings, we look back through the eyes of our “dream-makers”: revisiting the stories of partnership that empowered global innovation, and the unique spirit that continues to guide us toward the next chapter.

At the 2025 WuXi AppTec Investor Day, an analyst suspected she had misheard a number. She raised her hand and asked, “It typically takes two years for a brand-new production plant to move from construction completion to full utilization. You just said two months. Is that correct?”

“Yes, that is correct,” replied Dr. Minzhang Chen, co-CEO of WuXi AppTec.

The answer stirred a quiet murmur among the hundreds of investors and analysts in the room, a sound that signaled recognition more than surprise. This, after all, is the company that has spent years turning “WuXi Speed” into an operating system.

Chen’s own reaction, he admits, is often one of measured wonder. Not at a single headline or a quarterly crest, but at the cumulative force of teams that grind, iterate, and deliver day after day, year after year. If he could rewind time 17 years, even the confident scientist who joined WuXi AppTec in 2008 might not have predicted how far quality, capacity and efficiency would climb.

From Scientist to Platform Builder

When Chen arrived in 2008 to lead Process R&D, WuXi AppTec’s client roster numbered around 80, and the company employed around 100 process chemists focused mainly on early research. Today, WuXi AppTec is known for an integrated, end-to-end enabling platform spanning discovery through commercial manufacturing, serving thousands of global clients with more than 3,000 process chemists, matched by advances in technical depth.

The engine behind that expansion is the company’s CRDMO model, which links Research (R), Development (D), and Manufacturing (M) into a single accountable chain. Before CRDMOs, innovator companies typically stitched together different vendors across phases and modalities, a patchwork that raised complexity and failure risk. By contrast, an integrated platform concentrates dispersed steps, simplifies decision-making, and compresses time to market.

Chen often illustrates the model’s logic with a “funnel” he shares publicly each year. The top represents early discovery; the middle spans three clinical development stages; the base is commercial manufacturing. The numbers tell the story. In just 12 months, from October 2024 to September 2025, WuXi AppTec’s Chemistry platform synthesized and delivered more than 430,000 new small-molecule compounds, a figure that has exceeded 400,000 annually for three consecutive years.

Those “seed” candidates flow downstream. Project counts have risen not only at the front end but more steeply in later clinical phases. From 2023 to 2025, programs in preclinical/Phase 1, Phase 2, and Phase 3 climbed to 2,889; 374; and 87, respectively, with Phase 3 growing by 42.6%; commercial (M) projects expanded by more than 37%.

In a market where late-stage work attracts many suppliers, why do programs stay and expand on one platform? “Trust,” says Chen. Between early 2024 and mid-2025, 53 drugs involved in global licensing and/or acquisition deals were supported by WuXi AppTec’s platform. When rights changed hands, the new owners kept every one of those programs at WuXi AppTec, a retention rate of 100%.

In nine other cases, the acquiring company transferred projects from other providers onto WuXi AppTec. “They were brought by our former clients,” Chen notes. “They trust our quality.”

Early Bet, Built to Customize

Debate periodically flares up over whether small molecules are outdated. Chen’s view is pragmatic: For ease of use, small molecules will remain a patient-preferred option even as modalities diversify.

He also pointed out that beyond small molecules, there is a rise of complex modalities, particularly peptides, oligonucleotides, and their conjugates. Evaluate Pharma projected in August that the global peptide market will double in five years, with GLP-1 related therapies as a major driver, while oligonucleotides could quadruple by 2030 on the strength of central nervous system, cardiovascular, and rare disease applications.

Back in 2018, WuXi AppTec quietly placed a strategic bet: Building the WuXi TIDES CRDMO platform to support peptides and oligonucleotides across discovery, development, and commercial supply. In the past 12 months alone, the TIDES team completed more than 2,000 batches, achieving a 100% success rate in Process Performance Qualification (PPQ) for 10 New Chemical Entities (NCEs) since 2019.

The more subtle advantage is supply-chain depth. Leveraging synthetic chemistry know-how, the TIDES team pushes upstream to monomer-level building blocks and can manufacture special custom fragments, shrinking timelines and costs for clients.

One case stands out. A biotech developing a cardiovascular oligonucleotide therapeutic needed a highly specific GalNAc “guide” for targeted delivery. The variant was scarce and expensive. To meet the client’s need, WuXi AppTec mobilized six teams in parallel to synthesize the custom GalNAc, optimize routes to reduce byproducts, and develop drug product processes that keep line loss under 1%.

“Given the molecule’s complexity, you need many teams and capabilities to work together. This could not be found in any other companies.” Chen says.

In just 14 months, two candidates completed all preparations for clinical trial applications. That is “WuXi Speed,” applied to a new modality.

By the first half of 2025, 69 molecules are using the TIDES platform’s API and Drug Product services, more than double the number from two years ago.

Operational Excellence, Engineered

As partners and programs multiplied, day-to-day operations faced a different challenge: resource orchestration. Chen’s metaphor is delightfully un-corporate: Tetris?. Each batch has specific needs -- different reactors, different filtration requirements, different drying conditions, just like the blocks in the classic video game.

Total site capacity is finite, but the calculation is simple: Pack more “shapes” into the board with fewer idle gaps, and utilization rises.

Years ago, a single employee could schedule one plant with 36 reactors in an Excel? spreadsheet. Today, the network spans five sites, 40 plants, and more than 800 reactors producing over 30,000 batches a year; that’s roughly 100 a day. Neither Excel? nor any off-the-shelf system can manage that complexity, because few companies have ever worked at WuXi AppTec’s scale.

So, the Company built its own real-time planning system. The impact goes well beyond convenience. With high levels of automation and digital scheduling that squeezes out idle time, the platform turns planning discipline into real capacity.

Under the new system, API equipment utilization rose from 60% in 2020 to 72% in 2025. “If you factor in the downtime required for maintenance and cleaning, we are approaching the practical limit,” Chen explains.

Viewed another way, moving from 60% to 72% represents a 20% increase. With 40 production plants in operation, that gain is equivalent to adding eight more plants for clients, without pouring a single foundation.

As Chen puts it, “operational excellence” is a system, not a slogan.

And the system withstood the tests from global regulatory agencies. This March, WuXi AppTec’s two API manufacturing sites in Changzhou and Taixing passed FDA inspections with zero deficiencies. Notably, the Changzhou site underwent an unannounced inspection covering six major GMP systems and 21 FDA-approved products. The FDA completed the inspection a day ahead of schedule and found no observation.

Looking back at 2024, WuXi AppTec underwent 802 quality audits by global clients, regulators, and independent third parties, with a 100% pass rate and zero critical findings.

“Our sites are inspection ready any time.” Chen says. This simply reflects how WuXi AppTec adheres to global standards for quality, day after day, year after year.

The “Impossible Triangle,” Reframed

Looking ahead, WuXi AppTec is reinforcing capabilities where clients need them most. In Europe, the Couvet, Switzerland, site is adding spray-drying to address a clear market pain point. In the United States, the Middletown, Delaware, facility has completed its primary structure and is progressing through interior fit-out, with operations expected next year. In Asia, the first plant in the Tuas, Singapore, site is slated to be operational by 2027.

It’s not just about dots on a map. The task now, Chen says, is to infuse WuXi AppTec’s accumulated “operational excellence” at every site through redeploying experienced teams, standardizing training, and replicating best practices so new plants start up quickly and deliver consistently for global clients.

Conventional wisdom says you can pick two among quality, speed, and cost, but never all three. Chen does not agree. WuXi AppTec’s CRDMO model, combined with TIDES momentum, lean operations, and a global capacity network, aims to deliver all three simultaneously and, crucially, to keep learning so tomorrow’s baseline is higher than today’s.

He reaches for an old line from the

Dao De Jing